BTI 2014 LLC v Sequana S.A. & Ors

On 17 December 2008 the Second to Fifth Defendants, Antoine Courteault, Pierre Martinet, Clive Mountford and Martin Newell, attended a meeting of the board of a company called Arjo Wiggins Appleton Limited (‘AWA’). At that time, they were the four directors of AWA, now called Windward Prospects Limited, and AWA was a wholly-owned subsidiary of Sequana SA, the First Defendant (‘Sequana’).

The directors resolved at the meeting on 17 December 2008 to pay an interim dividend to the parent company Sequana of €443 million and resolved further that the payment of the dividend should be effected by way of set off against an equivalent amount of the intra-group receivable due to AWA from Sequana. The outstanding balance of the intra-group receivable after the payment of the dividend was €142.5 million.

On 18 May 2009 the directors of AWA held another board meeting. By this time there were only two directors, Mr Martinet and Mr Courteault. They held the meeting by telephone as Mr Courteault was in London and Mr Martinet was travelling to Turin. The directors resolved to pay a further interim dividend to Sequana to be satisfied by the release by AWA of €135 million of Sequana’s intra-group debt owed to AWA.

The payment by AWA of these two dividends is challenged in these proceedings by the Claimants on a number of grounds. The Claimants allege that the dividends contravened Part 23 of the Companies Act 2006 (‘CA 2006’). This is on the basis that the accounts on which the directors relied as showing there were sufficient distributable reserves to justify the payment of the dividends were incorrect and did not give a true and fair picture of the state of the company’s finances. It is also alleged that the decision to pay both dividends was a breach by the directors of their fiduciary duties towards the company. Finally it is alleged that the dividends constituted transactions which contravened section 423 of the Insolvency Act 1986.

The source of the Claimants’ complaint about the dividends lies almost 4000 miles away, in the sediment of the Lower Fox River in Wisconsin, USA. That sediment was heavily polluted during the 1950s and 1960s and the River has been subject to a complex and very expensive clean up operation pursuant to the US statute, the Comprehensive Environmental Response, Compensation and Liability Act 1980 (‘CERCLA’). Through a series of corporate acquisitions and asset transfers since then, the Claimant BAT Industries PLC (‘BAT’) is liable to pay for part of that clean up and AWA is liable to indemnify BAT for part of the monies BAT has to pay out.

That liability to indemnify BAT resulted in a provision being included for a number of years in AWA’s accounts to reflect the directors’ best estimate of the value of that liability. The Claimants assert that that provision was inadequate and further that the directors, when taking the various decisions under attack, should have taken more account of the possibility that, even if the provision in the accounts was a best estimate, that estimate could turn out to be wrong by a long way. They also assert that a large reduction in AWA’s capital which was approved two days before the payment of the December dividend was unlawful.

The trial which took place before me over 32 days was the trial of two sets of proceedings heard jointly:

In 1969 the National Cash Register Company (‘NCR’) purchased a company called Combined Paper Mills which owned a paper coating company operating in the Lower Fox River area. In 1970 NCR bought Appleton Coated Paper Company which also owned a paper coating facility there. These two businesses were later merged to become the Appleton Paper Division of NCR.

In 1978 the Appleton Paper Division of NCR was sold to BAT when it was acquired by a wholly-owned indirect subsidiary of BAT called Appleton Papers Inc (‘API’). Under the terms of that sale:

Thus, the first indemnity that is relevant for our purposes is this obligation entered into in 1978 by API and BAT to NCR.

When it acquired the Appleton Paper Division, BAT already had a paper business in the form of its subsidiary called Wiggins Teape Group Ltd. In 1990 the two businesses, API and Wiggins Teape Group Ltd were combined in a new company which had been incorporated in 1989 and was then named Wiggins Teape Appleton plc. It was demerged from BAT in 1990. Under the demerger agreement, Wiggins Teape Appleton plc and API agreed to indemnify BAT for liabilities arising from the demerged assets, that is for certain losses arising out of the neglect or default in the business or operations carried out by the de-merged group prior to the demerger, including any potential liability under the indemnity given by BAT in connection with the 1978 purchase.

Later in 1990, Wiggins Teape Appleton plc was merged with a French paper manufacturer called Arjomari Prioux SA and the company changed its name to Arjo Wiggins Appleton plc. In 1998 there was an agreement between NCR, API and BAT under which it was agreed that certain environmental liabilities of the three parties would be shared up to a total of $75 million as to 45% to be paid by NCR and as to 55% to be paid by API and BAT. The environmental liabilities covered by this allocation included not only the Lower Fox River but also ‘Future Sites’ as defined in the agreement. For our purposes the only Future Site relevant is the Kalamazoo River in the state of Michigan, USA.

Who would pay what share for costs above $75 million was still in dispute after the 1998 Agreement. In November 2005 there was an arbitration under the 1998 agreement by which it was determined that for costs above $75 million the split would be 60% to API and BAT and 40% to NCR.

In 2000 Arjo Wiggins Appleton plc was acquired by Sequana for €1.3 billion. Various restructurings within the Sequana group and sales of subsidiaries took place after 2000. One such transaction is significant in this narrative. In November 2001 API was sold by AWA to an employee buy-out vehicle Paperweight Development Corp. As part of this sale, AWA entered into an indemnity agreement whereby it indirectly indemnified API against all of its liabilities (save for a $25 million tranche which API had to cover), net of insurance recoveries, relating to the Lower Fox River (that is the liabilities that API had taken on under the 1998 settlement agreement). API in return assigned to AWA its rights to recover from any third parties. These rights against third parties included rights under certain historical insurance policies which had been taken out by BAT with a number of insurance companies between 1978 and 1986.

It was in the context of this 2001 sale and indemnity arrangement that AWA set up a Bermudan subsidiary called Arjo Wiggins Appleton (Bermuda) Limited which purchased a guaranteed investment contract referred to as the Maris Policy, discussed later. This was a policy issued with the insurer AIG on the deposit of $185 million in 2001. It provided AWA with a source of funds to pay for all aspects of the Fox River liability. The limits on the policy increased over the years but reached a maximum of $250 million by November 2008. The policy funds did not keep pace with inflation. The terms of the sale in 2001 also provided for AWA to have the right to control the defence of all claims within the indemnity given.

Following its disposal of its interest in API, AWA ceased to be a trading company. The movement of various businesses out of AWA and the loaning up to the parent Sequana of other receipts over the years resulted in AWA having on its balance sheet a very large, interest accruing receivable owed to it by Sequana.

The outcome of all this is that these proceedings were conducted on the basis that:

AWA had two other outstanding problematic liabilities in 2005. It had been found guilty of an infringement of the European Union competition rules and a substantial fine had been imposed, the precise amount of which was being contested on appeal to the court in Luxembourg. It also had responsibilities under the Wiggins Teape pension scheme for which it remained the principal employer despite no longer owning any businesses with employees. These were not resolved until early 2008.

The factual witnesses

All the witnesses giving factual evidence in the case gave evidence on behalf of the Defendants.

Mr Martinet started working for investment companies for the Agnelli family in 1992. Through that role he became a member of the supervising board of Sequana in 2004 and in May 2005 he took on the executive role of joint Deputy Managing Director of Sequana. Upon taking up that post, he became a director of various of Sequana’s subsidiaries including AWA. In July 2007 Mr Martinet ceased to perform an executive role at Sequana although he continued as a non-executive director of Sequana and as an executive director of AWA. He resigned from the board of AWA in May 2009 when AWA was sold by Sequana and from the board of Sequana in July 2014.

Mr Martinet was the principal witness for the Defendants and was cross-examined over eight and a half days during which he answered questions about very many meetings and documents from 2005 to 2009. He gave his evidence in excellent English although French is his native language. Mr Martinet struck me as an impressive, highly experienced business man and a careful and honest witness doing his best to answer questions to the best of his recollection, given that these events occurred a long time ago. Despite the very serious allegations being levelled against him in these proceedings, he did not become defensive and he resisted any temptation to embroider or improve the evidence he was able to give, distinguishing carefully between his actual recollection and speculation about what had happened. Indeed, it was striking that during the course of Mr Martinet’s long and tenacious cross-examination there were only one or two answers that Mr Smouha QC, acting for the Claimants, challenged as untruthful. In their closing submissions the Claimants did not invite me to reject anything that Mr Martinet said as being untruthful. I am confident that I can rely on Mr Martinet’s evidence as truthful both as to his recollection of events and as to his recollection of his thoughts and motivations at the time of these events.

Mr Courteault is Company Secretary of Sequana and General Counsel of the Sequana group. He has held those roles since May 2005. Between June 2005 and May 2009 he was a director of AWA. At the time of his appointment in 2005 he was primarily focused on various human resources and legal matters within the group, mainly arising out of a restructuring that took place in 2005. As the group at that time had about 300 subsidiaries all over the world, generating a large number of legal issues, his role was to provide strategic oversight and to liaise with external local legal advisers as necessary.

Following his initial briefings on the Fox River issue in mid 2005, Mr Courteault did not regularly engage with the issue and left the detailed numbers analysis to Mr Martinet and the other advisers involved. His understanding at the time of the figures used in the computations that I will describe later was much less thorough than Mr Martinet’s and his recollection of the detail correspondingly less certain. Mr Courteault also gave his evidence in very good English.

I found Mr Courteault also to be an honest witness. Again, I was not invited to disbelieve anything that he said in evidence and I find that both his written and oral evidence are truthfully given.

Mr Mountford is a qualified accountant who joined the tax department of the Wiggins Teape Group in January 1988 when it was a subsidiary of BAT. He became a director of AWA on 1 May 2002 and held that position until his resignation on 14 May 2009. Mr Newell is also a qualified accountant who joined the Wiggins Teape Group in 1982. He became a director of AWA at the end of April 2004 and held that post until he resigned on 14 May 2009. It is clear that before December 2008, the AWA board meetings had dealt only with formal matters – that is not surprising given that it was a non-trading, wholly-owned subsidiary.

Both men gave their evidence in an open and straightforward manner. I consider them entirely honest both in their evidence and more generally in their conduct as directors of AWA. They had a good understanding of what their duties were; when they were in doubt they made arrangements to get advice. They both emphasised in the course of their evidence that they were well aware of the significance of the transaction being proposed for December 2008 and the gravity of the criminal and civil consequences for them personally if they did not comply with the statutory requirements. I have no doubt that they are both cautious men and there was certainly no reason for them to put that caution aside in relation to these events.

Mr Lebard is currently the Chairman and CEO of Sequana which is the parent company of the Sequana group. He joined the Sequana group in 2002 and took on an executive role in 2004. Following a company reorganisation in 2005 he and Mr Martinet became joint Deputy Managing Directors of Sequana. In July 2007 Mr Lebard bought a 21.9% stake in Sequana with about €230 million from his family investment fund and took on the role of CEO. He therefore had a large personal stake in the success of the group.

Mr Lebard and Mr Martinet have known each other for many years and Mr Lebard describes Mr Martinet as his equal and trusted business confidante – he trusted his business judgement completely. Mr Lebard focused on the group’s day to day business needs and strategy whereas Mr Martinet focused on the group’s financial issues and also on particular issues facing the group such as the AWA liabilities. So far as he was concerned there was a large team of highly experienced and capable experts, including external advisers, tasked with managing the issue and he never had any reason to doubt that a prudent approach was being taken to drawing up the company’s accounts.

Mr Lebard came across in the witness box as a forceful personality who gave his evidence clearly and emphatically. It was clear to me that he prides himself on his business acumen and that he would be a tough negotiator. I accept his evidence as truthful.

Mr Thomas has been a certified chartered accountant since 1994 and began working for the Sequana group in late 2001. Between 2001 and May 2009, he was responsible for preparing the statutory accounts of AWA as well as being responsible for day to day accounting issues. He prepared the accounts that were used by AWA for the purpose of the capital reduction and the payment of the two dividends. His evidence described the spreadsheets which were used by the company to estimate the costs of the Fox River clean up for the purpose of arriving at the provision to be included in the accounts. He was clearly an honest witness and I accept his evidence as entirely truthful.

At the end of trial the Claimants did not submit to me that I should treat any of the evidence of any of the Defendants’ witnesses as unreliable or untruthful. The question is therefore whether the Claimants can prove their case on the basis of the evidence that was given.

The CERCLA experts

The Claimants’ CERCLA expert was Laurence S. Kirsch. Mr Kirsch is a partner and Chair of the Energy and Environmental Practice Group at Goodwin Procter LLP, a US based law firm. He has been practising in environmental law for over 30 years and a large part of his practice has involved representing large corporate clients in the area of ‘Superfund’ matters, that is the investigation and remediation of contaminated sites pursuant to CERCLA. From his experience in negotiating with governmental agencies and private parties and having litigated major cases under CERCLA, Mr Kirsch states that he has a good sense not only of his own practice but also of the practice of environmental law generally and the ways in which other environmental lawyers handle matters. Mr Kirsch prepared a main report and two supplementary reports.

The Defendants’ expert on CERCLA matters was Ronald J. Tenpas who is a partner in the Washington DC office of Morgan, Lewis & Bockius LLP. For the past eight or so years Mr Tenpas’ practice has focused nearly exclusively on matters relating to environmental law. From 2007 to 2009 he was Assistant Attorney-General for the Environment and Natural Resources Division of the Justice Department. In that role he supervised about 400 attorneys, including those who represent the Environmental Protection Agency. He states that between his duties in the Justice Department and his work for Morgan Lewis, he has been involved in dozens of CERCLA remediation matters. He has also regularly appeared in court or prepared filings for court proceedings. He also served a main report and two supplemental reports.

Mr Kirsch and Mr Tenpas met on 27 October 2015 to discuss their main reports. They produced a joint memorandum dated 13 November 2015 setting out the points on which they agreed and those on which they disagreed.

Both Mr Kirsch and Mr Tenpas undoubtedly have a wide ranging and profound knowledge of this complex area of US law and I found their evidence invaluable in getting to grips with many unfamiliar concepts.

The accounting experts

The Claimants’ expert accountant was David Lindsell. He has spent almost thirty years as a partner in Ernst & Young where he served as the lead audit partner for various FTSE 100 companies and as a member of the firm’s governing council for 15 years. He was appointed Ernst & Young’s first Global Director of International Financial Reporting Standards in 2003 and held that position until his retirement in 2007. He has held various positions in supervisory bodies and bodies responsible for devising and advising on the adoption and implementation of financial reporting standards. He was a member of the Auditing Practices Board for the UK and Ireland from 1994 to 2002. Mr Lindsell’s first report was served on 17 November 2015 and he served a supplementary report on 28 January 2016.

The Defendants’ expert accountant was Nigel Grummitt, currently a partner in Mazars LLP, the UK firm of the Mazars Group which is an international advisory and accountancy organisation. He is a fellow of the ICAEW and is a specialist in forensic and investigation services. Until recently he was also responsible for auditing a number of clients. He has worked on major group audits and has routinely been faced with the need to consider a company’s liabilities where the timing or quantum of those liabilities has been uncertain. He also states that he is responsible for the conduct of professional indemnity claims against the UK firm and recommends the appropriate provision for their own financial statements, liaising with Mazars’ auditors in their audit of each provision. Mr Grummitt’s first report was served on 17 November 2015 and his supplementary report was also served on 28 January 2016.

The accounting experts met on 30 November 2015 to discuss issues arising from their first reports and prepared a Joint Statement dated 15 January 2016 setting out their points of agreement and disagreement. Again, I found both experts to be very knowledgeable and candid witnesses. Their reports were thorough and helpful.

The documentary evidence

As is often the position in major litigation, much of the evidence took the form of email traffic, internal company memos, meeting minutes and informal notes. A large number of the emails in the bundles were between people neither of whom gave evidence in the proceedings, containing records of discussions with yet other people who were not witnesses in the case. When these emails were put to the witnesses during the trial they often could not remember what was recorded in the email or said that the email was not accurate. Counsel agreed that even though many of these documents contain multiple hearsay, I could treat the emails as evidence not only of what the author was thinking at the time but also of the truth of the matters recorded in the email, and could give as much or as little weight to it as I thought appropriate.

Other participants in the events giving rise to the claim

Although it would be going too far to say that holding the trial without evidence from Mr Christopher Gower was like performing Hamlet without the Prince of Denmark, Mr Gower was conspicuous by his absence from the proceedings and the trial. Mr Gower is a solicitor qualified in England and Australia. By the time that Sequana became interested in AWA and the Fox River liability issue, Mr Gower had been working as General Counsel for AWA for several years - since May 2001. In August 2005 he left AWA as an employee but continued in the same role as before in his capacity as an independent consultant to AWA.

Mr Gower was recognised by everyone involved at the time as the person with an in-depth understanding of every aspect of the complex Fox River liability issue and much of the advice of other people involved was mediated through him to the AWA and Sequana directors. It was Mr Gower who initially briefed Mr Martinet and Mr Courteault when they first became interested in this matter. Throughout the events that I shall relate, he was the source of advice and information of both the strategic and the most detailed kind for all the AWA directors.

On 18 May 2009 Mr Gower, with his colleague Mr Brian Tauscher, bought AWA from Sequana through their vehicle TMW Investments (Luxembourg) SARL (‘TMW’). In the sale agreement Mr Gower acknowledged that he had managed all aspects of the Fox River matter on behalf of AWA through the period 2001 to May 2009 and that he had ‘complete knowledge’ of the history of the Fox River matter, the litigation in which AWA was engaged in relation to that liability and the contingent liabilities it generated.

Both sides in this case criticised the other for having chosen not to call Mr Gower as a witness. The Claimants assert that the Defendants rely on Mr Gower’s expertise as a justification for the decisions that the directors took but that they have not called Mr Gower to confirm the advice he gave. The Defendants in turn point out that it now appears that Mr Gower is firmly within the Claimants’ camp because, as a result of agreements entered into between AWA and the Claimants in the light of this litigation, Mr Gower stands to benefit substantially from any monetary award that I make in the Claimants’ favour.

In all the circumstances I do not accept that it is appropriate to draw an adverse inference against either party for failing to call Mr Gower.

Brian Tauscher is in a similar position to Mr Gower although he is not so omnipresent in the contemporaneous documents as his colleague. From July 2001 he was counsel to AWA in relation to all insurance issues arising from the Fox River liability. Until 2006 he was a solicitor with McDermott Will & Emery LLP, in the US (‘MWE US’) and specialised in insurance law. In 2006 he set up his own law firm Brian M Tauscher PLLC and continued to provide advice to AWA. There is no criticism made of his competence or independence in giving the advice he gave.

Mr Bates is a senior lawyer in the US law firm MWE US. Like Mr Gower, he had been involved in advising on the Fox River dispute for some time before Sequana became interested in the matter. Mr Martinet referred frequently in his evidence to discussions he had had over the years with Mr Bates and clearly Mr Martinet and the other directors relied on Mr Bates as being a person with a detailed understanding of the Fox River liabilities issues. No criticism was made by either party of the other’s decision not to call Mr Bates as a witness.

The US legislation underlying this case is the Comprehensive Environmental Response, Compensation and Liability Act (‘CERCLA’) which was adopted in 1980. It confers on the US government wide ranging powers to deal with releases of hazardous substances. CERCLA is administered by the Environmental Protection Agency (‘EPA’). Clean ups are usually achieved by private parties undertaking the remediation work according to a design and plan of work which those private parties agree with the EPA. If the private parties are unwilling to organise and arrange for the remediation themselves, the EPA can commission the work and then seek reimbursement from the responsible persons.

Once the EPA has decided to clean up a particular site, it identifies potentially responsible persons or ‘PRPs’. Section 107 of CERCLA imposes liability on four categories of PRPs: (i) current owners of a facility even if the current owner made no contribution to the contamination; (ii) the past owner or operator of the facility at the time of the release of the hazardous substance; (iii) a party who arranged for the disposal or treatment of hazardous substances; and (iv) a party which transported hazardous substances to a disposal or treatment facility chosen by it.

The third category of person is important in this case and is generally referred to as a PRP having ‘arranger’ liability. The wording of the relevant provision of CERCLA is:

“(3)

any person who by contract, agreement, or otherwise arranged for disposal or treatment, or arranged with a transporter for transport for disposal or treatment, of hazardous substances owned or possessed by such person, by any other party or entity, at any facility or incineration vessel owned or operated by another party or entity and containing such hazardous substances.”

‘Disposal’ is defined elsewhere in CERCLA as “the discharge, deposit, injection, dumping, spilling, leaking, or placing of any solid waste or hazardous waste into or on any land or water.”

Liability under CERCLA is strict in the sense that there is no need for the EPA to show a lack of due care on the part of the polluter. While the process of identifying PRPs can be quite complex at ordinary disposal sites, it is unusually complicated for waterbody sites because of their size, the large number of parties – often including municipalities – discharging substances into the waterbody, and the long periods of time over which discharges to the waterbodies have typically occurred. The liability of PRPs is joint and several but even where the EPA identifies numerous PRPs, it may choose to pursue only some of them. The regime allows the EPA to compel any one or more PRPs to pay for the entire clean up, leaving the PRPs to seek contribution from other responsible parties in accordance with their equitable share of the costs. Therefore, in many cases it is necessary for the identified PRPs to sue other parties to contribute to their costs. However, as discussed further below, in some circumstances a PRP can attempt to limit its liability by showing that it should only be responsible for its own harm to the site because the overall damage is divisible or capable of apportionment.

The two main kinds of costs for which a PRP may be responsible are the costs of remedial action (initially the investigative costs and then the costs of the work itself) and damages for injury to, destruction of, or loss of natural resources. The second kind, natural resources damages (‘NRDs’), are an important element in this case. NRDs are intended to compensate the owner of the natural resources for a deprivation of that owner’s rights, for example for damage to fisheries or other amenities provided by a waterbody. The owner of those “common good” resources are the sovereign entities, which hold the resources on trust for the public. In the U.S., the sovereign entities claiming NRDs may include different agencies of the federal government, the state governments, and Indian tribes.

The EPA has information gathering powers, in particular under section 104(e) of CERCLA. A section 104(e) request may be directed at any party who has or may have relevant information about materials used or disposed of at a facility and the nature or extent of the release of hazardous material.

If the EPA arrives at an agreement with a PRP for the clean up of the site, the agreement is usually included in an Administrative Order on Consent or a Consent Decree. The settlement is usually included in a court order and before approving the order the court must come to a view as to whether the settlement is “reasonable, fair, and consistent with the purposes that CERCLA is intended to serve”.

In the absence of any such agreement, one of the statute’s most powerful tools is the federal government’s authority under section 106 of CERCLA to issue unilateral administrative orders. These can be issued without the participation of any neutral third party such as a court and can require parties to undertake investigative or clean up activity. Such orders are commonly referred to as “UAOs” or “Section 106 orders”. The courts have construed the relevant provisions of CERCLA as giving the EPA a broad discretion in issuing section 106 orders in a wide variety of circumstances. If a PRP refuses to comply with a UAO and is later found not to have had sufficient cause for that refusal, it may be liable to pay not only the EPA’s response costs but also damages up to three times the amount of those response costs together with daily penalties.

Often polluted sites are divided into “operable units” or “OUs,” sometimes designated by number, such as OU-1, OU-2, and so forth. Different operable units may have the same, different, or overlapping PRPs, and may be addressed collectively or separately for purposes of investigation and remediation. Dividing complex sites into OUs allows the EPA and the PRPs to address different situations (such as varying contaminants and upland versus waterbody sites) differently, and to handle these complex sites in more manageable chunks. In a river site like the Fox River, the location of a PRP’s facility along the river will determine for which OUs it is likely to be liable (otherwise than as an arranger) as any direct discharges can only have flowed downstream of the facility.

Once a contaminated site has been identified for action, the EPA will carry out a remedial investigation and feasibility study of the environmental issues at the site. This allows the lead agency to identify site conditions, determine the contaminants at the site, assess the risk to human health and the environment and conduct testing to determine how to treat the site. It also provides the mechanism for developing, screening, and evaluating alternative remedial actions. The EPA then selects the final remedy for the site. This occurs in two stages. First the EPA identifies a preferred remedy and carries out a public consultation and then the EPA selects the final remedy. At that point there is a Record of Decision (‘ROD’). The ROD may include cost estimates but the PRPs normally require their own contractors to prepare their own estimates of the costs of implementing the ROD selected remedy. If a PRP or group of PRPs agrees to implement the remedy or is ordered by the EPA to do so, the PRPs would retain their own contractor either to perform the work or to oversee subcontractors or some combination of both. The contractors to the PRPs would prepare estimates prior to and as part of the contracting process. RODs can also be amended over time.

Once a PRP has incurred costs it may pursue court action against another PRP in an attempt to recover these costs. There are two different avenues through which a party might seek to recover these costs, depending on the circumstances. First, a PRP that has voluntarily cleaned up a site may use section 107(a)(4)(B) of CERCLA – known as “cost recovery” – to bring claims against other parties. Cost recovery claims are only available if the PRP “has not been subjected to an enforcement or liability action, and ... is not party to a settlement”. Secondly, a PRP that has already been subject to civil liability or has settled its liability with the Government can seek contribution against other PRPs under section 113(f) of CERCLA. Section 113(f) contribution claims are subject to several important limitations. As part of CERCLA’s system of encouraging parties to reach settlements with the Government, CERCLA provides in section 113(f)(2) that a person who has resolved its liability to the United States or a State in an administrative or judicially approved settlement shall not be liable to other PRPs for claims for contribution regarding matters addressed in the settlement. The consequence of contribution protection is that, as parties reach settlements with the U.S., there are fewer remaining PRPs upon whom the remaining burdens can be imposed. Section 113(f) also provides that “[i]n resolving contribution claims, the court may allocate response costs among liable parties using such equitable factors as the court determines are appropriate”. Thus, defendants in section 113(f) contribution actions are only severally liable for their equitable share. There are certain, commonly-used factors for allocation determinations known as the “Gore factors” (named after former U.S. Vice President Al Gore). The Gore factors are:

While the Gore factors provide a helpful framework, they are not an exclusive or exhaustive list of the factors a court can consider in exercising its discretion to allocate liability in contribution proceedings among PRPs. Courts may employ only some of the Gore factors, a single Gore factor, or non-Gore factors like knowledge or the financial resources of the parties. The district court dealing with the claim enjoys broad latitude to decide which equitable factors will inform its decision in any given case, and that decision follows a fact-intensive inquiry.

Another important aspect of CERCLA liability is the divisibility of harm or the apportionment of harm. Section 107 of CERCLA generally imposes joint and several liability for the Government’s response costs on each PRP. However, under traditional tort principles, a tortfeasor can avoid being held jointly and severally liable for the entirety of damages if it can show that the harm is divisible, i.e. that there is a reasonable basis for apportionment. Because divisibility provides a defence to joint and several liability, the defence is typically asserted in response to a lawsuit brought under sections 106 or 107 by the government. But it can be important in contribution negotiations and proceedings too because if a PRP is only responsible for a divisible part of the loss, then it cannot be liable to other PRPs to pay a contribution in respect of other parts for which it is not jointly and severally liable.

The pollution of the Lower Fox River

The Lower Fox River is a river in eastern and central Wisconsin in the United States. It starts from an outlet at the north end of Lake Winnebago and flows north and then northeast until it reaches Green Bay in Lake Michigan. Unfortunately the high concentration of paper mills and other industry along the Lower Fox River over many years has been the source of much pollution of the river. The contamination which provides the background to this dispute arose from the manufacture of carbonless copy paper developed by NCR in the 1950s. The paper was coated with an emulsion incorporating microcapsules of ink. Instead of the traditional method of creating copies by placing a sheet of carbon paper between two sheets of plain paper, the NCR paper produced an identical copy because the pressure applied to one side of the top sheet by handwriting or typing caused the ink capsules on the underside of that sheet to rupture creating an identical copy on the lower sheet.

Polychlorinated biphenyls (‘PCBs’) were used in the creation of the emulsion and are highly toxic. Commercial production of carbonless copy paper using PCB emulsions began in the Fox River Valley in 1954 and continued until the early 1970s. NCR produced the emulsion and then sold it to other companies who applied it to paper (the “coating process”). Two paper coating companies operating in the Lower Fox River area were the Appleton Coated Paper Company and the Combined Locks Company. These two facilities were bought by NCR in 1969 and 1970, starting the chain of liability that has resulted in the indemnity owed by AWA to BAT. These companies then sold the coated paper to NCR, which in turn sold it to businesses that converted (by printing and sizing) the NCR paper into products for their customers (the “converting process”). The coating of the paper with the emulsion and the converting process both generated a by-product known as ‘broke’, that is damaged or off-specification excess paper which was sold to paper recycling mills. The converting process also generated trimmings. Broke and trimmings were then processed by the recycling mills so that the paper fibres could be re-pulped to make paper.

PCBs were discharged into the Lower Fox River in two ways. Some of the emulsion that was used to coat NCR paper in the coating process would be accidentally spilled or wasted and would mix with the paper mill’s wastewater during paper production and be discharged along with the wastewater, reaching the River. Secondly, when broke and trimmings were re-pulped by the recycling mills, they would wash the emulsion out of the usable paper fibres and discharge the contaminated waste water into the River.

The application of CERCLA to the Lower Fox River

The early stages

In September 1994 the State of Wisconsin notified NCR that it was designated as a PRP with respect to the Lower Fox River. In February 1996 the U.S. Fish and Wildlife Service notified API that it was potentially liable for NRDs in relation to the Lower Fox River. Then in July 1997, the EPA formally designated the following companies as PRPs:

Those companies formed themselves into the Fox River Group. Other PRPs were designated at a later stage.

For the purposes of the remediation, the EPA divided the Lower Fox River into five OUs. OU1 is Little Lake Butte des Morts; OU2 is Appleton to Little Rapids; OU3 is Little Rapids to De Pere; OU4 is De Pere to Green Bay; and OU5 is Green Bay. It is estimated that there are 11 million cubic yards of PCB-contaminated sediment spread over the 39 miles of the Lower Fox River and Green Bay. It is important to bear in mind that the direct discharges into the Lower Fox River for which AWA might ultimately be responsible were all down stream of OU1. So no contaminated water discharged by API’s mills could have affected the part of the Lower Fox River designated as OU1.

The WDNR Technical Memorandum 2d (February 1999): the Tech Memo

The Fox River Group and the Wisconsin Department of Natural Resources (‘WDNR’) carried out research jointly into the estimates of PCB concentrations in the Lower Fox River and the likely responsibility of each of the PRPs for the contamination. An early study was produced in June 1998 and a revised version published on 23 February 1999 (‘the Tech Memo’). The Tech Memo contains a careful explanation of the work that supports it and the purpose for which it was produced. It was undertaken in order to test how well models of the fate and transport of PCBs predict future conditions. The Tech Memo notes that while direct data do not exist for PCB concentrations and mass loads, a significant amount of information that can be used to create a reliable estimate of loading was found.

The specific objectives of the Tech Memo were to identify all major dischargers that operated at any time during the period of PCB discharge; to estimate the total suspended solids loads from all major dischargers to the Lower Fox River; to develop discharge volume estimates for all major dischargers; to develop estimates of production for all industrial dischargers during the entire period and to calculate, directly or by extrapolation, PCB loads from each discharger for the period 1954 to 1997.

The Tech Memo opened with a disclaimer:

“Many comments were received suggesting that this document can be used to “allocate liability”. As stated above, this document was prepared at the request of the [Fox River Group] for the sole purpose of conducting model hindcasts. Therefore, the Department believes this version of the document should not be used for an allocation of liability. If anyone reading this document wishes to pursue liability issues, they may contact the Wisconsin Department of Justice.”

However, it went on in the Note to Readers to say that:

“The information in this document reflects the most complete understanding of solids and PCB discharges to the Lower Fox River for the period 1954 to 1997. To the greatest extent possible, data from monitoring or discharger records were used to estimate discharges. However, complete records for this period no longer exist and not all existing records were necessarily available to the Department at the time this report was prepared. Should more accurate, verifiable information become available, these discharge estimates may be revised.

During the course of this effort, the Department received access to Confidential Business Information (CBI) provided by dischargers. This CBI contributed to the strength of this report. Although no CBI is presented in this report, CBI is present in many key spreadsheet files used to compute solids and PCB discharges to the Lower Fox River. The confidential status of this information prevents distribution of these files as open records.”

The Executive Summary of the Tech Memo sets out the following conclusions:

The Tech Memo set out a pie chart showing the shares of discharges that it attributed to each of the companies it considered as follows:

The Fox River Group’s agreement on interim allocation: 27 July 1999

On 27 July 1999 an agreement was reached between API, NCR, Georgia Pacific, Glatfelter, Riverside and WTM I on the outcome of an interim allocation mediation. The agreement covered all FRG assessments approved by those PRPs from 28 July 1999 until the end of Phase II of the Allocation Procedures, as defined in the agreement. They agreed that the costs would be attributed in the following proportions:

The Amendola Report

The United States Fish and Wildlife Service contracted with its own expert called Gary Amendola to estimate the weight of discharges of PCBs into the River. He produced his first report on 12 May 2000 (‘the Amendola Report’) setting out various scenarios. The author described the Amendola Report as a summary of preliminary estimates of PCB discharges for the period 1954 to 1985. Principal sources of information were said to be information provided by the paper companies and Fox River municipal authorities, the WDNR and published technical literature. The purpose of the report was to set out for review and comment by each of the PRPs the data used for their mills and the key factors and assumptions used to develop the estimates.

The preliminary estimates of PCB releases were shown in relation to four different scenarios ranging from a scenario assuming relatively low aggregate PCB discharges to a scenario assuming a high aggregate discharge. Mr Amendola also took 3% as the rate of loss of emulsion in the coating process. He also assumed that nearly all the PCBs in the River came from the production of carbonless copy paper. Amendola’s preliminary estimates were set out in a table which showed:

On 13 April 2001 the Amendola Report was revised and the estimate of NCR/API’s share by weight of the PCBs discharged into the river was reduced to 30% of the total. This seems to be because Mr Amendola reduced his assumption as to the amount of PCBs lost in the course of production of the coated paper. However, it seems to be accepted that this document was not in the hands of the parties to this litigation before 2008 and 2009 though they may have known of its existence.

US Government enforcement action

After identifying the principal PRPs, the Government moved forward with enforcement action. Over the years between 1994 and 2005 there were a number of funding agreements entered into between the Government and various PRPs, several consent decrees and other court orders made at the suit of the Government against various PRPs and funding agreements among the Fox River Group PRPs in addition to the one in 1999 that I have already described. These orders and agreements included the following:

A number of RODs were issued and amended by the WDNR. Each of these was a substantial document setting out the Government’s views as to what work needed to be carried out in the different OUs and setting out the very substantial costs likely to be incurred in performing that work. In addition to issuing the RODs, the Government began negotiations with the PRPs to get work started on remediation.

The Green Mediation

On 14 February 2007 the Government wrote to NCR and seven other PRPs to ask them to enter into settlement negotiations concerning a consent decree to implement the remediation programme. The PRPs together with two other entities Menasha Corporation (‘Menasha’) and the U.S. Army Corps of Engineers (‘COE’) started a non-binding mediation with an experienced and highly regarded mediator Eric Green (‘the Green Mediation’). Negotiations in the Green Mediation continued throughout the second half of 2007 and, with some stops and starts, through 2008. But the mediation was ultimately unsuccessful and no allocation agreement was achieved. However, the proposals which Mr Green put forward as a result of his detailed work play a part in the story as the directors regarded them as a useful reference point when considering what NCR/API’s share of the remediation costs and NRDs was likely to be.

Mr Green and his colleague Douglas Allen wrote a memo to the participating PRPs on 9 August 2007 giving an update of the status of the mediation. This was based on the 30 July 2007 mediation session and his follow up discussions with the parties’ representatives. Mr Green started with dire (and prescient) warnings as to what would happen if the parties failed to reach an agreement:

“In light of the time pressures we face, if we are to turn our substantial progress into an actual settlement, we have to work hard and fast. Thus, we are forgoing diplomacy in favor of frankness, candor and clarity so that all the parties to the mediation know exactly where things stand and precisely what is necessary for there to be a consensus based allocation in lieu of the unilateral administrative orders that will issue shortly and the litigation world war that will follow.”

He explained that they had been working on the assumption that they would need to provide $450 million new money for the OU2-5 remediation and $65-70 million new money for the NRD settlement. He described this and other assumptions as ‘subject to challenge and change’. He then said:

“There are an infinite number of permutations that get to $450 million and $65-70 million. Needless to say, each mediation participant has its own view as to how these numbers should be raised, and each of these views incorporates a large dose of subjective self-interest. For the avoidance of doubt, all parties should understand that we have faithfully listened, recorded, assimilated, and understood each of your arguments about how the targeted amounts should be allocated. There is a large measure of truth and rationality in all of your arguments, but they cannot all be right. As we have explained to you many times, including at the mediation sessions, we have formed our own impressions and reached our own conclusions based on the best understanding we have been able to reach of the technical, factual, operational, historical, geographical, hydrological, remedial, and equitable factors brought to our attention as part of the mediation. We recognize that there are other allocations that reasonable people could derive and defend and we respect the mediation parties' strongly held views on this subject.”

Mr Green then set out the shares which the mediators thought the PRPs should pay towards the $450 million total. The allocation he proposed was as follows – I have added in the percentages in italics:

On 24 September 2007 Mr Green and Mr Allen wrote again summarising their understanding of the progress made and the agreements reached at the end of the mediation session on 18 September 2007. The understanding related to the PRPs’ contribution to the remediation of OU2-5. They set out the following table:

“Summary of Fox River Parties' Contributions to the Remediation of OUs 2-5

Settling Party Contribution

NCR-API $240,000,000

GP	$90,000,000
USP	$40,000,000
WTM I	$13,333,000
Menasha	$10,000,000
Riverside*	$1,000,000
TOTAL:	$394,333,000

*Subject to agreement on payment terms”

This therefore allocated NCR/API a share of 60.8%. This proposal did not include any contribution from Glatfelter and did not include NRDs since the issue of NRDs had been put on temporary hold pending resolution of the remediation issues. However Mr Green noted that NCR/API, United States Paper and WTM I still had amounts on the table for NRDs of $25 million, $11.5 million and $5 million respectively. He also recorded that:

“The Settling Parties instructed the Mediator to inform US DOJ of the progress and the proposed way forward and next steps. Pursuant to these instructions, the Mediator talked with Randy Stone to inform him of these developments. Stone indicated he was disappointed with the lack of closure but that he appreciated the progress made so far. Discussions between DOJ and the Mediator are ongoing.”

