AC Ward & Son Ltd v Catlin (Five) Ltd & Ors

The claimant is a company in the Blueheath group of companies and is the owner and operator of a wholesale distribution warehouse at Victor House, Bay Manor Lane, Thurrock, Essex. The warehouse is occupied, with staff working in shifts from 22.00 hours on a Sunday to 22.00 hours on a Friday, but unoccupied from 22.00 hours on a Friday to 22.00 hours on a Sunday. On the weekend of 17/18 March 2007 whilst the warehouse was unoccupied, a gang of thieves broke into the warehouse and stole cigarettes and tobacco stored in a caged area on the mezzanine floor of the warehouse. The value of the goods stolen is agreed at £432,940.

The claimant now brings this claim for an indemnity in respect of that loss against its insurers, the defendants (to whom I will refer as “the insurers”) who insured it and the other companies in the Blueheath group, under a “Multiline” insurance policy covering material damage, business interruption, theft, loss of money, employer’s liability, public liability, computers and social housing all risks.

The defendants have resisted liability on the grounds that: (i) before the time of the loss, the claimant was already in breach of the Protection Maintenance Warranty and Burglar Alarm Maintenance Warranty in the Policy, so that the insurers had been discharged from liability from the date of breach of warranty; and/or (ii) a variation to the policy made on 20 February 2007 to remove Endorsement 6 (which contained an exclusion of cover for theft of cigarettes and tobacco outside business hours unless they were stored in a secure store on the ground floor of the warehouse) could be avoided for material non-disclosure or misrepresentation with the effect that the exclusion in Endorsement 6 remained in place; and/or (iii) the theft took place with the collusion of one or more employees of the claimant and was therefore excluded under the policy.

The insurance was underwritten pursuant to a binding authority granted by the insurers to underwriting agents, AT Docherty General Underwriting Agency Limited (“ATD”), with offices in Manchester and Leeds. They underwrote considerable numbers of composite multiline policies like the one under consideration on behalf of the insurers. The risk was broked to ATD by Henderson Insurance Brokers Limited in Leeds (to whom I will refer as “the brokers”).

The policy provided so far as relevant, as follows:

The main issues in the case as they remained at the end of the trial can be summarised as follows:

What has to be shown by insurers to establish collusion within the meaning of the exclusion clause in the Policy and was the theft carried out with the collusion of any employee of the claimant?

Does the Protection Maintenance Warranty (“PMW”) only apply to such security devices as were notified to the insurers at the time of placement of the insurance or prior to the theft or does it apply to whatever security devices were in fact present on the premises at the time of the theft?

On the true construction of the PMW what must the insurers establish for a defence of breach of warranty? Is it enough that the relevant security protections were not in “full and effective operation” (as the insurers contend) or must the insurers show that there was a defect in the protections which the claimant had failed to remedy as soon as reasonably practicable after it had discovered the defect (as the claimant contends)?

Was the claimant in breach of the PMW because (a) the guardwire fitted to the walls of the mezzanine area (including the outer wall) had not been connected to the intruder alarm system; (b) five of the motion detectors on the mezzanine had been angled upwards; (c) the ADSL line was disconnected between 13 February 2007 and 16 March 2007 or (d) there was a fault on that ADSL line during the period of the burglary?

Does the Burglar Alarm Maintenance Warranty (“BAMW”) only apply to burglar alarm systems which are specified in the Schedule to the Policy of Insurance and, if so, was the system at the warehouse so specified?

On the true construction of the BAMW what must the insurers establish for a defence of breach of warranty? Is it enough that the relevant burglar alarm system was not in “full and effective operation” (as the insurers contend) or must the insurers show that there was a defect in the system which the claimant had failed to remedy as soon as reasonably practicable after it had discovered the defect (as the claimant contends)?

Was the claimant in breach of the BAMW because (a) the guardwire fitted to the walls of the mezzanine area (including the outer wall) had not been connected to the intruder alarm system or (b) five of the motion detectors on the mezzanine had been angled upwards.

What representations were made by or on behalf of the claimant prior to the variation to the contract of insurance by which the insurers removed the exclusion in Endorsement 6?

Were the representations that were made material misrepresentations or, alternatively was there material non-disclosure by the claimant?

If so, did such material non-disclosure induce the insurers to agree to the variation?

Before going through the detailed chronology of events leading up to the theft, I shall set out my assessment of the various witnesses called on both sides, not least because a number of conflicts of evidence arise and in other respects issues arise as to whether particular witnesses were telling the truth.

The claimant called three witnesses from the brokers. The main witness was Mr Andrew White, a director of the brokers with primary responsibility for the claimant’s account. Mr White was an unsatisfactory witness in a number of respects. I formed the very firm view that his witness statements and to an extent his oral evidence were skewed, so as to show the brokers in as favourable a light as possible, whilst blackening the ATD underwriters as much as he could. For example, in his witness statement he made a wholly unwarranted attack on the underwriting of Mr Rusinek and Mr McAndrew (the ATD underwriters responsible for agreeing the addition of the claimant to the insurance and the removal of Endorsement 6 respectively) accusing them of a cavalier attitude to underwriting and in effect of having written business for premium income, without regard to the risk. Having seen Mr Rusinek and Mr McAndrew in the witness box and considered all the contemporaneous correspondence, including details of the number of risks declined, I am quite satisfied that there is nothing in Mr White’s criticisms.

Other aspects of Mr White’s evidence were unsatisfactory. For example, he gave evidence which was evasive about not having received an email of 2 June 2006 from Mr Sullivan of ATD. I was unable to accept that evidence which seemed to me designed to downplay as much as possible the brokers’ failure to inform the claimant about Endorsement 6 until eight months later in February 2007. Similarly, although initially in cross-examination Mr White accepted that the “earlier telephone conversation” referred to in an email from Mr McAndrew to him on 14 February 2007 had probably taken place that day, Mr White then went away and produced a third witness statement about which he was recalled and cross-examined on Day 5 of the trial. This further statement contained an elaborate reconstruction of his movements on 14 February 2007 with a view to bolstering his certainty on further reflection that the conversation had not taken place on that day.

Mr Richard Phillips was a Senior Commercial Broker reporting to Mr White. Overall he was a more straightforward witness than Mr White but his evidence was marred in places by an over-anxiety to support the evidence given by Mr White, which on occasions meant that Mr Phillips’ evidence on a particular point was not credible. Finally from the brokers, Mr Peter Wilkinson gave evidence about discussions with Mr McAndrew and others soon after the theft, from which he had the impression that the insurers considered cover was in place and about discussions with the Essex Police who did not consider the theft an “inside job”. I did not find his evidence of any assistance in relation to the issues I had to determine and propose to say no more about it.

The claimant called a number of its own employees to give evidence. Mr Sam Radford was the transport operations manager but was effectively the warehouse manager in 2006. He struck me as a truthful witness. Where his evidence conflicted with that of the insurers’ risk surveyor Mr Kenneth Thomas, I preferred Mr Radford’s evidence. Mr Graham Copper was a consultant Operations Manager employed for two months between the end of August and the end of October 2006 to oversee improvements of operations at the claimant, including security arrangements. He too was a straightforward witness, most of whose evidence I accepted.

Mr Simon Cloke became the Operations Director at the warehouse in November 2006 after Mr Copper had left and was in that position at the time of the theft. There were aspects of his evidence which troubled me, for example he said for the first time in cross-examination that he had carried out an additional walk-test after he first saw the Risk Improvement Requirements in January 2007, to see if the motion detector lights were working on the mezzanine floor. He had not mentioned this additional test in either witness statement or indeed in the Incident Investigation Report dated 24 May 2007, which he had prepared. Reluctantly, I have concluded that he made this evidence up.

Other aspects of his evidence were unsatisfactory, such as what he was told by the Initial engineer who visited the site immediately after the theft about the likelihood of motion detectors being left angled towards the ceiling. What he said in his witness statements was at odds with what he had said soon after the theft in the Incident Investigation Report dated 24 May 2007 and I considered his attempts to explain this away somewhat evasive. It is also surprising to say the least that he does not mention the Incident Investigation Report in either of his witness statements. These unsatisfactory aspects of his evidence led me to approach his evidence generally with some caution.

Mr Leigh Butler was responsible for IT at the Thurrock site from June 2005 to June 2007. He gave evidence about the problems with the ADSL line in February and March 2007. He struck me as a frank and truthful witness whose evidence I accepted. Mr Simon Mindham was the finance director of Blueheath Holdings at the time. He gave evidence about the problems with BT and with the ADSL line at the time. His involvement had been at one remove, so that his evidence was of limited assistance.

The defendants called a number of witnesses formerly employed by ATD. Mr Marek Rusinek was the Underwriting Manager at the Manchester office who wrote the original risk and agreed the variation to add the claimant to the policy. He was an engaging witness who seemed to me to be essentially telling the truth and who exhibited none of the cavalier attitude to underwriting of which he was accused by Mr White. Where his evidence was not all that convincing was in relation to whether the agreement to add the claimant had been subject to survey. It clearly had not been as a matter of objective contractual analysis, but Mr Rusinek seemed, as his evidence went on, to have convinced himself that it had been, possibly because he appreciated that a subject or condition should have been imposed but had not been. Ultimately however I did not consider his evidence central to the issues I have to decide.

Of much more importance to those issues was the evidence of Mr Paul McAndrew, the underwriter in ATD’s Leeds office whose underwriting decision it was to lift Endorsement 6 on 20 February 2007. I considered him an impressive and honest witness. Contrary to Mr White’s accusations, his attitude to underwriting was not cavalier. Both from his demeanour in the witness box and in his contemporaneous dealings with the brokers, it emerged that his approach to underwriting was a careful one. He was clearly concerned that the endorsement should not be lifted unless and until the Risk Improvement Requirements had been complied with. Although Miss Sabben-Clare, who appeared for the claimant, suggested that some of his evidence was a little exaggerated, for example in relation to how many telephone conversations he had with Mr White in January/February 2007, I did not consider that there was anything in this criticism. Where his evidence conflicted with that of Mr White, I preferred Mr McAndrew’s evidence.

Mr Andrew Sullivan was another underwriter in ATD’s Leeds office who was responsible for arranging the risk surveys of the warehouse and when those surveys raised serious concerns as to the security of cigarettes and tobacco stored on the mezzanine floor, he was responsible for imposing the exclusion which became Endorsement 6. He also struck me as straightforward and honest, willing to accept criticism over matters such as the late production of policy documentation. Where his evidence and that of Mr White was in conflict, particularly over whether a telephone conversation took place between them on 2 June 2006 during the course of which Mr Sullivan told Mr White about the imposition of the exclusion pending the implementation of risk improvements, I preferred the evidence of Mr Sullivan.

Mr Ken Thomas was an experienced insurance risk surveyor with more than 40 years experience in the industry. He is clearly competent and careful, although I suspect somewhat inflexible in his approach. In his oral evidence this manifested itself in a manner which was accepted by Mr David Turner QC who appeared on behalf of the insurers as having been somewhat defensive. Much of his evidence was not in truth controversial but where there were differences between his evidence and that of other witnesses, he would brook no suggestion that he was wrong or mistaken. I was not over-impressed by that rather inflexible approach and where his evidence differed from that of others, for example whether there had been a detailed discussion of his concerns with Mr Radford at the end of his visit to the warehouse on 21 June 2006, I was inclined to prefer that other evidence.

Mr Richard Heath is a forensic investigator employed by Hawkins & Associates Limited. He was instructed on 22 March 2007 immediately after the theft by the loss adjusters retained by the insurers, Carr Greenwood Smith, to investigate the security systems in place at the site which he visited on 26 March 2007. His evidence was not in truth controversial because, for example, by the time of his visit, the Initial engineers had attended and the motion detectors had been repositioned.

Finally the defendant called Mr Len Collins, an information disclosure manager employed by British Telecommunications plc (“BT”), on a witness summons to produce BT’s records relating to the disconnection of the copper wire landline and ADSL line at the premises in the period January to March 2007. Since he had had no involvement of any kind at the relevant time, his evidence was of limited assistance, although he was able to provide some useful opinions as to the interpretation of some of the puzzling aspects of the contemporaneous records relating to the telephone lines.

Both sides exchanged expert reports from security experts who then prepared a joint memorandum which essentially agreed the security issues they had been asked to consider, as a consequence of which it was agreed between the parties that it was not necessary to call either of them to give oral evidence. I will refer to what they agreed to the limited extent necessary when I deal with some of the security issues in more detail below.

The parties also served expert reports from underwriters on the issue of materiality. In his report, the claimant’s expert, Mr Allan Guest set out his views as to why the claimant’s various statements were not in his view material because they descended to a level of detail which a prudent underwriter would not require in deciding whether to remove the exclusion in Endorsement 6. There was a question mark over whether he had applied the correct test of materiality as opposed to a decisive influence test, but it is not necessary to decide that issue, since (for reasons I will come to in more detail when addressing the issue of materiality) it was quite apparent from his oral evidence in cross-examination that he accepted the materiality of the various matters said to have been misrepresented. In the light of the evidence of Mr Guest in cross-examination, the insurers decided not to call their expert underwriter, Mr Peter Sangster, so no further reliance was placed or could be placed by the insurers on his report.

The insurance was originally placed in September 2005 on behalf of Blueheath Holdings Limited and Blueheath Direct Limited alone (before Blueheath acquired the claimant). This was new business to these insurers, Blueheath previously having been insured elsewhere through different brokers. On 22 September 2005, Mr Marek Rusinek of ATD wrote to the brokers confirming that the insurers were holding covered subject to amongst other things, satisfactory proposal form and satisfactory survey and compliance with risk improvements. These subjects were standard at ATD for a new proposed insured. The proposal form was not immediately forthcoming and ATD chased the brokers on 27 October 2005, asking for its return within 14 days, otherwise cover might be jeopardised.

In fact the brokers had only sent the proposal form to their client on 17 October 2005. It was completed by Sue Pike, the financial controller of Blueheath, signed by her 17 November 2005 and returned to ATD by the brokers under cover of a letter of 21 November 2005. The proposal form was described on its face as “Proposal for Multiline Commercial Combined Insurance and Multiflex Traders Combined Policy”. Thus, as a comparison between the questions in the proposal form and the terms of the Multiline Policy eventually issued by the insurers makes clear, the proposal was specifically designed to be used with this policy wording. Section 4 of the proposal form, headed “Security on Premises” asks questions about matters such as the number and construction of perimeter doors, the locks and locking devices, whether there is a safe and whether there is a special “secure area” for high value or target goods and if so, the construction of the walls and ceiling and the doors. The section also asks whether there is an intruder alarm on the premises and if the installer is NACOSS approved, together with questions about the alarm system such as whether it is connected to the police or to a central station.

So far as Blueheath Holdings Plc is concerned, its premises were an office building on Upper Street, Islington and Blueheath Direct Ltd occupied rented space in the GIST Distribution Warehouse in Tamworth, so that the question about the secure area was answered “not applicable”. Although the question about the alarm system was answered affirmatively (clearly by reference to the Islington premises) Blueheath did not enclose the alarm specification as required by the proposal form. This was something for which the insurers chased at regular intervals thereafter, without success, and was a matter which had not been resolved at the time of the loss.

The proposal form ended with a declaration signed by Ms Pike with the standard provision that the proposal and the statements made in it formed the basis of the contract. It also included express agreement that: “the security of the Premises and all other protections and/or safeguards mentioned herein shall not be withdrawn or varied to the detriment of the Underwriters without their consent”, words which find their echo in provisions contained in the PMW and BAMW. Following receipt of the proposal form, ATD requested a risk control survey of the Islington premises to be carried out by one of its risk surveyors, on the basis that this was new business. Policy documentation was sent by ATD to Andrew White of the brokers under cover of a letter dated 21 December 2005.

At around this time in late 2005, Blueheath acquired AC Ward & Son Limited, the current claimant and CTM Wholesale Limited, a company with a warehouse in Wrexham. The claimant was coincidentally an existing client of the broker, insured through them with Royal & Sun Alliance (“RSA”) for the annual period from 1 April 2005 to 31 March 2006. For that reason, the insurance with RSA was left intact on acquisition and steps were only taken to bring the claimant and CTM onto the Blueheath insurance with the insurers with effect from 1 April 2006. The insurance with RSA included an intruder alarm warranty in RSA’s standard form.

A meeting took place at the brokers’ offices in Leeds on 7 March 2006, between Mr White and Mr Phillips of the brokers and Mr Rusinek and Mr Skidmore of ATD, to discuss the addition of the claimant and CTM to the Blueheath insurance. Mr Phillips prepared a note of that meeting, which states that a sum insured of £3 million split between cigarettes, wine and spirits and other goods was required. The note then sets out the premium quoted by Mr Rusinek and the commission split for the brokers for the different sections of the policy. Mr Rusinek also prepared a note which set out the same matters. He also had with him at the meeting schedules prepared by the brokers which contained summaries of the claimant’s existing insurance cover with RSA, thereby indicating what insurance it required.

There was something of a conflict of evidence about what else, if anything, was discussed at this meeting. Mr White, consistently with the thesis expounded in his witness statement that Mr Rusinek and other ATD underwriters were only interested in writing for premium income, insisted that the only discussion was about sum insured, premium and commission. The most he was prepared to concede was that there would have been some discussion about financial matters such as turnover, for the purposes of Mr Rusinek quoting the premium. He maintained that it was a short meeting limited to those matters. He said this notwithstanding that he accepted that, in all probability, before the meeting he would either have had discussions with the claimant or would have consulted the brokers’ file on the claimant about its insurance requirements and other matters, including security at the premises.

I was not convinced by Mr White’s evidence about the meeting. Mr Phillips was more realistic. He accepted that other matters not set out in his note had been discussed at the meeting, although he had no specific recollection beyond his note. He accepted that matters such as turnover and claims history would have been discussed. He also accepted that security at the Thurrock premises, including whether there was an intruder alarm, could have been discussed in passing. Mr Rusinek also gave evidence about the meeting. When Miss Sabben-Clare put to him in cross-examination that the discussions at the meeting had been limited to the sum insured, premium and commission, as set out in his and Mr Phillips’ notes, he insisted that there would have been a discussion about additional matters, including the security at the premises and the alarm systems in place. He went on to suggest that he would also have made it clear that the quotation given at the meeting was subject to survey.

I accept that in all probability there was a discussion in general terms at the meeting on 7 March 2006 about the security at the Thurrock premises, including the fact that there was an alarm system in place. However, I do not accept Mr Rusinek’s evidence that the quotation of premium he gave was subject to survey. If a survey was mentioned, which I doubt, it was not indicated that any insurance would be conditional upon a satisfactory survey. This is borne out by the fact that when Mr Rusinek confirmed the premiums and commissions in an email to Mr White of 27 March 2006, on the basis that the claimant and CTM Wholesale would be added to the existing Blueheath policy with effect from 1 April 2006, this was not stated to be subject to satisfactory proposal form or to satisfactory survey of the two sets of premises.

Mr Rusinek’s explanation in his evidence for not requiring a further proposal form was that these companies were becoming subsidiaries of an existing insured which had completed a satisfactory proposal form, which dealt with the financial side and standing of the risk and the moral hazard. In those circumstances, it was apparently not his or ATD’s practice to require a further proposal form to be completed. Whilst there was something slightly illogical in that practice, I reject any suggestion made on behalf of the claimant that the failure to require the claimant to complete a fresh proposal form was somehow indicative of indifference to the risk or to anything other than premium income on the part of Mr Rusinek.

Although, as I have said, Mr Rusinek said in his evidence that his quotations for insurance of the claimant and CTM Wholesale were subject to survey, they clearly were not. Again I reject any suggestion that this was attributable to an indifference to the risk on the part of Mr Rusinek. It appears from an internal note from Mr Rusinek to Mr Skidmore of 27 March 2006 that he recognised that a survey was required, but for whatever reason (possibly oversight) the quotation sent out that day was not subject to survey.

On 31 March 2006, Mr White sent a fax to Mr Rusinek confirming that, with effect from 1 April 2006, the claimant and CTM Wholesale should be added to the Blueheath policy, which should be amended to expire on 31 March 2007. The fax cover sheet stated that cover should be “as per the attached Register Sheets”. These documents which comprised the rest of the fax, set out a summary of the insurance for each section of insurance, derived from the existing Blueheath policy. Mr White accepted in his evidence that these Register Sheets served a similar function to that of a slip in the London market. The Register Sheets stated under Terms and Conditions inter alia: “Protection/Theft alarm warranty RedCARE”.

On 2 May 2006, Mr Andrew Sullivan, an underwriter in ATD’s Leeds office sent a Risk Control Survey Request to Mr Ken Thomas, an insurance risk surveyor with ATD, asking for a survey of the premises at Thurrock and Wrexham, on the basis that these were additional premises added at 1 April 2006. A survey of the Thurrock premises was undertaken on 17 May 2006 by one of ATD’s risk surveyors, Mr Joe Player. His survey report referred to Mr Sam Radford, the operations manager of the claimant having been in attendance, although it is fair to say that in evidence, Mr Radford could not really recall much of Mr Player’s visit.

Under the Theft section of his report, Mr Player reported that the main items of attraction were cigarettes and tobacco stock of which there was £1.5 million and wines and spirits of which there was £100,000. The report referred to the fact that the wines, spirits and tobacco were now all held on the first floor having recently been moved from the special secure store on the ground floor because of a lack of room. Mr Player then stated: “This area has not been fully alarm protected to the same standard as was installed on the ground floor and must be upgraded in view of the value of the stocks-see risk improvements.”

Mr Player’s risk improvements included that the intruder alarm signalling must be upgraded to Telecom Redcare GSM and that the intruder alarm protection must be extended and upgraded to provide adequate protection to the wines spirits and tobacco stocks held on the first floor. He stated that this must include high level detectors to detect entry through the roof and also vibration inertia sensors or similar protection to the perimeter walls. This report was received by Mr Sullivan on 30 May 2006. He was not happy with it, as he explained in evidence, because he considered the claimant’s security arrangements for high risk goods were inadequate but that Mr Player’s risk improvement requirements were not sufficiently hard hitting in view of the level of risk.

Mr Sullivan sent a copy of the report to Mr Ken Thomas and then spoke to Mr Thomas on both 1 and 2 June 2006. Mr Thomas said in evidence that he agreed with Mr Sullivan’s assessment that Mr Player had not got to grips with the solution to the security risk on the mezzanine floor, because he had not specified the need for a security cage, although Mr Thomas did not accept that that was the only thing which he and Mr Sullivan regarded as important. It was agreed between Mr Sullivan and Mr Thomas that Mr Thomas would conduct a further survey of the premises himself and that, in the meantime, Mr Sullivan would send an urgent email to Mr White at the brokers.

On 2 June 2006, Mr Sullivan drafted an email to go to Mr White in the following terms:

Mr Sullivan’s evidence was that, before sending that email to Mr White, he telephoned him on 2 June 2006 and discussed its contents. He said that he would not have sent such an email to a broker, seeking to impose a condition on cover, without discussing it with him first. That he did discuss it with Mr White is supported by the fact that on his copy of the email of 30 May 2006 by which he received Mr Player’s report, Mr Sullivan has written “NB e-mail dated 2/6/2006 discussed with Andrew White on 2/6/2006.” The email was then sent to Mr White on 2 June 2006 at 11.08 hours.

Mr White had no recollection that any such conversation had taken place and, whilst not denying that the email was sent, claimed in his evidence that he had never received the email. I found his evidence about this unsatisfactory and evasive. In support of the contention that he had never received the email, he said in his witness statement, not only that he had no recollection of ever seeing the email, but that it was not on the brokers’ email system and that Media Logic, who are an independent company in charge of the brokers’ email servers, checked and they could find no record of the email having arrived at the brokers. He repeated in cross-examination that Media Logic had been asked to check if they could find it and they could not. He explained that a lady called Emma Davis in the brokers’ Humberside office had been asked to contact Media Logic. This was evident from an email sent by her to Media Logic on 13 June 2008. She stated: “Bit of a shot in the dark I know. This has been deleted”. Although Mr White blustered somewhat at the point in his evidence when this was put to him, saying he did not know why she had used that terminology, it is perfectly obvious that the reason why she said that the email had been deleted is because that was what she had been told by Mr White or whoever else instructed her to contact Media Logic.

The response from Media Logic on 16 June 2008 was that it was not possible to locate the email on the brokers’ system, as deleted items had just been cleared out and various recovery options had been switched off. Despite Mr White’s protestations that the brokers had done everything possible to trace the email, that is not borne out by what Media Logic are saying. They are saying that it is not possible to trace a deleted email, not that they had actually looked for it among non-deleted items, which was the implication of what Mr White had said in his witness statement. It is striking that the brokers did not respond to Media Logic’s email and say: “no, it hasn’t been deleted, please look for it on the servers”.

Accordingly, in my judgment, contrary to what Mr White said in evidence, I consider that not only was there a telephone conversation between Mr Sullivan and Mr White on 2 June 2006 in which Mr Sullivan explained the email he was about to send, but that email was also received by Mr White, albeit that at some stage it was deleted. Of course by that email the insurers were seeking to impose what became Endorsement 6, as is clear from the bold words in the email. Leaving to one side whether the insurers were entitled in effect to impose such an exclusion unilaterally, this was plainly a serious matter which could negative the claimant’s theft cover. The contents of this email should have been communicated to the claimant immediately, but for whatever reason, they were not. This was clearly something which Mr White appreciated and about which he was embarrassed, which is no doubt why he was so keen to assert in his evidence that he had never received the email.

However as I have said, he did receive it, but for whatever reason, whether pressure of work or something else, never communicated its contents to the claimant. I should add that I was unimpressed by any suggestion that the fact that he had not communicated its contents to the claimant supported his case that he had not received it, not least because when he received Endorsement 6 in December 2006, it still took him two months before he told the claimant about it, an omission on his part to which I return below.

Mr Thomas visited the Thurrock warehouse to conduct his survey on 21 June 2006. He was met by Mr Radford, but escorted around the premises by another employee of the claimant because Mr Radford was busy. The visit lasted one and a half to two hours. There is not much between the parties about what he found during the visit. There was something of an issue as to how he gained access to the mezzanine floor. Access was restricted to managers, supervisors and the employees who actually worked picking goods on the mezzanine, all of whom would gain access via the offices at the front of the mezzanine at the same level. Mr Thomas insisted in evidence that he had got onto the mezzanine through some other entry point than via those offices. This issue may not matter very much and it may be that he gained access via a fire door at ground floor level.

Three aspects of what he observed on the mezzanine are particularly relevant for present purposes. First, there was a fenced area with wire mesh fencing along most of the northern and eastern edge of the mezzanine and then fencing running across the floor on the west side. The fencing was about 2.5 metres high. The back wall of the fenced area was the aluminium clad outer wall of the building. This left an unfenced area on the western side of the mezzanine in front of the offices which was not fenced or otherwise protected. As Mr Radford said in evidence during the day this area was used to store “picked” cigarettes, tobacco and other goods (in other words goods that had been picked ready for delivery or collection), although at night they were taken into the fenced area which was then locked. Mr Thomas was concerned that the fence was not an effective barrier to prevent thieves breaking and entering as it would be easy to cut if a thief could gain access to the mezzanine floor.

Second, Mr Thomas thought that there were no vibration sensors on the fencing on the mezzanine floor. It appears that there was guardwire along the inside of the southern external wall of the building and along the original western weldmesh wall, although he did not see it at the time, possibly because it was obscured by stored goods. If connected to the burglar alarm system this would have sounded the alarm when disturbed by movement of that fencing.