On 13 November 2007, the EPA issued a UAO against NCR, API and the six other PRPs (CBC Coating Inc., Georgia Pacific, Menasha, Glatfelter, U.S. Paper, and WTM I) directing the parties to implement the Government’s remedy for OUs 2-5. The UAO required full-scale remediation to begin in 2009. Paragraph 2 of the UAO provided that each PRP is jointly and severally responsible for carrying out all activities required by the Order. In other words, the Government did not divide or apportion any particular part of the contamination to any particular PRP but treated them all as responsible for all the contamination.

Parallel work on the NRDs

In parallel with this work on remediation costs the Fish and Wildlife Service and other trustees began to evaluate the potential natural resources damages associated with PCB discharges. On 20 June 2002 Georgia Pacific entered into a settlement agreement with the DoJ and the Fox River Trustees to pay $10.1 million to cover all NRD claims against Georgia Pacific for the Fox River. This was approved by the court in March 2004 thereby releasing Georgia Pacific from any further obligation to contribute either to the Government or to the other PRPs for NRDs.

On 3 June 2003, the Fox River Trustees issued the final Joint Restoration Plan for NRDs at the Fox River. The Joint Restoration Plan selected a proposed course of action which was estimated to cost $223-333 million.

On 9 July 2007, the Senior Attorney of the US Department of Justice, Randulph (Randy) Stone wrote a confidential settlement memorandum to the Fox River PRPs (‘the Stone Total NRD Offer’). This offer and a later offer to settle the NRD claim against NCR/API are relevant to the Claimants’ challenge to the provision in AWA’s accounts. The title of Mr Stone’s memorandum was “Recommended "Bottom Line" Settlement Position on Natural Resource Damages”. The memo explained:

“The settlement position outlined in this memorandum reflects the views of key legal and technical staff from several of the government agencies involved in natural resource damage assessment and restoration activities for the Lower Fox River and Green Bay Site. It incorporates their analysis of the most current information, including: (1) data on all past and ongoing restoration projects performed with interim NRD recoveries in this case, …; (2) all available information on the potential costs and benefits of prospective future projects, including project concepts advanced by PRP representatives in recent NRD Settlement Work Group Meetings; and (3) recent analyses that the Trustees and their consultant have performed for preparation of a formal Restoration Progress Report.”

Mr Stone said that he was prepared to recommend that the Trustees settle their remaining NRD claim on the terms summarised. He included the caveat that although the offer had been discussed with most of the staff-level Trustee representatives involved with the site, and with some senior-level officials, the offer had not been formally approved by the people who had ultimate authority to approve a settlement. He pointed out further that any proposed settlement was subject to Court approval.

Mr Stone outlined the work that had been carried out in the past to estimate the value of NRDs. The range used in 2001 as the starting point for any settlement negotiations was $176-333 million. That was based on two different possible remediation plans – an intensive remediation which would cost $176-256 million and an intermediate remediation plan which would cost $223-333 million. He referred to the Trustees’ Restoration Plan issued in 2003 which was based on the more expensive intermediate remediation plan but reflected the bottom end of the damages range for that plan, namely $223 million. However, he went on to say that since issuing the Restoration Plan, the Trustees had gained valuable experience in implementing actual restoration projects with interim NRD recoveries from partial settlements, and had worked toward accomplishing the overall goals in the most cost-effective manner. In many instances, the Trustees had been able to achieve high quality restoration at lower than expected costs; specific areas of wetland habitat had been restored or preserved at less than the per-acre cost originally estimated. That past experience led Mr Stone to believe that the Trustees could justify settling the total NRD claim now for less than the prior dollar estimates.

Mr Stone said that he also took into account the various payments that had already been made by the PRPs over the years under Georgia Pacific’s settlement and the other interim funding arrangements that had been concluded, including $25 million already paid over by NCR/API for NRD projects.

Taking everything into account, Mr Stone proposed a settlement figure of $93,575,000 to be paid over ten years and explained in detail how he had arrived at that figure. He then said that although that figure already included significant discounts, he was prepared to recommend ‘one more major concession to enhance the prospect of a near-term “global” settlement’. That concession was that he would accept immediate payment of the net present value of that payment stream into an interest-bearing account. That NPV was $75,647,794. The memorandum closed with the following comment:

“The settlement structure proposed here also would allow the PRPs to satisfy the remaining NRD claim for less than $76 million, and the PRPs' total past and future payments to settle the NRD claim would be less than $120 million. That total commitment by the PRPs would only be a fraction of the $223-333 million total claim amount estimated by the RCDP.”

The offer was rejected by the PRPs.

Later, in July 2008, there was a meeting between Mr Stone, Mr Gower, Mr Bates and Mr Tauscher with representatives of the DoJ, EPA and WDNR to discuss a form of consent decree. In an email to Mr Martinet, Mr Gower said that as regards NRDs:

“The Government made an outright demand on NCR/API for an additional $50 million for NRD if we wanted to settle the claim. Stone stated that in any event, a consent decree would need to address NRD, if only on the terms it proposed, to be sure that API/NCR were not able to use the proposed consent decree as a vehicle to insulate their resources from the NRD claim. Stone and John Carlucci (Department of the Interior) agreed that the Government's interest in this NRD claim was not driven purely by economic or financial considerations and that the Government would sue. Should such a suit be filed, it would be based on the full $333M assessment, and not their summer 2007 "bottom line" NPV offer of $76M. There did not appear to be much room for negotiation, but Stone also made clear that neither his proposal ($10M now and $10M next year) nor his $50M demand had been run by the NRD Trustees.”

Mr Gower described the tone of the meeting as having been ‘cooperative’.

As regards what share of the total NRD one could expect NCR/API to have to bear, the experts were agreed that this would reflect their allocated share of the remediation costs, subject to an important exception. That was that Georgia Pacific which was expected still to contribute to the remediation costs had settled its NRD claim in 2002 and so could not be pursued for any further contribution. Georgia Pacific’s share would have to be met by the other PRPs.

The Whiting litigation

On 7 January 2008, NCR/API commenced the Whiting litigation in the United States District Court, Eastern District of Wisconsin, Case 08-C-16, Appleton Papers Inc & NCR Corporation v George A. Whiting Paper Co. The initial claim was only against the George A. Whiting Paper Company, but NCR/API soon brought every significant potential contributor into the Whiting litigation as a defendant. In the Complaints, NCR/API asserted claims under CERCLA section 107 for cost recovery for remediation costs and NRDs, under section 113 for contribution for remediation costs, and sought a judicial declaration allocating costs and liabilities among NCR/API and the defendants.

The Whiting defendants filed motions to dismiss. On 20 August 2008, the court held that NCR/API could seek to recover clean up costs from other PRPs only under CERCLA section 113(f), not section 107(a). This meant that NCR/API would have the burden of showing that it had paid more than its equitable share before it could recover anything, and it could only recover from each defendant that defendant’s equitable share of the excess. NCR/API’s claims for contribution against other PRPs would be subject to equitable factor defences and the Gore factors weighed according to the discretion of the judge. Further, NCR/API would not be able to recover a contribution from any party that settled with the Government and obtained contribution protection. In September 2008, the Whiting defendants filed answers and counterclaims against NCR/API. The answers asserted numerous different defences to NCR/API’s claims. However, there were several defences asserted by virtually all of the PRPs. These asserted that all or substantially all of the PCBs that had been discharged into the Lower Fox River were a direct result of NCR’s manufacturing of carbonless copy paper either through direct discharges or the sale of NCR broke.

The judge docketed with the Whiting litigation was Judge Griesbach. On 28 September 2008 he made a case management decision and scheduling order (‘the Whiting CMO’) which had significant repercussions for the future shape of the litigation. By that stage there were twenty-three defendants who NCR/API claimed bore some degree of liability for some of the costs they had already incurred, or would incur, in cleaning up the River. In the Whiting CMO the Judge noted that all parties agreed that full preparation of the case for trial would require extensive document discovery, deposition of several hundred witnesses, and retention of expert witnesses at substantial expense to each party. Trial of all issues at once could take months. Judge Griesbach therefore divided the Whiting defendants into those who were party to the November 2007 UAO entered by EPA who had already incurred more than $100 million in clean up costs and those who were not. For those who were not a party to the UAO, Judge Griesbach stayed their disclosure obligation for four months to give them time to negotiate a settlement of their liability with the Government. If they succeeded then API/NCR’s claim against them for contribution under section 113 would be barred so they would drop out of the proceedings. As to the remaining parties, and the non-UAO defendants who did not reach early settlement, the Judge ordered a split trial on the issues (1) when each party knew, or should have known, that recycling NCR paper would result in the discharge of PCBs to a waterbody, thereby risking environmental damage; and (2) what, if any, action each party took upon acquiring such knowledge to avoid the risk of further PCB contamination. Judge Griesbach set 1 December 2009 as the date for the start of the trial on the knowledge issues.

The insurance position and the Green Bay Litigation

I have mentioned that AWA was by 2005 the beneficiary of the Historic Insurance Policies. There was extensive litigation in the US between insurers and insured over the timing and extent of the insurers’ liability to pay out under such policies for environmental damage costs. In July 2003 the Wisconsin Supreme Court overturned the existing law and held that comprehensive general liability policies of this kind did cover environmental claims. The costs recoverable under the policies include not only the costs that the insured had to pay towards remediation and NRDs but also defence costs, that is costs incurred by the PRP in litigation against the Government or the other PRPs.

Some of the insurers were keen to settle future claims by negotiating with AWA a cash sum that the insurer would pay up front in return for AWA releasing it from any further liability under the policy. By the time Sequana became involved in these events, some substantial settlements had already been achieved. AWA had loaned these receipts to Sequana and they formed part of the sum represented by the inter-company receivable in AWA’s accounts.

In January 2005 one of API’s insurers, a group of insurers referred to as CNA commenced proceedings in the Brown County Circuit Court of Wisconsin seeking a declaration that API did not have coverage under the polices. This was referred to by the parties as the ‘Green Bay Litigation’. On 16 March 2008, following a jury trial, it was held that API was entitled to insurance coverage in respect of the Lower Fox River clean up liabilities.

There were further legal proceedings to work out in more detail the nature of the insurers’ liability under the Historic Insurance Policies. The presiding judge in the Green Bay Litigation was Judge Zuidmulder. One of the main issues between the insurers and API was whether recovery was on what was called the ‘pro rata basis’ or the ‘all sums basis’. Under a pro rata basis, the insurer is responsible for only a pro rata share of the damages based upon the years that it provided coverage relative to years when no coverage was purchased. Thus, an insurer is liable for only the damages that accrue during a policy period. Under an “all sums” basis, the insurer is required to pay all sums that result from the occurrence that has triggered liability under the policy. The ‘all sums’ approach is therefore hugely more beneficial for the insured party than the pro rata approach. This was a significant point for the PRPs in the Fox River claim. The insurance policies which AWA had been assigned were from different insurers for different periods and covered different tranches of possible coverage.

On 12 November 2008, Judge Zuidmulder made three important rulings:

The first finding was very welcome news. Judge Zuidmulder’s decision meant that AWA could choose the order in which it claimed against its insurers and could claim against each of them up to the limit of each policy for loss arising, even if the damage to the River occurred before the policy started. The second ruling was immaterial since API expected to use up settlement monies fairly soon. The third point was very unwelcome since it might be many years before the final determination of API’s share and API was having to pay out costs on an ongoing basis. On 11 December 2008 the Company filed motions for reconsideration of the rulings regarding the timing of insurance payments.

The elements that go into the provision

AWA’s accounts were drawn up to the calendar year and were consolidated up into Sequana’s group accounts which were also drawn up to the calendar year. AWA’s accounts were initially drawn up showing amounts in sterling. However, in January 2007 AWA changed to show amounts in euros, making it easier to consolidate them with the Sequana group. It then changed its functional currency to US dollars in January 2009. In the years that are particularly relevant to these proceedings there was considerable delay in drawing up the end of year AWA accounts. The AWA accounts for the year ended 31 December 2007 were not finalised until 28 October 2008 and the accounts for the year ended 31 December 2008 were not finalised until 18 May 2009. It was not suggested that there was anything sinister in this – AWA was one of a very large number of companies being audited within the group.

The Sequana group accounts had however to be drawn up rather more quickly after the year end. PriceWaterhouse Coopers (‘PwC’) were the auditors for the group, the Paris office auditing the group accounts and the London office auditing AWA’s and many other subsidiaries’ accounts.

In order to arrive at the provision in respect of the Lower Fox River in AWA’s accounts, there are a large number of ‘moving parts’ to be calculated. I have mentioned below only those that are relevant to these proceedings. The calculation that had to be performed was in reality much more complicated than the explanation below.

Remediation costs The largest figure to be included in the calculation was the estimated total remediation costs of implementing the ROD ultimately agreed with the EPA. It was expected that the costs would be incurred over a 40 year period into the future. The source of the estimates for remediation costs was a company called Project Control Companies Inc (‘PCC’). They were acting for API and AWA in liaising with the contractors engaged by the PRPs in the Fox River Group to undertake the work. At the start of our involvement in events, the main contractor was called Shaw Environmental & Infrastructure Inc (‘Shaw’) but they were replaced by a company called Tetra Tech in 2007.

The Claimants accepted that the total remediation cost figures provided to PCC and used in the modelling of the provision were best estimates of those figures. But it is important to recognise, as was common ground, that the estimate figures varied widely over the course of the years; that the low and high estimates were very far apart and that there were factors at play that could cause the costs figures to change greatly over future years. Generally speaking, costs for this kind of work will tend to rise as work gets underway and problems that were not anticipated in the initial work plan emerge and have to be overcome. Conversely, sometimes contractors learn better ways of working as they go along and this enables them to save costs. Such savings were sometimes referred to by the parties as value engineering. A key factor in the amount of costs was the way the remediation work would be carried out. There were two ways of remediating the Lower Fox River, by dredging out the polluted sediment, transporting it away from the site and disposing of it elsewhere or by capping the sediment by pouring clean sand onto the river bed. Clearly the former was a much more expensive form of remediation than the latter so much of the negotiation over the RODs and plans of work focused on which parts of the River were so polluted that dredging was really needed and which parts were less polluted so that capping would suffice. These decisions were based on the sampling of sediment in the River. Programmes of sampling and analysis took place at various stages.

NCR/API’s share of remediation costs One of the most contentious issues was the best estimate of the share of the remediation costs that would ultimately be borne by NCR/API and hence by AWA. The estimate included in the accounts that were used to justify the May Dividend was challenged by the Claimants as being too low.

API/AWA’s share of NCR/API’s share As a result of the arbitration decision in November 2005 under the agreement between API, NCR and BAT, it was known that API and hence AWA would bear 60% of whatever costs NCR/API was liable to pay.

NRDs I have described the potential liability for NRDs and the Stone Total NRD Offer. A figure representing the best estimate of NCR/API’s likely liability for paying NRDs had to be factored into the computation of the provision. The Claimants challenged the figures used as the best estimate of NRD liability in both the December Dividend and the May Dividend.

Defence and other costs The costs incurred by AWA would include the costs of litigation on various fronts. There was no challenge by the Claimants to the best estimates of these costs.

Inflation and discounting rates. Since the provision was stated at present value but related to a liability likely to be spread over many years, the costs had to be inflated and then discounted to arrive at a present value. Although there was some criticism of the rates used, this was not an issue raised on the pleadings or dealt with by the experts.

Maris Policy The fund available under the Maris Policy did not increase by inflation once it had reached its maximum level of $250 million. Payments were being made out of the Maris Policy as AWA paid out sums under interim funding agreements or UAOs.

Currency conversion rates All the costs incurred in the Fox River liability were incurred in dollars and the Maris Policy was also expressed in dollars. Before January 2009 all these amounts had to be converted into euros to arrive at the provision for AWA’s accounts. This is important because some apparent changes in the amounts included in the accounts are due simply to currency fluctuations rather than to any change in the underlying dollar figure.

Calculating the provision The way that the provision was arrived at for AWA’s accounts was by constructing a series of spreadsheets in which all the figures were entered and formulae applied to arrive at an end result. The models were prepared by PCC. They obtained information about remediation costs from the contractors Shaw or Tetra Tech, they provided the appropriate inflation rates and discount rates though these would be checked with the auditors PwC. Mr Gower would provide them with other inputs, in particular the NCR/API share and the NRD figure, following all the discussions that I describe later. It is important to bear in mind that the provision in AWA’s accounts only had to reflect the best estimate of the amount which AWA was expected to have to contribute over and above the pot of money in the Maris Policy. When a dollar figure was arrived at as the estimate of AWA’s share of the clean up and NRD costs, the amount remaining in the Maris Policy would be deduced from that dollar figure and the result then be converted into euros before being included as the provision figure in the accounts.

The provision in the years to 2007

For several years after AWA took on the indemnity liability for the Lower Fox River it was thought that the funds in the Maris Policy would be sufficient to meet the whole liability into the future. No provision was therefore made in AWA’s accounts in the years 2001, 2002, 2003 and 2004. However, it seems that this assumption was based on the expectation that AWA/API’s share of the costs allocated to NCR/API would be less than the 60% that was ultimately determined in the November 2005 arbitration award.

In May 2005 Mr Martinet and Mr Lebard took on their executive roles as joint Deputy Managing Directors of Sequana and Mr Martinet became a director of AWA. At that time, as I have described, the striking feature of AWA’s accounts was that the balance sheet in the 2004 annual accounts showed a debt of over £450 million owed by its parent company and a called up share capital of over £200 million. All the Defendants’ witnesses and the accounting experts recognised that generally speaking it is not satisfactory to maintain in place such a substantial inter-company receivable between a parent and a wholly-owned, non-trading subsidiary. Other things being equal, it would be entirely proper and desirable for that inter-company receivable to be removed. It was also recognised by all the witnesses and the accounting experts that a way of removing the receivable and more generally ‘tidying up’ the corporate position would be to carry out the exercise that was carried out in this case, that is make a sizeable reduction in the company’s capital, declare a dividend in favour of the parent company and set that dividend off against the inter-company receivable. Ultimately, in general, the goal would be to reduce the share capital to a much smaller amount and remove the inter-company receivable altogether. The unusual feature which made any such tidying up exercise difficult in the case of AWA were the three liabilities to which it was, in 2005, still subject; the Lower Fox River indemnity, the EU competition infringement fine and the pension fund liability. The last two were familiar problems for the Sequana management and capable of computation. It was also fairly clear that their quantification would be resolved reasonably soon. But the Fox River liability was much more intractable. Sequana and the new AWA directors needed to understand much more about this problem before they could begin considering how the position of AWA within the group could be brought into line. In other words they needed to know how much money should be left in AWA to meet the Fox River liability and therefore how much money could properly be taken out by way of dividend to the parent Sequana.

Soon after his appointment, therefore, Mr Martinet started to ask for information about the Lower Fox River liability. On 11 July 2005 Mr Gower sent Mr Martinet a lengthy memo setting out the history of AWA’s indemnity, of the Fox River pollution and the enforcement activity so far by the US Government agencies, and of the potential liability for NRDs.

In reply Mr Martinet wrote back to Mr Gower acknowledging the complexity of the issue and asking him to provide a matrix of possible outcomes:

“Given this apparent complexity and as we discussed some time ago when we met (briefly) in Pascal's [Mr Lebard’s] office, may I ask you again on the basis of your intimate knowledge of these issues to please try and map it out under the form of some sort of a matrix/occurrence tree/prioritization of issues, if feasible, so that we could have your synthetic opinion at a glance as to what could be the worst case and best case scenarios looking forward. Am I right in assuming that the base case scenario (…) is some sort of a best guess between the worse and best case scenarios, and that it should be amended/refined as a continuous process on the basis of the latest developments (i.e. the $10.4 M insurance payments so far, etc .. )?”

In response to Mr Martinet’s request, Mr Gower emailed him again on 15 July 2005 with two documents. The first was an analysis which Mr Gower said attempted to quantify AWA's exposure based on the information available at that time. The second was the half yearly financial statement prepared by PCC that would be discussed at a meeting to be held in Chicago. The analysis described the various inputs and the uncertainties surrounding them. Mr Gower’s conclusion was (emphasis in the original):

“$104 million (API/NCR share of 40% of $500 million - API being allocated 50% of that share after the first $75 million) to $368 million (API/NCR share of 60% of $850 million - API being allocated 75% of that share after first $75 million. To this you then need to add the NRD liability resulting in a total range of possible liability of:

$124 million to $408 million in present value terms.”

It is helpful to unpack that a little to explain how the estimate was arrived at:

On 26 to 28 July 2005 Mr Martinet attended a three day meeting in Chicago. He went with Mr Courteault and they met there Mr Gower, Jeff Bates and Brian Tauscher from MWE US, Vincent Favier (the Sequana executive dealing with the provision in Sequana’s accounts for the Fox River liability) and Jeffrey Lawson and Susan O’Connell from PCC. The purpose of the meeting was to discuss in depth the background to AWA’s liabilities for the clean up of the River. Mr Martinet said it was very much a fact finding exercise and the first opportunity for him really to get to grips with the complexities of the issue. This was the first time that Mr Martinet met Mr Bates.

From this meeting Mr Martinet emerged with a clear perception of 60% as the most likely figure for the share that NCR/API would have to bear of the total remediation costs. This proved very important for all future calculations of the provision in the AWA accounts for the Lower Fox River liability. Mr Martinet’s evidence was that this 60% figure was discussed with Mr Bates at the Chicago meeting and Mr Bates advised that this was a prudent figure to use.

Mr Martinet’s understanding was that the 60% figure was derived as follows:

Mr Martinet recalled that Mr Gower’s advice had been that there was no need to take into account the possibility of arranger liability when estimating NCR/API’s share but Mr Bates was more cautious and advised that it should be added. Mr Bates’ view prevailed. Mr Martinet described the calculation above as ‘obviously not scientific’ but he believed that it involved two significant reference points, the two reports apportioning direct discharges into the Lower Fox River and the experience in paying for the OU1 clean up.

NRDs were also discussed at the Chicago meeting. Although the figures quoted by Mr Gower in his 11 July memo had been very large, Mr Martinet’s recollection was that at the meeting Mr Bates explained that there were ways of bringing down these damages significantly by taking steps, for example, to repair the river habitat for bird life.

On his return from Chicago Mr Martinet provided a memo to the Sequana board reporting on his discussions. The position he presented to the Sequana board was as follows (bearing in mind that the API share of the NCR/API share was still not settled by the 2005 arbitration at this time).

Mr Martinet noted that although this pot of money was higher than the median figure, it was less than half of the amount of $434 million estimated in the worst case scenario (that figure being 53% of $870 million). Mr Martinet also referred to the Historic Insurance Policies as providing about $940 million of potential coverage – something which he described as an absolutely fundamental point. He referred to the Green Bay litigation and noted that since there had already been several payments out by the insurers, there was reason for optimism.

In the memo to the Sequana Board Mr Martinet stressed the uncertainty of the outcome and described the strategy going forward. He praised ‘the professionalism and seriousness’ of the AWA teams working on the matter. He concluded that the Maris Policy might be adequate to cover the costs but that:

“… AWA is not impervious to a certain number of unpleasant surprises in the matter (the “worst case” falls within the range of $182 and $434 million) that would make this provision manifestly insufficient though it is not possible to evaluate the probability of this at this stage.”

The result of this was that it was agreed that the AWA accounts for the year ended 31 December 2005 would include provision over and above the Maris Policy. The provision included in respect of the Fox River liability was £50.8 million (sterling). The note to the 2005 annual accounts stated:

“During 2005 a clearer picture began to emerge on the future costs relating to the Fox River project. A subsequent review resulted in the creation of a provision which represents the anticipated excess discounted future expenditure on the project over the discounted available project insurance. The amounts are expected to be paid out over the next forty years.”

In December 2005 Mr Martinet brought a former colleague of his, Michael Bartolotta, onto the team. Mr Bartolotta was based in New York and had an accounting background. Mr Martinet valued his contribution because he was precise, thorough and good with numbers and able to scrutinise and challenge the figures being put forward by, amongst others, Mr Gower. Mr Bartolotta started analysing the PCC models, asking pertinent questions and finding some of the answers unsatisfactory. He emailed Mr Martinet copying him into the answers that he had received from MWE US and PCC to his comments on the model. He noted that some of the answers he had received ‘indicate errors or sloppy modelling which is being corrected’. He continued: ‘It is interesting to note that these items were not picked up by either PCC’s senior principal review or by Chris [Gower]’. Mr Martinet clearly regarded this stance of Mr Bartolotta as fully justifying his decision to bring him into the team. He emailed back:

“I believe this shows us that having somebody from within in the loop was both legitimate and presumably overdue. Good work!”

Mr Martinet was asked whether Mr Bartolotta’s fault finding knocked the confidence he had in Mr Gower. His evidence was that he recognised that Mr Gower was clearly ‘not an accounting expert and probably not a model expert either’. That did not mean that Mr Martinet did not trust him. Mr Martinet said:

“I needed extra comfort -- after Chicago, I thought it was a very complex issue and I thought it would not be useless to have an extra layer of checks and also on the issue, that is the -- not that I didn't trust Mr Gower. Frankly, I was rather impressed by what Mr Gower was doing. I thought his notes were very well done. He knew very well – you know when somebody knows an issue. You ask him questions, he responds immediately, gives you some – so I had no reason to distrust Mr Gower. But the numbers were significant. It was a significant issue. And I was more comfortable to have somebody that I would absolutely trust also to look at the numbers.”

In the AWA accounts to the year ended 31 December 2006, the same provision of £50.8 million was included for the Lower Fox River liability.

During the course of 2007, the uncertainty over the level of the fine for the EU competition infringement was resolved when the Luxembourg court reduced the fine payable and the fine was paid in full. Similarly the pension liability was resolved by an agreement at the end of March 2008 reached with the UK pension funds’ trustees whereby AWA ceased to be the pension funds’ principal employer.

Between the decision on the 2005 provision and the start of 2008 there were various other developments in the background that I have referred to earlier, in particular the Green Mediation and the replacement of Shaw with Tetra Tech as the main contractor for carrying out the Lower Fox River remediation work.

In October 2007 Mr Martinet prepared a file note for himself of the likely position at the end of 2007. He noted that budgetary slippage had occurred in the remediation of OU1 and might be repeated in OUs2-5. He set out a ‘Worst-case scenario’ of remediation costs of $750 million and NRDs of $70 million of which AWA would have to pay 36% (60% of 60%) less the $25 million that API was committed to pay, arriving at $286.2 million. His conclusion was that:

“While a successful mediation would have allowed us to entertain the possibility of recovering the provision of almost $50 million from 2007, the examination of the current situation allows us to think that such an outcome is now a good deal less likely and that, quite the contrary, we could be confronted with the need to increase the FOX provision from 2007 onwards by considerable proportions ($60 million in the worst-case scenario).”

Events in early 2008

On 16 January 2008, there was a meeting in New York between Mr Gower, Mr Martinet, Mr Bartolotta, Mr Bates and others including a construction expert and two people from PCC. The purpose of the meeting was, in part at least, to be able to fix the provision to be included in the Sequana and AWA accounts for the 2007 year just ended. At this point the total for remediation costs being discussed was $750 million.

The provision included in the Sequana group accounts drawn up shortly after the 2007 year end was the same as had been included in the AWA 2006 accounts.

In early 2008, attention was focused on the Green Bay trial concerning the ability of AWA to recover under the Historic Insurance Policies. There was also the resumption of the Green Mediation talks, prompted by the issue in November 2007 of the EPA’s section 106 order against the PRPs requiring them to execute the EPA’s proposed remediation work for OUs2-5.

There were also negotiations being conducted between NCR/API and Georgia Pacific aimed at the two of them arriving at an agreement. In early June 2008, Georgia Pacific put forward a proposal based on a total clean up cost for OU2-5 of $698 million with NCR/API paying $333 million, Georgia Pacific $156 million and the other PRPs $209 million. NCR’s chief legal officer wrote to Mr Gower in enthusiastic terms about the proposal. He stated that their experts’ current view was that NCR/API were responsible for 34% of the clean up costs in OUs 2-5 but that the experts for the other parties put NCR/API’s share at 60%. Mr Martinet said that what he took from this was not so much Georgia Pacific’s views of the precise numbers or their approach to calculating NCR/API’s share but their view that the other PRPs could be expected to contribute a significant amount of money to the clean up costs.

In the event NCR/API did not accept the proposal; the negotiations with Georgia Pacific continued throughout 2008 until they finally collapsed on 10 December 2008.

Also during the middle of 2008, following the successful outcome of the Green Bay litigation against the insurers, there were discussions about the possible sale of AWA. With this in mind, Mr Martinet approached various merchant banks to see if they would place a value on the company. He accepted that these discussions foundered because of the unquantified nature and complexity of the Fox River liability.

On 31 July 2008 Mr Martinet sent a memo to Mr Lebard copying Mr Courteault and Mr Bartolotta. He started by saying that:

“As customary, the news on the Fox River issue are contrasted: some good, some bad and still a lot of unknowns. Under the circumstances that I'll summarize below, urgent actions are now needed on AWA ltd so as to prepare for the future and control potential damages.”

Mr Martinet outlined the good news, that is the success in the Green Bay litigation jury trial in March 2008, but said that there was still uncertainty about the scale of remediation costs. There had been no settlement with the DoJ on NRDs. He described the Whiting litigation as having ‘triggered a multiple dog fights situation’. He also described the offers and counter offers as between NCR, API and Georgia Pacific, concluding that it was far from certain that this would lead to an acceptable compromise. He said:

“We therefore made clear, Mike [Bartolotta] and I, to our counsel (Jeff Bates) that AWA was very stretched, that he could not count on more money than what we had currently reserved for, and that either we could work a compromise with what we had on the table, or we would have to have a plan B, whatever such plan could be. None looks right at the time being.”

Mr Martinet noted in the memo that AWA’s two other issues (the competition infringement fine and the pension fund issue) were now resolved and “we need to recalibrate the company so as to transform it into a pure containment structure for Fox”. His recommendation was that work be undertaken with the assistance of Ms Georgia Quenby, a solicitor who worked in MWE’s London office, under the supervision of Mr Bates. He also stated his belief that they needed to include Mr Gower in the work. He continued:

“To calibrate the capital reduction, we obviously need to have a clear vision of the 2008 possible reserve adjustment, which is a function of what was explained above. And we will need to have a fairness opinion from a reputable third party that what is being done, when it will be done, is the best that could be done under the current circumstances.

By the way, the way to effect a capital reduction will change on October 1st, 2008. A judge decision will not be needed anymore, simply a letter by each member of AWA's board asserting that the company is in good financial condition for the upcoming 12 months following the capital reduction. Chris can help there, not with Antoine and myself, but with his former colleagues.

I would suggest that the maximum deadline for all this work be fixed at December 31, 2008, at the latest.”

When the memo was put to Mr Martinet in cross-examination he accepted that in the event, there was no “fairness opinion from a reputable third party” obtained. He was also asked why he had fixed a deadline and why he regarded the matter as urgent. His response was that Mr Bates had told him that AWA was seen around the Fox River as being a ‘deep pocket’ or ‘easy prey’ because of the very large net asset position on its balance sheet. In the AWA accounts for the year 2006, the inter-company receivable on AWA’s balance sheet was £464.6 million, up from £437.5 million in the year 2005. The advice Mr Martinet had received was that companies that appear to have deep pockets not surprisingly become the focus of the EPA’s attention when it comes to choosing which PRPs to make subject to further section 106 orders. Mr Martinet thought it would be better to resolve the question of how much money should be left in AWA by the end of the financial year so that the 2007 accounts that AWA would have to make available to the US authorities in the context of the negotiations would not show that it had several hundreds of millions of dollars available to spend on the clean up.

The September Internal Memorandum

After the summer 2008 break, the AWA executives turned their minds again to how to deal with the Lower Fox River provision.

Several emails in this period were put to Mr Martinet in cross-examination to suggest that the wording he used in them indicated two things. The first was that he was pushing Mr Gower to reduce the provision needed for the AWA accounts by all means possible, fair or foul. The second was that from at least July 2008 he had the fixed intention of eliminating the inter-company receivable in AWA’s accounts as quickly as possible because he knew that the ‘moving parts’ in the Fox River liability were all moving in the wrong direction and he feared it would soon become apparent that a much larger provision was needed in the AWA accounts to reflect the best estimate of the Fox River liability. For example, it was suggested to him that the Whiting CMO made by Judge Griesbach raised the possibility that NCR/API would be liable for 100% not just 60% of the remediation costs. Mr Martinet denied that this was his motivation. He insisted that throughout he was trying to take a professional and cautious approach to the issues with the intention, certainly, of reducing the €585 million receivable on AWA’s balance sheet but only so far as could prudently be done, having regard to a proper estimation of the likely liability. As to the Whiting CMO, Mr Martinet recalled that he had been told about the main elements of it and he remembered being told that there would be a first trial limited to the issue of knowledge. He also realised that since this first trial would not take place until 1 December 2009, the end of the Whiting litigation had been delayed. He said that the impression he got from his advisers was that they were not particularly worried because the issue of knowledge had already been explored during the Green Bay litigation against the insurers. He did not recall being told that one possible outcome of the knowledge issue trial would be that NCR/API would be adjudged responsible to pay 100% of the costs and unable to recover any contribution from the other PRPs. He said that the 100% assumption was never put to him as a realistic possibility. He accepted that some other moving parts might be moving in an adverse direction, for example with the possibility of remediation costs increasing if new sampling of the sediment indicated that more dredging than capping would be needed. But he was clear that any urgency from his point of view was to ensure that the 2007 audited annual accounts did not show the US Government any more money available for the clean up than was prudently necessary.

Mr Martinet prepared a confidential memorandum (‘the September Memo’) for his colleagues at Sequana and for Mr Bates and Ms Quenby, the lawyer in the London offices of MWE UK. He reiterated the point that he had made in July that that the other two issues had been resolved:

“As of now, therefore, AWA is in a position to be transformed into a single purpose entity responsible for the Fox River exposure, and be re-organized in such a way that it would have adequate resources on its own to meet its remaining long term obligations in the US.

To accomplish that, however, given the very significant, and largely excessive, amount of remaining inter-co accounts between AWA and Sequana, an additional, and presumably final, distribution of reserves coupled with an AWA capital reduction becomes mandatory before year end 2008.

This cannot be done without an in-depth re-assessment of the current Fox River exposure, coupled with a third party fairness opinion as to the adequacy of the thus actualized Fox River reserve and related available resources, as well as of the envisaged AWA capital reduction.”

He concluded: (emphasis in the original)

“The objective of rapidly transforming AWA into a viable Fox River containment structure is essential. Alternative scenarios therefore need to be rapidly envisaged so as to optimize such restructuration and best calibrate AWA's forthcoming capital reduction.”

Attached to the September Memo were various exhibits. Exhibit 3 set out the figures from AWA’s balance sheets in 2006, 2007 and F3Q 2008 and also showed a pro forma balance sheet showing the position if there were a capital reduction sufficient, when taken together with other reserves, to pay a dividend of €517 million. That would leave shareholders’ funds of €5 million. He commented:

“Under such a preliminary scenario, however, AWA would still remain with a 66 MEUR receivable from Sequana, corresponding roughly to the sums recouped by AWA from the Fox insurance carriers in 2006 and 2007, but still an unnecessarily large number, particularly in light of Sequana's currently available cash resources.”

He suggested some alternative courses, concluding:

“These different alternatives need to be urgently vetted, in light of the recent Fox River developments (which could impact AWA's reserve).”

Again Mr Martinet’s response as to why it was urgent was because of the ‘deep pocket’ point of which, he said, Mr Bates was well aware. It was put to him that the last sentence of Exhibit 3 was intended to convey the idea that the urgency arose because developments in the Fox River would inevitably result in a higher provision being required if the dividend was not dealt with by the end of the year. It was suggested that he wanted to ensure that the dividend was paid before new higher remediation costs estimates came in. Mr Martinet consistently denied that there was any such motivation. He described the intention behind the September Memo as follows:

“…our intention was not to strip AWA of all its resources. It was to make a proper work and to do a good work where we would be comfortable that what we had done was right. There were a number of constraints. One of the constraints was timing, because we needed to have a set of statements that could be presented to the authorities, showing that we were not a deep pocket. ….”

Mr Martinet said he had no settled commitment at this point as to what would ultimately happen to AWA. He accepted that Sequana’s objective was to reduce the capital of AWA and release part of the intercompany debt. The September Memo shows that it was Sequana’s wish to reach some finality in terms of its liability for Fox River by the end of the year if this was possible:

“And of course -- now, of course nobody wants to increase a provision, if you can avoid it. But it was not my intention. My intention was: let's have a hard look at this situation. Let's do the best we can. We now are at a stage where a number of things have happened over the past years, so on and so forth. It is time to look at it very seriously, try to do everything we can to come up with a solid opinion. And frankly, this is what -- and I spent the remaining weeks of 2008 to try and do that.”

It was also put to Mr Martinet that by this time the financial crisis that had started across the world was putting Sequana under financial pressure. Mr Martinet’s response was:

“But at that point in time I was not under the understanding that the situation of Sequana was such that we needed, at any cost, to do something improper. At that point in time I was not particularly worried by Sequana's position. I was worried in general by the economic situation, but I was not worried by Sequana's position, because it was not something that was of such an urgency that it would have worried me. That is what I'm trying to say.

I accept Mr Martinet’s evidence that the purpose of the September Memo was to start a process by which there would be a thorough reassessment of the Fox River liability so that a decision could then be made, in the light of that, as to what should be done. There was no decision taken by him or Sequana at this stage to eliminate the reserve in AWA’s accounts, regardless of what the reassessment revealed.

After the September Memo, there was a meeting in Paris on 30 September 2008 attended by Mr Martinet, Mr Lebard, Mr Courteault and Ms Quenby. This was to discuss the legal procedures for approving the capital reduction. Following the meeting Ms Quenby sent round an email to which she attached a table showing the different steps that would have to be taken.