Third, there were eight movement detectors covering the mezzanine area. Mr Thomas conducted a walk test and some of the lights on the detectors were not operating. He formed the view that these detectors did not fully cover the mezzanine area. However, significantly, when it was put to Mr Thomas by Miss Sabben-Clare that the fact that some of the detectors did not come on had nothing to do with the area they were covering, but showed that they were not working full stop, he accepted that as correct. He did not suggest in his evidence that the reason why the lights did not illuminate on some of the detectors was that the detectors were angled upwards and, indeed, gave no evidence at all about whether the detectors were angled upwards or downwards during his visit.

The extent to which these detectors did cover the area of the mezzanine (or would have done if properly angled) remains one of the issues in the case. The mezzanine (including the unfenced area to the west) was about 32 metres long and 8 metres wide. Although Mr Thomas did not identify the precise location of the detectors in his report, that is essentially common ground and can be summarised as follows. The detectors were located along the south external wall on or adjacent to the upright stanchions at about three metres in height. Detector 59 was on the west side of the stanchion closest to the offices, facing in that direction, with detector 60 on the north side of the same stanchion, facing towards the unfenced area on the west side of the mezzanine. Detector 71 was on the east side of the next stanchion moving eastwards. Since the wire fencing as it then was ran from just to the west of that stanchion in a northerly direction, it follows that detector 71 was within the original fenced area. This may be of some significance in determining to what extent the claimant complied with the Risk Improvement Requirements.

Detector 70 was then to the east side of the stanchion two to the east of the one on which detector 71 was located. There was no detector on or adjacent to the stanchion between the one on which detector 71 was located and the one on which detector 70 was located. Detector 69 was located on the east side of the next stanchion eastwards and detector 68 on the west side of the stanchion after that, which was on the corner before a slight return wall running north-south. Detectors 60, 71, 70, 69 and 68 thus faced in a generally northerly direction across the mezzanine. Another detector, 67, was around the corner of the same stanchion as detector 68, with detector 66 also on the same north-south wall.

Mr Thomas’ evidence was that after his tour of the warehouse, he went back and explained his concerns to Mr Radford, telling him that the security was unacceptable and that he would be speaking to insurers and recommending risk improvements which would be sent to the brokers. He said that Mr Radford’s reaction was one of disinterest. Mr Radford did not recall having any specific issues pointed out to him and thought it unlikely that they had been. All he recalled was being told by Mr Thomas that a report would be sent to head office. It is clear that Mr Radford did not take to Mr Thomas, whom he described as uncommunicative and aloof about his status within the claimant’s business. Having observed Mr Thomas’ demeanour in the witness box, I can see how he might have come over as uncommunicative and aloof.

Ultimately, it may not much matter who is right about this, since it is not suggested that Mr Thomas’ survey report was sent to the claimant and it is common ground that the claimant did not receive the Risk Improvement Requirements Mr Thomas set out, until they were sent by Mr Phillips of the brokers to Mr Paget of Blueheath under cover of a letter dated 15 December 2006. However, to the extent that it matters, I prefer Mr Radford’s evidence about this and consider that Mr Thomas did not spell out his concerns to Mr Radford. Furthermore, I formed the view that Mr Radford was a conscientious employee and that if Mr Thomas had spelt out his concerns, Mr Radford’s reaction would not have been one of disinterest. Rather he would have wanted to communicate those concerns to more senior management.

Following that visit Mr Thomas prepared his Risk Survey Report and a list of Risk Improvement Requirements. Although the Report is dated 17 July 2006, his evidence was that he had visited Mr Sullivan at ATD’s offices in Leeds on 4 July 2006. In his witness statement he did not recall the discussion they had but referred to Mr Sullivan’s witness statement where Mr Sullivan recalled Mr Thomas saying that a copy of the Risk Improvement Requirements had been sent to the brokers. In cross-examination Mr Thomas purported to recall producing the Risk Improvement Requirements and sending them in hard copy to Mr White of the brokers before 4 July 2006.

I was not impressed with that evidence, which, if correct, would have involved a surprising state of affairs whereby the insured’s broker received a copy of the Risk Improvement Requirements before they were received by the underwriter who was effectively the surveyor’s client. It is clear from the disclosed documents that the Risk Improvement Requirements were not sent to Mr Sullivan until 17 July 2006, when Mr Thomas sent the survey report and Risk Improvement Requirements to Mr Sullivan under cover of an internal memorandum of the same date to Mr Sullivan in Leeds. In that memorandum he referred to the fact that his risk improvements in place of those of Mr Player required a security cage to be built as well as upgrading the security alarm.

It is noteworthy that Mr Sullivan has written on his copy of the memorandum (no doubt following a telephone call with Mr Thomas on 18 July 2006): “NB Ken Thomas has sent a copy of the Risk Improvements to Andrew White. AM Sullivan 18/7/2006”. This suggests that the Risk Improvement Requirements were not sent to the brokers until on or about 18 July 2006. Mr White’s evidence was that he did not recall any conversation about the Requirements and did not receive them until 15 December 2006. However, I reject that evidence and conclude that the Risk Improvement Requirements were sent to the brokers by Mr Thomas on or about 18 July 2006, as is reflected in Mr Sullivan’s manuscript note.

The Risk Improvement Requirements included items about electrical equipment and installation, the safe and upgrading of the alarm signalling which had been in Mr Player’s Requirements, but went beyond Mr Player’s Requirements in significant respects. I should quote the relevant section of the document in full:

As those closing words suggest and as Mr McAndrew, who was the underwriter at ATD who dealt with the lifting of Endorsement 6, indicated in his evidence, in practical terms what one would expect to happen, if an insured was sent a list of Requirements such as this, is that the insured would consult its security company or alarm company as to the feasibility of what was required. If the Requirements were not feasible or too expensive, a dialogue would take place with underwriters as to whether certain Requirements could be varied or waived, as to which underwriters might consult the risk surveyor and then either agree such a variation or waiver or insist upon the Requirements being complied with.

In the present case that dialogue only took place to a very limited extent in February 2007 in relation to the ceiling of the cage. The reason for the absence of any further dialogue is not hard to discern: the brokers did not send the claimant the Risk Improvement Requirements until 15 December 2006. Even then the claimant would not have appreciated quite how important and significant these were because it was not told about Endorsement 6 at the same time and, indeed, was not informed about that Endorsement until it was set out in an email from Mr White to Mr Paget of 9 February 2007, to which I return below.

Mr Graham Copper acted as consultant Operations Manager of the claimant from the August bank holiday weekend of 2006 until the end of October 2006. Mr Copper was not a security expert as such, having learnt about security through his experience in his working life. He knew the then CEO of Blueheath, Mark Aylwin, who asked him to work at the claimant to overhaul its operational procedures. Part of this involved overseeing improvement works to the warehouse for which he was given a budget by Mr Aylwin of approximately £70,000. Health and safety was the priority as he explained in his witness statement, but that led directly to security improvements. Specifically, because the weldmesh fence on the mezzanine was only waist high in places, there was a risk of someone falling from the mezzanine to the warehouse floor below. Increasing the height of the fence not only made it safer but more secure.

The security improvements which he undertook and his evidence about them was as follows. The weldmesh fencing on the mezzanine was extended upwards so that it was at a consistent height throughout of three metres or so from the mezzanine floor and, on the western side, the fence was moved to the edge of the mezzanine, so that the whole mezzanine floor was now a caged area. In his witness statement Mr Thomas had said that the fencing installed was not weldmesh, but he abandoned that contention at the beginning of his oral evidence and accepted that it was of weldmesh construction.

Mr Copper said that he considered the need for a ceiling on the caged area because he recollected that at some stage around the time when he was working at Thurrock, someone had mentioned the installation of a roof. However, he decided that with the wall height, there was sufficient security without a ceiling. He also said that he did not arrange for a weldmesh wall to be installed along the inside of the external wall. His opinion was that the external wall was made of aluminium sheeting which was stronger than weldmesh. This opinion was based on common sense; he assumed that the external wall consisted of a semi-corrugated inner layer, then an insulation layer, then an outer layer. Although he was not 100% sure about that, the thickness of the outside wall was what he would have expected. He accepted that you could get an angle grinder through those layers but considered it would be difficult to do.

He also installed two security gates in the western and eastern weldmesh walls close to the northern edge of the mezzanine. Before their installation there was a form of pivotal gate which could be used for lowering or raising pallets from a fork lift truck on the ground floor below. However, he said that until the sliding gates were installed, if someone climbed up or used some means to get to the mezzanine level, there was no secure fence in way of the pivotal gate. Hence the installation of the sliding gates, so that outside business hours when they were shut, there was a continuous weldmesh wall including the closed gates.

He installed a further 10 metres of guardwire along the new western wall of the caged area and on the return on the south side where the gate into the new caged area from the offices was. The reason for installing the guardwire along the western side was that although there was a drop from the mezzanine to the ground floor, there was racking on the ground floor where goods were stored, which was in proximity to the western edge of the mezzanine. Accordingly, there was a risk of someone climbing up the racking to get to the height of the mezzanine floor. The guardwire was to prevent such a person from then breaching the weldmesh.

He did not install equivalent guardwire along the eastern or northern walls, because there was no racking in the vicinity of those walls up which an intruder could climb. It was put to Mr Copper by Mr Turner that a breach in the weldmesh in those areas might have been possible by intruders using forklift trucks. Mr Copper said that it would not be possible to use a forklift truck to climb over the top of the fence because such a truck was only tall enough to get to the level of the secure gate, but he accepted that it would be possible to use a forklift truck to get to the level of the mezzanine floor and then make a breach in the weldmesh. It seems to me that must be right and indeed it was that possibility which had led him to install the sliding gates.

So far as the guardwire which had run along the original western wall of the caged area was concerned, he said that the wall was left in place as long as possible to ensure as much security as possible during the upgrade works. When that wall was dismantled the guardwire was disconnected from the weldmesh, then the guardwire was extended and reconnected to the new western weldmesh wall. Initial’s engineers were on-site to carry out this process. Mr Copper could not say whether it was then connected or activated at the alarm panel, although he assumed it had been, which seems to me a fair assumption given that such re-connection and re-activation would be the responsibility of the Initial engineers.

Mr Copper recalled a discussion with one of the Initial engineers regarding the positioning of one of the motion detectors on the mezzanine level, the engineer suggesting that it be re-positioned to provide better cover over the extended mezzanine area. In his oral evidence Mr Copper recalled the two detectors (59 and 60) which had originally been outside the caged area but which were now within the extended caged area were directed to ensure that there was as much coverage as possible of the area, including looking north and south west. It seems therefore that it was to one of these detectors that the Initial engineer was referring.

Mr Copper also said in his witness statement that the movement detectors were angled and positioned by the Initial engineers and he was not involved in any discussion as to how that should be done. He recalled that Initial had suggested additional detectors but could not recall where they were to be located. In cross-examination he could not say whether these were to be located on the mezzanine or elsewhere in the warehouse. He said that he had to balance the suggestion of additional detectors against his budget. He was satisfied that there were sufficient detectors to cover the mezzanine area and that it was secure given the other measures. It is important to emphasise at this stage that whilst these security upgrade works were being carried out in September and October 2006, no-one at the claimant, including Mr Copper, had a copy of Mr Thomas’ Risk Improvement Requirements or even knew about them. Accordingly, there was no opportunity to discuss Mr Thomas’ requirement 2006/02 with Initial.

Mr Copper also gave evidence about ironing out various glitches with the CCTV camera system and supervising upgrading and re-wiring work by Security Design Services (SDS) with whom the claimant had a contract to monitor that the system was working. The claimant also had a contract with UK Business Watch (UKBW) to monitor the CCTV footage remotely from off site, particularly outside business hours. SDS provided new cameras and other equipment, including an ADSL line (i.e. a broadband connection) which carried the live feed on the CCTV cameras to UKBW’s central monitoring system. This was installed on about 28 September 2006. I will return to the significance of that in the context of the allegation by the insurers of breach of warranty by virtue of the ADSL line having been down between 13 February 2007 and 16 March 2007.

So far as the CCTV system was concerned, there were cameras around the outside of the building and protecting the ground floor interior of the warehouse. In relation to the mezzanine, there was a camera (camera 11) mounted on the same stanchion as detectors 59 and 60 but facing eastwards down the length of the mezzanine. There was also another camera, number 15, apparently on a rotating globe on the stanchion on which detectors 67 and 68 were located. That camera seems to have protected the northeast corner of the mezzanine.

Although the claimant was added to the insurance with effect from 1 April 2006, no policy schedule to that effect was produced by ATD until December 2006. Although, in his cross-examination of Mr White, Mr Turner was critical of the brokers for not chasing and, in her cross-examination of Mr Sullivan, Miss Sabben-Clare was critical of ATD for not having prepared the schedule sooner than they did, it seems to me nothing turns one way or the other on this delay. It is perhaps rarer to encounter such delays in the modern world with tighter regulatory regimes, but certainly until a few years ago delays of this kind or even longer in production of policy wording were all too common.

Mr McAndrew sent Mr White the Policy Schedule under cover of a letter of 4 December 2006, apparently received by the brokers on 15 December 2006. The Policy Schedule included Endorsement 6 which Mr McAndrew said in evidence he had added to confirm the email of 2 June 2006 that Mr Sullivan had written. Mr McAndrew’s letter of 4 December also said:

In the brokers’ reply dated 15 December 2006 Mr Phillips said that there were one or two queries which he needcd to resolve before issuing the documentation to the claimant, specifically:

In his letter of the same date to Mr Kevin Paget, Group Financial Controller of Blueheath, Mr Phillips refers to the Risk Improvement Requirements which he encloses and asks Mr Paget to let the brokers have confirmation when the necessary work had been carried so that they could notify the insurers accordingly. Mr Phillips did not inform Mr Paget about Endorsement 6, a surprising omission.