Following that meeting Mr Martinet asked Mr Gower to produce, in conjunction with PCC, some figures for a possible new provision. At this stage it was thought that the capital reduction and the dividend payment would be done in a single stage rather than, as ultimately happened, having a single capital reduction but two dividends. It was agreed that the matter would be discussed at a proposed meeting in New York towards the end of October 2008. Mr Martinet asked Mr Gower to prepare three scenarios, best case, mid case and high case. Mr Gower emailed PCC, copying Mr Bates and Mr Tauscher, asking them to produce the necessary models. He suggested that NCR/API share should range from 44% to 70%. As to the assumptions that should be made about liability for NRDs, Mr Gower said:

“For NRDs, we have two issues, amount and timing of payment. The high point should assume we pay 60% of 70% of the government's latest demand. JB — what was that number? The low point should assume 60% of 44% of the government's demand. The high point should assume payment in 2009. The low point should assume payment in tranches, say $5 million per year.”

He concluded ‘JB and Brian, please feel free to weigh in on any of these’; JB being Mr Bates.

On 6 October 2008 Mr Bates emailed Ms Quenby. He pointed out various factors that would need to be borne in mind. First, the cost estimates provided by PCC would be based on old sediment sampling data because there would not be any further sampling of the River bed until 2009. That new sampling might cause costs to go up in which event the 2009 provision would need to increase perhaps substantially. Mr Bates said:

“Based on this information, I think we have to contemplate that, regardless of the basis for the 2008 provision, we could see a significant increase a year later which, if the 2008 recap [sc. recapitalisation] is based on the 2008 provision, could be inadequate within 12 months.”

He also commented that the NCR/API share may change from the 60% assumed in the past and that if the Government’s offer to settle NRDs was rejected then ‘the Government’s litigation claim will be in the hundreds of millions of USD’. Mr Martinet recalled seeing this email when it was forwarded to him by Ms Quenby. He took it as a warning from Mr Bates that the company needed to be very prudent in its decisions because of the number of uncertainties. This was in part why Mr Martinet insisted that Mr Bates attend the meeting in New York. On 8 October 2008 there was a meeting between Ms Quenby and Mr Courteault to discuss the nuts and bolts of the reduction of capital mechanism. The point about the delays to the sediment sampling were discussed. Mr Courteault said that he regarded this as a good reason to be careful about the provision but not a good reason to postpone the whole transaction.

It was suggested by the Claimants that the extent of the uncertainty about the outcome of the Whiting litigation and the other moving parts was such that the AWA directors ought to have concluded at this stage that the process of making a decision to reduce capital and to pay a dividend was just too difficult and that they should have left the AWA accounts as they were. This was particularly true, the Claimants said, given that at the beginning of September 2008, in addition to the delays to the sampling, it was announced that the period for the design work on the remediation plan had been extended by the Government from the end of 2008 until May 2009. This meant a further delay on resolving the uncertainty about remediation costs. In my judgment it would be wrong to criticise the Defendants for undertaking the exercise that was outlined in the September Memo rather than leaving €580 million on the balance sheet indefinitely. Companies have to wrestle with uncertainties about the future and although the Fox River liability may be at the extreme end of the spectrum of uncertainty, it is not unique in that regard. They were fully entitled to investigate the position to see if some of the receivable could properly be released.

The New York meeting on 22 October 2008 and following

A meeting in New York was fixed for 22 October 2008. Before that meeting there were various emails back and forth in preparation. One was an email between Ms Quenby and Mr Bates dated 13 October in which she discussed what advice MWE could and could not give to the company regarding the capital reduction. She said that MWE could continue to advise the company on the law and the requirements without advising them actually to proceed. That must be a decision each director takes. Mr Bates suggested engaging PwC ‘to opine on the provision/reduction’. Ms Quenby replied that she would be happy with using PwC and would suggest this to Mr Courteault.

On 20 October, there was a further exchange of emails between Ms Quenby and Mr Bates outlining what they would say at the meeting. This included a description of the solvency statement procedure for reducing capital and the risk that any intra-group transfer of assets by AWA could be challenged in court by an allegation of fraudulent transfer if AWA failed to pay the indemnity at some point in the future. Mr Bates said at the end of the email:

“Basis for Supporting Solvency and Deferring or Defending Litigation

• Independent auditors have to conclude that AWA is solvent under UK law and that the intergroup transfer is not constructively fraudulent under UK and US law based on valuation of liabilities and assets, contingent and actual

• Confer with PWC regarding modality (e.g., possibly proposed IASB methodology regarding contingent liabilities and assets)

• No guarantee that this process will ultimately survive a court challenge in the UK or the US

• Should consider providing third-party or self-insurance or a guarantee for the litigation on the legacy insurance

• Should consider moving a trading company back into AWA

• Should consider moving the intergroup obligation in stages (130 Million euros now; another next year following river sampling; another following the insurance appeal)

• Should consider whether/when to notify or start arbitration with API over right to terminate indemnity”

The point made there about a possible claim for fraudulent transfer was spurred by the judgment in New Jersey Department of Environmental Protection et al v Occidental Chemical Corp. et al., case number L-009868-05, in the Superior Court of the State of New Jersey, Essex County (‘Repsol’) handed down on 5 September 2008. The issue determined by that ruling was whether the New Jersey court could assert personal jurisdiction over the Spanish company Repsol which had acquired two companies, Maxus and Tierra. Those companies were liable to the New Jersey Department for Environmental Protection for the clean up costs of the Passaic River. In addition to pointing to various links between Repsol’s business and New Jersey, the Government claimed that Repsol and YPF had depleted Maxus and Tierra of their assets, leaving them unable to satisfy their remediation obligations. The judge held that the New Jersey court did have jurisdiction. Amongst other reasons he held that to the extent that some wrongful conduct was required, the ‘extraordinary stripping of Maxus’ assets, if proven, [was] arguably sufficient, at least for jurisdictional purposes’.

The meeting on 22 October 2008 was attended by Mr Martinet, Mr Lebard, Mr Gower, Mr Tauscher, Mr Bates, Mr Bartolotta and Mr Lawson and Ms O’Connell from PCC. Several possible restructuring solutions for AWA were discussed, including moving a substantial trading company into AWA from elsewhere in the Sequana group. At the meeting PCC produced PowerPoint slides giving an update of the total cost estimate as at 22 October. As requested by Mr Gower, they set out cost estimate comparisons on the basis of various possible NCR/API shares ranging from 44% to 70% applied to remediation costs and NRDs. The numbers were discussed and PCC undertook to revise the figures taking the points raised into account.

In the meantime on 28 October 2008 the AWA board met to approve the 2007 Final Accounts. The balance sheet that was approved at the meeting showed a provision of €59.3 million (equivalent to $87,371,724) for the Fox River liability. It valued the inter-company receivable at €569.7 million.

The provision in the 2007 Final Accounts was based on:

On 2 November 2008 Mr Gower sent Mr Martinet some new modelling following the New York meeting. These were figures that PCC had sent to Mr Gower on 29 October 2008 for his comment. As regards the estimate of the costs to be borne by AWA (and hence the amount which would need to be provided for, once Maris Policy funds had been deducted), this set out the following figures:

The NCR/API share on which these figures were based was shown as 55.5% in the low estimate, 60% in the mid estimate and ‘variable’ in the high estimate because in the high estimate, different percentages of share were used for different OUs of the River. The background data sheets in the model show 64% was used for OU4 and 100% used for OU2 and 3. When asked about this, Mr Martinet could not remember why 100% had been used as the high estimate of NCR/API’s share but accepted that this must have resulted from something discussed at the 22 October meeting in New York. It may have reflected discussions going on with Georgia Pacific at the time, whereby Georgia Pacific was expected to contribute to the OU4 clean up but not to the OU2 and OU3 clean up.

As to the NRD costs included in the 2 November figures, the figure of $50 million was used as the cost for NCR/API for NRDs in all three scenarios, low, mid and high, giving API and hence AWA a figure of $30.9 million being 60% of $50 million. Mr Martinet’s evidence was that he considered that it was reasonable to use that figure even in the high, or worst case scenario, estimate, since it was derived from the Stone Total NRD Offer and so was, he said ‘something of certain solidity’.

There were some further email exchanges between Mr Martinet and Mr Gower in which Mr Martinet raised queries about the numbers. One question was whether there should be some variation in the figure used for the NRDs in the different estimates. Mr Martinet denied when asked about this that he was concerned that the numbers were higher than they had been previously. His concern was, he said, only that the assumptions had to be robust.

Also in early November 2008 Mr Martinet produced what he called his own ‘quick and dirty’ model to look at the sensitivity of the figures. He compared the figures that had been discussed at the New York meeting with the figures in the revised estimate sent to him on 2 November. But he made certain adjustments to the 2 November figures, for example by reducing the AWA share in the high estimate to 38.7% rather than 40%, including substantial sums for ‘value engineering’ (Footnote: ¹) in each of the estimates and reducing the NRD figure to $25 million in the low estimate, $35 million in the mid and $50 million in the high. The figures he arrived at for AWA’s liability on 4 November (before deducting Maris Policy funds) were:

Mr Martinet then took a further step and assigned a probability to each of the low, mid and high estimates, namely a 35% probability to the low estimate, 45% to the mid estimate and 20% to the high estimate. When these probabilities were applied to the 2 November figures as sent to him by Mr Gower, the overall average came out at $273,395,000. When the same percentages were applied to the figures as adjusted by him in the ‘quick and dirty’ model the overall average was $235.378 million, slightly lower than the figures that had been discussed at the 22 October meeting.

Mr Martinet was asked where he had got those probabilities from. His evidence was that these did not come from any source but were just dummy figures used in a working document to prompt further discussion with Mr Gower.

Mr Gower forwarded Mr Martinet’s ‘quick and dirty’ model to PCC. In the meantime, PCC had been working on their own revised figures which they sent to Mr Gower also on 4 November 2008. Those figures also included value engineering savings in all three scenarios and put in variable amounts for NRDs, albeit not the same figures that Mr Martinet had included in his model. The figures that PCC arrived on 4 November 2008 for AWA’s overall liability from their calculation before seeing Mr Martinet’s model were:

The low estimate figure was below the 2 November 2008 estimates and the mid and high not much changed. But they were all much higher than the figures in Mr Martinet’s ‘quick and dirty’ model. These figures were sent to Mr Martinet by Mr Gower on 5 November 2008. Mr Martinet’s response was to raise four issues about the net share, the scale of the value engineering benefits, why there was a variation in ‘past costs’ which seemed counterintuitive and why there was such a variation in defence costs. He concluded the email ‘In a nutshell, the mid case looks pretty bad to me as it requires a rather massive provision increase’.

The next run of figures was circulated on 10 November 2008. Changes had been made to many of the moving parts including value engineering, NRDs and the NCR/API share which was adjusted to 45% in the low estimate, 50% in the mid and 55% in the high. The figures resulting from these 10 November calculations were much lower:

Mr Martinet forwarded the revised 10 November calculations to Mr Lebard and Mr Bartolotta saying:

“Looks much better for us. Now, is it realistic?”

He also wrote back to Mr Gower saying:

“Chris, this obviously looks much better. Can you stand behind these numbers?”

Mr Martinet was cross-examined at some length about the chain of revisions to the numbers that I have described and his reactions to them. It was put to him that he was showing Mr Gower his determination to get the numbers down to a low figure so that the provision needed in the accounts was kept to a minimum. Mr Martinet denied that this was what was happening:

“Q. Now, Mr Martinet, as you have just accepted, these changes haven't happened with any input from the experts in relation to each of the assumptions. All you wanted at this stage was just to see whether you could get Mr Gower to say, "Yes, we can use these numbers."

A. But I'm asking, "Is it realistic?".

Q. Well, you knew it wasn't realistic.

A. I did not know whether it was realistic or not. I challenged the numbers. He changed the numbers and I am asking myself, and I am asking Pascal [Lebard] and Michael Bartolotta, "Is it realistic?" So I think basically I challenged some of the assumptions, I challenged the logic of the assumptions, came up with a number, numbers that are better. But I'm asking myself and my colleagues, I don't know if it is realistic so ...

Q. I see. Maybe I misread this, I'm sorry. You are asking Mr Lebard and Mr Bartolotta, "Is it realistic?”.

A. Yes, I'm pointing my question mark to my colleagues, I don't know if it is realistic, "Is it realistic?" It is a question mark.

Q. Because you knew perfectly well that, for example, the share has been changed, but there was no basis for changing the share, was there?

A. I think what I did was I challenged some of the assumptions which were not totally understandable from my standpoint. He changed them, the numbers came back low, and I was asking myself: is that correct? And I am asking my colleagues, you know, I don't know if it is correct.

Q. All you wanted was to -- having now got what you wanted in terms of lower numbers, was to just see whether you could protect your own back and see whether Mr Gower might say, "Yes, this is okay."

A. Protect my own back?

Q. Yes.

A. I don't think so.

Q. You have made the changes to assumptions in your quick and dirty model. You show it to Gower. Mr Gower passes that on to PCC. You discuss it with him on the phone. The changes are made. The numbers come back. They are good for you. You like them. Now the question is how to make it seem this is all okay.

A. It is not to protect my own back. I'm asking myself, you know: is it correct, is it -- are the assumptions okay? This is what I think this may suggest.”

Mr Martinet agreed that he would be happier with lower numbers but he did not ask Mr Gower ‘at any cost to change the numbers to fit my desires’ and he denied that the question to Mr Gower as to whether he could stand behind the numbers was ‘window dressing’.

Mr Martinet also pointed out that he sent the figures to Mr Bartolotta who had been brought in specifically to check the numbers. Mr Bartolotta wrote to Mr Martinet and Mr Gower on 10 November 2008 setting out various queries which arose because he had, he said, tried to put himself in the role of the auditor who was going to review the analysis, bearing in mind that auditors like consistency. In particular Mr Bartolotta challenged the move of the figure for NCR/API’s share away from the 60% that had been used up till then. He did not accept as valid the reason given for the move to lower shares, namely the ongoing discussions on a deal with Georgia Pacific.

The upshot of the points that Mr Bartolotta raised was that a fourth scenario was introduced into the modelling, referred to as ‘mid2’ in which 60% was used as the NCR/API share.

The 20 November 2008 estimates

On 20 November 2008, PCC sent Mr Gower some updated cost estimate scenarios and on 24 November Mr Gower forwarded them to Mr Martinet and Mr Bartolotta. The mid2 scenario figures in this estimate from PCC were ultimately the figures used to set the provision in the accounts used to support the capital reduction and the payment of the December Dividend. The four estimates given on this date are also significant in this case because this is the first set of figures provided to Mr Mountford and Mr Newell and are the figures on which they formed their opinion for the purpose of signing the solvency statement on 15 December 2008. The figures shown as AWA’s liability in the 20 November estimates were:

Mr Bartolotta responded to Mr Gower on 24 November 2008 with comments which merit being set out in full:

“Gentlemen - I have not worked my way through all the detail, - and I will look at it further tomorrow— however I have the same basic point I previously made regarding the approach and the auditors. In the past we did not use a range of outcomes we used a point. None of the email attachments has an analysis that shows the 12/31/07 reserve in comparison to your four outcomes — I think you need that for the auditors. Second the auditors will want you to explain all the year over year substantive changes by line item. This becomes much more complicated when you vary the various line elements in the different scenarios — which is why I have previously suggested that you only vary those elements where there truly is a reason for variability or multiple possible outcomes. For instance, in the past we picked a discount rate — based on a given methodology — now you have three different rates with a .20bps spread up and down — why? In the past we have used 60% of the full river costs as the NCR/API share — now we have a 45% and 50% case as well as 60% — what changed in 2008 to make the 60% less "correct" this year? - You see the point - where there is true variability — such as in the subsequent allocation to PRP's, or the contractor savings - new elements - I think "scenarios" may make sense — otherwise I think they are overly complicating.

On the other hand I know you have had more discussions on this issue, particularly with the new insurance valuation consultants and, in fact, preliminary discussions with the auditors and thus my comments may in fact be irrelevant, but I will be surprised if the auditor's first reaction does not bring similar comments.

Anyway — I think we need a comparative analysis to prior reserve calculations and I think we need a summary sheet that briefly explains why or supports the variance in each element that changes from case to case.”

Mr Martinet thought that this showed Mr Bartolotta asking all the right questions as he expected he would.

There were further iterations of the numbers in the four scenarios and questions being asked and answered about the factual underpinning of the numbers being used, including further discussion of what the other PRPs could be expected to contribute.

On 10 December 2008 Mr Gower sent through more provision calculations. These were sent not just to Mr Martinet and Mr Bartolotta but to Mr Mountford and Mr Newell as well. The covering email said that the mid2 range reflected Mr Gower’s view of ‘the reasonable and most-likely outcome’. The four estimates were:

Attached to the estimates were the spreadsheets showing a detailed breakdown of the figures and a reconciliation with the figures used in the 2007 Final Accounts that had been finalised on 28 October 2008.

The involvement of Mr Mountford and Mr Newell

I need now to backtrack a little to explain how Mr Mountford and Mr Newell became involved in the discussions about the Lower Fox River provision in the AWA accounts and about the proposed capital reduction and December Dividend.

Both Mr Mountford and Mr Newell knew Mr Gower and regarded him as the person who knew the most about the Lower Fox River liabilities. In his witness statement, Mr Newell says that he did not have a detailed understanding of the source of AWA’s Fox River liability or how it was estimated. He knew the matter was very complicated and that ‘it would take an incredible amount of time and effort … to understand the detail’. He did not consider that it was necessary for him to understand the detail because he relied on Mr Gower, whose job it was to manage the liability on behalf of AWA and to liaise with AWA’s professional advisers.

Both Mr Mountford and Mr Newell regarded it as uncontroversial that a subsidiary would remit excess assets to its parent by way of dividend. Mr Newell said he regarded it as bad financial management to leave significant assets in a non-operating subsidiary like AWA if they are not required. Having reached the point in mid-2008 when the competition infringement fine and the pension scheme issues had been resolved, it made sense to move any assets in excess of AWA’s liabilities to somewhere else in the group where better use could be made of them. Mr Mountford said that he had experience of this kind of capital reduction and dividend where a cash rich but inactive subsidiary passed cash up to the parent reducing or eliminating inter-company loans. Both men entirely understood that the key question for them to consider was the value of the assets that should be left in AWA to be sure that there would be sufficient funds to meet the future Fox River liabilities. The dividend paid to Sequana would have to be calibrated so as not to deplete AWA’s assets below that amount.

The first significant meeting so far as Mr Mountford and Mr Newell were concerned was on 15 October 2008 at MWE UK’s offices in London. The meeting was led by Ms Quenby, the MWE London partner responsible for the December capital reduction and dividend. Mr Mountford attended the meeting but not Mr Newell. At the meeting the steps that would need to be taken for the proposed capital reduction and dividend were explained.

From Mr Mountford and Mr Newell’s perspective the matter then went quiet until mid November 2008. On 19 November Mr Thomas told Mr Mountford that there was a meeting with PwC scheduled to take place at MWE’s offices on 21 November to discuss the Fox River issue. Mr Mountford was not invited to that meeting but he did attend a meeting a few days later on 26 November 2008. Mr Newell did not attend as he relied on Mr Mountford in effect to represent them both and to report back to him what happened. Mr Newell said he knew Mr Mountford to be ‘a very cautious and diligent person’.

The meeting on 26 November 2008 was attended by Mr Mountford, Mr Gower, Mr Thomas and two people from PwC, Mr Derbyshire and Mr O’Brien. The purpose of the meeting was for Mr Gower to brief PwC on the four scenarios that had by this time been devised by Mr Gower and PCC to estimate the potential costs that AWA might be liable to pay over and above what remained in the Maris Policy. The four scenarios presented were those that had been produced by PCC on 20 November 2008 as set out in paragraph 187, above: to recap, a low of $167,295,632; a mid1 of $202,792,591; a mid2 of $230,864,497 and a high of $264,289,286.

Mr Gower said at the meeting that the mid2 scenario was in his view the most realistic estimate. Mr Gower explained to the meeting some of the ‘big-picture developments’ in relation to the Lower Fox River issue, including the progress of the Whiting litigation, the recoveries by AWA of insurance indemnities under the Historic Insurance Policies and the identification of new PRPs who were added as defendants in the Whiting litigation. Mr Mountford’s evidence was that his overwhelming impression from Mr Gower’s presentation was that there were far more potential upsides for AWA in relation to issues such as insurance recoveries and API’s share of the clean up costs than there were downsides. His evidence was:

“Mr Gower was bullish. He was certainly – the impression I got from him at the meeting was that things were moving in our favour and that the likely liability would be moving downwards. Nevertheless I think in my role as a director it is necessary to be cautious, bearing in mind, of course, the risks I would be taking if we got this wrong, risks of a claim against me, risk of a criminal act. So yes, I certainly wanted always to be cautious.”

After the meeting, on 27 November 2008, Mr Gower emailed to Mr Mountford and Mr Newell the spreadsheets that had been presented to PwC at the meeting. Both Mr Mountford’s and Mr Newell’s evidence was that they focused on the high estimate in these spreadsheets because they wanted to be prudent by basing their assessment of how much money should be left in AWA (and hence how much could be safely paid by way of dividend to Sequana) on the high estimate rather than on the best estimate. They understood that the figures put into the model to generate the four scenarios came from Mr Gower and PCC. Mr Mountford carried out a computation that was, from that time forward, very significant for his and Mr Newell’s decision making. For all four scenarios he looked at the figure on the supporting spreadsheet for the Discounted Total Costs for AWA as per the table that Mr Gower had presented. That figure, it appeared from the supporting spreadsheet, is a discounted figure, that is to say, in order to reflect the fact that the costs are likely to be incurred over the next 40 years, the expected costs are increased for inflation over that period but then discounted back to obtain the net present value. However, Mr Mountford was not at this stage happy with the way the discounting had been done. He therefore removed the impact of discounting to net present value by adding back onto the costs the amount shown in the spreadsheet as ‘Discount impact’. For the high estimate this figure was $33,623,939. This gave him a total for undiscounted liability of AWA on the high estimate of $297,913,225 (instead of $264,289,286). From this he deducted the then value of the remaining Maris Policy funds which was $168 million to arrive at what he called AWA’s undiscounted cost of $129,913,225 or $130 million.

Mr Mountford wrote to Mr Courteault copying in Mr Newell and Mr Martinet on 28 November 2008. He set out what he considered were the pertinent facts that had emerged from the meeting with PwC and Mr Gower. He set out the table he had prepared of the range of AWA’s costs on both a discounted and undiscounted basis in the four scenarios, the highest of which was the $130 million as I have described. He noted ‘Clearly the figures are provisional at the moment and subject to further review by PwC - so they are just being used as illustrative numbers at the moment’. He went on:

“We also understand that it is the intention to reduce the share capital to a nominal amount on the 15th December. The amount of the reduction would be transferred to reserves. There would be a dividend payment out of reserves but a significant amount of reserves (€100m?) would be retained in the company. There may then be further distributions next year.

We believe that it would be prudent to retain some share capital in AWA on the 15th December. The amount of this capital would be the difference between the amount of the Fox River provision in AWA's books and the high undiscounted scenario - in the above illustration it would be the equivalent of $67m ($130m less $63m).

We consider this is prudent in view of the uncertainties in calculating the provision referred to above. We believe it should be the undiscounted provision as this will be AWA's actual liability.

If we reduce the share capital to a nominal sum, even if we retain a "cushion" in reserves, it would be easy for future directors' (who may not necessarily be ourselves and who we have no control over) to declare this as a dividend. This could leave AWA with insufficient capital to meet its liabilities if they actually exceed the estimated provision.”

Mr Newell’s evidence was that he agreed with all Mr Mountford’s comments except the suggestion that it was important to leave the ‘cushion’ in the company in the form of share capital rather than in the form of distributable reserves. That was Mr Martinet’s view too. Mr Martinet wrote back to Mr Mountford on 1 December 2008 saying that what mattered really was the amount of assets left in the company to meet its liability, regardless of whether those assets were in the form of capital or distributable reserves. He told them that he shared their concerns. He referred to the recent instruction of Aon Accuracy to provide ‘an additional layer of professional expertise’ and ‘fresh eyes’. He also noted the victory in the Green Bay litigation against the insurers. He went on: (emphasis in original)

“I therefore urge you not to jump to any conclusion before all the parameters are on the table, and we have had the time to analyse them thoroughly. Clearly, however, and despite all our efforts and diligences, we will still presumably be confronted at our next BOD with a not unsignificant level of uncertainties. On top of that, some new elements might emerge between Dec. 15th and the effective 2008 year end closing, which may somewhat change the picture and potentially reduce further the uncertainties.

Having said that, now that AWA has become a structure principally dedicated to the Fox River issue, its shareholders' worth is undoubtedly, and quite grossly, disproportionate. Hence the idea of first effecting AWA's capital reduction prior to year end, assorted with a dividend distribution in kind which would leave AWA with ample reserves to very comfortably cover even the worse case scenario, as will be apprehended on December 15th. We would then reconvene in early 2009 when the PWC diligences will have been fully conducted on the basis of the latest Fox River developments, if any, and the AWA's 2008 accounts will have been closed.

Only then shall we be in a position to determine the right level of shareholders' worth, at large, that AWA should be left with, and the definitive amount of reserves that should be distributed in kind to AWA's shareholder. In terms of methodology, therefore, this means that AWA's new capital should in our opinion be "nominal", and this is where we seem to differ.

In fact, we strongly believe that our responsibility as Directors is not so much with the form of AWA's capital reduction, as with the amount of resources that the company should be left with to best meet its API obligations regarding the Fox River matter. And we will not be in a position to express an educated judgement on this before Q1 2009.”

Mr Mountford and Mr Newell had an informal meeting with Mr Gower at a public house on 3 December 2008 to go through the material that Mr Gower had discussed with PwC and Mr Mountford on 26 November. This was the first meeting at which Mr Newell had had a proper opportunity to be briefed on the figures by Mr Gower. After that meeting Mr Mountford and Mr Newell decided to instruct the solicitors Eversheds to provide them with advice about their duties in respect of the capital reduction.

On 8 December Mr Martinet emailed Mr Mountford and Mr Newell explaining that the transactions planned for mid-December would (emphasis in original):

“calibrate AWA’s capital reduction so that the company would be left with €150 million of interco loan assets with Sequana …

[this] would mean that the company would without a doubt have more than ample resources to cover its future obligations to API even in the very worst case scenario.”

On 9 December 2008 Mr Mountford emailed Mr Gower saying that they were ‘getting more comfortable’ with the matters to be resolved on 15 December. He commented to Mr Gower that their meeting with him gave them a lot of reassurance since Mr Gower had commented that if he were a director, he would be ‘entirely comfortable with the proposed capital reduction’. Mr Mountford asked Mr Gower to produce a memo giving them, as much as he felt able to, his assurance that the proposed capital reduction and dividend was reasonable. Mr Gower provided the memo on 11 December 2008 and sent it to Mr Martinet and Mr Courteault as well. Mr Gower said:

“I have shared with you my financial analysis of the Liability and attach to this memo my most recent analysis. As we have discussed, there remain a number of uncertainties which may have a material impact on the amount AWA is called to ultimately pay pursuant to the indemnity it gave Appleton Papers Inc. in 2001. I believe the issue you need to consider is to what extent should the company retain assets in excess of the Maris policy to meet those indemnity obligations. In my opinion, the funds required in excess of the Maris policy, in 2009 dollars, range between $55 million and $90 million with my best estimate based on currently available information being approximately $66 million.

Consequently, I have no difficulty advising the board that, following the reduction of capital, a dividend which leaves approximately €150 million in distributable reserves will, in my opinion, have no impact on the company's ability to pay any sums due pursuant to the indemnity as they fall due. For that reason, I believe it is reasonable to proceed with the proposed capital reduction and payment of a dividend.”

Again, Mr Mountford and Mr Newell said that they did not pay much regard to the figures used by Mr Gower as they were working on the basis that the worst case scenario was $130 million. However, Mr Newell felt that the memo was ‘a big comfort’ to him.

The 15 December 2008 meeting: the reduction of capital

Mr Martinet chaired the board meeting on 15 December 2008 to approve the capital reduction. All four directors were present together with Ms Quenby and various other advisers at MWE UK’s offices at Bishopsgate in London.

The Board had in front of them:

I need to explain what some of these documents were.

the November 2008 Interim Accounts

The November 2008 Interim Accounts were prepared by Mr Thomas for the purpose of the capital reduction. They showed a provision for the Fox River liability of €62.8 million. This was derived from the estimates that PCC had sent on 20 November 2008 which showed that the mid2 estimate of AWA’s liability would be $230,864,497 – that figure had then had the remaining funds in the Maris Policy deducted and been converted from dollars to euros to arrive at €62.8 million. That figure was based on a 60% NCR/API share.

The treatment for NRDs had changed in the November 2008 Interim Accounts from the treatment in the 2007 Final Accounts. In the 2007 Final Accounts the NRD figure had derived from an estimate of total costs for NRDs of $196.6 million, multiplied by 60% to arrive at NCR/API’s share and then by a further 60% to arrive at AWA’s share of $70.8 million. In the 20 November model used for the November 2008 Interim Accounts the NRD figure was based on an assumption that NCR/API would settle with the US Government for $35 million payable over 5 years so that AWA’s share of that would be 60% of $35 million, that is $21 million.

There were other smaller liabilities shown so that the total provision for liabilities in the November 2008 Interim Accounts was €68.5 million. The balance sheet showed, under current assets, debts of €593.3 million of which €592 was the intra-group debt owed to AWA by Sequana. They showed called up share capital of €318.6 million; a share premium account of €69.8 million and profit of €128.6 million.

the December Aon Report

Aon Accuracy is a financial consultancy which performs valuations, financial modelling and forecasting and market analysis. It is part of the Aon group which is a global insurance specialist. They were instructed by Mr Martinet on 19 November 2008 to look at two aspects of the Fox River liability, the remediation costs and the likely receipts under the Historic Insurance Policies. So far as the costs were concerned, Aon looked at the raw material produced by Tetra Tech rather than only at the PCC figures.

Aon Accuracy produced a draft report on 10 December 2008 (‘the December Aon Report’). In the covering letter accompanying the report, Aon described the task they had undertaken as follows:

“Key matters for your attention

Our job consisted in providing you with an estimated value of the net liability related to the Litigation. This valuation was performed on the basis of information provided by Management and its counsel, which we considered as being reliable and that we did not seek to verify independently. In particular, a significant part of our job is based on Management views which are its sole responsibility.

In this context, our job as independent advisors has consisted in exercising our own judgment on the said valuation, challenging Management views, providing the valuation methodology, ensuring the overall coherence of the valuation model and performing the actual computation.”

The Claimants pointed to the reference there to Aon not verifying the information provided to them by management and counsel (by whom they meant primarily Mr Gower) as greatly diminishing any supposed usefulness of the Report as an independent assessment of the material set out in it. I do not accept that. There was no point engaging Aon just to tell AWA what Mr Gower had to say on any particular topic. They were already paying Mr Gower to tell them that. What Aon were engaged to provide was what is described in the second paragraph quoted above, namely their very extensive experience of how insurance and insurers work. It is clear from their reports that on various matters they did not simply accept management’s or counsel’s view but brought to bear their own analysis, usually to arrive at a more conservative figure. I also reject any suggestion, as put to Mr Martinet in cross examination, that he tried to influence the outcome of the Aon Accuracy report to improve the position on insurance recoveries.

The December Aon Report concluded as follows:

At the 15 December meeting, Mr Martinet told the directors that the AON work was not yet complete. He told them:

“in consequence, in assessing the assets and liabilities of the Company in order to reach their conclusions for the purpose of making the Solvency Statement, the directors of the Company were to assume that the value of the Relevant Policies was zero.”

The PwC Agreed Upon Procedures report

In addition to the auditing work carried out by PwC in Paris and London, a different team in PwC’s London office was engaged by Mr Martinet to carry out a specific task for the capital reduction and dividend meetings in December 2008. The letter of engagement dated 15 December 2008 stated that the directors were responsible, for, amongst other things “identifying the appropriate assumptions underlying the calculation of the Fox River provision”. It went on to provide:

“Our Services

You have engaged us to perform the specified limited scope procedures (the "Services") set out in Appendix A on the Fox River provision calculation (the "Schedule"). You are responsible for determining whether the scope of the Services is sufficient for your purposes. The Schedule has been prepared by, and remains the sole responsibility of the Company's directors.

Our work will be performed in accordance with the International Standard on Related Services (ISRS) 4400 "Engagements to perform agreed-upon procedures regarding financial information".

As noted in Clause 1.4 of the Terms of Business, the Services do not constitute assurance; you and we agree that we have not been engaged to, and will not, perform an audit or a review, the objective of which would be the expression of an opinion on the Schedule. Accordingly, we will not express such an opinion. If we were to perform additional procedures or if we were to perform an audit or review of the Schedule, other matters might come to our attention that would be reported to you.

You should understand that there is no guarantee that these procedures will result in the identification of all matters which may be of interest to you.”

It spelled out again on the following page what PwC were and were not doing and how this task was not the same as their audit work:

“Our work will be based on information provided to us by the Company and will be carried out on the assumption that information provided to us by the Company is reliable and, in all material respects, accurate and complete. As set out in Clause 1.4 of the Terms of Business, we will not subject the information provided by you or contained in our reports and letters to checking or verification procedures except to the extent expressly stated. This is normal practice when carrying out such limited scope procedures, but contrasts significantly with, for example, an audit.”

The work that they undertook to do was to obtain a copy of the Fox River provision calculation, that is the spreadsheet provided by PCC, to check its mathematical accuracy and, broadly, to look at sources from which the figures in the spreadsheet are said to have been drawn to check that the right figures have been correctly transposed from that source into the spreadsheet and that inflation and discount rates have been properly applied. They also undertook to compare the value of the "mid2" model provision, less the Maris Policy funds, to the value of the provision recorded in the interim unaudited accounts for AWA at 30 November 2008 and report any differences in value.

Clause 1.4 of the Terms of Business referred to in the engagement letter provided:

“No Assurance - Save as expressly stated in the Engagement Letter, we will rely on and will not verify the accuracy or completeness of any information provided to us. The Services do not constitute an audit or review carried out in accordance with generally accepted auditing standards and no assurance will be given by us.”

The reference to ISRS 4400 is to the International Standard on Related Services 4400 on Engagements to Perform Agreed-Upon Procedures Regarding Financial Information. The Standard sets out the ethical principles to which an auditor must adhere when carrying out AUPs, including integrity, confidentiality and technical competence. Beyond that, it emphasises the importance of ensuring that there is a clear understanding regarding the agreed procedures and the conditions of the engagement.

As well as the engagement letter there was also a representation letter sent by Mr Martinet to PwC on 15 December 2008 in which he confirmed that AWA was responsible for ensuring that the Schedule had been properly prepared and was complete and accurate in all respects, reflecting all matters of significance in relation to the provision.

The AUP Report was also dated 15 December 2008 and stated that PwC had checked the mathematical accuracy of the Schedule that had been provided and then ran through the different inputs into the computation indicating the document to which they had reconciled the figures in the spreadsheets. They concluded that the value of the total API (that is AWA) provision on the basis of the mid2 model was $225.6 million; deducting the remaining Maris Policy funds of $154.8 million from that resulted in a provision of $70.8 million equivalent to about €55.7 million at 30 November 2008 exchange rates.

At the 15 December meeting each of the directors signed a solvency statement under section 643 of the CA 2006 stating that they had formed the opinion that (i) there was no ground as at 15 December 2008 on which AWA could be found to be unable to pay or otherwise discharge its debts and that (ii) AWA would be able to pay or otherwise discharge its debts as they fall due during the year immediately following 15 December 2008. The capital of AWA was reduced, by special resolution of Sequana, as sole shareholder:

The 17 December 2008 meeting: the December Dividend

After the meeting on 15 December 2008 and the capital reduction resolution, Mr Thomas drew up another set of interim accounts for AWA, as at 16 December 2008 (‘the December 2008 Interim Accounts’). The 17 December meeting was held with Mr Mountford and Mr Newell in AWA’s offices in Basingstoke and Mr Martinet and Mr Courteault joined by telephone. It was common ground that this was a more formal meeting because the decision to declare the dividend had in effect been taken as part and parcel of the earlier meeting.

The December 2008 Interim Accounts showed the effect of the capital reduction but not, of course, the effect of the payment of the December Dividend. On the balance sheet:

The apparent change in the provision for the Lower Fox River from €62.8 million in the November 2008 Interim Accounts to €58.4 million in the December 2008 Interim Accounts was due only to currency fluctuations and not to any change in the underlying dollar figures used to compute AWA’s share of remediation costs or NRDs.

At the 17 December meeting the board resolved to pay the December Dividend of €443 million by way of set off against the inter-company receivable owed by Sequana. That left an outstanding balance of the inter-company debt of €142.5 million.

Following the capital reduction and the payment of the December Dividend there were various developments on the different fronts, some of them pointing to a lowering of the likely liability of AWA and some of them pointing to an increase in that liability. I describe in the following paragraphs the developments that seem to me the most pertinent to these proceedings.

The decision of the US Supreme Court in Burlington Northern

By far the most significant of these was that on 4 May 2009 the United States Supreme Court handed down its judgment in Burlington Northern and Santa Fe Railway Company et al v United States et al 129 S.Ct. 1870 (‘Burlington Northern’). Since this judgment played an important role in the decision of the AWA board to reduce the provision in the 2008 Final Accounts and pay the May Dividend, it is important to understand what the case decided.

Burlington Northern concerned land owned by the petitioning railway company (‘the Railway’) on which an agricultural chemical distributor, Brown & Bryant (‘B&B’), stored hazardous chemicals. B&B bought some of these chemicals from Shell Oil Company. Over time, some of these chemicals spilled during their transfer and delivery at the site and as a result of equipment failures. This resulted in significant soil and groundwater contamination. In 1989 the Government exercised its powers under CERCLA to clean up the site and by 1998 the clean up had cost $8 million. The Government sought to recover these costs from, amongst others, the Railway and Shell. B&B had become insolvent in 1989. The Railway was said to be liable as the owner of the land and Shell was liable because it had ‘arranged for disposal ... of hazardous substances’. The trial court found the parties liable but only for a small portion of the costs. The court found that the site contamination created a single harm but concluded that the harm was divisible and therefore capable of apportionment. The trial court found the Railway liable for 9% of the remediation costs, and Shell liable for 6%. On appeal the Ninth Circuit Court of Appeals upheld Shell’s liability but rejected the lower court’s apportionment and held that both Shell and the Railway were jointly and severally liable for the whole of the cost.

In the Supreme Court, Justice Stevens, giving the judgment of the Court, described the case as raising ‘the questions whether and to what extent a party associated with a contaminated site may be held responsible for the full costs of remediation’. It was not disputed that the Railway was a PRP because it owned the site at the time of the contamination and continued to own it. The more difficult questions were (i) whether Shell was an arranger because of its sales of the toxic chemicals to B&B: had Shell arranged for the disposal of the chemicals when it sold them and (ii) whether the harm was divisible so that the Railway was only liable for a portion of it.

As regards Shell’s liability as an arranger, Justice Stevens held that it was plain from the language of CERCLA that arranger liability would attach if an entity were to enter into a transaction for the sole purpose of discarding a used and no longer useful hazardous substance. It is similarly clear that an entity could not be held liable as an arranger merely for selling a new and useful product if the purchaser of that product later, and unbeknownst to the seller, disposed of the product in a way that led to contamination. Less clear is the liability attaching to the many permutations of “arrangements” that fall between these two extremes—cases in which the seller has some knowledge of the buyers' planned disposal or whose motives for the “sale” of a hazardous substance are less than clear. In such cases, Judges Stevens said, courts have concluded that the determination of whether an entity is an arranger requires a fact-intensive inquiry that looks beyond the parties' characterisation of the transaction as a “disposal” or a “sale” and seeks to discern whether the arrangement was one Congress intended to fall within the scope of CERCLA's strict-liability provisions. The Court held that the term ‘arrang[e] for’ disposal of a hazardous substance should be given its ordinary meaning. Consequently, under the plain language of the statute, an entity may qualify as an arranger when it takes intentional steps to dispose of a hazardous substance.

The US Government submitted to the Supreme Court that by including unintentional acts such as ‘spilling’ and ‘leaking’ in the definition of ‘disposal’, Congress intended to impose liability on entities not only when they directly dispose of waste products but also when they engage in legitimate sales of hazardous substances knowing that some disposal may occur as a collateral consequence of the sale itself. Shell knew when it sold the chemicals to B&B that this would result in the spilling of a portion of the hazardous substance by the purchaser or by the transporter. Since Shell expected this result, it was, the Government said, properly found to have arranged for the disposal of the chemicals. The Supreme Court disagreed:

“[8] While it is true that in some instances an entity's knowledge that its product will be leaked, spilled, dumped, or otherwise discarded may provide evidence of the entity's intent to dispose of its hazardous wastes, knowledge alone is insufficient to prove that an entity “planned for” the disposal, particularly when the disposal occurs as a peripheral result of the legitimate sale of an unused, useful product. In order to qualify as an arranger, Shell must have entered into the sale of [the chemicals] with the intention that at least a portion of the product be disposed of during the transfer process by one or more of the methods described in [the definition of ‘disposal’]. Here, the facts found by the District Court do not support such a conclusion.”

Thus although the facts found showed that Shell was aware that accidental spills took place, that did not support an inference that Shell intended such spills to occur. To the contrary there was evidence that Shell took steps to encourage its customers to reduce spills. Therefore ‘Shell’s mere knowledge that spills and leaks continued to occur is insufficient grounds for concluding that Shell “arranged for the disposal” of the chemicals. Shell was not therefore liable for any of the clean up costs.

The Supreme Court then turned to the question whether the Railway was properly held jointly and severally liable for the whole of the remediation costs. The Court referred to the Restatement (Second) of Torts which provides that when two or more persons acting independently cause a distinct or single harm for which there is a reasonable basis for division according to the contribution of each, then each is liable only for the portion of the total harm that he himself has caused. But where two persons cause a single and indivisible harm, each is subject to liability for the entire harm. The Court confirmed that CERCLA defendants seeking to avoid joint and several liability bear the burden of proving that a reasonable basis for apportionment exists – the courts will not ‘make an arbitrary apportionment for its own sake’. In a footnote, the Court drew a distinction between the principles applied in apportionment and the principles applied when determining contribution claims between defendants. Equitable considerations play no role in the apportionment analysis but apportionment is only possible where the evidence supports the divisibility of the damage. Contribution, however, allows jointly and severally liable PRPs to recover from each other on the basis of equitable considerations.

The question therefore was whether there was a reasonable basis for concluding that the Railway should be liable for only 9% of costs. The District Court’s calculation was based on the proportion that the Railway’s property bore to the surface area of the whole B&B site, the proportion that the period of the Railway’s lease bore to the whole period when the site was operated and the locations within the site that had been largely responsible for the leakages, which locations were not in fact on the Railway’s land. The Supreme Court rejected the Appeal Court’s criticisms of this exercise and found that it was reasonable for the court to use the size of the leased parcel and the duration of the lease as the starting point for its analysis, in particular given that the District Court had included a 50% margin of error in its calculation, that is to say the arithmetical product of the exercise resulted in a 6% share which the court had increased to 9%. The Supreme Court therefore reversed the judgment of the Appeals Court and upheld the District Court’s decision that the Railway was only liable for 9% of the clean up costs.

Other developments on the Lower Fox River

Settlement by Government with de minimis parties I referred earlier to the Whiting CMO made by Judge Griesbach in September 2008 in the PRP contribution proceedings. He had given the defendants who were not subject to the Government’s UAO a four month stay of their disclosure obligations in order for them to negotiate a settlement with the Government. Towards the end of December 2008 the Government indicated a shift in its negotiating stance whereby it would treat the non-UAO defendants as de minimis contributors with the exception of the two municipalities which had operated the waste water utilities at the Fox River. Those two municipality defendants would be expected to make more substantial payments towards the costs. At the end of January 2009, Mr Hermes (the lawyer with conduct of the Whiting litigation on behalf of API and AWA) reported to Mr Tauscher that the Government had settled with 11 de minimis parties for a total of $1,875,000. Those parties effectively dropped out of the contribution litigation as they were now immune from any further liability at the suit of the other PRPs. Mr Hermes noted that the DoJ used a $1.5 billion site cost number during negotiations, though that included NRDs of $250 million.

Funding for municipalities’ potential liability The upside of the Government’s de minimis settlement with 11 of the Whiting defendants was that it became clear that the two municipality defendants (the City of Appleton and Menasha) were not only expected to make a substantial contribution but were being put in funds by Federal Government to do so. Mr Gower reported that the State of Wisconsin expected to receive $50 million in federal funding to contribute to the clean up on behalf of the municipal PRPs.

Identification of non-NCR paper Aroclors Another point that emerged after the December Dividend was one that cast doubt on the assumption hitherto made by everyone that all the PCBs in the Lower Fox River came from NCR paper. Almost all PCBs were manufactured by Monsanto Chemical Company. Until 1971 Monsanto produced and sold eight PCB mixtures including Aroclors 1221, 1232, 1242, 1248, 1254, 1260 and 1268. The first two digits in the naming scheme indicate the number of carbon atoms in the ring structure, and the last two indicate the average chlorine content. The Aroclor used in NCR paper was Aroclor 1242.

Sediment from the Lower Fox River was analysed by the PRPs’ environmental consultant, Ann Arbor Technical Services Inc (‘ATS’). They had been commissioned by Hermes, the lawyers acting for API in the Whiting litigation. Their final report was produced on 4 May 2009. They analysed a total of 1,150 samples collected along the length of the Lower Fox River during the period from 2004 to 2008. The samples were consistent in indicating that Aroclors other than Aroclor 1242 were responsible for PCB contamination in Fox River sediments, namely Aroclors 1254, 1260 and 1268. The potential sources for these Aroclors included manufacturers and end users of a range of products including hydraulic fluids, adhesives, pesticide extenders, rubber plasticizers and cutting oils. It appears that AWA intended to deploy this information in the Whiting litigation to dissuade the Government from settling with a number of PRPs who had been regarded as de minimis contributors to the contamination in the river.

Financial difficulties of other PRPs One of the PRPs which NCR/API regarded as likely to contribute to the clean up costs was WTM I. The company filed for bankruptcy on 29 December 2008. There were also concerns that Georgia Pacific might be a candidate for Chapter 11 protection at some point. It had been assumed in the Green Mediation that WTM I would be a contributor of about $16 million. As I described earlier, WTM I had been allocated a 10% share in the Tech Memo and a 12% share in the Amendola Report.

Collapse of the negotiations with Georgia Pacific Before the December Dividend, AWA’s advisers had been very optimistic about the chances of NCR reaching a settlement with Georgia Pacific. However, on 10 December 2008 the discussions broke off without agreement. It appears that although such a settlement would have had the advantage of reducing the scope and hence the expense of the Whiting litigation, Georgia Pacific were not offering a substantial amount of money up front but rather were expecting their payments to be deferred until after NCR/API had paid its share under the agreement.

More information about NCR’s knowledge of the harmful effects of PCBs The Claimants assert that it also emerged at this time that there may be documents showing that NCR had more detailed knowledge about the toxicity of PCBs in the 1960s than previously thought. The other PRPs rejected settlement proposals from NCR/API on the grounds that they thought that NCR/API would end up paying for the bulk of the costs because of its greater knowledge. The question of knowledge had already been raised prior to the December meetings. Thus, on 15 October 2008, Mr Bates had emailed Mr Gower in the aftermath of the Whiting CMO. He reported that the Judge Griesbach had placed primary emphasis on fault as an equitable factor. He said:

“This "fault" factor has been a major concern during the insurance litigation and in the mediation. In addition, there was some focus on fault in the NCR-API arbitration, in light of NCR's cleaning up of PCB-contaminated sediment in Japan and documents from Wiggins Teape to NCR indicating concerns about the environmental effects of PCBs. We countered it in the mediation by producing evidence that the recycling mills should have known that PCBs were in carbonless paper around the time in the late 1960's when the environmental harms of PCBs became widely known, and that they did nothing to abate their wastewater discharges, despite numerous state orders to do so. The mediator gave NCR/API a 55.5% share, based, we think, largely on the arguments regarding this factor. As an aside, the mediator stated that Glatfelter had a harmful document with respect to NCR on this factor, but he was not able to tell us what it was.”

In December 2008 Georgia Pacific asserted that they had found a new document which was damaging to NCR/API on the knowledge point, but said that they were not prepared to disclose it to the NCR/API team. One document does seem to have emerged in January 2009, a memorandum dated 5 November 1964 from a research and development establishment in England. This memo raised the question of how much of the PCB in the NCR paper was removed during the repulping of waste paper and conversely how much might still be present in the new paper made from the recycled pulp. The concern arose from the fact that the new paper might still contain PCBs and might ultimately wind up being used for food wrapping. The concern expressed recognised that the PCB is toxic and that is why it was important to know how much, if any, remained in re-pulped paper and potentially came into contact with food for human consumption.

Formation of the LLC Another factor which the Claimants point to as indicating that the NCR/API share might increase rather than decrease between December 2008 and May 2009 was the formation of an LLC in the US by NCR and API. This LLC was set up to perform the remediation work at the site. Mr Kirsch (the Claimants’ CERCLA expert) argued that the fact that NCR and API took on this responsibility by themselves might be seen as an acceptance by them that out of the PRPs they were primarily responsible for the clean up. However it seems to me that this factor can work both ways. Given that an equitable factor in the attribution of share is the extent of the PRPs’ cooperation in the clean up, this might work in their favour. Further, it also appears that the costs paid out by the LLC were reimbursed to it in part by Georgia Pacific and not solely by NCR/API.

The decision in Plastics Engineering I referred earlier to the Green Bay litigation and Judge Zuidmulder’s ruling in the District Court in November 2008 upholding the ‘all sums’ approach as opposed to the ‘pro-rata’ approach to the insurers’ liability under the Historic Insurance Policies. On 29 January 2009, the Supreme Court of Wisconsin handed down its decision in Plastics Engineering Company v Liberty Mutual Insurance Company (2009) WI 13. This was a decision on a certified question from the Seventh Circuit. Nine amicus briefs were lodged with the Court. The issue was whether Wisconsin courts should adopt an ‘all sums’ or a ‘pro-rata’ approach to determining the liability of insurance companies when an injury spans multiple successive insurance policies. The Court noted that the issue turned on the specific wording of the insurance policies concerned but that this was standard language found in other insurance contracts governed by Wisconsin law. The question arose in the context of the claimant’s liability to people suffering from asbestos related diseases following exposure to the claimant’s product and the scope of the claimant’s insurance coverage against those claims under a policy with the defendant. The Court dealt with whether Liberty Mutual was liable to pay ‘all sums’ arising out of an occurrence, up to the maximum limit of the policy, or whether it was obliged to pay only a pro rata amount. The Court noted that different courts across the country had adopted different approaches to this question. The Court upheld the all sums approach so that Liberty Mutual was responsible for "all sums," up to policy limits, even where the compensation was for damage that occurred partly before and partly within the policy period.

Success in the Green Bay Litigation reconsideration Meanwhile, the Green Bay Litigation (between API/AWA and its insurers) was continuing. At the time of the December 2008 meetings, API’s application to Judge Zuidmulder for him to reconsider the negative aspects of his November 2008 rulings was pending. The hearing of the reconsideration took place on 21 January 2009 and the ruling appears to have been given immediately. The application to reconsider was successful. Judge Zuidmulder vacated the part of his order which had stipulated that AWA would have to wait until its share was ascertained before being able to claim and left this point open. He confirmed that settlement receipts would have to be used up before further claims could be made on the policies but, importantly, he confirmed that AWA was entitled to submit claims for costs, both defence costs and indemnification of remediation costs, to the insurers. Further, if the insurers refused to pay because they wanted to appeal to re-establish the principle that AWA could only claim once its share was allocated, they were on risk of having to pay interest at 12% on the claims if they ultimately lost that argument and it was found that they had been bound to pay the claims when submitted. The Plastics Engineering ruling of course made any appeal in the Green Bay litigation about ‘all sums’ a pointless exercise.

Developments within Sequana and AWA

Mr Mountford’s and Mr Newell’s resignations Mr Mountford and Mr Newell resigned as directors of AWA on 14 May 2009. The Claimants tried to portray this as indicating that the two men disapproved of the payment of the May Dividend and suggested that this showed they regretted their approval of the capital reduction and the December Dividend. However, Mr Mountford and Mr Newell both denied that that was their motivation. By the time that they were brought into the discussions about the size of the May Dividend there was very little time before the planned meeting for them to delve deep enough into the numbers to understand them fully and assess what additional dividend could properly be paid. I accept their evidence on this point. It would not be fair to infer from their resignation that there was in fact anything wrong with the May Dividend or that they were having second thoughts about the wisdom of what they had done in December 2008.

The downturn in Sequana’s turnover and the group’s alleged financial difficulties. In the course of 2008 the Sequana group overall suffered very substantial losses. Some of the evidence about a pending financial crisis within Sequana came from emails sent by the former Sequana Executive Vincent Favier to Mr Gower – the two men seem to have been close friends and the emails were sent in part in the context of making arrangements for family exchange visits between their young children. These comments suggested dire problems within Sequana and painted a very gloomy picture of the company’s position.

However, a different picture emerged in the evidence of Mr Lebard and Mr Courteault. Mr Courteault’s evidence was that there was no problem with cash flow for Sequana in Autumn 2008 and he was not aware of any negotiations with Sequana’s bankers regarding the credit facilities available to the group. Mr Lebard was also very firm in rejecting the idea that any financial problems within the group motivated any of the transactions challenged in these proceedings. He showed a very clear and comprehensive grasp, as one would expect, of the group’s financial figures. He said that there was no threat to the ratios in the covenants that Sequana had in its loans from its bank lenders. He accepted that the paper industry had descended into a shambles at this time and that there was a substantial drop in turnover. But his evidence was that the question of market turbulence and reduced volumes of sales had nothing to do with Sequana’s financial performance. Indeed, because of the robust steps that Sequana took to restructure the business, such as shutting down or selling off mills, laying off employees, entering into a factoring agreement to provide financing for the redundancy payments, they managed to protect their profit position, halving the number of mills they owned. There was no issue with the liquidity of Sequana despite the difficult market conditions and there was no issue with the banks about current or future lending.

Mr Lebard was cross-examined about the drastic drop in the Sequana share price from a monthly average of about €18.50 in January 2008 to €5 in December 2008. This, of course, had a very bad impact on his own substantial personal investment in the company – a blow about which he showed commendable stoicism. He denied that this drop reflected an accurate assessment by the market of Sequana’s future prospects. Rather, he explained that there was a hedge fund shorting Sequana stock and putting false rumours in the market about the difficulties facing Sequana to push the stock price down for the fund’s own advantage. As I understood his evidence it was, at base, that he had had to dismiss Mr Favier some time earlier for being disloyal, as Mr Lebard put it. Mr Favier had set up his own hedge fund and Mr Lebard suspected that Mr Favier was shorting their stock and spreading rumours of impending doom for Sequana out of spite. I have not, of course, heard Mr Favier’s response to Mr Lebard’s suspicions but in my judgment Mr Lebard’s clear evidence carries more weight than what is at best tittle tattle from a former executive, whatever his motives.

The Final Aon report. On 20 February 2009, Aon Accuracy produced their final report in the form of an addendum to the December Aon Report discussed earlier. Their conclusions were as follows:

It was, Aon Accuracy concluded, ‘virtually certain that AWA will receive indemnities from the carriers’.

On the basis of their calculations they concluded that the estimate of net liability (that is the amount of AWA’s liability for costs in excess of likely receipts under the Historic Insurance Policies) was $78 million compared with $88 million in the December Aon Report. On the basis that there was $162.4 million left in the Maris Policy, they concluded:

“The overview of outcomes presented in this graph shows that:

- there is an estimated 84% chance that net liability is lower than 64 m$,

- there is an estimated 93% chance that Maris policy is sufficient to cover AWA's net liability,

- there is a 7% chance that net liability amounts at an estimated 231 m$ (cases when API does not receive any indemnities from insurers after losing all appeals).”

AWA’s intentions after the December Dividend

Following the capital reduction and dividend in December 2008, AWA’s share capital was reduced to €1 million from €318.6 million, its distributable reserves stood at €77.8 million and the latest Fox River provision was €58.4 million.

Mr Martinet’s view at the time was, he said, that they would ‘have another hard look’ when the 2008 accounts came to be closed and see if it was possible to release the provision and distribute it together with the distributable reserves. He denied that at this stage there was a plan to do so come what may. In cross examination Mr Martinet was shown an email from Mr Gower to Mr Tauscher in which Mr Gower reported to Mr Tauscher that he had spoken to Mr Martinet that morning and that ‘His aim is to show that assets net off liabilities so that there is no need for any reserve to be held in AWA’. Mr Martinet did not remember saying this to Mr Gower and could not say why Mr Gower would have reported that to Mr Tauscher. Mr Martinet reiterated his understanding of what was happening at the time:

“A. My Lady, what I remember from that date was that we had done some work, we were okay, we wanted to see whether it could be done in a different way once the reports were finalised and so on and so forth. And we will see. Again, that is all I can say.

Q. You --

A. And frankly, I go one step further. Frankly, in January and February, I was not obsessed by Lower Fox River at all. We had other things to do and I must honestly say that this was -- we had done something. We had done something I think to the satisfaction of everyone. I thought that what we were leaving in terms of assets was perfectly palatable by Sequana, and that is it.

You know -- so -- and frankly, I remember that in January and February I had much other things on my plate rather than go again on this issue. I thought I treated it, with my fellow directors, in a reasonable way. So whatever Mr Gower was -- had in mind, you know, I don't know.”

At the start of 2009, the Sequana group accounts for the year ended 31 December 2008 had to be finalised and audited by PwC in Paris. This required a figure for the AWA provision in respect of its Lower Fox River liability. The figures provided to PwC Paris by Mr Gower as part of this process assumed an NCR/API share of 55% rather than 60%. This reduction was picked up and challenged both by Mr Bartolotta and PwC in Paris who pressed Mr Gower to justify the change. The justification was expressed in a note to the model spreadsheet:

“NCR/API's share of full river costs was changed from 60% last year to 55% this year based on the expanded list of defendants involved in the Whiting litigation case that was submitted to the Wisconsin court in 2008. In addition, the Cities of Appleton and Menasha were determined to NOT be de minimis. API expects the parties to the litigation will settle the complaint by 2012 and that when the other parties’ contributions are considered, API’s share will likely be 55% of total costs.”

Ultimately there was further modelling done and the provision for AWA’s liability for the Fox River clean up included in the audited Sequana group accounts for the year 2008 was based on an assumed NCR/API share of 47.5% not 55%.

Back with AWA, various options were explored in early 2009 to reduce the provision in the AWA accounts. Mr Martinet tried to persuade PwC to approve the treatment of the prospective insurance settlement proceeds as an asset on the strength of the Aon Final Report. PwC were firm that despite the optimism that had been expressed about likely future receipts under the Historic Insurance Policies, they would have to qualify the accounts if the directors insisted on this. That ruled out that idea so far as Mr Martinet and Sequana were concerned. Other ideas were floated but were not realistic for one reason or another.

There were also more serious discussions with Mr Gower and Mr Tauscher about the sale of AWA to TMW. On 24 February 2009 Mr Martinet and Mr Gower met for dinner and Mr Gower made an offer for the company which Mr Martinet then discussed with Mr Lebard. However, as discussions progressed it became clear that it was not possible to find an acceptable deal with Mr Gower without leaving a substantial hole of €35 million (about $50 million) in the balance sheet. This was because Sequana were not prepared to sell AWA to TMW with a substantial amount of the inter-company receivable still payable to AWA. TMW did not have the resources to take over that liability from Sequana to AWA and the inter-company receivable could not be eliminated by a payment of dividend by AWA unless the provision could be removed.

In February 2009 there was a further spreadsheet produced by PCC and Mr Gower which showed the NCR/API share at 47.5% and various other changes to different inputs. On 3 March 2009 Mr Gower sent Mr Martinet some more spreadsheets including, in the footnote, his justification for lowering the NCR/API share. The spreadsheet showed the share at 47.5% and the footnote set out the justification as based on (i) the fact that more defendants had been brought into the Whiting litigation; (ii) the indication from the US Department of Justice that it did not regard the Cities of Appleton and Menasha as being de minimis parties; (iii) the provision of $50 million federal funding to settle the municipalities’ liability; and (iv) the discovery by ATS of non-1242 Aroclors in the River.

The NRD figures in these spreadsheets remained the same as in the December 2008 Interim Accounts, namely $35 million spread over 2012-2016.

Reaction in the AWA team to Burlington Northern

On 4 May 2009, when the Supreme Court’s ruling in Burlington Northern was released, NCR’s counsel Mr Lieb sent a copy of the judgment to Mr Gower the same day. Mr Gower immediately forwarded this to Mr Martinet saying:

“The US Supreme Court handed down its decision in the Shell case today. The decision appears to be favourable to NCR and API but I will provide a more detailed analysis after I have discussed the decision with Jeff [Bates] and the NCR lawyers on Wednesday.”

Mr Martinet then spoke to Mr Bates late in the evening when Mr Bates was travelling. Mr Martinet said in cross examination:

“So I called Bates, I think. And I said, "Jeff, is it important?" I think I got him late and he said -- a very quick conversation, as I remember, but he said, "You will have -- I'm working on it. You will have my answer to that. But it is significant. It is significant on apportionment, it is significant on joint and several, it is significant." I think he also said for arranger and so on.

And so that struck me as something important indeed”

…

“A. I think I was probably in Paris. And I think he was travelling. So I got him late, very quickly. He said, "Look, it is an important decision. I am working on it, but it could be very significant and I am working on it." That is what I remember. And I think he said, "No apportionment is confirmed, no joint and several", I think he mentioned arrangers and so on. And I think I said, "Okay, maybe that is very good news indeed."

When pressed, Mr Martinet accepted that he was not certain that he could recall Mr Bates mentioning arranger liability in that call but he certainly got the impression that Mr Bates thought the decision was significant. Mr Martinet immediately appreciated the significance of this development in solving the problem of removing the provision, eliminating the inter-company receivable and selling AWA to Mr Gower and Mr Tauscher. Mr Martinet wrote to Mr Gower:

“it looks as if we do not have a solvency issue at all any more: the Shell decision of today (no joint and several liability, apportionment, etc...) should significantly reduce AWA's liability, presumably more than by the 35 M euros we have in mind. Too bad we can't book it retroactively for 2008.

Our call with PWC tomorrow therefore seems redundant. But I will participate out of courtesy while eating my morning croissant.”

Both Mr Gower and Mr Tauscher forwarded copies of the Supreme Court’s ruling to Mr Bates. Mr Tauscher described it as ‘the decision we’ve been awaiting…’. Mr Bates emailed back saying ‘Hooray!!!!’ and that the result was ‘bad’ for Randy Stone and some of the other PRPs. Mr Tauscher sent a copy to Mr Hermes describing it as ‘the latest word on arranger liability’ which, he said, did not look good for the defendants in Whiting. A few hours later, still on 4 May, Mr Bates emailed Mr Gower with his summary of the effect of the case. He opened the email saying:

“I have not done any back and forth on the counter-arguments in this ‘brief’.”

The email consisted of headings and then passages from the judgment cut and pasted into the email with certain important sentences in bold. The headings set out by Mr Bates were: (I have omitted the passages from the judgment)

“No arranger liability for API or NCR:

No joint and several liability for API and NCR, except for the Appleton POTW (although, API would still have successor liability exposure with respect to NCR):

• API liability:

◦

None for OU1 (including to WTM1), and a 107 Recovery for the $10 million put in the OU1 escrow

◦

None for non-1242 PCBs

◦

Successor liability with NCR

◦

Joint and several liability with Appleton POTW (API/NCR an arranger there)

◦

Apportioned liability for the OU2-5 and NRD

◦

No joint and several liability for bankrupt parties shares including WTM I, or "inability to pay" including USP and PHG.”

To decipher that a little – ‘Appleton POTW’ is the City of Appleton’s liability for its water treatment works through which much of the contaminated water was released into the River; the ‘non-1242 PCBs’ refers to the Aroclors which are not Aroclor 1242 according to the ATS report; and the ‘107 Recovery’ refers to API’s ability to claw back, pursuant to section 107 of CERCLA, the $10 million it had already paid towards the clean up of OU1 on the basis of arranger liability because it was now thought that there would be no arranger liability for that upstream section of the River.

On 5 May 2009, Mr Gower sent an email with a more considered response to Mr Courteault, PwC and the solicitors Freshfields who had been retained to draw up the documents for the sale of AWA to TMW. He said:

“The effect of the case, as it impacts API, appear to be:

1.

Arranger liability (effectively liability for some part of the other PRP's share of the cleanup costs) has been abolished in cases like this where the Appleton coating plant sold a useful product; and

2.

Where liability can be apportioned (as it surely can in this case where each company had its own distinct outlet to the river) then it should be which means that there is no "joint and several" liability and liability for the orphan shares (arising from bankruptcy or inability to pay) falls to the government.

This decision will have a significant impact on how we assess API's future liability. The government already has a number of reports it commissioned which apportioned liability based on assumed discharges which, in turn, were based on assumed production. Those reports allotted a 38% share to NCR and API. Setting aside what those companies have paid in the past and rounding up to 40% would result in API's future liability being:

Cost of cleanup and LTM:	604.5 million
Less 2008 and Jan/Feb 09 spend:	95.1 million
Remaining costs	509.4 million
40% for NCR/API	203.8 million
API share @ 60%	122.3 million

You will note that this amount is less than the current balance remaining in the Maris policy.

If we apply the same analysis to Natural Resource Damages and start with the government's "demand" of a further $76 million, the API share of that number would be $18.2 million resulting in a total future liability of $140.5 million. As at the end of March the limits remaining in the Maris policy were approximately $145 million.

This may moot any discussion regarding the transfer of the receivable as a reserve in excess of the Maris policy may no longer be necessary.

We can discuss during today's call.”

The NCR and API lawyers in the Whiting litigation (that is Mr Hermes for API) met on 6 May 2009. Mr Hermes reported back to Mr Gower and Mr Tauscher what had been discussed. It was clear from the meeting that Mr Hermes took a very different and more positive view of the Burlington Northern decision from the view expressed at the meeting by the NCR lawyers, Sidley Austin. He reported that the NCR lawyers did not seem to think that it would have much of an impact at all; they did not believe that it ended arranger liability. There had also been a discussion about how the issue of arranger liability could be brought into the Whiting litigation which was about the contributions of the PRPs amongst themselves. Mr Hermes had argued at the meeting that they should try to get the arranger claims against NCR/API dismissed because the equitable factors, including knowledge of the toxicity of PCBs, were only relevant to the sharing out of clean up costs for pollution for which NCR/API were responsible. If they could show that they were not responsible for some of the pollution, either because they had not ‘arranged’ it or because it was divisible from the pollution for which they were directly responsible, then the equitable factors were irrelevant as regards that pollution. But he reported that NCR did not seem to see it that way and warned Mr Gower, ahead of a meeting the next day that ‘it does not appear that we will start the conversation on the same page’.

Mr Martinet recalled that Mr Gower had reported this discussion with NCR to him but had reassured him that NCR’s analysis was wrong and that the API lawyers were confident that their analysis was correct.

Mr Bates, Mr Tauscher and Mr Hermes met on 8 May 2009 to discuss tactics in the Whiting litigation and in particular how to try to introduce the apportionment argument into the litigation and put it forward to the Government in the context of the continuing work being carried out pursuant to the UAO. At the end of his email to Mr Gower reporting this, Mr Bates noted that all the models so far - the Tech Memo, the Amendola Report and the Green Mediation - had put NCR/API responsibility for direct discharges into the River at around or below 40%. But he noted also that the Supreme Court took the trial court’s conservative ‘margin of error’ figure significantly into account so that AWA ‘will have to blend in a reasonably conservative apportionment margin of error’.

The run up to the 18 May 2009 board meetings

In the run up to the May board meetings there were various iterations of the spreadsheets passing between Mr Gower, PCC, Mr Martinet and Mr Thomas. Mr Thomas was tasked with both finalising AWA’s 2008 Final Accounts and drawing up interim accounts for the period 1 January 2009 to 15 May 2009.

On 7 May 2009 Mr Gower sent to Mr Martinet and Mr Courteault a revised provision calculation. The changes made were:

The file attached to that email set out the figures that had been attributed to the different PRPs:

The ‘FRG’ figure there came from the FRG companies' Agreement on Interim Allocation on 27 July 1999, which included a de minimis allocation to Riverside Paper Corporation. Under that agreement, the companies selected these interim allocation percentages but the percentages were not binding. The figures in the spreadsheet that Mr Gower attached to that email showed an AWA liability of $155,711,131 at a time when the Maris Policy funds left were $150,980,919. This was based on an NCR/API share of 40% and NCR/API liability for NRDs of $21.1 million payable over five years.

Mr Martinet responded to these new figures by emailing Mr Gower saying:

“I want to discuss this before you send it out. We are not there as far as we are concerned and I want to understand why. Could PCC also do the standard comparison analysis (by columns, comparing 31/12/08 to today's on a line by line basis), please.”

Mr Martinet was cross-examined on what he meant by ‘We are not there’. It was put to him that he meant that Mr Gower had not yet achieved the goal of reducing the estimate of AWA’s liability to the level of the Maris Policy so that the provision could be eliminated. Mr Martinet firmly denied that this is what he meant. What he meant by the email was, he said, that he did not yet understand the reasons why various numbers had been used and he wanted to be sure he understood it before the calculation was circulated more widely. The new figures also used inflation rates that were not consistent with the former numbers. I accept Mr Martinet’s evidence on this point – there is no justification for imputing some sinister slant to his request for clarification of the numbers and a line by line comparison with the previous figures.

Mr Gower replied setting out the comparison Mr Martinet had requested in an email on 8 May 2009. He commented that they could argue for a lower than 40% share because the Government reports on which it was based assumed that the Appleton facility discharged to the River on the same basis as the recycling mills, which it did not (in other words the discharges from the Appleton facility arose from accidental losses of emulsion during the paper production process rather than by deliberately washing the emulsion off recycled paper when recovering the fibres).

Mr Martinet replied with some suggestions as to how the figures could be reduced further. As regards the 40% figure he said that if this ‘is deemed to be too high’ then PCC and Mr Gower would ‘have to determine what could be a realistic and defendable share’. He also made points on the inflation and discount rates, the defence costs and the overall remediation costs perhaps all being too high. He concluded “All this suggests that we should in all likelihood be south, not north, of $150M”. He asked Mr Gower to consider and tell him what he thought.

On 8 May 2009, PCC sent Mr Gower some revised figures and a reconciliation of the current cost estimate to the 2008 year end model. They had reduced the NCR/API share to 40%, updated the allocation credit to reflect that change, removed agency past costs, reduced NRD costs to $21.1 million over 5 years and added a $10 million credit in 2013. The resulting estimate of liability was $153.9 million. Mr Gower forwarded this to Mr Martinet with the comment “If we lower the percentage to 38% (as per the government's own reports) the total falls to $146.7 million”. That 38% figure came, it appears, from the interim cost sharing agreement among the FRG in July 1999. On 11 May 2009, PwC emailed Mr Gower to say that they had no further questions and that the figures should also be supplied to Mr Thomas to draw up the 2008 Final Accounts. Mr Gower emailed a revised provision to Mr Derbyshire and Mr O’Brien at PwC, copying Mr Thomas, Mr Martinet and Mr Courteault. In that email Mr Gower said: (emphasis added)

“Following our discussion last week. I attach a revised provision which you will see (as described in Footnote 3) takes into account the impact of the Burlington Northern decision. I have also attached an analysis of the allocations suggested by the government's own apportionment reports and Fox River Group interim allocation which enabled me to arrive at the 38% figure (the page is taken from the analysis we did for API's insurers). You should note that those reports and the allocation made certain assumptions about the process at the Appleton facility which are likely to have led to a higher allocation to NCR and API than is actually justified. When arguing apportionment, the practice at the facility is crucial in determining the output to the river; all of the depositions of former Appleton employees who worked at the facility during the relevant period have confirmed that the emulsion containing the PCBs was handled carefully due to its cost and was not discarded but great lengths were taken to recover and re-use the emulsion. It is reasonable, therefore, to argue that the NCR/API share of liability on an apportionment basis could be as low as single figures. Furthermore, these reports and analysis, when prepared, only contemplated 7 responsible parties, which is no longer the case.”

Later on 11 May 2009 Mr Thomas wrote to Mr Gower that he had applied the appropriate calculations to the model and the provision required was now zero. Mr Martinet responded to Mr Thomas: (emphasis in the original)

“What we can infer from the recent developments is that the Maris should indeed be sufficient to cover the Fox liability as currently determined, and more. Out of prudence, though, I believe we should disregard the slight positive difference resulting from your calculus and satisfy ourselves strictly with "Maris = Fox liability".”

That email was copied to Mr Mountford and Mr Newell. Mr Newell in response emailed Mr Martinet and Mr Courteault saying:

“Putting on my director's hat I have to ask the question: in the light of these changes is it still in the best interests of the company to sell AWA Ltd at the price currently envisaged? There just seems to be a lot of potential upside & not much downside.”

Mr Martinet replied as follows:

“Martin, this is the KEY question, isn't it?

What we first have to understand here is that i) the good news we just received will also have the effect of decreasing whatever AWA/API might get from the carriers in the future, especially in light of what we have already gotten from them in the recent past, and ii) we wish we were perfectly sure that there was no downside at all. There is definitely going to be a lot of legal work needed to bend the Supreme Court decision to fit our case. We presumably have a very good chance of success there, but I cannot recount the number of times in the recent past where good news has been systematically mitigated with bad news (remember the NCR/API allocation mediation, etc...?).

We are currently working on the concept of getting something out of this transaction, while pondering whether we should try and maximize our share of the potential upside (and thus remain associated with the Fox problem for another number of years) or whether we should satisfy ourselves with mostly getting rid of a very hairy situation once and for all, without being held responsible for future downside, however remote this might be. It should be a basic risk/reward analysis, except there are many moving parts, numerous parties involved and much at stake, in either direction.

Hence the dilemma, which nevertheless has to be sorted out sooner rather than later.”

Mr Martinet said that at this point he was not totally convinced that selling the company to TMW was the right thing to do. But there was pressure from Mr Gower to agree the sale and Sequana wanted to move on to other issues that they thought were more pressing.

On 12 May 2009, PwC made a few comments on the model that Mr Gower had emailed to them the previous day. PwC raised various queries on the figures asking how they were supported and pointing out that the discounting previously used was no longer valid and should be set at zero. Mr Gower agreed with this, a change that increased the AWA liability from $146 million to $150 million. PwC did not query the change to the NCR/API share. They did query the NRD costs asking why they had changed, although acknowledging that there would be some reduction to reflect the change in NCR/API share. In a reply also on 12 May, Mr Gower explained the new NRD figure on the basis that he took the Stone Total NRD Offer of $76 million and added back what had been paid to date by the PRPs to arrive at a total NRD spend of $115.16 million. He took 38% of that (being the NCR/API share) to arrive at $43.8 million as the NCR/API share of NRDs and deducted the $25 million that they had already paid to get to $18.8 million. AWA’s share of that NCR/API liability would be 60% of $18.8 million being $11.3 million which he then spread over five years resulting in payments of $2.26 million per year.

In fact there were some mathematical errors in the spreadsheet sent round on 12 May 2009 because the NRD figure was double discounted, in the sense that the figure for AWA-only share of $11.3 million was wrongly treated as the NCR/API share so that the 60% allocation of that share to AWA was applied twice.

Finalising the AWA audited accounts for the year 2008: the emphasis of matter

The work on the provision was being undertaken not only for the purpose of considering whether a further dividend could be paid but also because a figure was needed for the AWA audited accounts for the year 2008. Early in the morning of Monday 18 May 2009, PwC, who were producing the 2008 Final Accounts to be approved, raised the point about the emphasis of matter. Mr O’Brien of PwC said in an email to Mr Courteault, Mr Martinet and others:

“By way of managing expectations, the good news is that the disclosure around LFR seems comprehensive but there remain some internal inconsistencies and outstanding disclosure (such as the detail of the capital reduction in December) which need to be processed.