Mr McAndrew responded to Mr Phillips’ letter of 15 December 2006 after the Christmas and New Year break, on 3 January 2007 in these terms, so far as relevant: “Endorsement 6 remains as the condition was applied with immediate effect as per Andrew Sullivan’s email to Andrew White of 2^nd June 2006”. From that response it is clear that far from being indifferent to the risk as Mr White suggested in his witness statement, Mr McAndrew was concerned about the security on the premises and was firm in his approach that the Endorsement should remain in place until the Risk Improvements required were in place.

Mr Phillips’ evidence was that at around this time he had asked Mr White about the email of 2 June 2006, which Mr White said he had not received. If that evidence were true, it is surprising to say the least that Mr Phillips did not respond to Mr McAndrew’s email by saying that the brokers had not received the email of 2 June 2006. In fact Mr Phillips did not respond directly or challenge Mr McAndrew about the imposition of Endorsement 6. I consider that his evidence about having been told by Mr White that he had not received the email of 2 June 2006 was an example of Mr Phillips being prepared to support Mr White’s evidence in a manner which was thoroughly unconvincing.

On 8 January 2007, Mr Phillips received from ATD an email which was one in a line of chasers about the alarm specification which had never been forthcoming from the insured. The email also asked about compliance with the Risk Improvement Requirements. As I mentioned above when referring to events after the proposal form was received from Blueheath, this request originally related to the Islington premises. However, when Mr Phillips received this email, he interpreted it as referring to the Thurrock warehouse, as he accepted in his evidence. This is also apparent from the letter he sent to Mr Paget on 12 January 2007 which stated:

Pausing there just to consider what was happening at the claimant, Mr Cloke had taken over as operations director in November 2006 after Mr Copper left. Mr Cloke’s evidence was that shortly before Christmas 2006 (before he saw the Risk Improvement Requirements which he said he did not see until January 2007) he conducted a walk test of the motion detectors on the mezzanine floor. He said that he found no dead spots when he conducted this test and that he did not notice whether or not the detectors were angled upwards. In his first witness statement he accepted that this test was less thorough than the one he conducted immediately after the theft when, although all the motion detectors were working, there were small dead spots in which it was possible to stand without activating the detectors. He could not say whether those dead spots had been present during the December walk test.

His oral evidence was that he received the Risk Improvement Requirements from Mr Paget who dealt with insurance matters in January 2007 and that, having read these, he then conducted another walk test, having picked up that Mr Thomas was saying that the lights had not illuminated on some of the movement detectors during Mr Thomas’ walk test. Mr Cloke claimed to have walked along the warehouse quite close to the detectors (although he accepted that because of the racking along the south wall he would have been a metre and a half away from the detectors) and then walked back a bit further away from the detectors and that all the detectors activated. This additional walk test in January was not mentioned in his Incident Investigation Report in May 2007 or in either of his witness statements and, albeit with some reluctance, I have concluded that this evidence about a second walk test after he received a copy of the Risk Improvement Requirements was not true.

In my judgment, the only walk test Mr Cloke conducted before the theft was the one in December 2006, which on his own evidence was before he received a copy of the Risk Improvement Requirements. As to why Mr Cloke has invented this second walk test, the most likely explanation would seem to be a concern on his part that the walk test he conducted in December 2006 may have been somewhat superficial and may not have been what Mr Thomas was contemplating in his Risk Improvement Requirements.

It was initially suggested by Mr Cloke in his witness statement that the Risk Improvement Requirements were vague and open to interpretation. However he accepted in cross-examination that if there was anything which required clarification, he would have sought such clarification via Mr Paget, but that he had not done so, apart from in relation to the issue of the ceiling of the cage. The fact that he did not seek any further clarification suggests that he did not in fact find the Risk Improvement Requirements vague or difficult to understand.

In relation to 2006/02 and the requirement for additional movement detectors in the mezzanine area where the cigarettes and tobacco were stored, his evidence was that he had considered this and thought that, as a consequence of the movement westwards of the weldmesh fencing, two additional detectors had been brought into that area, evidently a reference to detectors 59 and 60. In relation to the requirement for vibration sensors on the weldmesh walls, he realised that only the western wall had the guardwire. He said that he had discussed all the Risk Improvement Requirements with Mr Paget and that they had had a conversation about the height of the walls and the fact that there was a 10 metre drop to the ground floor below, so that it was not really necessary to have vibration sensors. The insurers did not require Mr Paget to attend for cross-examination, but he does not refer to any such discussion in his witness statement. All that he said in his witness statement was that after receipt of Mr Phillips’ letter of 12 January 2007, he thought he would have contacted Mr Cloke or Mr Radford for an update, but he did not recall. This is borne out by a telephone attendance note of 22 January 2007, stating that he told Mr Phillips that the Risk Improvement Requirements had been passed to Thurrock and that he would chase the relevant people and come back to the brokers, particularly in respect of the security cage.

On 5 February 2007, Mr White spoke to Mr Paget who informed him that a lot of work had been completed by the claimant, including the fact that the cigarettes were now in a cage (information which in all probability Mr Paget got from Mr Cloke). It was agreed that it would make sense for insurers to resurvey the premises to confirm everything was ok. On 8 February 2007, Mr White passed this information to Mr McAndrew with a view to having Endorsement 6 lifted, but although Mr McAndrew agreed to organise a resurvey as soon as possible, he said that in the meantime endorsement 6 would remain in place. Mr McAndrew’s evidence was that, in his experience, a resurvey would be organised to check what an insured has actually done.

It is accepted by the claimant that whatever might have been the status of the endorsement prior to that telephone conversation on 8 February 2007, it was in that conversation that it was agreed by the brokers on behalf of the claimant that Endorsement 6 was a term of the contract of insurance.

It is striking that it was not until after that conversation that the brokers informed the claimant for the first time about Endorsement 6. This was in an email from Mr White to Mr Paget of 9 February 2007 which stated that:

This provoked a telephone discussion the same day with Mr Cloke who informed Mr White that whilst the wines and spirits were kept in a secure area on the ground floor, the cigarettes and tobacco were kept in a secure area on the mezzanine floor. Mr White accepted in evidence that he was aware following that conversation that the claimant could not comply or was not complying with the exclusion in Endorsement 6 and that at that point he would have been in a state of anxiety. This is perfectly understandable given that, as matters stood, the high value cigarettes and tobacco at his client’s premises were not insured outside business hours.

Mr White then sent an email on 12 February 2007 to what he thought was Mr McAndrew’s email address, but it was mistyped so the email never reached him. This stated that: “I….would advise that Endorsement 6 relating to the West Thurrock premises needs to be amended, so that it reads that target stock has to be kept in the secure areas, either on the ground floor or on the mezzanine floor, outside business hours”. Obviously, since this was misdirected, Mr McAndrew never received it, so that two days later on 14 February 2007, Mr White had not received a response.

At 12.27 hours on 14 February 2007, Mr Paget sent Mr White and Mr Phillips an email setting out the position in relation to the Risk Improvement Requirements. His evidence in his witness statement was that so far as the position at the claimant was concerned, he cut and pasted this from information provided by Mr Cloke. So far as relevant, the email provided:

At 16.56 hours on the same day, 14 February 2007, Mr McAndrew sent Mr White an email in these terms:

An issue has arisen as to whether the reference to an “earlier telephone conversation” was to a telephone conversation earlier that day during which Mr White told Mr McAndrew the substance of what Mr Paget had said in his email earlier that day, not just in relation to 2006/01 but 2006/02. The reason why this matters is because it is the claimant’s case that, from this email onwards, all Mr McAndrew was interested in was the status of the cage and not compliance with the other Risk Improvement Requirements. The rival explanations for this are (on the part of the claimant) that it was only receiving information about the cage that induced Mr McAndrew to lift Endorsement 6 and (on the part of the insurers) that 2006/02 was not specifically mentioned in this email or subsequent emails sent by Mr McAndrew because he had received confirmation from Mr White, probably during this telephone conversation on 14 February 2007, that the Risk Improvement Requirements in 2006/02 had been complied with.

Mr McAndrew’s evidence in relation to the “earlier telephone conversation” was that he and Mr White had obviously had a telephone conversation that day and that he was reconfirming that if Mr White confirmed in writing what had been carried out by the insured then he would be considering providing cover. It is fair to say that he was not specific that it was in that conversation that Mr White told him that 2006/02 had been complied with. However, in my judgment, considering his evidence about the lifting of Endorsement 6 as a whole, he said more than once that he had been advised that 2006/02 had been complied with and an obvious candidate for when he was told this is on 14 February 2007, after Mr White received the email from Mr Paget.

Mr White accepted in evidence on Day 2 of the trial that before Mr Paget’s email he had tried on a number of occasions to persuade Mr McAndrew to remove Endorsement 6 but that Mr McAndrew was implacable. He then gave the rather surprising evidence that when he read the word “Actioned” he would have understood that to mean the same thing as “Completed” (to which I will return below) and that this email from Mr Paget was good news which he would have been eager to share with ATD. He also accepted that the reference to “an earlier telephone conversation” in Mr McAndrew’s email of 14 February 2007 “must be” to a telephone conversation earlier the same day. Although at first he did not think that in that telephone conversation he would have passed on that good news in Mr Paget’s email, he then accepted that he would have passed on all the information in Mr Paget’s email.

I propose to set out this passage in his evidence in full:

In terms of Mr White’s evidence what then happened is that on 11 November 2009, which was Day 3 of the trial and the day after Mr white had given evidence, a third witness statement from him was signed, seeking to retract his evidence that he had had a telephone conversation with Mr McAndrew on 14 February 2007 and stating that, having checked his diary for that day, he was now certain he had not had such a telephone conversation that day. He was recalled for further cross-examination on 16 November 2009, Day 5 of the trial.

The thrust of the evidence he now gave as set out in that witness statement was that, having consulted his diary on 14 February 2007, he had been due to have a meeting in Barnsley at 10 am with a client, a Mr Fred Brooker, who ran a building company, but that he had had carpet fitters at home that day, so that he had pushed back the meeting with Mr Booker. He would not have left home until late morning, it would have taken an hour to drive to Barnsley and the meeting would have lasted at least a couple of hours. He would then have had some lunch and left Barnsley about 3pm arriving at his office about 4pm. Since he was leaving the office that day to go on holiday for several days, he was sure he would have left the office by 5pm at the latest, because his partner’s parents were due to arrive to stay over before going on holiday together. He would have been in the office sorting things out, not picking up new tasks, the implication being that he would not have bothered with Mr Paget’s email. He was now sure that he had not had a telephone conversation with Mr McAndrew on 14 February 2007.

There were a number of flaws in this reconstruction of events which were exposed in cross-examination. First as Mr White said, Mr Booker was fairly elderly and he did not arrange insurance meetings lightly, because he had other things to do. This was not some routine meeting such as for renewal, but a mid-term meeting which Mr White said must have been arranged to discuss an important issue. It seems to me inconceivable that Mr White would have put back the meeting with Mr Booker for his own domestic personal convenience. I reject any suggestion that the meeting was put back.

Second, there was some discussion between Mr Turner and Mr White about journey times, Mr Turner putting to Mr White various matters by reference to “Google maps”. Whilst the Google journey times may have been on the optimistic side, not really allowing for traffic congestion, I considered that Mr White had exaggerated the length of time the journeys would take.

Third, although in the witness statement he had rather given the impression of a leisurely lunch in Barnsley, in cross-examination he accepted that since he was going on holiday the next day, he would have wanted to get back to the office and would probably have got a sandwich from the White Rose shopping centre and taken it back to his office. On this basis it seems to me that in all probability, he was back at his office soon after 1pm and certainly no later than 2pm.

Fourth, as he accepted in cross-examination the resolution of the issue concerning Endorsement 6 was not a new task on 14 February 2007. He accepted that there was a long gap in which the brokers had not told the claimant about the Endorsement and, as he eventually accepted, in view of the fact that on 9 February 2007, when told of the Endorsement, Mr Cloke had made it clear the claimant could not or at least was not complying with the Endorsement, it was an urgent imperative, “very very important” as he put it, to get the Endorsement lifted as soon as possible. In view of that admission, although he still maintained that there had not been a conversation with Mr McAndrew on 14 February 2007, I simply cannot accept that change of evidence.

In my judgment, once Mr White went back to his office in the early afternoon and checked his emails, which he would almost certainly have done, even if (which he could not now remember) he did not have a Blackberry, he would have seen the email from Mr Paget. The urgency of the situation (particularly since the absence of any response from Mr McAndrew to Mr White’s misdirected email of 12 February 2007 must have concerned Mr White) was such that Mr White would have wanted to convey this good news to Mr McAndrew on the telephone as soon as possible. This is what he had ultimately accepted he had done when cross-examined on Day 2 of the trial and that is what I find he did. Whether at that time he really believed that “actioned” was synonymous with “completed” is open to question (for reasons set out hereafter), but I suspect that, in his eagerness to get the matter resolved, he would have left Mr McAndrew with the impression that 2006/02 had been complied with.

As to why Mr White sought to change his evidence, he said when recalled that he had been very tired during the cross-examination on Day 2 and had personal problems. Whilst all that may be true, it does not begin to explain the complete volte-face. I consider that the real reason was an appreciation on Mr White’s part, after he had given evidence on Day 2, that as a consequence of his admission that he had had a telephone conversation with Mr McAndrew on 14 February 2007 in which he relayed all the information in Mr Paget’s email, he had assisted the insurers’ case on misrepresentation and inducement, something he was not anxious to do, as demonstrated by the lengths to which he was prepared to go in criticising the underwriters in his first witness statement.