The one matter that I want to flag at this stage is that our opinion will include an "emphasis of matter" paragraph which will refer to the LFR disclosure in the accounts. This very clearly states it is not a qualification but given the level of judgement and major change from the previous year appears appropriate. Very happy to discuss further when you have a chance to see the words.”

The PwC audit report on the 2008 Final Accounts included the following statements:

“Opinion

In our opinion:

• the financial statements give a true and fair view, in accordance with United Kingdom Generally Accepted Accounting Practice, of the state of the Company's affairs as at 31 December 2008 and of its profit for the year then ended;

• the financial statements have been properly prepared in accordance with the Companies Act 1985; and

• the information given in the Directors' Report is consistent with the financial statements.

Emphasis of matter

Without qualifying our opinion we draw attention to note 15 to the financial statements, which describes how the Company has indemnified a former subsidiary company, Appleton Papers Inc, for costs in connection with the costs of investigation, remediation of and other costs related to the alleged contamination of the Lower Fox River in Wisconsin, USA. The valuation of this liability and its settlement date, together with the realisation of potential contingent insurance policy assets, involve significant judgements by the Company. While the Directors have carried out an assessment of the position at 31 December 2008, this matter will still depend on the rulings of court cases and other agreements with relevant other parties in the future, the outcome of which are not certain at the date of these financial statements, nor necessarily under the control of the company.”

The source of this idea of an emphasis of matter is the International Standard on Auditing (UK and Ireland) 700. That contains the following guidance to auditors:

“30.

In certain circumstances, an auditor’s report may be modified by adding an emphasis of matter paragraph to highlight a matter affecting the financial statements which is included in a note to the financial statements that more extensively discusses the matter. The addition of such an emphasis of matter paragraph does not affect the auditor’s opinion. The paragraph would preferably be included after the opinion paragraph and would ordinarily refer to the fact that the auditor’s opinion is not qualified in this respect.

…

32-1. The emphasis of matter paragraph describes the matter giving rise to the significant uncertainty and its possible effects on the financial statements, including (where practicable) quantification. Where it is not possible to quantify the potential effects of the resolution of the uncertainty, the auditor includes a statement to that effect. Reference may be made to notes in the financial statements but such a reference is not a substitute for sufficient description of the significant uncertainty so that a reader can appreciate the principal points at issue and their implications.

…

32-3. In determining whether an uncertainty is significant, the auditor considers:

(a)

the risk that the estimate included in financial statements may be subject to change;

(b)

the range of possible outcomes; and

(c)

the consequences of those outcomes on the view shown in the financial statements.

32-4. Uncertainties are regarded as significant when they involve a significant level of concern about the validity of the going concern basis or other matters whose potential effect on the financial statements is unusually great. A common example of a significant uncertainty is the outcome of major litigation.”

The 18 May 2009 meetings and the sale of AWA to TMW

The 18 May 2009 meetings were held by telephone call between Mr Martinet and Mr Courteault who were by then the only two directors. In fact there were three meetings held that evening. By the time of the first meeting, held at 9:10 pm the directors had seen:

At the first 18 May 2009 meeting Mr Martinet and Mr Courteault resolved:

There were two other board meetings held by telephone by Mr Martinet and Mr Courteault on the evening of 18 May 2009. The second meeting (following the meeting at which the May Dividend was approved) was timed as starting at 9:30pm and approved the sale and purchase agreement between Sequana and TMW. This involved a series of legal steps including the creation of a ‘golden share’ retained by Sequana, discussed below. The third meeting of the evening started at 10:30 pm and approved the sale of the company to TMW.

An important feature of the sale agreement between Sequana and TMW was the retention by Sequana of a ‘golden share’ which would enable Sequana to benefit from any future ‘upside’. The ‘upside’ could arise from successful negotiation of settlements with the Historic Insurance Policies insurers. If Mr Gower and Mr Tauscher could negotiate large lump sum settlements for AWA with the insurers and then minimise AWA’s actual future liabilities for clean-up costs they would enjoy the balance of the insurance pay outs as pure profit.

Mr Lebard’s evidence was that it was important for him that the deal with Mr Gower allowed Sequana to share in any ‘upside’ of AWA’s likely future position:

“A. … But again, my feeling, my understanding as a businessman, was that between the Maris Policy and the potential insurance recoup, actually there were more assets than liabilities in this case.

Q. That is the upside?

A. That is my feeling.

Q. That was the potential upside?

A. Yes, that was my feeling. It was more assets than liability. And so that in the end also I also mention that to you, my Lady, yesterday, is the fact that Mr Gower wanted to make the transaction. He was himself absolutely convinced that it was the case. So I didn't want to look stupid or foolish vis à vis my shareholders in my board in having Mr Gower make so much money that Sequana, in the end, had made a wrong decision.

Q. Of course Mr Gower wasn't going to be exposed to the downside, was he, the downside risk, personally?

A. Personally, I don't think so, yes. But again, you know, he was thinking himself, and I was sure of that, that he will make a fortune in taking over the Fox liability. And because he knew so well the case, I was pretty much convinced that there was pretty good news to come, especially vis à vis some court decisions and in the end he was really thinking he will make millions of euros.”

On 18 May 2009, Mr Martinet sent a confidential memo to the senior executives of the Exor company which owned Sequana, explaining the rationale behind the sale of AWA. In the memo Mr Martinet described the outcome of the Green Bay litigation, the large number of defendants in the Whiting litigation and the Burlington Northern Supreme Court decision. He said in the memo that the consequence of this was that ‘it can be reasonably demonstrated’ that the Maris Policy monies will be enough to cover AWA’s liabilities and indeed that there may be excess monies available both in the Maris Policy and from the recoveries under the Historic Insurance Policies. AWA was now in a position to be sold to Mr Gower and Mr Tauscher whom he described as ‘the natural purchasers’. He then described the main features of the proposed transaction:

The memo concludes with Mr Martinet’s comments:

“Altogether, we believe it's a good deal which, for all intents and purposes, allows Sequana to get rid of a very hairy issue under favourable terms, while greatly limiting its future exposure to the Fox matter, if any at all.

Time will tell, bearing in mind that the remaining Maris monies, anyway, are expected to cover API's Fox River obligations for a minimum of 2.5 to 3 years at the pre Burlington Northern (Latest Supreme Court decision) run rate.

A minimum monitoring will nevertheless remain necessary in the future, if only to secure Sequana's earn-out rights which could be significant. In the short run, Sequana will generate a windfall profit of — €35 million by getting rid of its Fox River reserve.”

The board meeting of Sequana took place on 27 May 2009. That described the effect of the sale as follows:

“The immediate effect of this operation was to externalise a significant underlying risk that was difficult to control from the scope of the group, with Sequana having expressly excluded any guarantee under the sale contracts of API and the Fox River risk. The CEO drew the Board’s attention to the strictly confidential nature of the conditions of this operation, which needed to be maintained in the group’s own interests.”

AWA was incorporated under the Companies Acts and was subject at all material times to the English statutory and common law provisions and principles restricting distributions and requiring capital to be maintained. Part 23 of the CA 2006 (‘Part 23’) sets out the detailed rules restricting the circumstances in which a distribution to members can be made lawfully. These provisions came into force on 6 April 2008.

There are two bases of claim alleging that the Dividends are unlawful. The first and primary claim is that the dividends were paid in contravention of Part 23 and the second that the dividends were unlawful at common law as a distribution of capital. However, the common law claim was not pursued in closing submissions and it was not suggested that the result would be any different from the result reached by applying the statutory scheme. I have set out the statutory provisions in an Annex to this judgment and summarise them here.

Section 829 of the CA 2006 defines a ‘distribution’ as every description of distribution of a company’s assets to its members whether in cash or otherwise, subject to some exceptions which are not relevant here.

Section 830 provides that a company may only make a distribution out of profits available for the purpose and that a company's profits available for distribution are its accumulated, realised profits less its accumulated, realised losses.

Section 836 then deals with how one determines what those profits and losses are and that is by reference to specified items in the company’s accounts, namely profits, losses, assets and liabilities, and provisions of various kinds; for our purposes, provisions of the kinds mentioned in paragraphs 88 and 89 of Schedule 4 to the Companies Act 1985.

Where the accounts relied on are the company’s last annual accounts, section 837 sets out the requirements that must be met by those annual accounts. In the sections as they were modified by transitional provisions applicable during the relevant period, the company’s accounts had to have been ‘properly prepared’ in accordance with the Companies Act 1985 or to have been so prepared subject only to matters that are not material for determining whether the distribution would contravene Part 23. Section 837 also provides that if the auditors of the accounts have qualified their report, they must have stated in writing that the qualification is not in respect of matters that are relevant to the application of Part 23.

Section 838(1) provides that where interim accounts are used to justify a dividend, they must be accounts that enable a reasonable judgment to be made as to the amounts of the items on which the justification of the dividend depends.

Section 847 provides that where a distribution is made in contravention of Part 23 then in certain circumstances the company can claim back the dividend from its member. This is a key provision in this case as it forms one of the main causes of action relied on by BTI in its action against Sequana.

BTI’s claim that the December and May Dividends contravened Part 23 are at base that the accounts by reference to which the directors justified the distributions for the purposes of section 836 either were not ‘properly prepared in accordance with the Companies Act 1985’ as required by section 837(2) in so far as they were AWA’s last annual accounts or that they did not enable a reasonable judgment to be made as to the amounts of the specified items as required by section 838(1) in so far as they were interim accounts. There are three aspects of the accounts that are challenged:

The Defendants say that the capital reduction was effective but they accept that if they are wrong about this, then both dividends did contravene Part 23. They also contend that the provision for the Lower Fox River liability was appropriate, but again they accept that if they are wrong on this, then the dividends did contravene Part 23. As regards the alleged failure to disclose details of contingent liabilities, the Defendants argue that they did comply with the relevant obligations but they also argue that even if they did not, these are not defects that result in the dividends contravening Part 23.

The reduction of capital: legal issues

Until 1 October 2008 a decision by a company to reduce the company’s capital required the approval of the court. On that date, section 641 in Chapter 10 of the CA 2006 came into force and provided two alternative procedures for a company to reduce its share capital. A private company limited by shares can reduce its share capital by special resolution supported by a solvency statement. Any company can reduce its capital by special resolution confirmed by the court. The relevant statutory provisions are set out in the Annex to this judgment.

Section 642 describes the circumstances in which a reduction of capital is supported by a solvency statement. The directors must make the statement not more than 15 days before the date of the special resolution. The requirements for a solvency statement are set out in section 643. A solvency statement is a statement that each of the directors has formed two different opinions about the company. The first is the opinion, as regards the company’s situation at the date of the statement, that “there is no ground on which the company could then be found to be unable to pay (or otherwise discharge) its debts”. The second opinion is that the company will be able to pay (or otherwise discharge) its debts as they fall due during the year immediately following the date of the statement. According to section 643(2), in forming those opinions the directors must take into account all of the company's liabilities, including any contingent or prospective liabilities. Section 643(3) prescribes the form that the solvency statement must take. It also provides that if a director makes a solvency statement without having reasonable grounds for the opinions expressed in it, he commits a criminal offence punishable by 12 months imprisonment or a fine or both.

Section 644 (as modified by transitional provisions in this case) deals with the registration of the special resolution and supporting documents. Within 15 days of the special resolution being passed, the company must deliver to the registrar the special resolution together with a copy of the solvency statement and a memorandum of capital. The memorandum of capital must show the amount of capital, the number and value of shares and the amount deemed to be paid up on each share. The registrar must register the documents delivered to him and subsection (4) provides that the resolution does not take effect until those documents are registered.

There are a number of issues raised by the parties on the construction of these provisions that are questions of law. The first seven issues relate to whether the solvency statement signed by the four directors on 15 December 2008 complied with section 643. The eighth issue concerns the memorandum of capital sent by AWA to the company registrar after the meeting on 15 December 2008. The legal issues can be summarised as follows:

There is also the factual issue as to what test the directors did in fact apply.

Before embarking on a discussion of these legal points there are some preliminary points to make. First, it is accepted that the directors all had a genuine belief that they were entitled to make the solvency statement in the sense that they were not acting dishonestly in making the statement despite knowing that they could not properly do so. Secondly, it is only section 643(1)(a) that is relevant in these proceedings. BTI does not assert that the directors were wrong to form the opinion required in section 643(1)(b)(ii); namely that AWA would be able to pay or otherwise discharge its debts as they fell due during the year following the reduction of capital. Thirdly, the allegation that the solvency statements were invalid does not depend on the Claimants showing that there was anything wrong with the provision made in the accounts. This is an allegation they make even if, which of course they deny, the accounts were drawn up properly. Fourthly, it is accepted by Sequana that all four directors had to have the right state of mind in signing the solvency statement in order for the capital reduction to be valid. Finally, it is worth bearing in mind that there is no cause of action relied on by BAT or BTI arising from the capital reduction per se – the relevance of the question whether the capital reduction was valid or not is that if it was not valid then it is common ground that the distributions of the December Dividend and May Dividend were unlawful because they were based on improperly drawn up accounts.

The correct approach to construing these provisions

The Claimants submit that I must construe the solvency statement route in the context that it operates in parallel with the court approval mechanism. The court approval mechanism contains strong protection for creditors both in its statutory terms and in the way it has been applied in practice by the courts over many years. The Claimants emphasise that it is a fundamental proposition of company law that a company’s creditors should be able to rely on the fact that the company has received, or is entitled to receive, a fund of assets represented by its issued share capital. Rules designed to ensure that the creditors can rely on that fund of assets include the rules preventing the company from reducing its share capital, repurchasing or redeeming its own shares or returning its capital to shareholders. Although these rules have been modified by statute, there are still important restrictions on the company’s freedom of action in respect of its capital. Parliament cannot have intended to leave creditors’ interests without adequate protection. The Claimants say there is no reason why creditors should be in any different position when the company uses the solvency statement route from the position they would be in if court approval were sought. The solvency statement route provisions should be interpreted in a way which enables them to be used only in clear, cast-iron cases where there is no threat to creditors’ interests.

The Defendants argue that the solvency statement route provisions must be construed in the context that it is vital to have certainty about what the value of the company’s capital is. To allow an unravelling of a capital reduction many years later on the basis of the kind of analysis that we have undertaken in this trial is the wrong approach. They rely on the comments of Megarry J in De Courcy v Clement and another [1970] 1 Ch 693 (‘De Courcy’). In De Courcy the claimant sought a declaration from the court that the directors’ declarations of solvency in a winding up made over four years earlier were of no effect because of the omission from the statement of assets and liabilities of an alleged loan. Megarry J held that it would be wrong to imply into the statute a requirement that the declarations of solvency be substantially or entirely correct. He considered that a declaration of solvency was a watershed, determining that the liquidator would be dealing with a members’ rather than a creditors’ voluntary winding up. A liquidator must, the judge held, be able to feel a degree of assurance that he was acting in accordance with the law when he relied on a declaration of solvency. Megarry J described a requirement of perfection for the declaration of solvency as creating:

“… something of a concealed trap, taking effect if it proves to be substantially complete and perfect, and retrospectively destroying the operation and effect of the declaration of solvency if, at whatever distance of time, error in the statement can be detected. The implication of a retrospective effect is, in my judgment, a significant pointer against a strict construction.”

Mr Foxton QC acting for the Defendants also relies on the criminal sanction imposed on directors under section 643(4) if they make a statement for which they have no reasonable grounds. The existence of the criminal sanction indicates, he submits, that a clear and coherent test which is easy for directors to understand is what is needed.

My conclusion on this point is that there are competing policy considerations here that Parliament must have taken into account when introducing this new mechanism for allowing a reduction of capital without the need for judicial scrutiny and without the protection for creditors that the court process has always provided. The key feature of this procedure is that there is no mechanism whereby the creditors of the company are notified of the proposed reduction or have an opportunity to object. This feature and the absence of any third party oversight struck Mr Courteault as remarkable since it contrasted with the French provisions with which he was familiar. However, that is what Parliament decided to enact. The concerns expressed by Megarry J in De Courcy about the dangers of retrospectively unravelling declarations of solvency years after the event apply equally here. I do not regard policy considerations as requiring me to approach the task of construing the provisions in any way that is out of the ordinary. The words must be given their natural meaning.

Is it necessary to show that the directors applied the right test and in fact formed the opinion stipulated?

Mr Foxton submits that once it is accepted, as it is here, that the directors acted honestly and that the solvency statement meets the requirements in section 643(3) as to its form, that is the end of the inquiry. The statement must be taken at face value and there is no justification for examining what opinion the directors in fact formed.

I do not accept that submission. Subsection (2) of section 643 states that in forming the opinions required by subsection (1) the directors must take certain things into account. That demonstrates in my judgment that the directors must actually have formed the opinions set out in the solvency statement. It is not enough that they make a solvency statement that says that they have formed those opinions if in fact they have not – for example because they misunderstood what the correct test was. I agree with the Claimants that it is necessary for the court to be satisfied that the directors have in fact formed the opinions required and hence that they have applied the correct test in coming to those opinions.

What is the correct test under section 643(1)(a)?

This was the most important area of dispute between the parties. It turns on what is meant by the requirement that there is ‘no ground’ on which the company could be found at the date of the statement to be unable to pay its debts. The Claimants submit that the directors have to form the opinion that there is no possible basis, even assuming that the worst case scenario arises, for concluding that AWA is unable to pay its debts. In their written opening submissions they expressed the test as a requirement that the directors come to the conclusion that there is no basis on which it could possibly be thought that the company is unable to pay its debts. In their closing written submissions the Claimants submitted that the provision requires the directors to consider not just their best estimate of the likely quantum of the contingent/prospective liability but the range of possible estimates at the time. If the directors recognised that a higher estimate than their best estimate could be arrived at, then they must consider whether the company’s presently available assets would be sufficient to cover that higher estimate. If the assets would not be sufficient, then there is a ‘ground’ on which the company could be found to be unable to pay its debts, and the directors cannot make the solvency statement. During oral closing argument, Mr Thompson QC, acting for the Claimants, expressed the test slightly differently saying that the provision requires a high degree of confidence on the part of the directors because in order to give sufficient protection for creditors corresponding to the court sanction route, the directors must be satisfied to a high degree of confidence that the company is able to pay its debts. ‘No ground’, Mr Thompson argued, should be given its natural meaning of ‘no basis’. On the facts of this case, it was easy for the directors to work out, adapting the PCC spreadsheets, that even if only one or two of the moving parts in the calculation of the provision moved against NCR/API, the Lower Fox River liability could quickly overtop the assets available as at 15 December 2008. In those circumstances the directors could not form the necessary opinion under section 643(1)(a).

Mr Foxton proposed a very different construction. He pointed out that the provision refers to the company being ‘found to be unable to pay’ its debts. This must, he argued, refer to the company being found by a court to be unable to pay its debts. This is in effect a cross reference to the grounds in section 123 of the Insolvency Act 1986. Section 123 sets out the grounds on which a company is deemed to be ‘unable to pay its debts’ for the purpose of conferring jurisdiction on the court to wind up the company under section 122(1)(f).

I do not accept Mr Foxton’s submission as a possible construction of section 643(1)(a). There is nothing in the wording of that subsection to suggest that a cross-reference to section 123 was intended. The circumstances in which the deeming provision in section 123(1)(a) arises do not fit easily into the mechanism for the solvency statement route. For example, a healthy and profitable company at no risk of either balance sheet or cash flow insolvency may for whatever reason neglect to pay a statutory demand for three weeks and hence be deemed unable to pay its debts. That does not greatly affect the company unless and until someone petitions to wind it up on that ground, at which point it can, of course, avoid being wound up by paying the debt. I do not see why the mere existence of an unsatisfied statutory demand should prevent the directors from making a solvency statement for the entirely separate purpose of effecting a capital reduction. Indeed, Parliament cannot have intended to import into section 643 an obligation on the directors to address their minds to the issues that frequently arise under section 123(1)(a), such as whether there is a bona fide dispute as to the amount claimed in the statutory demand. Further, section 123(1)(e) refers to the court being satisfied that the company is unable to pay its debts as they fall due. If Mr Foxton’s arguments were right, there would be a curious overlap between the directors having to form an opinion on that ground under section 643(1)(a) and also form the opinion required under section 643(1)(b).

However, I also do not regard the test that the Claimants propose as a workable one either. That would, it appears, require the directors of a company to go through every contingent or prospective liability facing the company, work out what is the worst that can happen, attempt to put a value on the liability in that scenario and then measure the company’s assets against that figure. Further it requires, on the Claimants’ argument, the directors to assume that any insurance that it has arranged will not pay out, since there is always a possibility in a worst case scenario that the insurer may find some way of disclaiming liability under the policy. Such a test would render the solvency statement mechanism unusable in most cases. It is not right to adopt a principle that only comfortably applies in the relatively unusual case of a non-trading company faced with a single major liability, the test has to be one that directors can apply in the whole range of circumstances facing a company. In a different context in this case, Mr Foxton made a point as regards BAT’s own contingent liabilities disclosed in its accounts. For example, in the Notes to its 2009 annual accounts BAT disclosed contingent liabilities in relation to tax audits, product liability litigation in which significant compensatory and punitive damages were being sought, civil actions by government healthcare agencies to recover the costs of treating people with smoking related illnesses, eight separate class actions by groups of smokers, claims relating to asbestos, a claim by the US DoJ for disgorgement of profits pursuant to the RICO statute and tobacco related claims in many other countries in the world. The notes conclude that “At least in the aggregate, and despite the quality of defences available to the Group, it is not impossible that the Group’s results of operations or cash flows in particular quarterly or annual periods could be materially affected by the final outcome of any particular litigation”. Even if a company’s position was somewhere between the two extremes of AWA with one liability and BAT with multiple liabilities, I do not see how the directors could work out what the worst case scenario is for each liability and put a value on it in order then to be able to form an opinion as to whether there are any grounds on which the company could be found to be unable to pay its debts.

I hold that the opinion that the directors must form is not whether, if calamity were to strike on some or all fronts, the company might be unable to pay its debts nor is it whether the court would have jurisdiction to wind up the company under section 123 of the Insolvency Act on a petition issued on the day the solvency statement was signed. The test is not a technical one but a straightforward one applying the words of the section. The directors must look at the situation of the company at the date of the statement and, taking into account contingent or prospective liabilities, form an opinion as to whether the company is able to pay its debts.

Taking into account contingent and prospective liabilities

This issue is linked with the previous question. The Claimants submit that the directors must take account of the contingent and prospective liabilities by assuming that they will have to be paid in full. They must consider the possible quantum of the liability when it is ultimately payable. The Defendants submit that this taking into account refers only to the need to ensure that proper provision is made in accordance with the accounting rules for potential liabilities. If a prospective liability is not sufficiently certain to merit a provision in the accounts, then there is no need for the directors to consider it.

Again, I do not accept that either of these constructions is right. I consider that the test to be applied is the same as was adopted by Sir Andrew Morritt C when he was considering the construction of the same wording in section 123(2) of the Insolvency Act 1986 in BNY Corporate Trustee Services Ltd and others v Neuberger Berman Europe Ltd (on behalf of Sealink Funding Ltd) and others [2010] EWHC 2005 (Ch). Andrew Morritt C said:

“31.

… the requirement 'to take account of contingent and prospective liabilities' cannot require such liabilities to be aggregated at their face value with debts presently due. Such inclusion would be commercially illogical; an obligation to pay £100 today has a higher present value than an obligation to pay £100 in five years. Had the simple aggregation of present and prospective liabilities been intended the subsection would have provided that the amount of its liabilities ‘include its contingent and prospective liabilities’.

32.

Third, subject to the foregoing, the subsection is silent as to what ‘taking into account’ a prospective liability involves. On the one hand a prospective liability cannot simply be added at its face value to the present liabilities of the company; on the other it cannot be ignored. In my view, the content of ‘taking account of” must be recognised in the context of the overall question posed by the subsection, namely whether the company is to be deemed to be insolvent because the amount of its liabilities exceeds the value of its assets. This will involve consideration of the relevant facts of the case, including when the prospective liability falls due, whether it is payable in sterling or some other currency, what assets will be available to meet it and what if any provision is made for the allocation of losses in relation to those assets."

The test is not a technical, all or nothing one, but a question to be posed by the directors to themselves considering the nature of the contingent and prospective liabilities, what assets will be available to meet them and what provision (in a non-technical sense) has been made for that purpose.

Must directors have reasonable grounds for the opinion that they form?

The fourth issue arising on the construction of the section 643 is whether the directors have to have reasonable grounds for the opinion that they form.

I find that this is not an element in subsection (1) but only in the definition of the criminal offence in subsection (4). The absence of reasonable grounds is a necessary ingredient of the criminal offence but it does not render the solvency statement invalid. I recognise that this creates a legal position which at first sight seems counterintuitive. If a director makes a solvency statement having no reasonable grounds for forming the opinion he in fact forms then he commits a criminal offence but the reduction of capital is still valid. But Mr Foxton has persuaded me that this must be right. The actus reus of the criminal offence involves not just the making of the solvency statement but the bringing about of the capital reduction. That is why the criminal offence is only committed when the statement is delivered to the registry because that step is needed to bring about the capital reduction. If the lack of reasonable grounds prevented the capital reduction from taking effect, there would be no actus reus but only ever an attempt to effect a capital reduction.

Mr Thompson pointed to section 642(1)(a) which provides that the resolution reducing the capital must be supported by a solvency statement ‘in accordance with section 643’. He points out that section 643 includes subsection (5). However, if reasonable grounds were part of the requirement for the solvency statement, one would expect it to be referred to in subsection (1) or (3) and not incorporated rather obliquely by inclusion in the criminal offence. I do not consider, therefore, that there is a requirement that the directors show that they had reasonable grounds for forming their opinion. Of course, the issue may arise on the facts of any given case since if it were entirely unreasonable to form the necessary opinion, a court may not be persuaded that the directors did in fact form the opinion. But it is not a separate element in the validity of the solvency statement.

When forming their view, should the directors assume that all the distributable reserves created by a reduction in capital will be distributed by way of dividend?

I can deal with this point fairly shortly as on the facts the issue does not arise for decision. The Claimants submit that where the purpose of the capital reduction is the creation of distributable profits, the directors must assume, when forming their opinion about the company’s situation as at the date of the statement, that the funds released to be distributable reserves will all be paid out as dividends by the company. The Defendants argue that such an approach wrongly elides the test in section 643 with the test for the distribution of dividends under Part 23. However Mr Foxton rightly accepted that the evidence given by all four directors showed that they regarded the capital reduction approved on 15 December 2008 and the distribution of the December Dividend two days later as being effectively one transaction. The directors had therefore assessed the company’s solvency on the basis that that dividend would be paid.

The issue might have arisen in relation to the May Dividend since the directors did not accept that the decision to declare that dividend had been taken by 15 December 2008. But in fact all of the capital reduction monies were paid out in the December Dividend. The capital reduction was €317.6 million plus the €69.8 million in the share premium account and the December Dividend was €443 million. The December Dividend distributed not only all the capital released pursuant to the solvency statement but some of the distributable reserves that had been on the balance sheet before the capital reduction. The May Dividend did not therefore involve the distribution of any of the capital released pursuant to the solvency statement.

The relevance of the Historic Insurance Policies

The Claimants submit that when taking account of the contingent and prospective liabilities of the company pursuant to section 643(2), the directors were not entitled to take into account the possibility that AWA would receive monies from its insurers under the Historic Insurance Policies either in settlement on cashing out the policies or by way of reimbursement for payments made by AWA under its indemnity.

This is part and parcel of their submission, which I have rejected, that the directors are required by section 643 to consider the worst case scenario when forming their opinion. I consider that it flies in the face of commercial sense to require directors to disregard the existence of insurance that they have put in place precisely to cover these contingent and prospective liabilities when they are considering the ability of the company to meet them in the future. An analogous point was considered by the Supreme Court in BNY Corporate Trustee Services Ltd and others v Neuberger Berman Europe Ltd (on behalf of Sealink Funding Ltd) and others [2013] UKSC 28 (‘Eurosail’). In that case, the conditions of issue of certain loan notes provided, as one of the specified events of default, a test mirroring the test in section 123 of the Insolvency Act 1986. It was alleged that the default provision had been triggered because the company was unable to pay its debts. This depended on how the requirement to take into account the company’s contingent and prospective liabilities should be applied when determining whether the value of its assets was less than the amount of its liabilities within the meaning of section 123(2). Lord Walker of Gestingthorpe (with whom Lord Mance, Lord Sumption and Lord Carnwath JJSC agreed) stated that: “Whether or not the test of balance-sheet insolvency is satisfied must depend on the available evidence as to the circumstances of the particular case”: paragraph 38. He noted that the future prospects of the company were dependent almost entirely on matters outside the control of its management, namely currency fluctuations, movements in LIBOR and the performance of the UK economy generally. Over the period of more than thirty years until the final redemption date of the loan notes, these factors were “a matter of speculation rather than calculation and prediction on any scientific basis.”.

At paragraph 42 of his judgment Lord Walker approved the statement of Toulson LJ (as he then was) in the court below in describing the court’s task as follows:

“Essentially, section 123(2) requires the court to make a judgment whether it has been established that, looking at the company’s assets and making proper allowance for its prospective and contingent liabilities, it cannot reasonably be expected to be able to meet those liabilities. If so, it will be deemed insolvent although it is currently able to pay its debts as they fall due. The more distant the liabilities, the harder this will be to establish.”

Of particular relevance to this issue is the fact that at first instance in Eurosail ([2010] EWHC 2005 (Ch)) Sir Andrew Morritt C considered whether it was possible to take into account the claims of the issuer in the liquidation of Lehman Brothers as present assets, even though the claims had not been admitted by the liquidators, and if so what was their value. The Chancellor recognised that the claims were not included in the financial statements because it is normal accounting practice not to recognise sums that may be recovered from ongoing litigation but have not yet been recovered:

“34.

… But the exercise required by section 123(2) is not the production of an annual balance sheet but a comparison of the value of assets with the amount of liabilities in order to ascertain solvency. There is no doubt that the claim in the liquidations of the Lehmans companies is a present asset of the issuer. Similarly there is no doubt that it is of considerable value, whether admitted or not. If the secondary market is indicative of value it is worth some 35% to 37% of US$221m. This asset and its value should not be ignored.”

Mr Thompson submitted that the Eurosail case can be distinguished from the instant case because the reason why the unadmitted claims in the Lehman insolvency could be taken into account was because they were tradeable on an established secondary market unlike AWA’s claims under the Historic Insurance Policies. I do not agree. The relevance of the fact that the claims against the Lehman liquidators were tradeable was only that this established that the claims had some value – not always the case with a claim against a liquidator. I consider that in line with Eurosail the directors were fully entitled to take into account the likely receipts by AWA under the Historic Insurance Policies when considering solvency.

The error in the memorandum of capital

Section 644 of the CA 2006 provides that within 15 days of the resolution to reduce the company’s capital being passed, the company must deliver to the registrar a memorandum which shows the amount of the share capital, the number of shares into which it is to be divided, the amount of each share and the amount at the date of the registration deemed to be paid up on each share. A memorandum purporting to comply with this requirement for AWA was sent to the company registrar within 15 days but it contained an error. The memorandum said that the share capital as reduced by a special resolution passed on 15 December 2008 was as follows:

This was wrong. It was true that the nominal share capital had been reduced by the capital reduction from 417,869,520 shares to 100,248,828 (a reduction of 317,620,692) but the paid up share capital had only been €318,620,413 so the capital reduction resulted in that paid up share capital being reduced to €999,721. Mr Thompson submitted that this was a gross error – it was wrong by a factor of 100 and so not a technical defect. According to section 644(4), the resolution does not take effect until documents under subsection (1) are registered and since the document registered could not be described as a memorandum complying with subsection (2), that had never happened. The resolution did not, therefore, take effect.

On this point I accept the submissions of the Defendants that to interpret the statutory provision as the Claimants contend would impose a stringency that cannot have been Parliament’s intention. The De Courcy decision discussed earlier (paragraph 318, above) shows that inadvertent error of this kind does not prevent a document from complying with statutory requirements and does not prevent the consequences of the adoption or registration of the document from taking effect. In De Courcy Megarry J rejected a similarly strict construction of the requirement for a statement of the company’s assets and liabilities saying: (pages 698H – 699)

“There must be something which can be reasonably and fairly described as "a statement of the company's assets and liabilities": but if there is, then even if it subsequently appears that there are errors and omissions, these will not prevent the statement from being a statement within the subsection. I do not think that I ought to impute to Parliament an intention to require perfection in a provision which contains no words to indicate this super-human standard.”

The same reasoning applies here. Despite the mistake mixing up the authorised share capital with the paid up share capital, the memorandum registered was still in my judgment sufficiently compliant with section 644 to avoid the extreme consequence that the resolution was invalid and hence all the audited company accounts filed since that date have been seriously misstating the company’s capital position. Mr Thompson argued that the less stringent meaning given to the statutory wording at issue in De Courcy was justified because it is very difficult for company directors to be able to state with complete certainty what the assets and liabilities are whereas it is very easy to state correctly what the company’s paid up capital is. However, that was not one of the several reasons relied on by Megarry J for his conclusion. Many of the reasons he did rely on apply equally here, in particular the fact that a criminal offence is committed by an officer of a company for a default in compliance and the serious impact of a retrospective destruction of the operation and effect of the special resolution.

I therefore reject this challenge to the efficacy of the capital reduction.

The reduction of capital: the factual issues

I have set out the history of the events leading up to the capital reduction on 15 December 2008. I now consider what opinion the directors in fact formed and whether this was an opinion that entitled them to sign the solvency statement.

Mr Martinet’s and Mr Courteault’s solvency statements

Even a best estimate can be wrong and, as I have explained earlier, I accept that the obligation under section 643(2) to take account of all the company’s contingent and prospective liabilities requires a director to consider, not the worst case scenario on all fronts, but whether there is going to be enough money in the company to meet those liabilities beyond the extent to which they are provided for in the accounts. Mr Martinet was cross-examined extensively over whether he took into account not only the provision that was included in the November 2008 Interim Accounts but also the possibility that that provision might be wrong. He was taken to the many documents, for example, in which NCR/API shares of more than the 60% used in the November 2008 Interim Accounts provision calculation had been posited and in which low, mid and high estimates had been used for the different moveable parts.

Mr Martinet pointed out that so far as he was aware that 60% share already had a margin built into it for error: see the explanation in paragraph 126, above. His evidence was that he had taken into account what he thought was a worst case scenario; he knew that there were a number of uncertainties that were extremely difficult to deal with so the way they dealt with those uncertainties was to leave enough assets in the company. His approach was therefore not to try to compute with any precision what the value of the worst case scenario was but rather to work out the best estimate of the liability and then leave a substantial margin in the company’s asset base. In my judgment that approach was entirely fair and, indeed, inevitable in the circumstances.

In an important passage in his evidence Mr Martinet described his thought process when he was considering the effect of the capital reduction on AWA’s liability. He said that as at September 2008, the inter-company debt owed by Sequana to AWA, converted from euros to dollars would amount to about $766 million. There was also about $160 million left in the Maris Policy. So AWA had assets of almost $1 billion to meet its potential liability. Given that AWA’s likely liability was only about 36% of the total remediation costs (60% of NCR/API’s share of 60%), $1 billion would only be needed to cover AWA’s share if the overall liability for all the PRPs together was over $2.5 billion. No one had ever mentioned that kind of figure as the likely overall costs. The company needed much less than that since the figures generally quoted for total remediation and NRD costs were about $650 million.

The capital reduction proposed would leave about $350 - $360 million in the company, even ignoring possible receipts from the Historic Insurance Policies. This comprised what would be left of the inter-company receivable; €149 million, equating to about $190 -200 million and the left over funds in the Maris Policy of about $160 million. Grossing that amount up, total costs would need to be close to $1 billion in order for AWA’s 36% share to be $350 million. So even after the capital reduction (and the distribution of the December Dividend) there would be enough money in the company to meet its liability even if there was a substantial cost overrun (i.e. total costs for all PRPs were close to $1 billion) or if AWA’s share in fact rose from 36% to 50%. His evidence was therefore that at the 15 December 2008 meeting he and his fellow directors understood that the company had ‘a comfortable buffer to take care, even of the worst-case scenario, and much further than that’. This was because according to the calculations he made, the cushion left in the company was more than the amount included in the provision itself and was enough to enable AWA to meet its liabilities if total costs increased by 50% or if AWA’s share of costs of $650 million increased from 36% to 50%.

I find that Mr Martinet’s approach to the difficult question he was faced with was entirely reasonable. I do not accept that he was obliged to carry out some more detailed ‘flexing’ of the models that he had been provided with to see where the numbers ended up. The models provided costs estimates of a net present value of sums to be expended over 40 years. It is of the nature of the use of spreadsheets and models that they generate figures exact to the last dollar – a spurious precision in these circumstances. That was needed in order to arrive at the figure for the provision because that needs to be a precise figure capable of audit. As to the solvency statement, Mr Martinet was well aware of the wide ranges quoted by the experts for possible remediation and other costs and the uncertainty about the NCR/API share. He took these uncertainties into account by leaving in AWA not only the margin of error built into the 60% NCR/API share but a very substantial amount of money to cope with the provision turning out to be wrong. Again, that was even ignoring the possible recoveries under the Historic Insurance Policies.

Mr Courteault’s evidence was less precise than this since he was largely relying on Mr Martinet to have worked out the numbers. Mr Courteault’s evidence as to what he thought going into the 15 December 2008 meeting was as follows. He understood the difference between the balance sheet insolvency and the cash flow insolvency tests in section 643(a) and (b)(ii) of the CA 2006. He also realised that he had to consider AWA’s ability to meet its share of the Fox River liability over the whole 40 year period, not just any demands under the indemnity that happened to come in over the following few years. He did not himself dig into the numbers but relied on what Mr Martinet and the experts they had engaged told him. When he signed the solvency statement he thought that there were enough assets in the company for it to meet its liabilities and had no reason to doubt this. He also had regard to the Historic Insurance Policies because the December Aon Report said that there was a very high probability of substantial recoveries to help AWA in future.

It was put to Mr Courteault that the meeting on 15 December 2008 should have been postponed because the documents were not ready until very shortly beforehand; they had been prepared in a rush and the directors had not had time to consider them. Mr Courteault denied this, saying that there was plenty of time at the meeting itself for them to consider the documents and the issues had been fully discussed beforehand. He said that there had been a full discussion at the meeting and that the directors were free to decide what they wanted to decide.

Mr Courteault accepted that by October 2008 it was Mr Lebard’s and hence Sequana’s wish to eliminate the receivable and move the money standing on AWA’s balance sheet up into the group. But Mr Courteault denied that this aim would have led any of them to achieve this goal ‘at any cost or at any price’:

“Q …Mr Lebard's view was that he wanted to take everything out, he wanted to keep any value in Sequana, not in AWA?

A. That was not my view. I mean, my view was that we had to achieve a transaction where we could reorganise AWA with only the Fox River liabilities in it. And then I was here to achieve this in a right way, I mean. I am the lawyer of the company. So I'm trying to -- what I am trying to do here is probably to make sure that we comply with the rules and that we are taking care of everything. That is the only thing.

Now, I was -- my target here, which has been assigned to me, was to achieve a transaction where we could reduce the receivable, so through capital reduction and the distribution of dividend, and maybe at some stage we would try to sell the company to somebody else. But that is all that what we had in mind here.”

In so far as Mr Courteault’s opinion was in effect the same opinion that Mr Martinet formed when signing the solvency statement, I find that it was the necessary opinion for the purposes of section 643.

Mr Mountford’s and Mr Newell’s solvency statements

Mr Mountford’s evidence was very clear as to how he and Mr Newell satisfied themselves that they were justified in signing the solvency statement. As I have described, by late 2008 both men were aware that Sequana wanted to move excess assets out of AWA by declaring a dividend that would pay off a large portion of the existing inter-company receivable. AWA would then be left as a vehicle solely responsible for the Fox River liability.

Mr Mountford and Mr Newell thought that there needed to be a ‘buffer’ not only over and above the provision in the accounts but also over and above the worst case scenario figure they had in mind. Mr Mountford’s calculation going into the 15 December 2008 meeting was that the worst case scenario was $130 million over the Maris policy of $160 million. I have described how he arrived at this figure in paragraph 200, above. The capital reduction and December Dividend would still leave €150 million of assets (about $200 million) (in the form of the provision and the remaining intercompany loan) plus the Maris Policy funds in the company.

Both Mr Mountford and Mr Newell also bore in mind the December Aon Report that indicated that there was a very high probability of recovering substantial sums under Historic Insurance Policies to meet the future liability, particularly given that AWA had already received over $100 million in payments under the policies. There were, they concluded, sufficient assets in the company without the insurance recoveries to cover the worst case scenario and there were more assets plus the insurance recoveries (which they thought of as amounting to a further $100 million) to provide a buffer even above that. All these calculations, of course, were made by Mr Mountford on the basis that the December Dividend of €443 million would be paid in two days’ time.

At the meeting on 15 December 2008 Mr Mountford recognised that when deciding whether to sign the solvency statement he had to consider both cash flow insolvency and balance sheet insolvency. He also recalled that Mr Martinet told the meeting that because the December Aon Report was not in final form, they should assume that there would be no monies coming in from the Historic Insurance Policies. Although the figures for the worst case scenario referred to in the minutes of the meeting were rather lower than those Mr Mountford had in his own mind, his recollection was very firm as to his own thought process in coming to the decision to sign the solvency statement:

“Q. The reality is, isn't it, Mr Mountford, that you knew very well in that meeting on 15 December that there was a serious risk that if this dividend was paid, the company might ultimately be unable to meet its indemnity liability if it turned out to be larger than estimated?

A. I was very cautious. I did not want to expose – as I said yesterday, my motivation here was to make sure I did everything legally, did not expose myself to a criminal act or indeed a claim, a personal claim against me. So I reviewed -- I believed that the amount of assets that we were leaving in the company were more than sufficient to meet the worst case scenario that we foresaw. So in my mind there was no question of insolvency. If there had been any issue in my mind on that, I simply would not have taken the risk of doing it.”

Mr Mountford was asked whether he felt under pressure from Sequana at the 15 December 2008 meeting to approve the transaction. His reply was that the main pressure for him was always the thought that they needed to be prudent, that he was putting himself at risk, and he was potentially committing a criminal act if he did not behave prudently. That, he said, was the main pressure he always felt in the process.

Mr Newell’s evidence as to his decision to sign the solvency statement was:

“I knew that the solvency statement was something that each individual director has to sign. So it is up to each individual director to make their own judgment as to whether they should sign it or not.

For me, I had my worst case scenario in mind, I had my assets in mind, and that is why I signed the solvency statement.”

At the meeting on 15 December 2008, Mr Newell described how Ms Quenby from MWE UK stood up in front of them and explained about the solvency statement and then came round to each of the directors in turn to get them to sign. He described his state of mind:

“A. Well, all I can say is that at that meeting that is what I genuinely believed. I did -- I thought there was no basis on which the company could not meet its liabilities in the future. And that is why I signed the solvency statement. I mean, I knew this was a significant transaction. I'm not going to do anything that I thought was incorrect. We had had McDermott there, giving us advice. We had PwC, they were at the meeting. They had vetted the accounts. They had looked again at the provision. As I say, from my point of view, when I signed that solvency statement, I believed there was no risk of the company becoming insolvent. That is all I can say.

Q. But you didn't, at that stage, have enough material to allow you to reach that conclusion, did you?

A. Well, I believe that I had. We had a range of estimates from PCC and Mr Gower. We had had PwC to look at the provision, review the provision. We had had McDermotts providing legal advice to us. I can sit here and say to you, my Lady, when I was at that meeting I had no doubt in my mind that there was no -- that the company -- there was no risk of insolvency.”

Although Mr Mountford and Mr Newell went about their calculation in a different way from Mr Martinet and Mr Courteault they were in effect doing the same exercise. They bore in mind the uncertainties surrounding the different moving parts within the AWA liability for the Fox River clean up; they were fully aware that the provision was only a best estimate and might be wrong by a substantial margin and they formed the honest opinion that even so, taking into account all the circumstances surrounding that liability, there was plenty of money left in the company after the capital reduction and the payment of the December Dividend to be sure that AWA would be able to meet its indemnity obligations.

In my judgment all four directors formed the necessary opinion for the purpose of section 643 and the challenge to the solvency statement and to the capital reduction fails. I therefore find that the December 2008 Interim Accounts (used to justify the December Dividend) and the 2008 Final Accounts (used to justify the May Dividend) were not defective by showing that the capital reduction had taken place.

The adequacy of the Lower Fox River provision in AWA’s accounts

It is alleged by the Claimants that the provision made in AWA’s accounts for the Lower Fox River liability was inadequate. This is another reason why it is said that the accounts did not justify the distribution of the dividends and hence why the dividends contravened Part 23. The Claimants’ case changed significantly during the course of the trial. It was initially alleged that the provision should have been based, at least as at December 2008, on the possibility that NCR/API would be liable for 100% of the clean up costs. This was supported by Mr Kirsch’s opinion relying particularly on the Whiting CMO. The Claimants submit that the decision embodied in the Whiting CMO shows that Judge Griesbach thought that knowledge was going to be a very important equitable factor which might override all other factors. However, after the close of evidence from the factual witnesses, the Claimants accepted that the figure of 60% used for NCR/API’s share in the provision for the purposes of the capital reduction and the December Dividend was a reasonable best estimate. However, the Claimants maintained their challenge to the NCR/API share used in the 2008 Final Accounts relied on to justify the May Dividend. As I have described, the best estimate of the NCR/API share in those accounts was assumed to be 38% (60% of which was then taken to be AWA’s share).

In relation to both the December Dividend and the May Dividend, the Claimants challenged the accuracy of the NRD estimate. This was a figure of $35 million used in the December 2008 Interim Accounts as the estimate of NCR/API’s share and a figure of $18.8 million in the 2008 Final Accounts (both of which were then multiplied by 60% to find AWA’s share).

The relevant statutory provisions and accounting standards

This raises a large number of factual issues but here I summarise the relevant provisions (set out in full in the Annex) and the legal issues that arise.

Part 23 allows a dividend to be justified by reference either to audited accounts or interim accounts. Where audited accounts are relied on they must have been properly prepared in accordance with the Companies Act 1985, or have been so prepared subject only to matters that are not material: section 837(2) of the CA 2006. Where interim accounts are relied on they must enable a reasonable judgment to be made as to the amounts of the relevant items in section 836(1): see section 838(1).

For the December Dividend the accounts relied on were the December 2008 Interim Accounts: see paragraphs 226 and 227, above. For the May Dividend the accounts relied on were the 2008 Final Accounts: see paragraph 293 above.

Section 226A of the CA 1985 sets out how a company’s annual accounts must be drawn up. They must include a balance sheet and that balance sheet must give ‘a true and fair view’ of the state of affairs of the company as at the end of the financial year. Subsection (3) provides that the accounts must comply with the provisions of Schedule 4 to the Companies Act 1985 as to the form and content of the balance sheet and the additional information to be provided by way of notes to the accounts. The relevant paragraphs of Schedule 4 for our purposes include paragraph 89 which provides that references to provisions for liabilities are to any amount retained as reasonably necessary for the purposes of providing for any liability the nature of which is clearly defined and which is either likely to be incurred, or certain to be incurred but uncertain as to amount or as to the date on which it will arise.

The question here is therefore how to determine whether accounts give ‘a true and fair view of the state of affairs of the company’. It was common ground between the parties that this was determined by reference to the relevant accounting standards applicable to the preparation of accounts. For this purpose the Defendants accepted that the same standard of probity in relation to the computation of provisions applied in interim accounts as applied for the company’s annual accounts. In other words, there is no difference between the requirement that an audited balance sheet show a true and fair view of the state of affairs of the company within the meaning of section 226A(2) of the Companies Act 1985, brought in by section 837(2) of the CA 2006, and the requirement that interim accounts enable a reasonable judgment to be made for the purposes of section 838(1).

As to the content of that requirement, the parties referred to the Opinion dated 21 April 2008 by Martin Moore QC who was instructed to advise on, amongst other things, whether the earlier Opinions from former eminent Chancery silks on the relationship between accounting standards and the requirement in the Companies Act 1985 that accounts give a true and fair view still held good. He concluded that they did. The requirement to prepare accounts which show a true and fair view is a legal requirement, the satisfaction of which is a question of law for the courts to determine. In determining that question, the Courts will rely very heavily upon the ordinary practices of professional accountants in determining whether accounts show a true and fair view. That is because those practices reflect the accumulation of experience and good professional practice and mould the expectations of the users of accounts as to the sufficiency and utility of the information in terms of quantity and quality.

The main accounting standard relevant to the computation of the Lower Fox River provision is Financial Reporting Standard 12: provisions, contingent liabilities and contingent assets issued by the Accounting Standards Board in 1998 (‘FRS 12’). The important distinction so far as making provision in financial accounts is the difference between recognising provisions and disclosing contingent liabilities. In FRS 12 a provision is a liability that is of uncertain timing or amount, to be settled by the transfer of economic benefits. A contingent liability is defined in paragraph 2 as either:

According to FRS 12, a provision should be recognised when an entity has a present obligation (legal or constructive) as a result of a past event, it is probable that a transfer of economic benefits will be required to settle the obligation, and a reliable estimate can be made of the amount of the obligation. Unless these conditions are met, no provision should be recognised.

There is no doubt here that the obligations of AWA under its indemnity to API counts as a present obligation arising as a result of a past event. Indeed, FRS 12 gives clean up costs for unlawful environmental damage as an example of an ‘obligating event’ which gives rise to a provision. FRS 12 acknowledges that the use of estimates is an essential part of the preparation of financial statements and does not undermine their reliability. This is especially true in the case of provisions, which by their nature are more uncertain than most other balance sheet items. FRS 12 provides that an entity will normally be able to determine a range of possible outcomes and can therefore make an estimate of the obligation that is sufficiently reliable to use in recognising a provision. If it is not possible to do so, then the liability is contingent only and should not be recognised in the accounts.

FRS 12 gives guidance as to how the company should measure the provision and what should be disclosed in the accounts:

FRS 12 draws a distinction between the case where the obligation which the provision recognises is one for which the company is jointly and severally liable with other companies and one where it may be entitled to reimbursement from other companies for the funds it has to pay to settle the obligation. According to paragraph 29, where an entity is jointly and severally liable for an obligation, the part of the obligation that is expected to be met by other parties is treated as a contingent liability and so not included in the provision sum. The entity recognises a provision for the part of the obligation for which a transfer of economic benefits is probable. Paragraph 56 deals with the position where some or all of the expenditure required to settle a provision is expected to be reimbursed by another party, for example under an insurance contract. In most cases, the company will remain liable for the whole of the amount in question so that the company would have to settle the full amount if the third party failed to pay for any reason. In this situation, a provision is recognised for the full amount of the liability and a separate asset for the expected reimbursement is recognised when it is ‘virtually certain’ that reimbursement will be received if the company settles the liability. The reimbursement should be treated as a separate asset.

A duty to obtain professional advice

There is one legal issue that arises about the application of these standards. This relates to the obligation to obtain proper professional advice. Both accounting experts Mr Lindsell and Mr Grummitt agreed that, faced with the need to make a provision for the Fox River liability, the directors should have recognised that expert help was needed in understanding the legal implications of CERCLA, working out AWA’s likely exposure and assessing what insurance cover was available. This applied to the drawing up of the interim accounts as well as to the final audited accounts.

Both experts also relied on ISA 620, an international standard on auditing which contains guidance to auditors on using the work of an expert as audit evidence. Mr Grummitt accepted that it was reasonable to treat the relevant parts of ISA 620 as the appropriate criteria for selecting and appointing expert advisers to advise a company in relation to a liability that is or may be a material item in its accounts. The main appointment factors are competence, objectivity (ensuring that the expert does not have a conflict of interest), the scope of the work the expert is required to perform (ensuring that the expert’s task reflects the need for assurance) and an assessment of the actual work of the expert, in light of the auditor’s overall knowledge of the business. Both experts applied these criteria to Mr Gower and the other experts but arrived at different results in respect of some of them.

The legal issue which arose is whether this requirement to seek expert advice is a free-standing requirement that must be satisfied before the accounts are considered to provide a true and fair view (as the Claimants contended) or whether the issue of the experts’ suitability is only relevant if the accounts themselves are determined by the court not to provide a true and fair view (as the Defendants contended).

I do not regard this as a separate requirement. If the figures themselves provide a true and fair view then it does not matter how they were arrived at by the directors. In my judgment, any imperfections in the process of producing the figures cannot taint the accounts if the figures themselves are justified.

The challenge to the provision in the December 2008 Interim Accounts

To recap, the provision for the Fox River liability included in the December 2008 Interim Accounts used to support the December Dividend was €58.4 million. This was based on the figures in a PCC spreadsheet that computed the AWA liability in excess of the Maris Policy at $79,911,234. It assumed (amongst many other things) that:

The only element I need to consider here is the challenge to the reasonableness in this computation of the NRD figure of $35 million. The Claimant’s challenge to this was primarily two fold. First they say that it was wrong to base the NRD input on an assumption that there would be a settlement between NCR/API and the Government. It should have been based on the share of the total amount of NRDs that the Government was likely to claim from all the PRPs, as appears to have been done in the 2007 Final Accounts. They relied on the evidence of Mr Kirsch that absent an actual settlement, any offer from Randy Stone did not fix the upper limit of the PRPs’ liability and when the offer was rejected, the Government was free to pursue the PRPs for a much larger sum. The second challenge is that there was no real justification for the figure of $35 million given that the lowest figure so far offered by Mr Stone to settle with NCR/API was $50 million.

As to the first point, it is true that Mr Stone indicated in the memorandum setting out the Stone Total NRD Offer that the claim that the Government would pursue if the PRPs did not accept this offer would be for $223 – 333 million: see paragraphs 88 onwards, above. However, that does not mean that the provision should be based on those much larger figures for a number of reasons. First, as Mr Martinet pointed out in cross-examination, the Stone Total NRD Offer itself recognised that since the plans to which those costings related had been formulated the Trustees responsible for repairing NRD had gained valuable experience in implementing actual restoration projects:

“In many instances, the Trustees have been able to achieve high quality restoration at lower than expected costs (specific areas of wetland habitat have been restored or preserved at less than the per-acre cost originally estimated by the RCDP). That past experience with interim recoveries — coupled with a reasonable expectation that a near-term final settlement could fund other cost-effective restoration projects — leads me to believe that the Trustees could justify settling the total NRD claim now for less than the prior dollar estimates, as discussed below.”

Those cost figures also did not take into account the credit for the money the PRPs had already spent on NRDs.

Secondly, in fact we know that all the PRPs rejected this offer indicating that the considered view of all the parties was that they could do better than that in further negotiations. It seems to me unlikely that the PRPs thought there was a serious risk that they would ultimately have to pay a very much higher sum at the end of costly litigation. Thirdly, the CERCLA experts’ evidence was that NRD claims generally tend to settle rather than to fight. It is true that, as Mr Stone made clear in his offer, the Trustees who would ultimately make the decision whether to settle at this level had not yet committed to the sum he put forward. However, Mr Kirsch and Mr Tenpas both also agreed that Mr Stone would not have put forward the offer unless he had been confident that it would be acceptable to those within the Government who had to approve any proposed consent decree. Further, Mr Lindsell very fairly accepted in his evidence that a reserve could be based on expected settlement if this was the best estimate.

In my judgment there was plenty of evidence on which the directors could conclude that the best estimate of likely liability for NRDs should be based on the likelihood of settlement with the Government rather than on the pessimistic view that the case would fight, the Government would revert to claiming a much higher figure and a court would ultimately award a much higher figure than had been discussed in the settlement negotiations. I do not see, therefore, that the provision in the accounts is wrong simply because it was based on a possible settlement figure rather than on the figure that the Government might claim if the issue were litigated.

Where did the $35 million come from? The Stone Total NRD Offer to settle for an upfront lump sum of $76 million was made in July 2007. That offer was itself a substantial reduction from an earlier offer, on which the 2005 and 2007 provision was based, of $125 million in 2004. There is evidence that by September 2007, the Government had reduced this to $67.3 million. Mr Gower emailed Mr Martinet and Mr Bartolotta on 19 September 2007 reporting on a meeting apparently with those parties taking part in the mediation and this is one of the items he mentions. Then on 30 May 2008 Ms Zurlo, a lawyer at MWE US, wrote to Mr Bates, Mr Gower and Mr Tauscher reporting on a conversation with Mr Stone. She said:

“First, (after noting we appreciated his comments and efforts) API would be looking for closure on the NRD issue. Without missing a beat, Randy said that closure would be $50M (for both API/NCR). He knew I wasn't calling to negotiate, and that I was only passing the info along to you all.”

It appears therefore that although Mr Stone had described his offer as the bottom line, there was still room for negotiation both as regards a global settlement for all the PRPs and for an individual settlement for NCR/API.

There is also evidence that Mr Bates had suggested a figure of $35 – 40 million as the best estimate of NCR/API’s likely liability and this was in fact an increase of $10 million over the amount estimated in 2006. In October 2007, Ms Dupuis from PCC emailed Mr Gower about her work on the 2007 reserve calculation, seeking his advice on some of the numbers. She said:

“1)

Shared NCR/API NRD payment- Last year we listed a $25 million payment for 2007. Should we use this same figure for 2008? Jeff Bates suggested $35-$40 million. We also would like to ask Laura Moore [of MWE US] to provide us with a short justification for the amount you give. We would insert her justification in the backup documentation for the reserve. She provided us with justification for last year's figure and, based on her time and knowledge on the subject, we do not believe it would be a difficult task.”

In February 2008, PwC in Paris were working on the provision to be included for the purposes of the Sequana group accounts. Mr Basset of PwC wrote to Mr Gower asking various questions and Mr Gower interpolated his answer into the email in bold italic font:

“3/ Estimation of NRD costs

As of December 2007, we understand that your best estimate is that API/NCR liability for NRD costs will be $35 million (+ $10 million compared to 2006, no support for that amount in the PCC report)

We understand that in a 2000, potential natural resource damages ("NRDs") were estimated by the U.S. Fish and Wildlife Service ("FWS") to fall in the range of $176 million to $333 million for all PRPs in the aggregate. In its 10-Q report, dated 5/10/2007, NCR Corp. mentions "for total natural resource damages (NRD), NCR uses a best estimate of $131 million. The range of reasonably possible outcomes is between $10 million and $176 million."

— > Could you please explain what are the 35MUSD derived from and why that amount has increased compared to 2006? How will you share these expenses with NCR (same agreement 60/40?)?

— > What is the status of the negotiations between IGP (Inter Governmental partnership) and API/NCR on that matter?

The companies are not negotiating the government's NRD claim at this time. It may be deferred until after the cleanup is complete. The $35million figure reflects our reasonable estimate of the NCR/API share of the government's demand for a further $75million which it made last year. Any NRD settlement will be shared with NCR according to the terms of the arbitration agreement. The 2007 reserve forecasts a $35M NRD settlement in 2009. By then, the $75M threshold with NCR will have been met and API's share will be 60%. I do not believe, however, that we will reach a settlement of NRDs in 2009.”

On the likely settlement figure for the NRDs, I also bear in mind the experience of Georgia Pacific in settling its own NRD liability. This shows that the courts are prepared to approve settlements that represent a small portion of the PRP’s potential liability. In 2001 Georgia Pacific had settled its liability for NRDs for $11 million. There was a legal challenge to the fairness of that settlement. In a ruling on 19 March 2004 (US and State of Wisconsin v Fort James Operating Company 313 F.Supp.2d 902 (2004)), the US District Court in the Eastern Division of Wisconsin held that the deal was substantively fair and reasonable and that it was consistent with the statutory goals of CERCLA. The court noted that the most recent estimate suggested that Fort James (that is, Georgia Pacific) discharged in the range of 15-20% of the total PCB mass into the River. The Government’s estimate for the total NRDs was said in that case to be $176 - $333 million so the court recognised that the settlement figure was far below 15% of even the bottom of this range.

However, the court acknowledged that there were good reasons to discount their claim against Georgia Pacific: the relative costs and benefits of litigation versus settlement including the risk that the Government’s estimate “greatly overestimates the total potential damages, particularly in view of the state's previous damages assessment”. There was also a value in discounting the share of an ‘early settler’ to encourage others to settle the claims against them. The court therefore held that while the consent decree did not mandate that Fort James pay its full equitable share of the natural resource damages, when viewed in light of the benefits of early settlement and CERCLA's joint and several liability scheme, it appeared to compensate the public satisfactorily for Fort James's share of the estimated natural resource damages.

I accept that the ‘early settler’ point would no longer be relevant to any settlement entered into between NCR/API and the Government. But it seems clear from this case that individual settlements at a sum far below what might be considered the PRP’s share of the Government’s total claim are agreed by the Government and approved by the court.

On the other hand, the Claimants point to what the December Aon Report said about NRDs. After recounting the various contributions so far made to NRDs by different PRPs and the offers from Mr Stone ($76 million from all PRPs and $50 million from NCR/API) Aon analysed five scenarios. These ranged from Scenario 1 which was that there was ‘a small chance’ that API would settle with the Government for a $12 million one off payment in 2009 (npv $11.8 million) to Scenario 5 which was that litigation which would take place after remediation was complete would result in a claim for $125 million from all PRPs (npv $44.3 million). They attached different probabilities to the five different scenarios and arrived at a compound average of these values and probabilities. This produced an average value of NRDs amounting to $30.4 million for API/AWA alone (not for NCR/API). This was about $11 million more than the figure included in the model that was used for the November 2008 Interim Accounts.

It is unfortunate that there does not seem to be a document in the very voluminous disclosure available at trial in which there is set out a calculation of how the $35 million was arrived at from the various settlement figures being discussed. We do not, of course, have Mr Gower’s evidence as to how he arrived at that figure. However, there is plenty of evidence, as I have described, showing the $35 million figure being discussed. Although there are other figures also referred to, it was open to the management of the company in arriving at a best estimate of NRDs to conclude that $35 million was the best estimate for the NCR/API share and hence that 60% of that, $21 million, was a reasonable figure to include for AWA’s share of NRDs.

I note that Mr Bartolotta, MWE US and the PwC Paris people were challenging this number to find out why it was proposed for inclusion in the provision. It appears that Mr Bates did support the $35 million figure and Mr Bartolotta and PwC must have been satisfied with the answers they were given. The Claimants made much of the point that the advice that Mr Martinet claims he was given by Mr Bates was not recorded in formal memos, even if it went to very important matters. Mr Martinet acknowledged that on occasion MWE had provided a formal written opinion, for example on the US law relating to fraudulent transfers in the context of the proposed sale of AWA or on the merits of a settlement proposal from Georgia Pacific. But Mr Martinet’s evidence was that he dealt with Mr Bates mostly on the telephone or at meetings and that is how legal advice was sought and provided most of the time. He felt comfortable ringing up Mr Bates whenever he wanted to discuss anything.

The Claimants also submitted that Mr Martinet and the other directors were wrong to place such reliance on Mr Gower’s opinion. This was for three reasons. First, it was said that he was not a US qualified lawyer and so could not properly advise on the many elements of the provision that called for expertise on US litigation. Secondly, he was not independent because he was engaged as a consultant by the company and thirdly, the Claimants point to incidents recorded in the contemporaneous documents that show Mr Gower to be prone to put an optimistic slant on things, making him unreliable. I consider these criticisms were misplaced. By the time Mr Martinet and Sequana became interested in the Lower Fox River liability Mr Gower had already been working full time on all aspects of the issue on behalf of API and AWA for several years. It is apparent from the number of memos and emails Mr Gower sent over the relevant period that he had a comprehensive and detailed knowledge of the litigation and the legal issues despite his lack of any formal qualification as a US litigator. There was no reason for Mr Martinet to jettison the ‘sunk cost’ of Mr Gower’s knowledge and expertise and bring on board a US lawyer who would have to start familiarising him or herself with the complex history of the matter. It is true that Mr Gower was engaged by AWA and so was not independent of the company in a formal sense. But he was a professional person and I am not prepared to assume that he would give false advice or put his own interests over those of AWA for some personal advancement.

Finally, Mr Martinet accepted that Mr Gower was not perfect. There are various points in the email chains where Mr Martinet is rather critical of Mr Gower. I find that Mr Martinet did what any astute and careful businessman would do when presented with an adviser who is not of his own choosing but who is already deeply involved with the issue to which Mr Martinet was coming afresh. He made an assessment of Mr Gower’s strengths and weaknesses and where he identified weaknesses he supplemented Mr Gower, as it were, with people he trusted. That was why he engaged Mr Bartolotta to oversee and challenge Mr Gower’s work and that was why he spoke directly to Mr Bates to get his input first hand rather than always relying on Mr Gower’s reports of Mr Bates’ advice. His conclusion about Mr Gower was as follows:

“I always said that I believed that Mr Gower was extremely knowledgeable of all the issues in Fox River. He had spent seven years or eight years, whatever, working only on that. And although I understand that he was not a CERCLA expert, but he was -- he had a legal mind. He was an intelligent person. He was hard-working. And I -- and it was my view was that he knew much more than me and that his advice was reliable. I also knew that Mr Bates was very much in the background and so -- and Mr Bates being a CERCLA expert. And yes, I mean, I thought that Mr Gower was as good as any, and more probably better than any to provide us with solid advice. That is what I thought.”

I therefore reject the criticism of the $35 million figure in so far as the complaint is that it was based on the advice of Mr Gower. I conclude that there is no reason to doubt that the NRD figure of $35 million was the best estimate of all those involved in calculating the provision on the basis that this was the likely settlement figure that would ultimately be negotiated with the Government to settle NCR/API’s liability.

The challenge to the provision in the 2008 Final Accounts (used for the May Dividend)

There are two challenges to the 2008 Final Accounts by reference to which the May Dividend was justified. In these accounts the provision was eliminated because it had been concluded that the best estimate of AWA’s liability would not exceed the funds left in the Maris Policy. The spreadsheet from which this estimate of liability was drawn showed that:

The reduction of NCR/API’s share from 60% to 38%

The reduction of the NCR/API share from 60% to 38% was one of the most contentious issues in these proceedings. I have described earlier the events which occurred between December 2008 and May 2009 in the Lower Fox River and how the provision calculation evolved over that time. I now turn to consider whether those events justified a reduction in the NCR/API share from the 60% used in the November and December 2008 Interim Accounts to the 38% share used in the 2008 Final Accounts.

Although the Claimants did not, in the end, challenge the 60% NCR/API share used for the provision on which the December Dividend relied, Mr Martinet was cross-examined as to whether it was appropriate by May 2009 to rely on the three very old reports that Mr Gower referred to as the basis for the 40% direct discharges which were the starting point for the 60% NCR/API share computation. It was put to him that they were unsuitable bases for PCC’s computations because of the purpose for which they had been drawn up and because they were caveated by the authors in various ways. Mr Martinet’s response in relation to the FRG cost sharing report could equally apply to the Tech Memo and the Amendola report, namely that ‘it had the virtue of existing’. He thought that it represented work that had been done seriously by the various parties involved. I agree with Mr Martinet’s approach here. There are of course many uncertainties, but his obligation as a director to come up with a figure for the provision did not permit him to throw up his hands in despair. He and those advising him had to rely on the work done by their advisers and on reports that had ‘the virtue of existing’ and that is what he did.

As to the move from 60% to 38%, the spreadsheet from PCC which set out the calculation on which the elimination of the provision in the 2008 Final Accounts was based contained various Notes on the different moving parts. The note explaining the NCR/API share said as follows:

“NCR/API's share of full river costs was lowered from 60% last year to 38% this year in light of the following developments which have occurred since 1 Jan 2008: (i) the list of defendants named in the Whiting litigation (filed by NCR and API in January 2008) now stands at 23, which number includes the 6 other PRPs together with a number of other companies which are not expected to be de minimis parties; (ii) in seeking to reach a settlement with 11 of the defendants in the Whiting case, the US Department of Justice has indicated that it does not consider the Cities of Appleton and Neenah-Menasha to be de minimis parties suggesting that NCR and API may expect the City of Appleton to pay a material proportion of those companies’ liability; (iii) NCR has informed API that, as a result of its lobbying, it expects the state of Wisconsin to receive at least $50 million in funding from the federal government to settle the cities and municipalities liability; and (iv) as a result of testing conducted by API late in 2008, it appears that a material amount of the PCBs in the river have resulted from the discharge of Aroclors other than 1242, suggesting that the universe of parties liable to pay for the remediation will further increase; and (v) the Supreme Court's ruling in Burlington Northern which the company's legal adviser has stated (i) removed the likelihood that NCR and API would bear any arranger liability and (ii) suggests that liability for the cleanup will be apportioned resulting in no joint or several liability falling on any of the PRPs. For these reasons, API believes that a more realistic estimate of the combined NCR/API share of the remediation costs is now 38% of the total costs, which number is supported by the government's own apportionment analyses although that number is the upper end of the range and, therefore, remains prudent.”

Mr Martinet confirmed in evidence that these were the factors that he took into account when concluding that the new spreadsheet gave the best estimate. I will consider each in turn and the reasons why the Claimants say it was wrong for the directors to rely on these points to reduce the NCR/API share.

The increase in defendants in the Whiting litigation The Claimants did not accept that the mere increase in the number of defendants to the Whiting litigation was any reason to conclude that NCR/API’s share of the clean up costs would ultimately be lower than 60%. They also make the point that the number of defendants increased before the December 2008 meetings and yet the 60% figure was still used in the November and December 2008 Interim Accounts. This second point is not a valid one because it was clear from the evidence that the 60% figure was retained from earlier years as a prudent step because the work on the figures was still on-going. It was not based on a conclusion that nothing material had happened to change the share but rather that it was too soon to work out what any such changes might mean.

I consider that it was reasonable to assume that if there were 23 defendants in the Whiting litigation this was likely to lead to a larger pool of people being held liable to pay something towards the costs of clean up. This would in turn reduce the amount that NCR/API would have to pay. Mr Martinet’s evidence was that his recollection of this development was that although some Whiting defendants were being considered de minimis, there were still something like 18 or 19 defendants that were still very seriously being considered as potential PRPs. It was not simply a matter of the number of defendants increasing. More significant to my mind is the fact that the 60% share previously used was based on the direct discharge share of 40% in the Tech Memo and the Amendola Report: see the explanation of the 60% in paragraph 126, above. Those reports had shared the direct discharges out as between only a handful of PRPs. If many more companies were being brought into the dispute than had been considered in those two early reports then it is certainly likely that when the pie is shared out among this larger group, a smaller piece will land on NCR/API’s plate.

The likely contribution from the Appleton and Menasha municipalities Mr Martinet’s evidence was that more significant in his mind than the settlement with the de minimis parties was that the City of Appleton and the Menasha Sewerage Commission had been put in funds to pay $50 million towards the clean up. This was, in my judgment, an important factor. It is not surprising that once it becomes known that $50 million has been allocated to the municipalities to enable them to pay their share of the costs, people thereafter assume that $50 million will indeed be required from those parties. That is consistent with the CERCLA experts’ evidence that parties with funds available are often targeted by the Government. I accept that this development overrode the apparently negative development of the settlement by the Government with 11 de minimis defendants. The 60% share that had been used up to and including the spreadsheets relied on for the December Dividend had not allocated any share to these 11 defendants because they had not been included in the Tech Memo and the Amendola Report. So their exclusion as de minimis participants now could not have signalled the need for an increase in share.

The identification of the non-NCR paper Aroclors in the Lower Fox River I referred earlier to the ATS Report which identified the presence of other Aroclors in the Lower Fox River sediment, indicating that sources other than NCR paper were responsible for the pollution: see paragraph 242, above. The Claimants, supported by Mr Kirsch, dismissed the ATS Report as irrelevant because it was commissioned by AWA for the purpose of persuading the Government not to settle with the smaller PRPs for de minimis amounts. I reject that approach. The presence or absence of Aroclors in the sediment is an empirical fact and there would be little point in faking or exaggerating test results, even supposing that a consultant would be willing to do so. The work was commissioned to see if anything useful could be found to support NCR/API’s case. Presumably if it had confirmed the assumptions underlying the Tech Memo and the Amendola Report that all the PCBs came from NCR paper, the Report would have been shelved.

As it turned out, ATS’s research did find something useful. Its significance was increased when viewed together with the Burlington Northern ruling since the ATS finding pointed to the possibility of the harm to the River being divisible as between one industrial sector and another. Mr Bates was certainly alive to the significance of these two developments taken together. In the email that I quote from in paragraph 269, above, Mr Bates said that the consequence of Burlington Northern was no liability for non-1242 PCBs. The conclusion that the ATS Report would, when explored further, lead to additional PRPs outside the carbonless paper industry being required to make a contribution either by the EPA or by the Whiting claimants seems to me entirely logical and to justify a reduction in the assumed NCR/API contribution.

Burlington Northern I have described the ruling of the Supreme Court on 4 May 2009 and the very positive reaction of AWA’s advisers, including Mr Bates and Mr Gower. It was primarily this which in Mr Martinet’s mind justified the reduction from the 47.5% share that had been used in the Sequana group accounts in early 2009 to the 38% share in the AWA 2008 Final Accounts. He accepted that his thoughts about the best estimate were all based on the assumption that Burlington Northern applied – he had been very strongly advised that Burlington Northern would be applied and that it represented ‘a major change in the CERCLA world’ as regards arranger liability and apportionment. He did not take into account the percentage share if Burlington Northern did not apply.

The Claimants suggested that it was inappropriate to make such a significant deduction from the NCR/API so soon after the Supreme Court ruling was delivered. I do not accept that criticism. All those involved in these matters were well aware of the Burlington Northern case and of the decisions in the lower courts. They had followed the progress of the case once the Supreme Court accepted the appeal in July 2008 and Mr Bates attended the oral hearing before the Court on 24 February 2009 and so knew the issues that were being argued. He had had plenty of time to think about the implications for the Fox River case pending the handing down of the judgment and had discussed its implications with Mr Gower in February. It appears that during the hearing the Supreme Court Justices gave some indication of the way their thoughts were tending as Mr Bates correctly predicted the outcome in an email at that time.

I have described the email that Mr Bates wrote which Mr Gower forwarded to Mr Martinet and Mr Courteault amongst others on 5 May 2009. It was put to Mr Martinet that all that Mr Bates was doing here was putting forward the arguments that could be made about the effect of the Burlington Northern, not that he was saying that this was his conclusion of its effect. Mr Martinet said that was not how he read it – he took it ‘as something absolutely as a sea change’. I do not accept that this document was as limited as the Claimants suggest. It was described by Mr Gower as Mr Bates’ analysis of the decision. Clearly what Mr Martinet and his colleagues wanted was exactly that; Mr Bates’ thoughts on what the impact of the decision was on their position, not simply what arguments could be put forward. I do not read the comment about ‘back and forth on the counter-arguments’ as indicating anything more than that the passages quoted in the email from the judgment were limited to the Supreme Court’s conclusions and did not set out what counsel in the case had argued.

The Claimants also criticise Mr Martinet for not obtaining formal advice from the lawyer acting for API in the Whiting litigation, Mr Hermes, as to the real impact. But Mr Martinet knew that all the advisers were in touch with each other and there was no reason to think that Mr Bates’ and Mr Gower’s view was strikingly different from Mr Hermes’ view.

I also recognise, as was discussed in some of the email traffic, that there were difficulties in ensuring that the benefit of the ruling in Burlington Northern translated into a lowering of liability for NCR/API. Apportionment or divisibility of harm was not an issue in the Whiting litigation because that claim was concerned only with the allocation of shares of the loss for which PRPs were jointly and severally liable. As Mr Kirsch explained, the only way to push the issue of divisibility was for NCR/API to stop complying with the Government’s UAO order to force the Government to bring enforcement proceedings. In defending those proceedings, NCR/API could rely on the Burlington Northern ruling to argue that they should not be required to pay so much towards the clean up costs. The experts differed as to how risky a strategy that might be. Mr Kirsch said that if the court ultimately rejected NCR/API’s reliance on Burlington Northern on the basis that it was not for some reason applicable to the Lower Fox River site, then NCR/API might be held liable to pay triple damages and daily penalties for failing to comply with the UAO. NCR/API might also be branded as uncooperative for the purposes of assessing the Gore factors in the Whiting litigation. Mr Tenpas acknowledged that risk but thought that it was unlikely that a court would treat NCR/API has having had no sufficient cause to disobey the UAO order even if the court decided that Burlington Northern did not justify a removal of arranger liability or a division of the harm.