This case has undoubtedly placed Mr White under a great deal of strain personally, as he accepted at the outset of the further cross-examination on Day 5. It is not necessary to decide why that is, but at least in part it is likely to be as a consequence of a concern about the brokers’ failure to inform their client of the imposition of Endorsement 6 for eight months from June 2006 until February 2007. This meant that the constructive dialogue between insurer and insured through the broker concerning security measures at the premises, which Mr McAndrew recognised as what would often happen in practice, simply did not take place.

It appears that on 16 February 2007, someone at the brokers (although he could not remember if it was him, this must have been Mr Phillips since Mr White was away on holiday) spoke to ATD (probably Mr McAndrew) about the fact that the cage did not have a ceiling. The insurers asked whether the walls went to the ceiling of the warehouse or whether it was possible for intruders to climb over the sides of the cage to gain entry to the secure area. This question was relayed to Mr Paget in an email from Mr Phillips on 16 February 2007. Mr Paget forwarded the email to Mr Cloke for response.

Mr Cloke responded the same day, stating:

This response was forwarded to Mr Phillips in an email of 16 February 2007 and he informed Mr White. On 20 February 2007, Mr White then had a telephone conversation with Mr McAndrew, the substance of which was recorded by Mr White in a handwritten note on a copy of the email string:

Miss Sabben-Clare put to Mr McAndrew that he had been saying that the deal was, tell me as much as you can about this cage, if you do that we will amend the endorsement, the implication of this question being that by this stage all he was interested in was the cage, not compliance with 2006/02. Mr McAndrew’s response clearly refuted that implication:

Following that telephone conversation, Mr Phillips sent Mr McAndrew an email at 11.23 hours on 20 February 2007 in the following terms:

It was in response to this email that Mr McAndrew sent an email at 15.15 hours on 20 February 2007 confirming that the wording would be amended to lift Endorsement 6, albeit on terms:

When, by reference to this email, Miss Sabben-Clare put to Mr McAndrew that it was the nature of the caged area on the mezzanine which was important to him, he said:

Miss Sabben-Clare then put to Mr McAndrew that, as his agreement to a cage demonstrated, what had mattered to him was that he was satisfied about the security arrangements, not dotting the Is and crossing the Ts of the risk improvements. His answer was:

Sorry, my Lord. I believe the risk improvements are all interlinking. So I agreed to delete endorsement six even though the cage did not have a roof on it because I was under the understanding that the alarm had been extended as per risk improvement, I think it was 2006/02, with the additional sensors.”

I accept that evidence, which it seems to me is borne out by the terms of Mr McAndrew’s email. The reference to “where the cage on the mezzanine does not meet our survey requirements (other than the lack of ceiling)” goes beyond the physical structure of the cage and includes requirements a) and b) in 2006/02.

I will deal with the question whether the statement that 2006/02 had been complied was substantially true or not in the later section of the judgment where I deal with misrepresentation, but I should just deal with the change in wording between Mr Paget’s email at 12.27 hours on 14 February 2007 which referred to 2006/02 as “actioned” and Mr Phillips’ email of 20 February 2007 which referred to it as having been “complied with”. As I see it, as a matter of ordinary language, there is a difference between “actioned” and “complied with” or “completed”, the former importing that something will be dealt or complied with or is in the course of being dealt or complied with, rather than that it has been complied with. The difference between “actioned” and “completed” is also apparent from the text of Mr Paget’s email, which, when something has been completed says so, for example in relation to 2006/05.

Furthermore, the fact that “actioned” did not mean completed was recognised in Mr Phillips’ witness statement where, referring to Mr Paget’s email of 14 February 2007 he said:

However, the other evidence called by the claimant on this issue was unsatisfactory. I have already referred to Mr White’s assertion in cross-examination that he had always considered “actioned” meant the same thing as “completed”. That evidence was frankly incredible and was belied by the fact that soon after he gave that evidence, he used the word “actioned” in re-examination in the sense of “will be dealt with” rather that “has been dealt with” or “completed”. He was asked what he would have done if he had received the email of 2 June 2006 and said they would have “actioned it” by saying to insurers in effect that the Endorsement could not be imposed.

Equally surprisingly, in the light of what he had said in his witness statement, Mr Phillips said in cross-examination that at the time he did not see any difference between the two words “actioned” and “completed”. When he was asked to explain how in the light of that evidence, he could explain the first sentence of the paragraph from his witness statement which I have quoted above, he could not explain it. In my judgment his evidence in cross-examination was another example of his making things up to support Mr White.

Perhaps most surprisingly of all, Mr Cloke who was, after all, the author of the text set out in Mr Paget’s email, said in cross-examination that when he said “actioned” he meant “completed”. If this evidence had really been true, it is a complete mystery why elsewhere in the same text he refers to Risk Improvement Requirement 2006/05 as “completed”.

It is also striking that this evidence is inconsistent with what the claimant’s solicitor, Miss Georgina Squire of Rosling King said, presumably on instructions, in a witness statement dated 3 October 2008 served in opposition to the insurers’ summary judgment application. Referring to the email from Mr Paget which had said 2006/02 had been “actioned” she said: “This supports the claimant’s contention that the [brokers] did not tell ATD that the Risk Improvements had been completed. It is unlikely that [the brokers] said this when that was not what they had been told.”

In my judgment, all three of the claimant’s witnesses (Mr White, Mr Phillips and Mr Cloke) were being less than frank with the court on this point. In the case of the two witnesses from the brokers, the motive for this may be obvious: to avoid any criticism for having overstated the true state of affairs. In the case of Mr Cloke the motive is less obvious, but it may be that he too wanted to avoid the suggestion that the brokers had misrepresented the position to the insurers.

One of the matters alleged by insurers to have been a breach of the Protection Maintenance Warranty is the fact that the ADSL line providing the link for the remote monitoring of the CCTV system was disconnected from 13 February 2007 to 16 March 2007. Before dealing later in the judgment with whether this was a breach of warranty, I should set out my findings as to how that disconnection came about and whether steps were taken “promptly” by the claimant to remedy this defect.

It is apparent from the claimant’s phone bill and BT’s records that BT thought that it had disconnected the copper cable or land line (without which the ADSL line cannot operate) on 31 January 2007, for non-payment of a bill. Mr Mindham’s evidence, although he was not directly involved, was that he remembered being told by one of the accounts clerks in the claimant’s finance department, probably in early February 2007, that there had been an issue with BT about billing which was “down to BT” and that, once the accounts staff learnt of the problem, they dealt with it immediately. Although he had not been directly involved, he trusted the accounts staff and had no reason to suppose that they were misleading him. That the matter was dealt with immediately is borne out by the fact that, according to the claimant’s phone bill and BT’s own records, the land line was reconnected on 8 February 2007.

Mr Cloke’s evidence was that the problem with BT was that a direct debit was not collected by BT. This is certainly consistent with Mr Mindham’s recollection of having been told that the problem was down to BT, but inconsistent with the claimant’s own telephone bills, which do not suggest that a direct debit was in place. Perhaps one was in the process of being set up, but that is speculation on my part and the claimant did not call evidence such as would enable the court to resolve the issue as to the cause of the problem.

Fortunately, it is not necessary to resolve that issue because, despite what the phone bills and BT’s records say, the land line was never in fact disconnected, between 31 January 2007 and 8 February 2007 or at all. This is apparent from the email records of the ADSL line. As part of the service provided to the claimant by SDS, whenever the ADSL went down or came back up again, an automatic email was generated, which was sent to SDS and thus recorded in their system. It is essentially common ground that this is the best record available of when the ADSL line was operating. Mr Heath conducted an analysis of those emails and prepared a summary which shows that, albeit with periods of downtime, the ADSL line was operational throughout the period from 31 January to 8 February 2007.

Since the ADSL line could not operate unless the land line was connected, it necessarily follows that the land line was never in fact disconnected between those dates. As to why BT’s records, reflected in the phone bills, showed that it was disconnected between those dates, Mr Collins, the BT employee who gave evidence, advanced the tentative opinion, which it seems to me is likely to be right, that although the records might show a disconnection, it may be that an engineer had not in fact gone to the exchange to disconnect the line before the issue with the bill was resolved, so that in fact the line was never physically disconnected.

What the SDS email records do show is that the ADSL line was disconnected between 13 February and 16 March 2007. Although the insurers maintained their case that this was a consequence of the non-payment of the BT bill, I have some doubt whether this was so, given that (a) the BT line was never in fact disconnected and (b) the problem with BT seems to have been resolved by 8 February 2007. It is true that in an email of 1 June 2007 explaining what had occurred, Mr Cloke refers to a direct debit having been set up and BT having failed to request a collection and the claimant’s accounts department not learning of this he thought until the week commencing 19 February, but this simply does not tally with BT’s records which do not show any problem, let alone a disconnection, after 8 February. It seems to me that the disconnection of the ADSL line may have been attributable to some other unidentified problem.

At all events, whatever the cause, it would seem that UKBW picked up that the line was down soon after it occurred and notified the claimant. In his email of 1 June 2007, Mr Cloke says that line problems were notified to the claimant some time in the week commencing 12 February. There is then something of a gap in the claimant’s evidence, since Mr Butler, who dealt with IT at the claimant, did not become aware of the problem (according to Mr Cloke’s email) until the week commencing 19 February 2007. Mr Butler’s evidence was that UKBW would have contacted Mr Ted Sergeant at the claimant, who would then have contacted him. Mr Sergeant was not called to give evidence.

Mr Butler thought that it was likely that he learnt of the problem on 23 February 2007, which was the day he sent an email to Mr Critchlow of SDS saying that the link allowing the security system to be monitored from an external internet connection had been lost. He thought that he spoke to Mr Critchlow over the telephone soon after that. On 28 February 2007, he sent Mr Critchlow an email asking him to send the link that allowed for external monitoring. In that email he also said that in relation to the billing issues with the PST number (i.e. the land line) he had been told by accounts that everything was up to date and that there should not be any outstanding issues. He said in evidence that he had tested the land line by ringing it and it was connected.

On 1 March 2007, Mr Critchlow responded to this email saying that he believed the line had been cut by BT but now that everything was up and running again, SDS would have to re-apply to effect the broadband, which they had done that day. On 8 March 2007, Mr Critchlow told Mr Butler that broadband would be live again on 15 March. Mr Butler’s evidence was that it normally took about five days to reconnect a BT land line but fourteen days to reconnect an ADSL line. In the event, the ADSL line was operational again on 16 March 2007. On that day, Mr Butler was contacted by SDS to say the line was up again. He physically restarted the router, which was in a cabinet with other equipment in the transport office at the claimant’s premises. He was certain that the ADSL line was active when he restarted the router as there was a continuous green light on the router.

The intruder alarm was set on the evening of Friday 16 March 2007 by Jamie Chandler, the warehouse supervisor, when the staff left at 22.00 hours. On the afternoon of 17 March 2007 at 16.55 hours one of the employees, Dean Champion, attended the premises to drop off a vehicle key and unset the alarm. He was only in the premises for a matter of minutes and reset the alarm. His evidence, which was agreed without any need for him to be called, was that if there had been a problem with any section of the alarm, this would be apparent from the alarm panel. Mr Cloke said that any fault would have to be rectified before the alarm could be reset. The alarm was fully reset on both these occasions, indicating that no fault was showing on the alarm system.

The circumstances of the theft were reconstructed by Mr Cloke and then Mr Heath of Hawkins from the CCTV footage viewed after the theft. At about 01.22 hours on Sunday 18 March 2007, a gang of at least five thieves breached the perimeter chain link fence of the premises. They first breached the envelope of the building by cutting through a ground floor roller shutter door with an angle grinder. Four of them entered the warehouse but then exited again. This is because immediately behind the door was shelving stacked with goods, not permitting further access into the warehouse and another means of access evidently had to be found.

At 02.14 hours the thieves breached the aluminium cladding of the wall, again using an angle grinder, at the mezzanine level above loading bay 2 to the right of detector 60. It appears that no-one entered the building but a torch was shone around inside. It may be that the thieves saw the detector and decided against entry at that point. At about 03.14 hours the gang cut through the cladding with an angle grinder again at the mezzanine level but further eastwards, above loading bay 4, just to the left of a vertical stanchion which did not have a motion detector on it. Through that hole they successfully entered the building on the mezzanine level.

Between then and 05.21 hours, the thieves loaded stock through this hole in the external wall into a van and then left the site. At 06.03 hours they returned with a larger van and loaded that until 07.20 hours when it began to get light. During that second visit, there was sufficient light in the warehouse to discern something of the thieves’ modus operandi from the CCTV footage taken by camera 11, which was mounted pointing eastwards on the same stanchion as detector 60 (which was pointing approximately northwards). Down the middle of the mezzanine lengthwise there was a picking bench or table with rollers on it to enable goods to be moved sideways along the bench or table. This ran from approximately in front of detector 68 to just eastwards of a point in front of where the successful breach in the external wall had been made.

The stock which was stolen was mainly located at the eastern end of the picking bench, in the area of the southern external wall between detectors 69 and 68, apart from picked stock stolen from nearer the point of entry, near detector 71. The thieves used small trolleys or “dollies” on which the goods were loaded. Some of the stolen goods (presumably those already picked) were in blue “tote” boxes. The thieves appear to have pushed the dollies westwards along the “corridor” between the southern external wall, which had racking or shelving against it, and the picking roller bench.

In so far as the thieves themselves can be seen in the CCTV footage, it is only as shadowy blurred shapes, but it appears that, when pushing the dollies along that corridor, the thieves were crouching down, possibly to avoid being caught by the movement detectors. Again it seems that it was only when they were in the vicinity of the hole they had cut in the wall, through which they were passing the stolen goods to be loaded in the van, that they were standing up. It may be that the fact that they were crouching or crawling along that corridor to avoid being caught by the movement detectors explains why, as Mr Cloke said, it took the thieves so long to remove from the premises what was physically a relatively small amount of stock.