Despite these difficulties, in my judgment the prize for NCR/API of persuading a court that NCR/API should not carry arranger liability for the recycling mills’ discharges and/or that it should not be jointly and severally liable for a substantial part of the River’s contamination was a very valuable one. I consider it was reasonable of Mr Martinet and Mr Courteault to assume, given the very positive response to the ruling that they received from their advisers, that the ingenuity of US litigation attorneys acting for the PRPs would find a way of bringing the matter to court. Similarly, although the factual scenario of the pollution from spilled and leaking chemicals in an asphalt covered goods yard in the Burlington Northern case was very different from the Lower Fox River, the significance of the ruling was to signal to courts that they were entitled to take a broad brush approach to divisibility. On arranger liability, I can see strong similarities in Shell’s position selling useful chemicals to B&B knowing that some of them will be spilled with NCR/API’s position selling useful brokes and trimmings to recycling mills knowing that they would wash the emulsion out of the pulp before reusing it. I do not see that it was overly optimistic to assume that arranger liability was no longer on the table. Mr Kirsch also challenged reliance on Burlington Northern to remove arranger liability on the basis that although the Supreme Court ruling signalled a substantial change in approach in relation to arranger liability for some states in the US, the courts in Wisconsin had been applying the ‘useful product’ concept already. But the important point is that the 60% NCR/API used earlier did, despite that, include a proportion for arranger liability in relation the whole of the River and it was this proportion that was now being removed.

Taking all these factors into account, I find that it was reasonable for Mr Martinet and Mr Courteault to base their best estimate of the NCR/API share on the assumption that NCR/API would no longer be liable as an arranger of other PRPs’ discharges and that a way would be found to cut down the scale of the pollution for which it was jointly and severally liable.

The Claimants pointed to various factors which they say should have caused the directors to regard an increase in the NCR/API share rather than a decrease as being likely. I consider here the factors that appear to me to be the most significant ones.

Financial difficulties of the PRPs WTM I filed for bankruptcy on 29 December 2008. There was a difference of opinion between Mr Kirsch and Mr Tenpas about whether this meant that any claim against it by the Government or by PRPs was discharged (that is, whether or not it could be maintained against WTM I if and when the company emerged from the insolvency proceedings). Other evidence of different PRPs, including Georgia Pacific, either being in financial difficulty or stating in their financial accounts that they would not be prepared to contribute more than a given figure to the Lower Fox River costs were put to the Defendants’ witnesses. The Claimants contended that these indicated great uncertainty about the assumption that there would be other PRPs still available at the end of the day to pay the balance of the costs beyond the NCR/API share. Mr Martinet denied that he thought it possible that Georgia Pacific would have foundered by the end of the process. He could not remember whether any of the other financial difficulties were factored into the final number given for NCR/API’s share. He also made the point first that some of the PRPs might well, like API, have insurance policies or indemnities from other substantial companies which could still be required to pay out in these circumstances and secondly, that the 60% NCR/API share used in the earlier provision had included a margin of error designed to take account of other PRPs dropping out of the picture.

I agree that it was important to bear in mind that a best estimate that NCR/API will pay only 38% of the costs assumes to a certain extent that other PRPs will be able to meet the remaining 62%. If some of them have become insolvent then other PRPs may have to pick up their share. But it is also important to bear in mind that much of the drop from 60% to 38% was based on the Burlington Northern decision. That, it was thought, removed arranger liability and allowed AWA to limit its liability to those divisible parts of the River’s contamination for which it was jointly and severally responsible. If this is right then the demise of other PRPs does not necessarily mean that part of its share will then fall to NCR/API. If the failed PRP is responsible for direct discharges for which NCR/API is neither an arranger nor jointly and severally liable directly, then my understanding is that NCR/API’s liability would not necessarily increase. Although I accept therefore that the actual or possible bankruptcy of some of the PRPs was a factor that needed to be borne in mind, it is far from clear to me that at this stage it pointed clearly to a likely increase in NCR/API share.

NCR/API’s knowledge of PCB toxicity I referred earlier to statements from other PRPs in January 2009 that they had uncovered evidence to show that NCR’s state of knowledge in the 1960s of the toxicity of PCBs was greater than had previously been thought: see paragraphs 245 onwards, above. I do not accept the Claimants’ submission that this was a significant factor pointing towards a higher share for NCR/API than had previously been thought appropriate. There is evidence that Mr Bates reviewed these documents but no one appears to have regarded them as sounding any alarm bells, given the extent to which the knowledge issue had already been investigated in earlier litigation and negotiations.

Mr Martinet’s evidence about the May Dividend

I have discussed the reasons that were put forward in the Notes to the PCC spreadsheet for reducing the NCR/API share to 38%. Mr Martinet also gave evidence about how he approached the question of whether to pay the May Dividend. Mr Martinet accepted that by the time the 18 May 2009 meeting took place, the decision to pay the dividend had in effect been taken because they had come to the conclusion a few days earlier that the inter-company receivable could be eliminated, releasing the money previously used in the provision into distributable reserves. It was put to Mr Martinet that the only question he and Mr Courteault had considered at the first 18 May 2009 meeting was whether the company had sufficient distributable reserves on its balance sheet to cover the proposed May Dividend. He did not accept that that was the only point:

“ That was to me the most important point, of course. But there was also the matter of the solvency which we had -- which we had to consider and which we had considered carefully. Again, prior to that day. So we knew it was an important decision.

…

Q. And the solvency point that you mention, was what? What was the question that you say had already been considered?

A. Well, the solvency point dates back to the long work we had had on this issue in the preceding month, and of course we needed to be comfortable that this would not put the company in jeopardy, looking forward. We were leaving some resources in the company. We thought those resources were enough to cover its liabilities in the long run, and that is what we were considering.

Q. Can you just identify what resources you are referring to?

A. Two resources, one which was the Maris, which was at that point in time, $151 million, if my recollection is correct. And the second was of a more contingent nature, but on which we were very confident, which was the possibilities of very, very significant settlements or indemnities from the insurance carriers.”

Mr Martinet thought the estimate of $146,688,383 for AWA’s liability for the Lower Fox River ‘was the result of a lot of hard work, a lot of process, looking at those numbers, looking at the cost, the remediation costs that were provided by a number of experts’. He also realised that the figures used would be reviewed by the company’s auditors and that there were a number of people working on the numbers, checking their consistency with previous figures used. I accept that he honestly thought it was indeed the best estimate.

The 38% figure was criticised by the Claimants on the basis that it removed not only arranger liability from the previous 60% estimate but also any margin for error that had been built into the 60% figure. Mr Martinet’s evidence was that he believed that the 38% NCR/API share did contain ‘an element of prudence’. Mr Martinet knew that Mr Gower had said in an email to PwC that the right number was in the range from as low as a single digit share to 38% or 40% (see paragraph 283, above). He explained the calculation that he made in his own mind at the time. He understood that the 60% figure that had been used for a long time, including in the November and December 2008 Interim Accounts, had been based on an assumption of direct discharges of 40%, that is to say, two thirds of the 60% were for direct discharges, one sixth for arranger liability and one sixth for margin of error. But even before Burlington Northern, the NCR/API share had been brought down to 47.5% in the models used for, amongst other things, the provision in the Sequana consolidated accounts for the year ending 31 December 2008. That pre-Burlington Northern figure of 47.5% therefore included arranger liability so the reduction from 60% must be the result of factors that decreased the direct liability to below the 40% that had been posited in the Tech Memo and Amendola reports. He thought that applying the same ²/₃ ¹/₆ ¹/₆ approach, the figure of 47.5% must have been based on direct discharges of about 31.5%, plus arranger liability of 8% and a margin of error of 8%. Therefore the figure of 38% ultimately used, having stripped out arranger liability, had a built in margin of about 6.5% if liability for direct discharges was in fact only 31.5%. His view was therefore that the 38% they used resulted in a conservative or prudent figure because it effectively only stripped out arranger liability.

By this time Mr Martinet was aware of course that there had been the favourable reconsideration by Judge Zuidmulder of his ruling in the Green Bay Litigation, the confirmation of the ‘all sums’ approach in the Plastics Engineering decision and the Final Aon Report. Mr Martinet was justified in feeling more confident in assuming that there would be recovery of insurance proceeds under the Historic Insurance Policies. He considered that the likely recovery figure was about $100 million. There was therefore still a substantial buffer taking into account the Maris Policy funds and the Historic Insurance Policies in case the best estimate turned out to be wrong.

The Claimants cross-examined Mr Martinet on the basis that the changes made to the figures after the December Dividend were all part of a plan between him and Mr Gower to push the figures down by whatever means possible to remove the provision. On the seventh day of his cross-examination, when it was suggested to him that he had no basis for accepting one particular figure amongst many Mr Martinet said this:

“At the end of the day we were proposed a number of numbers which we thought were solid or robust and that is what was used. And that is what I remember. I never --You see, my Lady, for the past few days I have been -- the alternative is either an I am an idiot and unprofessional, or I am a manipulator, with tortuous ideas. Well, you know, maybe I'm an idiot, but I'm not a manipulator, I'm not trying to doctor anything there. On the contrary, it is ironic, because all of my life I have been considered as a straight arrow, so I understand what you are saying, Mr Smouha, but again, I was trying to do a professional job, trying to determine whether we should change the numbers or not, and the conclusion was that the numbers we had used, there was no compelling reason to change them. Rightly or wrongly, I don't know.”

I accept that evidence and having seen Mr Martinet in the witness box I reject the suggestion that he approached the issue in a biased or improper manner. Of course it was clear to everyone that Sequana would prefer not to have to include a provision in AWA’s accounts on the grounds that its liability would not exceed the funds in the Maris Policy. But there is a big jump from that position to a position in which two directors and their adviser set about to manipulate the figures to arrive at that goal, knowing that there is no real basis for changing the numbers. I am sure that Mr Martinet did not make that jump and that nothing he said to Mr Gower suggested to Mr Gower that he was being asked or was expected to make that jump either. I have not had the advantage of seeing Mr Gower in the witness box. But whatever his inclinations may have been, he knew very well that Mr Bartolotta and other professional people such as PwC and the US lawyers would be scrutinising the figures, asking difficult questions and making sure that there was an adequate explanation for any change made.

As Mr Gower has not given evidence we do not know how he and those he consulted arrived at the conclusion that all the factors listed in the Note to the spreadsheet should lead to a reduction of 22% rather than, say, of 15% or 30%. But the need to make a provision requires a hard number to be given to soft factors. I have not seen anything in all the evidence before me which suggests either that that reduction was unreasonable or that any of the many other people involved in this exercise expressed a view that the figure should be something different. Mr Martinet did accept that he knew that there had been substantial extra work done on sediment sampling since the reports referred to by Mr Gower in that email. But it was not suggested to him by any of the people working on the issue and advising him that this work justified a movement away from the 40% direct discharge figure that had been the basis for the provision as from 2005.

I find therefore that there was plenty of evidence which justified Mr Martinet’s and Mr Courteault’s conclusion that 38% was the best estimate as at 18 May 2009 of NCR/API’s likely ultimate share of the costs of the Fox River remediation. There was therefore nothing wrong with the 2008 Final Accounts on that score.

The NRD value of $18.8 million for NCR/API

I have already explained why I consider that it was reasonable for the NRD figure included in the calculation of the provision to be based on a likely settlement figure rather than to assume that the Government would pursue and ultimately win a claim for a much higher figure. I have also set out how Mr Gower explained to PwC on 12 May 2009 the calculation of the $18.8 million used for the purposes of the 2008 Final Accounts: see paragraph 288, above. He seems to have gone back to the Stone Total NRD Offer of $76 million, taken into account the amounts already paid and applied the 38% NCR/API share to the total. I accept that there is no automatic read across from the appropriate share of remediation costs to the share of NRDs because it was expected that Georgia Pacific would contribute to remediation costs but it was immune from any further contribution for NRDs because of its settlement with the Government. However Mr Tenpas’ evidence (and it appears Mr Bates’ advice at the time) was that NRDs would be apportioned on the basis of direct discharges. This is the 15% to 20% figure that the court referred to in the ruling approving the settlement of the Georgia Pacific NRD liability. On that basis $18.8 million as the best estimate for NCR/API’s share of NRDs has plenty of support in the evidence.

I therefore find also that the $18.8 million figure used for the NCR/API share of NRDs (leading to an $11.3 million share for AWA alone) was reasonable as a best estimate for NRD liability.

The disclosures made in AWA’s accounts regarding contingent liabilities

The third challenge to the accounts (the first being to the capital reduction and the second being to the value of the provision) is to the lack of disclosure in the accounts about contingent liabilities.

The obligations about disclosure are also found in Accounting Standard FRS 12. Paragraph 90 provides that:

“90.

An entity should disclose the following for each class of provision:

(a)

a brief description of the nature of the obligation, and the expected timing of any resulting transfers of economic benefits;

(b)

an indication of the uncertainties about the amount or timing of those transfers of economic benefits. Where necessary to provide adequate information, an entity should disclose the major assumptions made concerning future events, as addressed in paragraph 51; and

(c)

the amount of any expected reimbursement, stating the amount of any asset that has been recognised for that expected reimbursement.”

Where the liability is a contingent liability either because it is only a possible and not a present obligation or because, although it is a present obligation, it is not probable that it will give rise to a transfer of economic benefit or it is not possible to arrive at a reliable estimate of the amount needed to settle it, then no provision should be made. There must however be disclosure of the contingent liability: (emphasis added)

“91.

Unless the possibility of any transfer in settlement is remote, an entity should disclose for each class of contingent liability at the balance sheet date a brief description of the nature of the contingent liability and, where practicable:

(a)

an estimate of its financial effect, …;

(b)

an indication of the uncertainties relating to the amount or timing of any outflow; and

(c)

the possibility of any reimbursement.”

There is an exception to the need to make disclosure of a contingent liability that may be relevant here:

“97.

In extremely rare cases, disclosure of some or all of the information required by paragraphs 89-94 can be expected to prejudice seriously the position of the entity in a dispute with other parties on the subject matter of the provision, contingent liability or contingent asset. In such cases an entity need not disclose the information, unless its disclosure is required by law; but should disclose the general nature of the dispute, together with the fact that, and reason why, the information has not been disclosed.”

Disclosure actually made in the accounts

In the December 2008 Interim Accounts there was no disclosure of the nature of the Lower Fox River liability or any reference to the element of the liability that was contingent and not included in the provision because it was expected that the other jointly and severally liable PRPs would meet it.

In the 2008 Final Accounts there was of course no provision made for the Fox River liability. In note 15 in the accounts, headed ‘Contingent liabilities’ there was a brief description of the source of AWA’s indemnity liability, the two RODs issued by the EPA, the UAO issued in November 2007, the Maris Policy and the Historic Insurance Policies, and the Green Bay litigation. It recorded the conclusion of the Final AON Report that the probability of obtaining significant recoveries from the insurers was very high. Note 15 then continued:

“15.3: the LFR reserve:

For the purposes of creating its reserve, the Company using third party experts (notably PCC) has taken a reasonable, prudent view as to API's likely share of the cost of remediating the remainder of the river as proposed in the ROD2 amendment and as further modified by information received from the contractors during the bid process and subsequent review.

A significant unknown remains the precise percentage of the total costs of the remediation and natural resource damages which will be borne by NCR Corporation and API, which two companies the Agencies treat as a single share. The final cost to the Company is unlikely to be determined until all allocation issues are finally determined by a court or through some form of dispute resolution.

Nevertheless, as between NCR and API, costs are shared pursuant to a 1998 settlement between these two companies and a November 2005 arbitration which found API liable for 60% of all costs in excess of $75 million. For the purposes of creating its reserve, the Company has taken a prudent view as to API's likely overall percentage allocation, without taking into account the existence of highly likely contingent asset (see 15.2 above [sc. the Historic Insurance Policies]).

On 4 May 2009, the US Supreme Court delivered its decision in the Burlington Northern case. In that case the court considered three issues relevant to the Fox River matter (i) arranger liability (ii) joint and several liability, and (iii) apportionment. The impact of the decision in that case on the Company's liability for the Fox River may mean that API and NCR have no arranger liability; that their liability may be apportioned and, if so, no joint and several liability exists. For these reasons management has re-assessed the Company's exposure and determined that a reduction in the share for which API may be found liable is more likely than not.

15.4: other contingent liabilities:

There are no other contingent liabilities arising in the ordinary course of business in respect of litigation and guarantees, which the directors believe will have a significant effect on the financial position of the Company and its subsidiary undertakings.”

The May 2009 Interim Accounts were drawn up for the period 1 January 2009 to 15 May 2009 and also did not include any provision for the Fox River liability. They contained a shortened version of the background paragraphs of the 2008 Final Accounts including the reference to Burlington Northern and continued:

“In light of the Burlington Northern decision and based on the Company's own scientific and technical investigation regarding the processes utilised at the facility now owned by API, the directors have determined that the Company has sufficient financial assets to meet its indemnity obligations to API on the basis that those obligations are limited. It is the view of the directors that it is reasonably likely that the Company, on behalf of API, has already paid more than the amount that API will ultimately be held legally obligated to pay. Accordingly, whilst the Company intends to continue to fund the ongoing cleanup work for the time being, resolution of API's share will be sought through appropriate avenues including the courts.”

There were no disclosures about Future Sites in general or the Kalamazoo River in particular in any of the accounts.

The relevance of defective disclosure in the accounts

The accounting experts agreed that the notes complied with some aspects of FRS 12 but they disagreed on a number of issues, including about whether it is possible that AWA relied on the exception in paragraph 97 of FRS 12 despite the absence of any reference to this in the contemporaneous documents. In my judgment, however, any defects in the disclosure made in the accounts are not relevant to the question whether the accounts gave a true and fair view or enabled a reasonable judgment to be made for the purposes of Part 23 of the CA 2006. The wording of section 836 focuses on the numbers used in the accounts as opposed to the other narrative parts. Thus, section 836(1) provides that the lawfulness of a distribution is determined ‘by reference to the following items as stated in the relevant accounts’. For our purposes, these items are those in section 836(1)(a), that is profits, losses, assets, liabilities, those in paragraph 88 of Schedule 4 to the Companies Act 1985 (namely provisions relating to depreciation or diminution in the value of assets) and those in paragraph 89 of Schedule 4 (namely references to any amount retained as reasonably necessary for the purposes of providing for any liability the nature of which requires provision for accounting purposes). This focus on amounts rather than narrative is apparent too from section 837(2) which qualifies the requirement that the accounts must be properly prepared by stipulating that this is ‘subject only to matters that are not material for determining (by reference to the items mentioned in section 836(1)) whether the distribution would contravene this Part’. This requirement that any defects in the accounts must relate to material items is further reinforced by section 837(4) which envisages that even a report qualified by an auditor can support a lawful distribution if the auditor states in writing that the matters in respect of which his report is qualified are not material for determining whether a distribution would contravene Part 23.

I also agree with the Defendants’ submission that the fact that directors can rely on interim accounts to determine whether a distribution is lawful or not is an indication that disclosure notes are not relevant to this exercise. In my judgment there are no accounting standards, practices or guidelines that impose disclosure obligations on those drawing up interim accounts. The absence of disclosure in the December 2008 Interim Accounts which supported the December Dividend did not result in those accounts being defective in any way and so cannot have prevented those accounts from enabling a reasonable judgment to be made of the company’s affairs for the purpose of Part 23.

Further, I conclude that even if there were defects in the 2008 Final Accounts which supported the May Dividend they are not defects which result in the distribution contravening Part 23. However in case I am wrong on the legal point I have decided, I have already set out above the disclosure that was made. I now set out my conclusions on the main factual dispute arising from this issue, by way of a short diversion to the banks of the Kalamazoo River.

Disclosure: the Kalamazoo River

The Kalamazoo River is a CERCLA site that stretches for approximately 80 miles in the state of Michigan. From the 1950s to 1970s, it was contaminated with PCBs due to the activities of paper mills and other industrial operations. Although there were no mills actually making carbonless paper on the Kalamazoo, the paper mills there bought broke and trimmings and washed the emulsion into the Kalamazoo when they were recycling and de-inking it to recover the pulp fibres. After investigations by federal and state agencies, the Kalamazoo River was designated as a Superfund site in 1990 (that is before the Lower Fox River). The PRPs designated included Georgia Pacific but not NCR or API. By 1995 there were already contribution proceedings on foot among the PRPs and the first ROD was issued in 1998, four years before the first ROD for the Lower Fox River.

In 2000 a Michigan District Court linked PCB contamination in the Kalamazoo River to PCBs from NCR carbonless copy paper in its decision in Kalamazoo River Study Group v Rockwell International. But it was not clear how the PCBs would have got there. Mr Kirsch points out that the Revised Amendola Report in April 2001 stated Mr Amendola’s belief that roughly 30% of the broke sold from the NCR mills was probably recycled in the Kalamazoo River watershed. This is in part what caused him to lower his estimate of the direct discharges from those mills into the Fox River.

In 2003 NCR/API received a section 104(e) request for information from the EPA concerning the Kalamazoo River. The request did not indicate that NCR or API were being considered as PRPs but asked for information about historic sales of broke and production of NCR paper. NCR and API’s response stated that they had not identified any documents showing sales of NCR paper broke from their predecessor facilities to mills in the Kalamazoo River area. NCR and API also stated that there were no such sales, since its records indicated that all broke was sold to and used by mills on the Lower Fox River.

Nothing further was heard about the Kalamazoo River by NCR or API until after the events with which these claims are concerned. Thus, by the time of the dividends in December 2008 and May 2009, the US government had not issued letters to NCR/API asserting that they were liable for contamination at the Kalamazoo River. The Kalamazoo River parties had not sued NCR/API for contribution, nor otherwise threatened or given informal notice of such litigation. Moreover, Georgia Pacific had pursued contribution litigation against other parties without naming NCR or API in its suits.

Mr. Kirsch's view is that AWA should have been on notice that NCR/API’s potential liability was not remote because (i) NCR Paper appeared to be the primary source of PCBs in the Kalamazoo River; (ii) NCR/API had received section 104(e) requests from the EPA concerning the Kalamazoo River; and (iii) after receiving its section 104(e) request, API notified AWA of its indemnity obligation with respect to any claim at the Kalamazoo River. However, he accepted that there was no good reason why if the PRPs arranging the clean up of the Kalamazoo River had thought that they had a credible arranger liability claim against NCR, they did not bring proceedings before May 2009.

Mr. Tenpas' view is that NCR/API should not have anticipated any liability for the Kalamazoo River. In the absence of NCR/API paper-coating operations near the site, there was no reason to expect that NCR/API would have had liability as direct dischargers. Further, there was no reason to believe NCR/API would be held liable as arrangers of hazardous substances because there was no evidence that NCR or API sold broke to the recyclers on the Kalamazoo River. Even if they had, broke was likely to be regarded as a "useful product" whose subsequent disposal would not give rise to arranger liability. He did not regard the receipt of the section 104(e) letter as significant given that subsequently other parties were named and pursued by the Government as PRPs but not NCR or API.

Mr Martinet’s evidence was emphatic that he had never been told that there was a possibility of AWA having a liability for the Kalamazoo River. He remembered the name being mentioned but he said ‘it was never, never presented to me as even a remote prospect’. He did recall looking at the question of liability for Future Sites more generally but Mr Gower told him that ‘it was a non-issue’ and that in all the seven years he (Mr Gower) had been working on Fox River issues, he had never been confronted with a Future Sites issue. Mr Martinet also said that he did not realise that API did not have records of the plants to which they had sold broke or of the fact that broke from the Fox River paper coating plants might have been sent to the Kalamazoo River. Nor was he aware of the fact that API and NCR had been sent information requests by the Government in respect of the Kalamazoo River in 2003. He did recall that there had been attempts to obtain insurance for AWA’s liability for Future Sites in general (rather than the Kalamazoo River in particular). A possible carrier was identified but this had not been pursued because Mr Martinet was advised that nothing had happened in relation to Future Sites over the past seven years and it was therefore ‘a very remote point’. The cost of the proposed insurance was high and he concluded that it was not worth paying.

Mr Courteault, Mr Mountford and Mr Newell all gave evidence that they had never been aware of any potential liability for the Kalamazoo River and this was not challenged when they were cross-examined.

The key question is therefore whether the possibility of AWA having to pay out money for the Kalamazoo River clean up was ‘remote’. If it was remote then there was no need under paragraph 91 of FRS 12 to disclose it as a contingent liability. If it was not ‘remote’ then disclosure should have been made in the 2008 Final Accounts.

In my judgment there was no reason for the directors to conclude that the possibility of AWA being liable for any part of the costs of cleaning up the Kalamazoo River was more than remote. There is no evidence that it was raised with Mr Martinet at any of the initial briefing meetings to discuss the problems underlying the AWA provision. If the US lawyers thought there was a risk of this liability arising they would have drawn it to Mr Martinet’s attention. I accept Mr Martinet’s evidence that Mr Gower told him that there was nothing to worry about regarding potential liability for Future Sites. The directors and their advisers were entitled, in my judgment, to infer from the fact that no action was taken for several years against NCR or API after they responded to the section 104(e) letters, even though there was considerable clean up activity and accompanying litigation, that they would not be pursued for these costs.

I would therefore find, if it were relevant, that there was no breach of FRS 12 paragraph 91 arising from the absence of any disclosure relating to the Kalamazoo River.

Conclusion on the “could not” claims

I have therefore concluded that the December and May Dividends did not contravene Part 23 of the CA 2006:

Having reached that conclusion I do not need to resolve the many interesting legal questions which would arise if either or both of the dividends did contravene Part 23. The Claimants sought relief for such contravention via four routes; the statutory remedy under section 847, a remedy in equity against Sequana for unconscionable receipt, a claim in restitution for unjust enrichment and the avoidance of the release of the inter-company receivable on the grounds of mutual mistake as to the lawfulness of the distribution. In relation to each of these routes, the parties made rival submissions as to the state of knowledge on the part of Sequana that the Claimants had to show, whether the amount recoverable would be the whole dividend or only that part that could not have been lawfully paid and the application of any change of position defence. Although I heard argument on all these points, I have decided not to lengthen this judgment by considering them, particularly since in respect of some of the points, the Defendants reserved the right to argue the point differently if the case went further, so that there may be little to be gained from my analysis of case law which binds me but does not bind a higher court.

BTI argue that even if the provision was based on a best estimate of the liability, the AWA directors were still in breach of their fiduciary duties in declaring a dividend of the distributable reserves shown in the accounts. The directors were aware that the estimate was surrounded by great uncertainty and there was a risk that the liability would ultimately be much greater than that estimate. In those circumstances, the Claimants submit, it was a breach of the directors’ duty to AWA to pay away the December and May Dividends to Sequana. Even though they could declare the dividend in the light of the available distributable reserves they should not have done so.

The Claimants allege that the directors have acted in breach of a number of the duties that are imposed on them by sections 171 – 174 of the CA 2006. The main provision relied on is section 172 and in particular the proviso in section 172(3). BTI also referred to the duty in section 174 (to exercise reasonable care, skill and diligence), that in section 171 (to exercise powers for the purpose for which they were conferred) and that in section 173 (duty to exercise independent judgment). But it was not suggested that the result of the case could be any different depending on which duty was breached. I will therefore focus on section 172.

Section 172 provides:

“172 Duty to promote the success of the company

(1)

A director of a company must act in the way he considers, in good faith, would be most likely to promote the success of the company for the benefit of its members as a whole, and in doing so have regard (amongst other matters) to--

(a)

the likely consequences of any decision in the long term,

(b)

the interests of the company's employees,

(c)

the need to foster the company's business relationships with suppliers, customers and others,

(d)

the impact of the company's operations on the community and the environment,

(e)

the desirability of the company maintaining a reputation for high standards of business conduct, and

(f)

the need to act fairly as between members of the company.

(2)

…

(3)

The duty imposed by this section has effect subject to any enactment or rule of law requiring directors, in certain circumstances, to consider or act in the interests of creditors of the company.”

BTI accept that, given that Sequana is the sole shareholder, a breach of fiduciary duty can only arise – or at least is only incapable of ratification by Sequana – if at the time of declaring the dividends, the directors were bound to consider the interests of the creditors of the company and not, or not only, the interests of Sequana. The principal legal issue for the ‘should not’ case is therefore whether the duty to act in what the directors believe to be in the best interest of creditors had arisen here at the time of the December Dividend or the May Dividend. I shall refer to this duty as ‘the creditors' interests duty’.

Some preliminary points

There was some common ground between the parties as to the following principles. First, the effect of subsection (3) of section 172 is to retain the common law principles as to when the creditors' interests duty arises; it does not attempt to codify those principles. Secondly, once the creditors' interests duty has arisen, the duty is still the duty owed to the company. There is no cause of action conferred on the creditors by a breach of the duty. The claim against the directors under this head is brought by BTI as the assignee of AWA’s rights against the directors, not by BAT as creditor.

Thirdly, it is accepted by the Claimants that there is a single threshold for when the creditors' interests duty arises for all decisions taken by the directors. BTI were not submitting that the duty could arise in respect of a decision whether or not to pay a dividend but not in relation to the directors’ other decisions about the running of the company. However, BTI do say that the fact that the decision in dispute here is a decision to pay a dividend, rather than any other kind of decision about the future of the business, is significant because it is a decision that benefits only the interests of the shareholder and not at all the interests of the creditors.

Fourthly, it is agreed that the content of the duty - whether it is to give paramount consideration to the interests of creditors or only to take their interests into account in some lesser way - does not vary according to the degree of risk of insolvency that has arisen.

Fifthly, if the court decides that the creditors’ interests duty did arise but the directors did not in fact take the interests of the creditors into account in their decision making, that is not of itself a breach of fiduciary duty, invalidating everything done automatically: see the judgment of Pennycuick J in Charterbridge Corporation Ltd v Lloyds Bank Ltd and another [1970] 1 Ch 62 as applied in Colin Gwyer & Associates Ltd & Anor v London Wharf (Limehouse) Ltd & Ors [2002] EWHC 2748 (Ch).

When does the creditors' interests duty arise?

The Defendants accepted that the creditors' interests duty can arise in circumstances where the company is not actually insolvent. Something short of actual insolvency is sufficient. The question is, how close to insolvency does the company have to be? The Defendants argue that it has to be very close to insolvency, the Claimants contend that it is enough if there is a real, as opposed to a remote, risk of insolvency.

The list of authorities in which this test has been expressed and applied is a long one, yet the final position seems not to be at all clear. What does seem to me important is to look at the facts of the cases as well as the formulation of the test in the judgments. Statements of principle may be expressed more broadly than is warranted by the facts of the particular case. But where a formula extending the creditors' interests duty to a situation short of insolvency is cited by the court in a case where the company was in fact found to be insolvent or to be very close indeed to insolvency, or conversely where the court was satisfied that there was no problem with the company’s solvency, it is unlikely that the court had turned its mind to the precise point at which a solvent company crosses some threshold which causes the creditors' interests duty to arise.

The earliest case to which I was referred was In Re Horsley & Weight Ltd [1982] 1 Ch 442. There a liquidator sought to set aside a pension policy acquired by the company for the benefit of a retiring director. The question was whether the decision to acquire the pension had been ratified by the shareholders. The Court of Appeal upheld the finding that there had been no misfeasance by the directors. Templeman LJ said that if the company ‘had been doubtfully solvent at the date of the grant to the knowledge of the directors’ then the directors would have been liable, but, he said the good faith of the directors was not impugned. Cumming-Bruce LJ referred to a suspicion that at the time of the decision the company was not in a position to pay the money to the respondent. But that evidence ‘fell far short of proof that the directors should at the time have appreciated that the payment was likely to cause loss to the creditors’. Buckley LJ also stressed that the good faith of the directors was not questioned and there was no suggestion that they had failed to apply their minds honestly to the question whether the decision was a fair and proper thing for the company to do in the light of the company’s financial state as known to them at the time.

Horsley & Weight was considered by the Court of Appeal in New Zealand in Nicholson and others v Permakraft (N.Z.) Ltd. (in liq) [1985] 1 NZLR 242; (1985) 3 ACLC 453 (‘Permakraft’). There the court considered a claim by the liquidators of Permakraft against the former directors to recover a dividend paid to themselves as shareholders from distributable profits as part of a plan to restructure the company. The honesty of the directors was not in issue. White J found at first instance that the directors were in breach of duty and should refund the whole amount. The argument was advanced that Permakraft was in a state of near insolvency at the relevant time and that the capital profit should have been retained in the interest of creditors. The appeal was allowed. Each judge on appeal expressed his conclusions differently. Cooke LJ said:

“(iii)

The duties of directors are owed to the company. On the facts of particular cases this may require the directors to consider inter alia the interests of creditors. For instance, creditors are entitled to consideration, in my opinion, if the company is insolvent, or near insolvent, or of doubtful solvency, or if a contemplated payment or other course of action would jeopardise its solvency.

The criterion should not be simply whether the step will leave a state of ultimate solvency according to the balance sheet, in that total assets will exceed total liabilities. Nor should it be decisive that on the balance sheet the subscribed capital will remain intact, so that a capital dividend can be paid without returning capital to shareholders. Balance sheet solvency and the ability to pay a capital dividend are certainly important factors tending to justify proposed action. But as a matter of business ethics it is appropriate for directors to consider also whether what they do will prejudice their company's practical ability to discharge promptly debts owed to current and likely continuing trade creditors.

…

… If the company's financial position is precarious the fortunes of such suppliers may be so linked with those of the company as to bring them within the reasonable scope of the directors' duties. They may continue to give credit in ignorance of a change damaging to their prospects of payment.”

Richardson LJ stated (page 463) that:

“If a company is solvent in the sense of its assets exceeding its liabilities there can, I think, be no question of a separate duty to creditors: they have their ordinary remedies if their accounts are not paid. If it is insolvent the creditors have an interest in the company and the directors might be said to have a duty to them for creditors' money is then at stake. It is in the intermediate situation of near insolvency or doubtful insolvency that greater difficulties of legal principle arise.”

Richardson LJ held that the first instance judge had made no findings as to solvency and on his assessment of the evidence, ‘that starting premise of near insolvency is not justified’. Somers J also stated that ‘it has been suggested that when the solvency of a company is doubtful or marginal it will be a misfeasance (probably not capable of being ratified or exonerated by shareholders) to enter into a transaction which directors ought to know is likely to cause a loss to creditors’. However, he concluded that he did not have to decide the issue since he was satisfied the company was solvent at the material times (page 464).

Permakraft was applied by the Court of Appeal in Australia in Kinsela & another v Russell Kinsela Pty Ltd (in liq) (1986) 10 ACLR 395 (‘Kinsela’). There the first instance court had set aside a lease granted by the company to its directors as lessees. The Court described the company’s financial position at the time the lease was granted as ‘to say the least, precarious’ and it was wound up as insolvent three months later. The judge had found that the company had sustained regular and perhaps increasing trading losses and its accounts showed a substantial excess of liabilities over assets. Street CJ, giving the judgment of the court, described the company as an ‘insolvent company in a state of imminent and foreseen collapse’ (page 399). He held that where a company is insolvent, ‘the interests of the creditors intrude’:

“They become prospectively entitled, through the mechanism of liquidation, to displace the power of the shareholders and directors to deal with the company’s assets. It is in a practical sense their assets and not the shareholders’ assets that, through the medium of the company, are under the management of the directors pending either liquidation, return to solvency, or the imposition of some alternative administration.”

Street CJ said it was not necessary in Kinsela to consider the degree of financial instability which would impose the creditors' interests duty on directors since in that case the company was plainly insolvent and about to collapse. The judgment of Street CJ in Kinsela was cited with approval by the English Court of Appeal in Liquidator of West Mercia Safetywear Ltd v Dodd & anor (1988) 4 BCC 30 (‘West Mercia’). In West Mercia there was no doubt that both the company which made the challenged payment and the parent company which received the money were insolvent at the time the payment was made. Similarly in Facia Footwear Ltd (in administration) and another v Hinchliffe and another [1998] 1 BCLC 218 where these cases were cited, there was uncontested evidence that the company was ‘hopelessly insolvent’ at the time that the payments were made (page 225f-g).

The Claimants also referred me to the Australian case of Grove v Flavel (1986) 11 ACLR 161. This is the first case in which the formulation referring to a real as opposed to remote risk was used. The issue arose in a different context, an appeal from a criminal conviction of a former director for the offence of making improper use of information acquired from his position as an officer of the company for the purpose of gaining an indirect advantage. He had written out cheques on the company’s bank account to himself and other creditors within the group whose debts were thereby discharged. The information he was alleged to have obtained was that the company was experiencing liquidity problems and had been refused a bank loan. The court noted that there was no doubt that at the time of the cheques, the company was in serious financial difficulty and that liquidation ‘became a real possibility’ if the creditors sought to enforce payment. The Court quoted from Permakraft and Kinsela. In construing the word ‘improper’ in the statutory offence in issue, Jacobs J (with whom the other judges agreed) held: (page 170)

“1.

A director of a company X Ltd who, upon acquiring information which leads him to believe that the company faces a risk of liquidation, whether voluntary and because it cannot pay its debts as they fall due or at the suit of creditors, which is a real and not a remote risk, thereupon acts to protect himself and other companies of which he is a director from the consequences of such liquidation, to the possible detriment of the creditors of X Ltd, is acting "improperly" as a director of X Ltd because:

(a)

There can be no doubt of such possible detriment when the action taken involves a disposition of the assets of X Ltd, in this case debts owing to X Ltd, which would be part of the fund available to creditors generally in the event of liquidation. It is in the words of Richardson J in Nicholson & Ors v Permakraft (NZ) Ltd (in liq), supra, "the creditors' money that is at stake".

(b)

If that is the principle which dictates the "duty" of a director to have regard to the interest of creditors when the company is known to be insolvent there can be no reason in principle why knowledge of a real risk of insolvency should not attract the same duty.

2.

Whether there is such a real and perceived risk of insolvency must depend upon the facts of the particular case.”