Having left the premises at 07.20 hours on 18 March 2007, the thieves returned at about 19.20 hours that evening and re-entered the warehouse through the hole they had made, evidently with a view to stealing further stock. However at 19.46 hours the intruder alarm was activated by motion detector 60. The thieves left the premises. Chubb Security as key holder were called by the alarm company’s receiving centre and attended outside the premises at around 20.40 hours, about the same time as the night shift arrived. Mr Sergeant was called and he in turn called Mr Cloke who arrived at the premises at about 23.30 hours having travelled from his home in Folkestone. The police arrived at about midnight.

After his arrival at the premises, Mr Sergeant had telephoned UKBW to query whether they had seen anything of the theft from remote monitoring of the CCTV system. A representative of UKBW told him that they had not been able to see any events at the premises due to a “temperamental connection”. During the call they checked the system and found they did have a connection. No-one at UKBW had telephoned the claimant over the weekend to say that there was a problem with the connection. It was subsequently established that there was a fault on the ADSL line caused by a noisy BT line or some other faulty BT equipment. Mr Heath accepted this in cross-examination.

The thieves were never caught. In discussions which Melissa Padoa, a solicitor with the claimant’s solicitors, (whose statement was the subject of a Civil Evidence Act notice since she is on maternity leave in South Africa) had with the Essex Police, she was told that the theft had been very well organised and the thieves had known exactly what they were doing. She was also told that the police had no proof that this was an “inside job”.

As was accepted by Mr Thomas in cross-examination, the fencing comprising the inner walls of the cage on the mezzanine was constructed of weldmesh. According to Mr Copper, after the improvements he supervised in September/October 2006, the weldmesh was consistently three metres or so high. The outside wall of the building, which was made of aluminium cladding, formed the fourth wall of the cage.

On the mezzanine, there was guardwire in place (i) along the short run of weldmesh on the south side where the cage was entered via a gate from the area where the offices were located and (ii) along the length of the weldmesh on the western side of the mezzanine. There was also guardwire attached to the inside of the southern external wall at mezzanine level. From a photograph which shows this, it appears that it ran at approximately one and a half metres in height from the mezzanine floor.

Towards the end of the trial an issue arose as to whether, at the time of the theft, there was also other guardwire along the inside of the external walls of the warehouse at ground floor level. This led to a flurry of further statements after the hearing had ended, with the insurers serving a third witness statement from Mr Heath, exhibiting photographs taken during his visit on 26 March 2007 and more recently in September 2009, in support of a case that at ground floor level, the guardwire did not extend around the whole of the perimeter wall of the warehouse. The claimant responded with witness statements from Martin Dunn, a Service Manager with Initial, now known as Chubb Electronic Security Systems Limited, and a further, third statement from Mr Copper dated 25 November 2009.

Further written submissions were also served by both counsel from which it is apparent that the extent of the guardwire on the premises at the time of the theft is agreed to have been as follows:

on the mezzanine level as set out in paragraph 137 above;

at ground floor level along the northern wall of warehouse;

at ground floor level around part of the area described as “accounts storage offices”; and

at ground floor level around the returns room and transport office, a U shaped area immediately to the west of the accounts storage offices.

After the theft, it was found that the guardwire at the mezzanine level was not connected to the intruder alarm system and had not been connected at the time of the theft. Mr Copper’s evidence was that the original guardwire had been disconnected at the time of the works he supervised, when the western wall was moved outwards to the edge of the mezzanine and the length of the guardwire was extended. Initial engineers were on site at the conclusion of those works to reconnect the guardwire to the intruder alarm system, but they did not do so, which is the explanation for the guardwire being disconnected at the time of the theft. It would seem that, through the fault of the Initial engineers, all the guardwire at mezzanine level had been left disconnected, because otherwise the alarm should have been triggered by each of the two breaches of the envelope of the building at that level by the thieves, as was agreed by the security experts.

Mr Copper had no reason to know or suspect that the guardwire had not been reconnected to the system and I am satisfied that no-one else at the claimant either knew or should have known that the guardwire was disconnected prior to the theft. This was first discovered on the morning after the theft, Monday 19 March 2007, when an Initial engineer attended the premises. He was unable to tell Mr Cloke why the guardwire had not been connected and said he would have to call another engineer/supervisor to the site. Another Initial engineer attended a few days later and said that he could not see the guardwire anywhere on the alarm panel. The guardwire was connected by Initial on 22 March 2007. According to Mr Heath, Mr Cloke told him that an Initial engineer had admitted that the non-activation of the guardwire during the theft might have been their fault and, ultimately, in August 2007, Initial admitted liability, but under their terms and conditions this was limited to some £18,000.

In Mr Copper’s third statement dated 25 November 2009, the claimant raised for the first time the fact that there were three infra-red beams protecting the ground floor of the warehouse. This was a matter also dealt with in the parties’ further written submissions. It was agreed by both counsel that I should deal with the matter on paper without the need for a further hearing. The insurers accept that there were such infra-red beams but deny their relevance to the issues for decision. In his further written submissions Mr Turner states, it seems to me with some force, that if these were relevant to the claimant’s case that the security protections at the time of the theft were sufficient, they would surely have assumed some prominence at the trial. As he also pointed out, their presence was never put to Mr McAndrew or Mr Thomas in cross-examination.

I shall return when I deal later in the judgment with materiality and inducement to the question what, if any, relevance their presence has to the issues in the case. For the present I simply note two matters. First, in her further written submissions, Miss Sabben-Clare said that the claimant would not rely upon a sentence in Mr Copper’s latest statement which asserted that if an intruder broke into the warehouse at ground floor level from the eastern or western walls, he would trigger the alarm (because of the infra-red beams) if he moved level with the perimeter of the mezzanine cage. The reason for not relying on this sentence is said to be because of the tiny chink of the cage which extends west of beam 1. Second, as Mr Turner pointed out, there is no evidence before the court as to the effectiveness of the beams and the extent to which they could be circumvented by determined thieves.

It is common ground that, at all material times from when the claimant was added to the policy of insurance, there were eight motion detectors on the mezzanine floor which formed part of the intruder alarm system. The only change made was that when the western weldmesh wall was rebuilt at the edge of the mezzanine, detectors 59 and 60 which had formerly been outside the fenced area were now within that area, albeit that detector 59 was pointing south west, protecting the gated entrance to the enlarged fenced area from the offices. Contrary to what was suggested at one point of the trial, detector 71 was always within the original fence area.

After the theft, when the Initial engineer attended the premises on the morning of Monday 19 March 2007, it was ascertained that it was detector 60 that had been activated and had triggered the alarm. In his Incident Investigation Report dated 24 May 2007, Mr Cloke says that he inspected the point at which the thieves had entered the building and noticed that the other motion detectors covering the mezzanine area (which he identified in discussions with Mr Heath as detectors 71, 70, 69, 68 and 67 but not 66) were angled towards the ceiling. He then conducted a walk test which indicated several dead spots in the area. In his evidence he explained this in a little more detail. The detectors were not angled downwards as they are in more recent photographs showing how they were positioned by Initial after the theft, but were behind the vertical in the opposite direction, so angled upwards, although not pointing straight up at the ceiling.

He accepted that he had spent more time doing the walk test immediately after the theft than he had on the walk test conducted in December 2006 because he was trying to work out how the theft had happened. He did not think that during that earlier walk test, he would have noticed that the detectors were angled upwards, because he was merely looking to see if the lights on the detectors illuminated. During the walk test after the theft, Mr Cloke found a number of what he described in his witness statement as “small dead spots ..in which a person could stand without activating a [detector]”. He could not say whether those dead spots had been present during his walk test in December 2006 since, as he accepted, he was checking much more carefully after the theft. Despite some fencing in cross-examination, in my judgment it was the discovery of those dead spots after the theft which he had not found during his earlier walk test which alerted Mr Cloke to the possibility that the detectors had been repositioned between the two walk tests.

In his Incident Investigation Report Mr Cloke says that he asked the Initial engineer who was on site on 19 March 2007 about the positioning of the detectors (in other words their angling upwards) and the engineer said that he thought they were unlikely to have been left at that angle. In his second witness statement dated 14 October 2009, Mr Cloke said that the idea that the detectors might have been repositioned came from the Initial engineer. When he asked the engineer if it was normal for detectors to be angled towards the roof, the engineer replied that he was not sure. This evidence is very different from what he reported the engineer as having said in the Incident Investigation Report (to which neither of his statements referred at all).

In cross-examination he provided a yet further version of what the engineer had said when he asked whether it was normal to leave detectors angled towards the ceiling to the effect that the engineer had said “no, I don’t think so but I’m not sure”. He could not explain why that was not in his second witness statement (which he had signed only a month before giving evidence). I was unimpressed with that evidence which seemed to me designed to bridge the gap between his two previous inconsistent accounts of what the Initial engineer had said. It seems to me that in all probability, what the engineer said is accurately recorded in his Incident Investigation Report compiled two months after the theft rather than in his subsequent embellishments in evidence.

Mr Cloke reviewed the CCTV footage which was available for about 28 days before the theft and could see nothing to indicate that the detectors had been moved or who by. In my judgment, if the detectors had been moved by an employee of the claimant during that period of a month prior to the theft (whether for innocent or nefarious reasons) that would have been apparent from the CCTV footage. The detectors were at about three metres height from the mezzanine floor and could probably only have been moved by climbing a ladder or step ladder or, more hazardously, by clambering up shelving. Furthermore, access to the mezzanine caged area was limited to the three managers (consisting of Mr Cloke himself, Mr Radford and Andy Sutton, the warehouse manager), together with three supervisors (two transport supervisors and a stock manager) and the six employees who worked on the mezzanine.

If the detectors had been moved by an employee at a time earlier than the available CCTV footage, whilst that would no longer be captured on CCTV, the greater length of time before the theft makes it all the more unlikely that such an employee would have moved the detectors to assist the theft, given that more than a month then passed, during which there would be a risk that someone might notice the angle of the detectors. Furthermore, given that it is difficult to see how such repositioning could have been effected without the use of some form of ladder or by clambering up the shelving, surely one or more of the limited number of employees who had access to the mezzanine caged area would have recalled after the theft that so-and-so had been seen climbing up and altering the angle of the detectors.

It is essentially common ground between the parties that it is extremely unlikely that the detectors were repositioned upwards by the thieves during the theft. Any such repositioning would have taken place whilst it was dark in the warehouse so torches would have to be used and it is difficult to see how a major repositioning of the detectors could (a) have taken place without some sign of that appearing on the CCTV footage and (b) have realistically have occurred without a tamper signal being transmitted to the alarm panel, thereby triggering the alarm. This was also agreed between the security experts.

Thus, despite the fact that the Initial engineer thought that it was unlikely that the detectors had been left angled upwards by Initial themselves during one of their visits, the other candidates, namely an employee or the thieves seem to me so unlikely that the probability is that it was Initial engineers who, for whatever reason, repositioned the detectors at some stage. In any event, whoever did reposition the detectors, this evidently went undetected until after the theft.

In closing submissions Mr Turner provided the court with a number of trigonometric calculations of the extent to which, despite the theoretical length and depth of the beam from a particular motion detector whether alone or in combination with others, there would be areas where the detectors would not pick up the presence of a person, particularly if the detectors were angled upwards. I did not find the detail of these calculations of much assistance, because they necessarily involved a series of assumptions which could not be checked by reference to direct evidence; in particular there is no evidence as to the precise upward angling of the detectors at the time of the theft.

In general terms, what those calculations did demonstrate is that with the detectors angled upwards, however slightly, there would in all probability have been areas in the “corridor” between the picking bench and the racking against the southern external wall where the thieves could move without the detectors being activated. However, since the thieves did go undetected until detector 60 (which was further to the west and not angled upwards) was activated, in a sense this all goes without saying. What is unclear is the extent to which this gang of thieves managed to avoid activating the other detectors by crouching or crawling along that “corridor”.

So far as the CCTV system is concerned, in my judgment the ADSL line had clearly been reconnected on 16 March 2007, before the theft, and whatever problems occurred thereafter which prevented UKBW from detecting the theft via remote monitoring of the CCTV system were attributable to faults with BT’s line or equipment and not the fault of the claimant.

The insurers contend that exclusion (a) in Section C-Theft Insurance applies on the basis that the theft was carried out in collusion with one or more employees of the claimant. The insurers rely in this regard on the fact that the detectors on the mezzanine (other than 59, 60 and 66) were all angled upwards towards the ceiling, which they contend is likely to have been carried out by one or more employees, since realistically only such an employee will have had an opportunity to reposition the detectors in this way.

The allegation of collusion is a serious allegation of dishonesty upon which the burden of proof is on the insurers. The correct approach to such allegations is that set out by Stuart-Smith LJ giving the judgment of the Court of Appeal in National Justice Compania Naviera v Prudential Assurance Ltd (“The Ikarian Reefer”) (No 1) [1995] I Lloyd’s Rep 455 at 459:

In the light of the conclusion I have already reached that the most probable explanation for the upwards angling of the detectors is that they were left in that position at some stage by Initial engineers, the insurers simply cannot discharge that burden of proof, because the evidence fails to exclude a substantial possibility that the loss was accidental, in the sense that the theft occurred other than with the collusion of the claimant’s employees.

In any event, irrespective of the question of repositioning, I agree with Miss Sabben-Clare that there are a number of other features of this theft which make it unlikely that it was committed with the collusion of any employee. First and foremost, there is the fact that the thieves only got inside the warehouse at the third attempt and that the first entry on the ground floor was into a space where further access was blocked by shelving. I agree with Miss Sabben-Clare that, if this had been done with collusion, it indicates an Inspector Clouseau like bungling as to how to enter the building, which is belied by the professionalism of other aspects of the theft. This is unlikely to have been a piece of double bluff to disguise the fact of collusion.

Ultimately at the end of the trial, Mr Turner fairly and realistically did not press the case on collusion very hard, but he did make the perfectly valid point that if the repositioning of the detectors was not carried out by employees and was present at the time of Mr Cloke’s walk test in December 2006, this must cast some doubt over the thoroughness of that walk test, a matter to which I return below.

As is apparent from the summary of the issues set out above, the insurers’ defence of breach of warranty raises a number of sub-issues. As originally pleaded there was an issue as to whether the PMW and the BAMW were on their true construction strict promissory warranties or suspensive conditions. The insurers, no doubt in order to avoid arguments about the harsh consequences of their argument, were contending that these two warranties were, on their true construction, only suspensive conditions, pursuant to which cover was only suspended for as long as the condition or warranty was not complied with.

The insurers applied for summary judgment on that basis, as well as on a number of other grounds, but HHJ Mackie QC rejected insurers’ application and held, in a judgment dated 19 November 2008, that the two warranties were on their true construction promissory warranties and not merely suspensive conditions. Although the insurers appealed, in the event unsuccessfully, various other aspects of their summary judgment application to the Court of Appeal, they did not appeal that conclusion that the relevant provisions were promissory warranties.

At one point, Mr Turner suggested, by reference to the decision of Morland J in Kler Knitwear v Lombard General Insurance [2000] Lloyd’s Rep IR 47, that HHJ Mackie’s conclusion that the provisions were promissory warranties was wrong. In the light of the provision in Additional Memorandum 3 as to the effect of warranties and their breach, it seems to me that any argument that the decision of HHJ Mackie QC was wrong would be doomed to failure. The unequivocal terms of that provision make it clear that the relevant provisions are indeed promissory warranties, and HHJ Mackie QC was right to reach that conclusion. In any event, for the purposes of the present trial, that conclusion is binding on the parties.

The principles of construction that the court should adopt in relation to continuing or promissory warranties in contracts of insurance were recently restated by Sir Anthony Clarke MR (as he then was) in Pratt v Aigaion Insurance Co SA (“The Resolute”) [2008] EWCA Civ 1314; [2009] 1 Lloyd’s Rep 225. Having summarised the principles of general application to contractual construction he went on at paragraphs 13 and 14 of his judgment:

The first set of submissions which Miss Sabben-Clare advances are to do with the fact that the PMW and BAMW are both to be found in ATD’s standard policy wording which is for use with the form of Proposal Form which was completed by Blueheath when the insurance was first taken out, but not by the claimant when the insurance was varied with effect from 1 April 2006. She submits that (i) in the case of the PMW, “the protections provided for the safety of the insured property” is a cross-reference to whatever answers the insured has given to the questions in Section 4 of the Proposal Form concerning security on the premises, which defines what those protections are. If the insurers have not required the insured to fill in a Proposal Form, the phrase “the protections provided for the safety of the insured property” is too vague and open-ended to be given any meaning particularly given that a warranty is involved and (ii) in the case of the BAMW, sub-clause (a) refers to “burglar alarm system stated in the Schedule, which has been approved by the Insurers” and the rest of the provision refers back to that burglar alarm system. Since there was never a “burglar alarm system stated in the Schedule, which has been approved by the Insurers” in relation to the Thurrock warehouse, the BAMW simply does not bite in this case.

I am unable to accept these submissions. It is clear both from the Register Sheets sent by the brokers on behalf of the claimant to the insurers and from the wording of the Policy Schedule that the parties intended that both warranties should apply to the insurance to be provided to the claimant. I would be reluctant to adopt a construction which frustrated that intention. In my judgment, Miss Sabben-Clare’s approach adopts far too literal a construction.

In relation to the PMW, the phrase “the protections provided for the safety of the insured property” is not on the face of the warranty tied to protections identified in the Proposal Form. As a matter of commercial common sense, it seems to me that what this is referring to is whatever security devices or protections the insured has in place at the insured property at the time of inception of the insurance. There is nothing vague or ambiguous about that: even if the insurers do not know what the protections are (because no Proposal Form has been completed), the insured does.

However I do agree with Miss Sabben-Clare that, applying the principles set out in The Resolute the warranty is only given in relation to the maintenance of the protections which were in place at the time of inception of the insurance which in the case of the claimant is 1 April 2006. There is nothing in the wording of the warranty to suggest it applies to future protections which the insured may put in place after inception of the insurance. That would be potentially open-ended and onerous and would require clear words to that effect, which are absent here.

In relation to the BAMW, it seems to me that there is some force in Mr Turner’s submission that, although the Policy Schedule states “Make & type of Burglar Alarm System: Not provided”, that is sufficient to amount to a “burglar alarm system stated in the Schedule”. The fact that it had not been approved by the insurers should not prevent the warranty applying. Even if this submission by Mr Turner were wrong, I consider that the warranty should be construed so as to apply to the burglar alarm system which was in place at the Thurrock warehouse and that any other construction would be commercially absurd.

In the present case, the insured had failed repeatedly to provide the specification of the burglar alarm systems on its premises. Initially the insurers were seeking that specification in relation to Blueheath’s Islington premises but by the time the further reminder came to Mr Phillips of the brokers on 8 January 2007, he interpreted the request for the specification as relating to the Thurrock warehouse. That is precisely what he asked Mr Paget for in his letter of 12 January 2007. However, no specification was forthcoming. It seems to me that if one had been, so that the make and type of alarm had now been provided, it would be an absurdly literal construction to say that the BAMW did not apply because the system was not “stated in the Schedule” and I do not see why the insured should be in a better position by having failed to provide the specification of the alarm system.

In my judgment, to give the BAMW a commercially sensible construction the reference to “the burglar alarm system” should be construed as a reference to the system in place at the Insured property, but, as in the case of the PMW, only such system as was in place at the inception of the insurance on 1 April 2006, for the same reason as stated above in relation to the PMW.

Miss Sabben-Clare’s primary submission, if the PMW and BAMW did apply to the protections and burglar alarm system at Thurrock, is that on their true construction there is only a breach of warranty if the claimant fails to remedy a defect falling within the scope of either warranty which it knows of but fails to remedy promptly. This is said to be for a number of reasons which can be summarised as follows:

Any other construction fails to give any effect to the closing words of each provision: “All defects occurring in any protections must be promptly remedied”. If the insurers were, as they contend, automatically discharged from liability because a defect of which the claimant was unaware amounted to a breach of warranty, then those words would always be redundant.

The claimant’s construction is the only one which makes commercial sense since on the insurers’ construction the insured would lose its protection even if it was unaware of the problem and could not reasonably have known it. This is an absurd and unreasonable construction of a contract of insurance.

The insurers’ construction is inconsistent with the approach adopted by courts in earlier cases, specifically the decisions of Donaldson J in de Maurier v Bastion Insurance [1967] 2 Lloyd’s Rep 550 at 559 that a burglar alarm was “in operation” provided it was switched on even if, unbeknownst to the insured, there was a fault in the system and of Woolf J in Melik v Norwich Union [1980] 1 Lloyd’s Rep 523 at 531 that a burglar alarm was “kept in full operation at all times when the property was unattended” notwithstanding a defect of which the insured was unaware and had had no opportunity to remedy.

The BAMW contains words of action on the part of the insured that the alarm system “shall have been put into full and effective operation” the word “put” importing something within the claimant’s knowledge and control. The system is “put” into operation by switching it on and provided that is done, the existence of a defect in the system of which the insured is unaware (and could not reasonably have been aware) should not amount to a breach of warranty.

Although the PMW does not use the word “put” it is equally directed at action on the part of the insured and should be construed in the same way as the BAMW.

Mr Turner submitted that both warranties should be construed as strict obligations so that if, objectively, the burglar alarm system or the protections were not in “full and effective operation” because for example the guardwire was disconnected, then there was a breach of warranty even if the insured was not aware of the relevant problem.

Mr Turner next submitted that Melik v Norwich Union [1980] 1 Lloyd’s Rep 523 is the only reported case on similar wording to the present and has stood unchallenged for more than 25 years. Far from supporting the claimant’s case, Mr Turner submitted that it supported the insurers’ case. In that case the warranty was that the burglar alarm be “kept in efficient working order” and “kept in full operation at all times when the premises are unattended”. Prior to carrying out a burglary of the insured’s premises, thieves cut all outgoing telephone lines and thereby disconnected the alarm from the control centre. In rejecting the insurers’ argument that the alarm had not been kept in full working order or in full operation, Woolf J drew a distinction between a requirement that an alarm “be” in efficient order as opposed to being “kept” in efficient order. The word “kept” implied that before there was a breach, the insured had to be aware of the defect and given an opportunity to remedy it.

The relevant part of his judgment at 530-1 provides as follows:

Mr Turner submitted that since this interpretation of well-known wording had been acted on for a considerable period of time, the court should be chary of interfering with that interpretation: see In re an arbitration between Hooley Hill Rubber and Royal Insurance [1920] 1 KB 257 at 269 per Bankes LJ. This principle applied in relation to similar, not merely identical, wording: see Toomey v Eagle Star [1994] 1 Lloyd’s Rep 516 at 520 per Hobhouse LJ. Neither the PMW nor the BAMW used the word “kept” and had the parties wanted to qualify the insured’s obligation as in Melik, they could and should have done so by using the word “kept”. The PMW said that the protections “shall be in full and effective operation” words of strict obligation as Woolf J had recognised in contrasting “be” in efficient order with “kept” in efficient order. Similarly, the word “put” in the BAMW was not synonymous with “kept” but “switched on” and imported a strict obligation.

Mr Turner sought to distinguish de Maurier on the basis that the obligation there was that the alarm was “in operation” so there was an unqualified obligation to put something into operation but no requirement as to the quality of the operation that was required. Donaldson J declined to construe the words as meaning objectively, fully operative. In contrast, both warranties here use the words “full and effective operation”.

Mr Turner next submitted that the effect of Miss Sabben-Clare’s construction was to deprive the body of the text of each warranty of any effective meaning, qualifying each warranty by reference to the closing words which were not themselves warranted. This was an example of an impermissible construction, in effect allowing “the tail to wag the dog”. It would have the effect of putting a red line though the operative provisions of the warranty. Mr Turner accepted that his construction did not give any real effect to the closing words of each warranty and that those words, which were not themselves warranted, were effectively surplusage.

I have formed the very firm view that the claimant is right that each of these warranties is a qualified one, in the sense that the insured is only in breach of warranty if there is some defect in the particular protection or the burglar alarm system, of which the insured becomes aware or should reasonably have become aware and the insured has then failed to remedy the defect promptly. Each of these provisions has to be construed as a whole and the claimant’s construction does so. It also gives effect to the important closing words of each provision “All defects occurring in any protections must be promptly remedied”, which form part of each of the warranties.

The insurers’ construction gives no effect whatsoever to these words, since if the obligation to be or to put in full and effective operation is a strict one, the insured will always be in breach of warranty and the insurers relieved from liability, before the insured knows of the defect or has any opportunity to remedy it. That was of course the very construction of such a warranty which Woolf J sought to avoid in Melik. It seems to me that the fact that the insurers were reduced to arguing that the closing words were mere surplusage (which they clearly are not) demonstrates the fallacy of their argument, which fails to give any effect to these words. Furthermore, that construction produces a result inconsistent with a reasonable and businesslike interpretation of the warranties and in accordance with the principles in The Resolute should yield to the construction which is reasonable and businesslike, which is the claimant’s construction.

Furthermore, contrary to Mr Turner’s submission, I do not consider that the claimant’s construction emasculates the body of each of the warranties. As Miss Sabben-Clare accepted, the insured would be in breach of warranty if it was reckless as to whether or not the protections or the burglar alarm system were in full and effective operation, adopting a “don’t care” attitude to whether or not there were defects. In such a situation, the insured would be in breach of warranty if it transpired that there were defects, even though the insured had not actually known of them. Miss Sabben-Clare puts the test as one of recklessness, it seems to me correctly, on the basis that an obligation on an insured to take reasonable steps to avoid a loss, here in Special Condition 1 of Section C of the Policy to take “all reasonable precautions for the safety of the property insured”, has consistently been interpreted by the courts as requiring the insured to refrain from reckless (as opposed to negligent) conduct: see Fraser v Furman [1967] 2 Lloyd’s Rep 1 at 12 per Diplock LJ and Sofi v Prudential Assurance [1993] 2 Lloyd’s Rep 559, where the principle previously applied to liability insurance was extended to property insurance.

It also seems to me that there is considerable force in Miss Sabben-Clare’s submission that since both provisions use words of action by the insured, it is implicit that, provided that the insured does what is required of it and remedies the defects of which it knows promptly, it will not be in breach of warranty. The warranties are after all “maintenance” warranties and contain within their provisions express obligations to maintain the protections and the burglar alarm system in good order. An insured can hardly be in breach of a maintenance obligation if he maintains the system but unbeknownst to him there is some latent defect in the system. However it is not necessary to decide this point finally, since I have concluded that the warranties are qualified by the closing words in each.

Mr Turner’s submissions about Melik run the risk of interpreting the words of a judgment as if they were a statute. I do not consider that Woolf J was intending to lay down some immutable rule that using the word “kept” was the only way in which a strict obligation in a warranty could be qualified or that a provision that a protection or a burglar alarm “shall be in full and efficient operation” or is “put in full and effective operation” must inevitably be a strict obligation, particularly if, as in the present case, the words are subsequently qualified by the closing words “All defects occurring in any protections must be promptly remedied”. Taking those two provisions together, the one qualifies the other in the same way as the use of the word “kept” did in Melik and I have little doubt that Woolf J would have construed the PMW and the BAMW in the same way as I have.