Mr Thompson also referred me to Hilton International Ltd v Hilton [1989] 1 NZLR 442, a case in the New Zealand High Court where Tipping J stated that even if the company is solvent in both balance sheet and cash flow terms, a dividend cannot be paid if the payment ‘is likely to jeopardise either form of solvency’. He also held that the directors are at fault if they pay when they ought to appreciate that risk; and if the company's solvency is doubtful the directors must be able to show that they acted in good faith and on reasonable grounds. On the facts of the case, the company had been cash flow insolvent at the date of the dividend. The judge found that the company was either in a state of balance sheet insolvency ‘or very close to it’ as a result of and following the declaration and crediting of the capital dividend: page 462. He found that if the directors had commissioned accounts prior to the declaration of the dividend it would have emerged that the passing of the dividend would have taken the company into a state of balance sheet insolvency. Even if he was wrong in taking that view, the dividend ‘would have seriously jeopardised the company’s balance sheet solvency’ – the company was put ‘on a knife edge’ by the payment of the dividend.

Other cases in which the court has referred to a test of ‘doubtful solvency’ or of the company being ‘on the verge of insolvency’ have also, on their facts, concerned companies that were in fact insolvent or very close to collapse:

These authorities were reviewed by John Randall QC (sitting as a deputy High Court Judge) in Re HLC Environmental Projects Ltd (in liq.) [2013] EWHC 2876 (Ch) (‘HLC Environmental’). The company in that case was a building development company. The judge set out the figures included in the company’s accounts, showing that throughout the relevant period the company had substantial net current liabilities, overall liabilities and trade creditors and had made substantial trading losses annually. There was a large shareholder deficit throughout the period: paragraph 77. This provided strong prima facie evidence of insolvency, even though the company had failed to include any provision in the accounts for liabilities that should have been provided for. He found as a fact that the company was both balance sheet and cash flow insolvent: para 86. The judge then set out the various references in the case law to a company being ‘doubtfully solvent’ or to the ‘parlous financial state’ of the company or the company being ‘on the verge of insolvency’. He also cited the passage from Giles JA’s judgment in Kalls included in Newey J’s judgment in Vivendi and continued:

“For my part, I do not detect any difference in principle behind these varying verbal formulations. It is clear that established, definite insolvency before the transaction or dealing in question is not a pre-requisite for a duty to consider the interests of creditors to arise. The underlying principle is that directors are not free to take action which puts at real (as opposed to remote) risk the creditors' prospects of being paid, without first having considered their interests rather than those of the company and its shareholders. If, on the other hand, a company is going to be able to pay its creditors in any event, ex hypothesi there need be no such constraint on the directors. Exactly when the risk to creditors' interests becomes real for these purposes will ultimately have to be judged on a case-by-case basis. Different verbal formulations may fit more comfortably with different factual circumstances.”

The Claimants relied on this passage to submit that it is sufficient for them to show that AWA was at a real risk – as opposed to a remote risk – of insolvency for the creditors' interests duty to have arisen. They say that the risk here was ‘real’ as opposed to remote. They emphasise the unusual and extreme circumstances of (i) AWA’s position as a non-trading company; (ii) the very great uncertainties, outside AWA’s control, surrounding the scale of the indemnity liability; (iii) the wide range of possible outcomes for each of the moving parts making up the best estimate, such that the final liability might be several multiples of the best estimate and (iv) the fact that the decision to pay the interim dividends benefited only the shareholder and brought no possible benefit to the creditors of the company. The Claimants produced models which showed how flexing even one of the moving parts involved in the computation of the likely liability resulted in a very large deficit in AWA’s balance sheet.

To say that my house is on the verge of burning down seems to me to describe a much more worrying situation compared to one in which there is a risk which is something more than a remote risk of my house burning down. Similarly, giving the words their natural meaning, a test set at the level of ‘a real (as opposed) to remote risk of insolvency’ would appear to set a much lower threshold than a test set at the level of being ‘on the verge of insolvency’ or of ‘doubtful’ or ‘marginal’ solvency. But I agree with the conclusion of Mr Randall QC in HLC Environmental that the authorities appear to treat these and all the other formulations as different expressions of the same test. Having reviewed the authorities I do not accept that they establish that whenever a company is ‘at risk’ of becoming insolvent at some indefinite point in the future, then the creditors' interests duty arises unless that risk can be described as ‘remote’. That is not what the cases say and there is no case where, on the facts, the company could not also be accurately described in much more pessimistic terms, as actually insolvent or ‘on the verge of insolvency’, ‘precarious’, ‘in a parlous financial state’ etc.

The essence of the test is that the directors ought in their conduct of the company’s business to be anticipating the insolvency of the company because when that occurs, the creditors have a greater claim to the assets of the company than the shareholders. This case is very different from the other cases in which the triggering of the creditors' interests duty has been considered. AWA’s balance sheet showed no deficit of liabilities over assets and there were no unpaid creditors knocking at AWA’s door. It was not in the downward spiral of accumulating trading losses, with no income and no prospect of any income that is typical of the companies where the duty has been held to have arisen. I agree with the statement of Norris J in Frohlich that the underlying principle is that:

“The acts which a competent director might justifiably undertake in relation to a solvent company may be wholly inappropriate in relation to a company of doubtful solvency where a long term view is unrealistic”. (emphasis added)

In the instant case, there was a real possibility that AWA would never become insolvent or even close to insolvent. The best estimate of the Fox River liability might turn out to be accurate in which case the company’s assets would be sufficient to meet the liability even without the need to rely on proceeds from the Historic Insurance Policies. It cannot be right that whenever a company has on its balance sheet a provision in respect of a long term liability which might turn out to be larger than the provision made, the creditors' interests duty applies for the whole period during which there is a risk that there will be insufficient assets to meet that liability. That would result in directors having to take account of creditors’ rather than shareholders’ interests when running a business over an extended period. This would be a significant inroad into the normal application of directors’ duties. To hold that the creditors' interests duty arises in a situation where the directors make proper provision for a liability in the company’s accounts but where there is a real risk that that provision will turn out to be inadequate would be a significant lowering of the threshold as currently described and applied in the cases to which I have referred. I can see no justification in principle for such a change.

The Claimants submitted that there were a number of ‘red flags’ waved at Mr Martinet and Mr Courteault before the 18 May 2009 meetings which should have alerted them to the risks of AWA tipping over into insolvency. One of these was the emphasis of matter that PwC included in the 2008 Final Accounts. Mr Smouha submitted that the inclusion of this showed that PwC were uncomfortable with the removal of the provision for the Fox River liability in those accounts. I do not accept that anything can be inferred from the emphasis of matter about PwC’s views, other than they had concluded that the conditions set out in the accounting standard for when an emphasis of matter is appropriate had been satisfied. Mr Martinet’s understanding of why this had been included was slightly different from the accounting standard’s guidance. He thought it was more to do with drawing the readers’ attention to a significant change in assumptions from those underlying the previous year’s accounts whereas the standard suggests that an emphasis is important where there is a significant degree of uncertainty about a provision. Whichever is the case, it would not be fair to read into this some subtle signal from PwC that they disapproved of what the company had done. If they had been unhappy they could have qualified their audit report, as they threatened to do if the directors insisted on treating expected insurance receipts as an asset on the balance sheet.

Another factor that the Claimants referred to as a red flag as to the risk of insolvency was a letter dated 22 April 2009 from Mr Mark Richards, the Chief Executive Officer of API to Mr Martinet raising concerns about the sale. That letter showed that Mr Richards had slightly misunderstood the nature of the transaction proposed, because it appears that he thought that AWA was going to assign its rights and obligations under the indemnity agreement to another entity. In his response on 11 May, Mr Martinet explained the nature of the proposed deal and pointed out that the indemnity agreement did not include any provision forbidding the sale of AWA to a third party.

Although I can understand Mr Richards’ concern, I do not see that this should have alerted AWA’s directors to any problem with the company’s finances that they did not already appreciate.

Taking all these factors into account, I do not think that AWA could be described as on the verge of insolvency or of doubtful insolvency, or as being in a precarious or parlous financial state. The risk it faced that the best estimate would turn out to be wrong and that the company might not have enough money, when called upon in the future, is a risk that faces many companies that have provisions and contingent liabilities reflected in their accounts. It is not enough in my judgment to create a situation where the directors are required to run the company in the interests of the creditors rather than the shareholders of the company.

I therefore hold that the creditors' interests duty had not arisen at the time of the directors’ decision to pay the December Dividend or the May Dividend. There can therefore have been no actionable breach of fiduciary duty by them in making the payments. As with my findings in relation to Part 23 of the CA 2006, and for the same reasons I will not seek to resolve the many additional legal issues that would arise as to the appropriate remedy against Sequana or the directors if I had come to the opposite conclusion.

So far the claims discussed have been claims brought by BTI as assignee of AWA’s rights as against the directors for breach of fiduciary duty and as against Sequana under section 847 CA 2006 or the other non-statutory routes described. There is in addition a claim brought by BAT in its own capacity under section 423 of the Insolvency Act 1986.

Section 423: the law

Section 423 provides:

“423 Transactions defrauding creditors

(1)

This section relates to transactions entered into at an undervalue; and a person enters into such a transaction with another person if—

(a)

he makes a gift to the other person or he otherwise enters into a transaction with the other on terms that provide for him to receive no consideration;

(b)

…

(c)

he enters into a transaction with the other for a consideration the value of which, in money or money’s worth, is significantly less than the value, in money or money’s worth, of the consideration provided by himself.

(2)

Where a person has entered into such a transaction, the court may, if satisfied under the next subsection, make such order as it thinks fit for—

(a)

restoring the position to what it would have been if the transaction had not been entered into, and

(b)

protecting the interests of persons who are victims of the transaction.

(3)

In the case of a person entering into such a transaction, an order shall only be made if the court is satisfied that it was entered into by him for the purpose—

(a)

of putting assets beyond the reach of a person who is making, or may at some time make, a claim against him, or

(b)

of otherwise prejudicing the interests of such a person in relation to the claim which he is making or may make.

(4)

…

(5)

In relation to a transaction at an undervalue, references here and below to a victim of the transaction are to a person who is, or is capable of being, prejudiced by it; …”.

I shall refer to the purposes set out in section 423(3) as ‘the s 423 purpose’.

Section 424 limits the kinds of claimant who can apply for an order under section 423 to someone who is a victim of the transaction (except in circumstances which do not apply here). An application made for an order is treated as made on behalf of every victim of the transaction. Section 425 sets out the kinds of orders that the court can make if there has been a transfer, including an order requiring the property transferred by the transaction to be vested in any person or to require any person to pay to any other person in respect of benefits received from the debtor such sums as the court may direct.

The question of whether the s 423 purpose must be the predominant purpose of the transaction was considered in Inland Revenue Commissioners v Hashmi [2002] EWCA Civ 981 (‘Hashmi’). There the court was asked to set aside a written declaration of trust by Mr Ghauri of the beneficial interest in a business property for the benefit of his son. The claim under section 423 was brought by the Revenue asserting that the trust was aimed at avoiding a very large assessment for tax on undeclared profits in the business. The judge at first instance held that Mr Ghauri had deliberately and dishonestly under-declared his taxes and he knew that his other assets would be insufficient to meet the tax payments were his dishonesty ever discovered. The judge further held that Mr Ghauri had had two purposes, the first was that of a caring parent who wanted to secure the property for his son and the second was to put the property beyond the reach of creditors should they emerge. The judge made the declaration sought by the Revenue. On appeal, Arden LJ having described the history of section 423 held:

“23.

… Accordingly, in my judgment, the section does not require the inquiry to be made whether the purpose was a dominant purpose. It is sufficient if the statutory purpose can properly be described as a purpose and not merely as a consequence, rather than something which was indeed positively intended. Moreover, I agree with the observation of the judge that it will often be the case that the motive to defeat creditors and the motive to secure family protection will co-exist in such a way that even the transferor himself may be unable to say what was uppermost in his mind.

…

25.

… [F]or something to be a purpose it must be a real substantial purpose; it is not sufficient to quote something which is a by-product of the transaction under consideration, or to show that it was simply a result of it, …, or an element which made no contribution of importance to the debtor's purpose of carrying out the transaction under consideration. I agree with the point made by Lord Justice Laws in argument, that trivial purposes must be excluded.”

Arden LJ went on to uphold the judge’s finding that “the intent to defraud the Revenue was the dominant purpose”.

Laws LJ agreed with that judgment and expressed the test in the following terms:

“32 ... It is clear that the statutory purpose referred to in section 423(3) of the Insolvency Act 1986 need not be the only purpose for which the impugned transaction was entered into. Moreover, there is in my judgment no warrant for a construction of the statute which would qualify the term "purpose" by the adjective "dominant". No such qualification is required to make sense of the Act or to give it pragmatic efficacy. On the contrary, it is easy to envisage cases where more than one purpose is at hand between whose weight or influence it is on the evidence impossible to distinguish in practical terms.

33.

In such a case, in my judgment, the application of section 423(3) is by no means necessarily excluded. What in my judgment is required is that the claimant show that the donor, vendor or settler was substantially motivated by one or other of the aims set out in section 423(3)(a) and (b) in entering into the transaction in question. There may be cases in which, even absent the statutory purpose, the transaction would or might have been entered into anyway. That would not necessarily negate the section's application; but the fact-finding judge on an application made to him under section 423 must be alert to see that he is satisfied that the statutory purpose has in truth substantially motivated the donor if he is to find that the section bites.”

Simon Brown LJ considered but rejected a test which turned on the question whether the transaction would have been entered into “but for the debtor’s wish to put his assets beyond his creditors’ (or prospective creditors’) reach”. Instead he said that in each case the question to be asked is: can the court be satisfied that a substantial purpose of the debtor's transaction was (putting it in shorthand) to escape his liabilities? However, he added the caveat that if the judge were to find in any given case that the transaction is one which the debtor might well have entered into in any event, he should not then too readily infer that the debtor also had the substantial purpose of escaping his liabilities.

Hill v Spread Trustee [2006] EWCA Civ 542 was a claim brought by the trustee in bankruptcy of Mr Nurkowski to set aside a settlement trust made by Mr Nurkowski in favour of his daughter. The judge at first instance had found Mr Nurkowski to be ‘a dreadful witness’ and completely dishonest. He was not able to rely on anything Mr Nurkowski said. He found that one of the purposes of making the settlement was to prejudice the interests of the Revenue. This was not the sole or dominant purpose but it was positively intended and a factor that substantially motivated Mr Nurkowski. Arden LJ noted that an unusual feature of the case was that the settlement alone could not prejudice the Revenue. It was Mr Nurkowski’s failure to reveal an offer to acquire the land at a high value with the effect that the land valued in the settlement was too low that prejudiced the Revenue, because it led to the Revenue accepted a lower value for the purpose of computing tax. On the issues of law, Arden LJ (with whom, on this issue, the other two members of the court agreed) stated that the test whether Mr Nurkowski had the necessary intention is a subjective test: the judge had to be satisfied that he actually had the purpose, not that a reasonable person in his position would have it: paragraph 86.

Section 423: some preliminary points

First, it is not disputed here that BAT is a ‘victim’ within the meaning of section 423(5) and so is entitled to bring this claim. Secondly, this claim can only be brought by BAT and only as against Sequana as the recipient of the dividends. Thirdly, although this section is found in the Insolvency Act, there is no requirement that the company be insolvent or on the verge of insolvency. However, questions about the risk to creditors created by the payment of the dividend may arise when considering whether the victim’s interests were prejudiced by the transaction for the purposes of section 423(3)(b). Fourthly, it is accepted that when considering whether AWA acted with the s 423 purpose, it is enough if the majority of the directors acted with that purpose in declaring the dividend. This of course is only really relevant in relation to the December Dividend since it was not suggested that Mr Martinet and Mr Courteault could have had different purposes by the time of the May Dividend. Fifthly, although the section is headed ‘Transactions defrauding creditors’ there is no requirement of fraud in the sense of a finding of dishonesty on the part of the transferor.

The legal issues arising can be summarised as follows:

As to the application of the test to this case, BAT say that the claim is made out on the basis of the clear evidence of the witnesses, frankly given, as to what they intended to achieve by the 17 December 2008 and 18 May 2009 meetings.

Is a dividend a transaction falling within section 423(1)(a) or (c)?

Sequana submit that a dividend paid by a company to its member is not a gift, nor is it properly characterised as a transaction for no consideration or at an undervalue. The transaction pursuant to which the dividend is paid is not the decision of the directors to pay it but rather the antecedent contract deemed to exist by virtue of section 33 of the CA 2006. That section provides that the provisions of the company’s constitution bind the company and its members to the same extent as if there were covenants on the part of the company and of each member to observe those provisions.

Mr Foxton relies on Inland Revenue Commissioners v Laird Group plc [2003] UKHL 54 (‘Laird’) in which the issue before the House was whether the payment of a dividend was a ‘transaction in securities’ for the purposes of the Incomes and Taxes Act 1988. Lord Millett, with whom all their Lordships agreed, cited the definition of the term ‘transaction in securities’ and identified the question in issue as whether the payment of a dividend is a transaction in or relating to the shares in respect of which it is paid. He referred to a deeming provision in the legislation which made clear that without that provision, a distribution to shareholders upon a liquidation of the company would not be a transaction in securities. Accordingly the critical questions were why the payment of undistributed profits of the company to the shareholders in the course of a liquidation was not a transaction relating to their shares, and what, if any, was the difference between such a payment and a distribution to shareholders by way of dividend while the company is a going concern. Lord Millett examined the juridical nature of a share quoting at paragraph 35 of his speech the description by Lord Russell of Killowen in Inland Revenue Commrs v Crossman [1937] AC 26:

“It is the interest of a person in the company, that interest being composed of rights and obligations which are defined by the Companies Act and by the memorandum and articles of association of the company.”

The rights of the shareholders in a company are set out in its articles of association and are the rights to receive dividends if declared, rights to vote and rights in a liquidation to receive a share of surplus assets after discharge of liabilities. Lord Millett went on to say:

“37.

… The distribution of the undistributed profits of a company in liquidation to its shareholders is not a transaction relating to securities because neither the shares themselves nor the rights attached to them are affected by a payment which merely gives effect to the shareholders’ rights; they receive only what is already theirs. Distributions are made to shareholders in respect of the shares, but the shares of the individual shareholder are nothing more than the measure of the proportion of the total which is due to him.

38.

In my opinion the position is not materially different if part of the undistributed profits is paid to the shareholders by way of dividend while the company is a going concern. The Court of Appeal seized on the fact that until a dividend, if final, is declared or, if interim, is paid, the shareholders have no right to it: [2002] STC 722. Accordingly, the Court of Appeal concluded, the declaration and payment do not merely give effect to pre-existing rights.

…

40.

The right to receive a dividend does not arise until the conditions laid down in the company’s articles of association are satisfied. Any requirement that a dividend must be declared by the directors before the shareholders are entitled to receive it must be found in the company’s articles of association: there is nothing in the Companies Acts save an obligation not to pay dividends out of capital. Constraints on the shareholders’ rights to receive dividends contained in the articles are self-imposed.

41.

In the early days articles of association commonly left the declaration of dividends to the company in general meeting, that is to say to the shareholders themselves. Gradually, however, it became the general practice to require the dividend to be declared by the directors. The change was a response to the increasing separation of ownership and management. The shareholders own the company, but they entrust the management of its undertaking to the directors. To enable the directors to carry out their functions, shareholders give them a discretion to decide how much of the company’s funds should be retained to pay creditors and carry on the business and how much can safely be returned to shareholders by way of dividend. By declaring a dividend, the directors effectively release funds due to the shareholders from their power to retain them in the business.

42.

Whether the company is in liquidation or continuing to carry on business as a going concern, therefore, the distribution of the undistributed profits of a company to the shareholders entitled thereto merely gives effect to the rights attached to the shares. The funds are released, in the one case from the liquidator’s discretion to retain them for the purpose of the winding up, and in the other from the directors’ discretion to retain them for the purposes of the undertaking. Given that the former is not “a transaction relating to securities’’, neither in my opinion is the latter. The relationship between the payment and the shares in respect of which it is paid is the same in both cases.”

Although I see the force in Mr Foxton’s submission I do not consider that the analysis of the nature of the dividend in Laird for the purposes of statutory provisions at issue there can be read across to the construction of section 423. A key step in Lord Millett’s reasoning that the payment of the dividend was not a transaction relating to shares was the statutory provision which made it clear that Parliament did not regard the liquidation of company in itself as a transaction relating to its shares. That is not the position here. In construing section 423 I accept Mr Thompson’s submission that the wording of section 423 is deliberately wide in order to protect creditors from assets being moved from the potential debtor out of their reach. Subsection (1) is drafted to exclude transactions only where the consideration received by the potential debtor is not significantly less in value than the consideration that the debtor receives. Where the consideration provided by the debtor and the other party to the transaction are roughly the same, there can be no detriment to the creditor because the debtor’s assets are not depleted. But where they are not, then the creditor is less likely to be able to recover what is owed to him. There may be a situation where the consideration paid by the debtor to the third party is fixed in a contract but payment is delayed. Provided that the initial contract was not entered into with the s 423 purpose, that delayed payment is not a transaction with the s 423 purpose because the purpose is to fulfil the contractual obligation to make payment. The payment may have the consequence of depriving other creditors of money later, but as Arden LJ emphasised in Hashmi, consequences are different from purposes. The payment of the dividend is not, in my judgment, the satisfaction of an earlier obligation in the same way. It is true that the reason why the member of the company, rather than any other person, receives the dividend is because of the pre-existing relationship of company and shareholder. But the decision to pay the dividend and choice of its value is not the consequence of that relationship because it is discretionary not only in its amount but in whether it is paid at all.

Mr Foxton argued that there is no lacuna left in the protection of creditors if dividends are not transactions within section 423 because creditors are already adequately protected by Part 23 of the CA 2006 and by the fiduciary duties owed by directors when the creditors’ interests duty arises, as discussed earlier. I do not accept that. First, the only possible claimant in respect of those other causes of action is the company itself whereas the class of claimants here is much wider under section 424. Secondly, the powers of the court to put matters right under sections 423(2) and 425 are much broader and more flexible than the remedies available under Part 23. It is not difficult to see that a blanket exclusion of dividend payments from the scope of section 423 will quickly reduce the efficacy of the provision given the many instances where the directors and shareholders of a company are the same or linked individuals.

There is no reason in the wording of the section to exclude the payment of a dividend from the scope of section 423 if the payment is made with the s 423 purpose.

Did the directors of AWA have the s 423 purpose in relation to the December Dividend?

I have described earlier the evidence of the four directors as to their intentions in declaring the December Dividend. I am fully satisfied that the directors cannot have had the s 423 purpose in declaring that dividend. At that stage there was no settled intention of selling AWA to someone outside the Sequana group. Certainly it was Mr Martinet’s and Mr Courteault’s wish to declare a further dividend so as to remove the inter-company receivable if that became possible later on. But this was long before the developments I have described in section VII of this judgment and there was no reason to think that it would become possible. There was, as I have found, no intention to declare a further dividend come what may and no intention to sell the company to someone outside the group unless the receivable could be reduced to very much less than it stood as at that date.

It is also relevant that Mr Lebard’s evidence was that Sequana’s policy was to stand behind its subsidiaries. Whatever the position on AWA’s accounts, if AWA was owned by the Sequana group, they would have injected whatever cash was needed into the company to enable it to meet its obligations, even if they had no legal obligation to do so. If you own a business, Mr Lebard said, “you have obligations and if you need cash to do it, you just do it”. He said:

“If I can elaborate, I will give you some reason for that. Never, ever Sequana didn't fulfil its obligations vis à vis any subsidiaries. And also we had some subsidiaries in the paper industry that you can imagine which had a tough, tough time, and very big difficulties. We had injected the money which had to be needed to find ways to help our business to survive and to work. That is the first point.

The second is we also had a board of directors and maybe I will come back to that, which are prestigious and very well-known people, including senior people from the Agnelli family. And in the end, remember that the Agnelli family still owned close to 30% of Sequana. Sequana was consolidated into the IFIL group; the Agnelli family. For the Agnelli family, can you imagine that they would accept that one of their subsidiaries would let fall bankrupt a business in the US, with obligations vis à vis the US Government? It is just impossible.

So in my mind, in the mind of the board, it was not question at all not to fulfil our obligations vis à vis AWA. As simple as that.”

Mr Martinet and Mr Courteault were presumably aware of this as senior executives of Sequana. Since they are part of the guiding mind of AWA for this purpose then I do not see that AWA can have had the s 423 purpose at that point. It was not suggested that Mr Mountford and Mr Newell had that purpose if Mr Martinet and Mr Courteault did not.

Did the directors of AWA have the s 423 purpose in relation to the May Dividend?

There is, in contrast, plenty of evidence to show that the intention of AWA, through the governing minds of Mr Martinet and Mr Courteault, in declaring the May Dividend was to remove from the Sequana group the risk that the indemnity liability to BAT for the Lower Fox River clean up might turn out to be much more than the amount available from the Maris Policy plus the Historic Insurance Policies receipts.

So far as Mr Martinet is concerned, I have already referred to the email on 11 May 2009 when he was responding to Mr Newell’s question about whether they were foregoing a good opportunity to enjoy a ‘potential upside’ by selling AWA to TMW: see paragraph 286, above. Mr Martinet’s view was that AWA should be satisfied with “mostly getting rid of a very hairy situation once and for all, without being held responsible for future downside, however remote this might be”. In a more formal setting, Mr Martinet’s report to the Sequana board meeting on 27 May 2009 referred to the immediate effect of the operation being “to externalise a significant underlying risk that was difficult to control from the scope of the group, with Sequana having expressly excluded any guarantee under the sale contracts of API and the Fox River risk”.

Mr Courteault’s evidence was to similar effect when he was cross-examined about the terms of the sale of AWA to TMW and the absence in that sale of any indemnity or guarantee from Sequana to TMW in relation to AWA’s existing liabilities:

“Q. … in this case you very carefully, in a number of places in the documentation … inserted clauses to make it absolutely clear that the seller, Sequana, would have absolutely no post-sale liabilities in relation to the indemnity liabilities, Fox River, Future Sites and so on; correct?

A. It was absolutely the purpose of -- it was a part of this transaction. I also took care –

Q. It was the reason for the transaction?

A. Well, the reason for the transaction was a sale. But we did not intend to guarantee any of the liabilities regarding the Fox River or any of the Future Sites.

Q. Because if you had that would have cut across … the objective of removing Sequana from …

A. From the risk, yes, of course.

Q. Thank you.

A.

And that is the reason why I also took care of including in the agreement some wording about the fact that Mr Gower and Mr Tauscher were fully aware of what they were purchasing. … It was important for me that these two persons, who were the purchasers of the company, recognised that they knew exactly what they were purchasing. … I wanted Mr Gower and Mr Tauscher to give a personal guarantee that they were fully aware of what they were purchasing and so on. It was important for us.”

Although Mr Courteault there spoke about removing the risk of any liability being owed to Mr Gower and Mr Tauscher, he must have realised that in fact the absence of any guarantee from Sequana to TMW was likely to be of much greater concern to BAT that it would be to TMW.

Mr Lebard explained the importance for Sequana of removing the risk from AWA’s and Sequana’s balance sheet. Sequana was trying to present itself to investors as a ‘pure player in the paper business’. Investors were discouraged from investing in the Sequana group if they could see environmental issues in the U.S. on the balance sheet:

“… it is immediately scary, and they don’t want to spend the time in trying to understand the case and they invest in another company. It is a marketing and communications thing.”

There may well have been legitimate business reasons for Sequana wishing to rid itself of the risk that the Maris Policy and the Historic Insurance Policies would ultimately not be enough to enable AWA to fulfil its indemnity obligations. But there is no requirement in section 423 for the transferor to be motivated by some ill will towards a particular creditor or to be acting dishonestly (although many of the cases in which section 423 is relied on are cases of dishonesty). The removal of the ‘scary’ item on Sequana’s balance sheet could only be achieved if the May Dividend could be paid by off-setting it against the remaining inter-company receivable so that the company could be sold. That was the purpose of the transaction.

Mr Foxton argued that the purpose of the sale of AWA might have been to remove the liability from Sequana’s account but the transaction being challenged here under section 423 is not the sale as such but the payment of the dividend. Further, he submitted, there is no evidence that the dividend payment was motivated by the desire to remove any risk of further liability from Sequana. I do not accept that one can distinguish between the purpose of paying the dividend and the purpose of selling the company in that way by the time the May Dividend came to be paid. Both Mr Martinet and Mr Courteault knew what the sequence of events would be on the evening of 18 May 2009 during the series of board meetings that they held over the telephone. It is clear from Hill v Spread Trustee referred to earlier that it is enough if the impugned transaction is entered into with the s 423 purpose; it does not have to achieve that purpose by itself:

“102.

… If the transaction is entered into with the requisite purpose, the fact that some other event needs to occur does not mean that the transaction cannot itself be within section 423(3). I consider that this is what the judge meant by his test of whether the transaction was an essential part of the purpose. … The right approach in my judgment is to apply the statutory wording. It is enough if the transaction sought to be impugned was entered into with the requisite purpose. It is entry into the transaction, not the transaction itself, which has to have the necessary purpose.”

I have no doubt here that the payment of the May Dividend was entered into with the purpose of eliminating the receivable which then cleared the way to AWA being sold and to Sequana removing any risk of having to fund the indemnity itself if the funds left in AWA proved to be inadequate.

The Defendants say that provided that AWA had made proper provision for the Fox River liability in its accounts then there is no room for a finding that the payment of a dividend which was not contrary to Part 23 and was not in breach of the directors’ fiduciary duty was nonetheless made with the s 423 purpose. The Defendants say in effect that if a provision has been properly made in relation to the claim, the company only needs to keep back that money. It is free to spend any additional monies or dispose of any additional assets and it cannot be said to have the s 423 purpose. There is no justification, Sequana submits, for overlaying on those protections the application of section 423.

Sequana also points to the difficulties that would arise in practice if companies have to take into account the possible application of section 423 when considering whether to declare a dividend. I referred earlier to BAT’s own complex and high value list of contingent liabilities. Do those actual and potential claims mean that a company in BAT’s position risks entering into a transaction at an undervalue with the s 423 purpose if it declares a dividend without checking not only that it has made adequate provision in its accounts in accordance with the relevant accounting standards but that it has enough assets left to meet the totality of the amount claimed if that turns out to be higher than the best estimate on which the provision is based?

The Claimants rely on the very particular circumstances of this case. AWA was a non-trading company and a wholly owned subsidiary. Its only function was as a containment vehicle for the Fox River liability. There is clear evidence that the purpose of the declaration of the May Dividend and the sale of AWA to TMW clearly was to remove from Sequana the risk that the Maris Policy plus the insurance proceeds might not be enough to meet the indemnity. Such evidence of the subjective intention of those in control of the company when making the decision to pay the dividend will distinguish this case from other cases where directors declare dividends for their shareholders for the usual reasons for which dividends are paid, without turning their minds to whether this leaves enough money for potential creditors. Here there is no doubt that the subjective intention of the directors at the time of the May Dividend and the sale was to prevent AWA having any legal or moral call upon its parent company to meet its creditors’ claims. After the declaration of the dividend and the sale to TMW, the creditors were prejudiced because the assets of AWA had been depleted and it no longer had any call on Sequana to that extent.

After some hesitation I have concluded that the Claimants are right on this point. Section 423 does not distinguish between companies and individuals. The first limb of the s 423 purpose - putting assets beyond the reach of a person who is making or may at some time make a claim against him - has inherent in it the assumption that following the transaction, the person does not have sufficient funds remaining with him to satisfy the actual or potential claim made against him. If a person or a company has plenty of assets left with which to meet the claim, then however many additional assets are gifted to people, he or it cannot have the s 423 purpose. This must be inherent in the wording of section 423(3)(a), and is confirmed by the second limb which refers to action “otherwise prejudicing the interests of” the claimant, implying that the transaction in the first limb must prejudice those interests too.

If an individual (‘A’) comes under a liability to another person (‘B’), for example by negligently injuring B in a car accident, there may be a great deal of uncertainty about how serious B’s injuries will turn out to be in the long term. A may receive advice from the experts he engages that it is likely that B will make a full recovery in which case the best estimate of the value of B’s claim is £50,000. But the advisers may also warn that, if it turns out that B’s injuries are much worse and likely to be permanent, the value of the claim could be very much higher, perhaps up to £500,000. Suppose that A has savings of £50,000 in the bank and a house worth £1 million. A decides to leave the money in the savings account to meet B’s claim but to gift the house to his daughter C just in case it turns out that B’s injuries are much more serious than expected. Does A transfer the house to C with the s 423 purpose? I think he does. If that is right then the same reasoning applies to AWA. AWA received advice and, as I have held, properly took the view that the best estimate of its liability under the indemnity was $143 million, covered by the Maris Policy. But Mr Martinet and Mr Courteault were well aware of the great uncertainties that existed over the ultimate level of remediation and NRD costs. The size of NCR/API’s share would be determined perhaps only after many years of the further litigation. It may have been an unlikely scenario and I have held that it was not sufficiently likely to generate a duty to take account of BAT’s interests when declaring the May Dividend. But the evidence shows that it was precisely the scenario they had in mind when they paid the May Dividend in order then to be able to sell the company and move the risk out of the group. The transaction was undertaken with the intention of putting assets beyond the reach of BAT in the event that the Maris Policy and the Historic Insurance Policies receipts were not enough to meet the indemnity claim. This prejudiced BAT because, as Mr Martinet and Mr Courteault knew, the new owners of AWA would not have any other funds to make good any shortfall.

I therefore find that AWA, through its directors, did have the s 423 purpose when paying the May Dividend.

The Defendants’ final point on section 423 is that Sequana has changed its position as a result of the payment of the dividend and that this constitutes a defence to the claim under section 423. For the existence of this defence Mr Foxton relies on the judgment of Sales J (as he then was) in 4Eng Ltd v Harper [2009] EWHC 2633 (Ch) the leading case on the scope of the powers of the court under section 425. In that case a director had transferred his house into his wife’s name with the s 423 purpose of avoiding having to satisfy a judgment against him. Sales J found that the wife did not know of the illegitimate purpose for the transfer. He noted that the statute does not specify any particular mental state or action on the part of the transferee as an ingredient of the conditions for liability. That does not mean however that such matters are irrelevant for defining the extent of the liability to be imposed, or the order to be made, at the next stage in the analysis, when the court considers the question of remedy under s.423(2) and s.425. Sales J held:

“13.

ln my judgment, the nature of any order and the extent of the relief granted by the court under s.423(2) and s.425 should take into account the mental state of the transferee of property under a relevant transaction (or of any other person against whom an order is sought) and the degree of their involvement in the fraudulent scheme of the debtor/transferor to put assets out of the reach of his creditors. The principles in the application of this statutory regime should reflect in this respect general principles inherent in other areas of the law, which treat the mental state and degree of involvement of a defendant in wrongdoing as relevant to the extent of recovery available against him (compare, as one example among many, Seager v Copydex Ltd [1967] 1 W.L.R. 923, 932 — no order of an account of profits ordered against an innocent wrongdoer in respect of a breach of confidence). Although the trigger conditions for liability to make restoration under s.423 set out the basic balance to be struck between the interests of the creditors and of a transferee as established by Parliament, the making of an order under s.423(2) and s.425 necessarily requires some further balancing of the interests of the transferor's creditors and of the transferee to be determined by the court, since by the time the court has to take action events will have moved on from the transfer and the balance of the equities between creditors and transferee may well have been affected by changes in circumstances over time.”

After giving examples of different situations, Sales J went on:

“16.

In choosing what relief is appropriate in a given case, a great deal will depend upon the particular facts. One of the reasons the court is given such a wide jurisdiction as to remedy under this regime is to allow it flexibility in fashioning relief which is carefully tailored to the justice of the particular case. Helpful analogies may be drawn with other areas of the law to guide the court in reaching its conclusion, but given the wide range of situations which the statutory regime is intended to deal with it would be wrong to be unduly prescriptive in trying to lay down hard and fast rules for the application of these provisions.”

The Defendants submit that Sequana as transferee of the May Dividend did change its position after the transactions on 18 May 2009 by selling AWA to TMW and thereby losing control over the management of the company’s exposure to the Lower Fox River and the various pieces of litigation that AWA was involved in against the other PRPs or against the insurers under the Historic Insurance Policies. If the May Dividend had not been paid, the inter-company receivable would not have been released, the company would not have been sold and Sequana would have retained control of AWA. They point also to the complex funding agreement that was concluded between BAT/BTI, NCR, API and AWA in 2014 under which, they say, AWA’s exposure to the various risks involved had been fundamentally re-written.

In my judgment the 4Eng decision establishes that these points are relevant to the question of what is the appropriate relief to be granted and do not provide a complete defence to the claim under section 423. I do not accept the submission that Sequana must be shown to have acted in bad faith in the sense of having engaged in sharp practice or recklessness before it is appropriate to fashion a remedy under section 425.

In any event Sequana’s position is very different from the position of the transferee in 4Eng. First, Sequana fully shared the s 423 purpose of AWA declaring the May Dividend and in fact was the intended beneficiary of that purpose since it was Sequana which no longer bore the risk of the liability exceeding the Maris Policy. AWA and Sequana had always relied on the abilities of Mr Gower and Mr Tauscher to conduct those matters for their benefit; their interests were fully aligned and remained so, as a result of the golden share, after the sale of AWA to TMW. I do not therefore accept that there was a significant detriment on which Sequana can rely in giving up control of the conduct of the Whiting litigation or the control of the Historic Insurance Policies settlement negotiations.

The parties were agreed that, in the event that I found that the claim under section 423 succeeded, I should not attempt to fashion a remedy at this stage. This was in part because in order to do so it is important to understand what has in fact happened in respect of the Lower Fox River and Future Sites liabilities since May 2009. During the course of these proceedings, we have all carefully shielded our eyes from what has happened since then in order to ensure that the assessment of what the directors knew and thought at the relevant times was not coloured by hindsight. However it may be helpful for me to give the following indication. I do not currently have in mind a remedy whereby the May Dividend is simply repaid by Sequana to AWA. In 4Eng Sales J contemplated such a payment in a situation where the transferor has become insolvent so that the transferred money is repaid to the transferor’s liquidator or trustee in order to be distributed to the class of creditors. That is not the position here.

For the reasons set out above, I therefore dismiss BTI’s claims. Neither the December Dividend nor the May Dividend contravened Part 23 of the CA 2006 because:

As regards BAT’s claim under section 423 of the Insolvency Act 1986:

I thank all those involved in this case which was very complex both factually and legally.