I turn to consider whether in the light of what I have held to be the correct construction of the Policy, the claimant was in breach of either of the two warranties. It will be apparent from the summary of the issues which I set out in the judgment that there is a considerable overlap between the two warranties, in the sense that the insurers contend that the fact that the guardwire was disconnected and that five of the motion detectors on the mezzanine were angled upwards was a breach of both warranties.

I propose to consider first the issues about the ADSL line. The short answer to the allegation that the disconnection of this line between 13 February and 16 March 2007 amounted to a breach of the PMW is that, since the line was not installed until September 2006, it was not a “protection” which was in place at the inception of the insurance on 1 April 2006 and accordingly is not covered by the warranty at all.

Even if my construction of the warranties, that they only apply to protections or a burglar alarm system in place at the time of inception of the insurance were wrong, I would not consider that there was a breach of warranty here. Whilst it is true that the claimant (probably through Mr Sergeant) does not seem to have done anything about the problem of disconnection of the ADSL line for a week or so from when UKBW informed the claimant of a problem (probably soon after the line was disconnected on 13 February 2007) until Mr Butler was told of the problem on 23 February 2007 and contacted SDS, I accept that this was a relatively short delay such as occurs in even the most well-run businesses. Thereafter, I consider that the claimant (through Mr Butler) acted fairly quickly and once the request for reconnection had been made, it took about the usual length of time for that to take place. The line had been reconnected by the time of the theft. Viewed overall, I consider that the claimant did promptly remedy any defect through the disconnection of the ADSL line and was not in breach of warranty.

So far as concerns the intermittent fault on the ADSL line during the theft which seems to have prevented UKBW from viewing the CCTV coverage, this was a fault for which BT not the claimant was responsible and since UKBW did not tell the claimant about the problem until after the theft, there is no question of the claimant having known about any defect or had the means of knowledge. No question of breach of warranty arises.

Turning to the guardwire and the fact that it seems to have been disconnected at the time of the theft, two points arise. First, since the warranties only apply to protections or elements of a burglar alarm system which were in place at the time of inception of the insurance, the warranties do not bite in respect of the additional guardwire installed during the works supervised by Mr Copper. Second, the failure to reconnect the guardwire was the fault of Initial’s engineers and not the claimant. As I have already said in my detailed analysis of the facts, so far as the employees of the claimant (and specifically Mr Copper) are concerned, they had no reason to suppose that Initial’s engineers had not done their job properly and they did not know there was a defect.

I also do not consider that this is something the claimant ought reasonably to have known. Since the guardwire was disconnected, it would not show up as a fault on the alarm panel and it is expecting too much of the employees of the claimant who set and unset the alarm (by definition in circumstances where they did not know there was a problem) that they should somehow have picked up that the guardwire was disconnected, by going through the handbook or other alarm documentation. They simply had no reason to do so. I do not consider that there was any breach of warranty in relation to the guardwire.

So far as the angling of the motion detectors is concerned, to the extent that this is said to have been a breach of either warranty, the burden of proof is clearly on the insurers. I have indicated that I consider the most likely explanation for the angling of the detectors (despite what the Initial engineer said to Mr Cloke on the morning after the theft) is that this was done by Initial engineers at some stage. However, it is not possible to say when this was done and therefore whether it was done before or after inception of the insurance. In that context, it is noteworthy, as I said at an earlier stage of this judgment, that Mr Thomas did not give any evidence about the angle of the detectors at the time of his visit to the premises on 21 June 2006. In those circumstances, it seems to me that the insurers simply cannot establish that there had been a variation of the protections or the burglar alarm system without their consent since inception. Accordingly, insurers fail at first base in seeking to establish a breach of warranty.

In those circumstances it is not strictly necessary to go on to consider the other matters which insurers would have to prove to establish a breach of warranty, namely that this was a defect of which the claimant was aware (or about which the claimant ought reasonably to have known) and which the claimant had failed to remedy promptly, but I shall indicate my conclusion on those matters.

First, on the question of knowledge, since the insurers’ case on collusion fails, it seems to me that they must fail on actual knowledge of any employee that the detectors had been angled upwards. Mr Cloke clearly did not know and did not notice the angle of the detectors during his walk test in December 2006. As to whether employees ought to have known, it seems to me that Mr Cloke is right that one would not expect the employees working on the mezzanine to notice whether detectors had been moved or what angle they were at.

The question which arises is whether Mr Cloke should have picked up the positioning of the detectors on his walk test in December 2006. By his own admission, this was less thorough than the walk test he conducted immediately after the theft. Furthermore, the fact that he saw fit to invent in his evidence a yet further walk test in January or February 2007, after he had seen the Risk Improvement Requirements, suggests that he was not confident as to the thoroughness of his original walk test.

However, on balance I consider that Mr Cloke is not to be criticised for failing to pick up that the detectors were angled upwards. In December 2006, the walk test was conducted to check that the lights were illuminating and he said that they were. In those circumstances, he had no reason to pay any particular attention to the angle of the detectors. The walk test carried out immediately after the theft was conducted in circumstances where he knew that the alarm had not activated for some time during the theft and he was seeking to identify what the problem was. Even then, he said that there were only small dead spots where the lights did not illuminate, which suggests that the angling upwards was not severe, borne out by his own evidence about the detectors being “behind the vertical”, suggesting that the angle may not have been prominent unless you were examining the detectors in the sort of detail which Mr Cloke did after the theft. In all the circumstances, I do not consider that the claimant was in breach of warranty in relation to the positioning of the motion detectors. Overall, the defence of breach of warranty fails.

In order to avoid a contract of insurance or, here, a variation of such a contract, the insurers have to satisfy the two stage test laid down by the House of Lords in Pan Atlantic Insurance v Pine Top Insurance [1994] 1 AC 501:

That the particular fact misrepresented or not disclosed would have influenced the mind of a reasonably prudent underwriter in exercising his underwriting judgment as to whether to write the risk or here to extend cover by agreeing the variation: see e.g. per Lord Goff at 517E-G and Lord Mustill at 550C.

That the underwriter who agreed the risk or in this case the variation, here Mr McAndrew, was induced to agree it by the fact(s) misrepresented or not disclosed. To establish inducement the insurers must satisfy the “effective cause” test as clarified by the majority of the Court of Appeal in Assicurazioni Generali SpA v Arab Insurance Group [2003] Lloyd’s Rep IR 131 per Clarke LJ at para. 59:

In considering whether the insurers can satisfy this two-stage test, the first question is what representations were made by or on behalf of the claimant in relation to the security measures and the burglar alarm system, prior to the insurers agreeing to remove the exclusion in Endorsement 6. On the basis of the detailed findings of fact I have reached, it seems to me that there were two sets of representations.

First, as I have found, contrary to Mr White’s evidence, at some point in the afternoon of 14 February 2007 after receipt of Mr Paget’s email timed at 12.27 hours that day and before Mr McAndrew’s email to Mr White at 16.56 hours Mr White communicated to Mr McAndrew the substance of Mr Paget’s email to Mr White, probably telling him as part of that conversation, that the Risk Improvement Requirement 2006/02 had been complied with. It is fair to say that this conversation was not specifically pleaded as a representation in the Amended Defence and Counterclaim, but that it was insurers’ case that such a representation was made during the conversation emerged clearly during the evidence.

Second, by the email of 20 February 2007 from Mr Phillips to Mr McAndrew, the following representations were made:

The insurers contended that the statement that Risk Improvement Requirement 2006/02 had been complied with involved the following representations:

That there was a cage constructed of substantial weldmesh on all four walls;

That additional motion detectors had been installed on the mezzanine level;

That there was guardwire round all four walls of the cage, which had been commissioned;

That the existing detectors had been checked and found to provide full coverage.

The insurers alleged that these representations were misrepresentations. In the alternative, if these were not representations, there was material non-disclosure in that the claimant was alleged to have failed to disclose the true position as regards those matters.

The insurers persisted until the conclusion of the trial in a case that the claimant had misrepresented the position about the “cage” because (a) the weldmesh of which it was constructed was not “substantial”; and (b) it was not in truth a cage because one side of it was bounded by the southern external wall, rather than by another weldmesh wall within that external boundary, so that the cage was free-standing within the mezzanine.

So far as the first point was concerned, there was no representation made by the claimant as to the construction or structural strength of the weldmesh, so there is nothing in this point. So far as the second point is concerned, the insurers never specified that the “cage” had to be a free-standing structure within the mezzanine with four weldmesh walls. At the outset of the trial it seemed to me that an enclosed area with three weldmesh walls on the western, northern and eastern sides and a solid rear wall on the southern side comprising the external wall of the building was as much a “cage” as is a cage at the zoo enclosing a tiger, which has bars or wire mesh on three sides but a solid rear wall, behind which in that example, the tiger’s sleeping quarters lie. I did not change that view as the trial went on.

I should add that the reference to the cage being 10 metres high in places was to the fact that the distance from the floor of the warehouse to the top of the weldmash fencing was 10 metres. It was never suggested by the insurers that they had misunderstood this, nor was any misrepresentation in that regard alleged by the insurers. In the circumstances, I am firmly of the view that there was no misrepresentation about the cage on the mezzanine floor.

However, in my judgment, the statement that Risk Improvement Requirement 2006/02 had been complied with is a different proposition. The Risk Improvement Requirement had three elements. First, that “additional movement detectors [be installed] in the area of the mezzanine where cigarettes & tobacco and wines & spirits are stored and packed”, this to be done “to provide more comprehensive coverage to detect intruders, who may break in through the roof” (my emphasis). In my judgment there can be no doubt that what Mr Thomas and the insurers were requiring was further movement detectors in addition to the eight already on the mezzanine, to provide more comprehensive coverage. If there had been any doubt about that, it was resolved conclusively by the contrast with “existing” detectors referred to in the “NB”.

All that happened during Mr Copper’s upgrade works in September and October 2006 is that the cage was extended westwards so that all the detectors were now within the cage (including 59 and 60), albeit that detector 59 continued to point outwards in the direction of the offices. However, there were no additional detectors installed and, try as Miss Sabben-Clare might in her submissions, I do not see how that particular circle could be squared. There simply were no additional detectors, and in representing that the Risk Improvement Requirement had been complied with, the brokers were misrepresenting the position.

Miss Sabben-Clare relied upon the principle enshrined in section 20(4) of the Marine Insurance Act 1906, that a representation of fact is true if it is substantially correct, that is to say if the difference between what is represented and what is actually correct would not be considered material by a prudent insurer. She relied upon the evidence of Mr Cloke in his witness statement:

As Miss Sabben-Clare said, this was not challenged in cross-examination, but it does not follow that the court has to accept it. As I have said, there were various unsatisfactory aspects of Mr Cloke’s evidence which led me to approach his evidence generally with caution. This conversation with the Initial engineer is not mentioned in Mr Cloke’s Incident Investigation Report and it seems to me that it would have been if it had occurred. In any event, even if this was said, it was not said after Initial had carried out a thorough survey of the mezzanine and the existing detectors to determine whether additional detectors would provide more comprehensive coverage. It would be quite another matter if the claimant had asked Initial to conduct such a survey, in response to the Risk Improvement Requirements, and there had been a report from Initial that no additional detectors were necessary. That would then have been reported to insurers who would no doubt have consulted Mr Thomas to see what his reaction was.

However, none of that happened and I am not over-impressed by the statement by the Initial engineer, even if it was said, which was made without a full survey. Furthermore, Mr Cloke’s evidence about this conversation was, on one view, inconsistent with Mr Copper’s evidence that during the upgrade woks he supervised, the Initial engineers had suggested additional detectors be installed. He could not remember whether this was on the mezzanine, but it could possibly have been.

In any event, common sense would suggest that additional movement detectors would have provided more comprehensive coverage, specifically, a detector on the stanchion between those on which detectors 71 and 70 were located. This stanchion was in fact closest to where the thieves successfully broke into the building. It also seems likely that some additional detectors on the northern weldmesh wall of the cage pointing south or on the side walls of the cage facing across the length of the mezzanine in the east-west direction would have provided more comprehensive coverage.

The second element of Risk Improvement Requirement 2006/02 was the requirement for “vibration inertia detectors [to be installed] on the walls and ceiling of the mezzanine where cigarettes & tobacco are to be stored on the mezzanine” this again to be done “to provide more comprehensive coverage to detect intruders, who may break in through the roof”. Of course, Mr McAndrew withdrew the requirement for a ceiling, but the requirement for vibration inertia detectors on the walls remained. This would have been satisfied by the presence of guardwire on all the walls of the cage, but there was not. The guardwire ran only along the inside of the external southern wall, along the southern run of weldmesh by the gate to the area from the offices and along the length of the western weldmesh wall. There was no guardwire on the eastern or northern weldmesh walls. By representing that the Risk Improvement Requirement had been complied with, the brokers were representing that there were vibration inertia detectors on all the walls of the cage and this was a misrepresentation.

Once again, Miss Sabben-Clare sought to argue that the representation was substantially correct because the absence of guardwire on those two walls was not material, since there was a clean drop from there to the warehouse floor. Miss Sabben-Clare relied upon Mr Copper’s evidence in his witness statement that the risk of a thief accessing the cage via the northern or eastern walls was “minimal”. However, in the light of Mr Copper’s acceptance in cross-examination that it would be possible to use a forklift truck to get to the level of the mezzanine floor and then breach the weldmesh, it seems to me that his evidence that the risk of access was minimal cannot stand. In the circumstances, the absence of guardwire protecting those walls of the cage was material. That conclusion is not altered by the reliance by the claimant, very late in the day, on the presence of infra-red beams on the ground floor. This was never explored with any witness at trial and I accept Mr Turner’s submission that it is simply impossible to assess how effective those beams would have been in deterring determined thieves.

Whilst I am satisfied that, by the representation that 2006/02 had been complied with, there was a representation that the guardwire extended round all the walls of the cage and that this was a misrepresentation, I do not consider that the insurers’ case that there was a further misrepresentation that the guardwire had been commissioned, is made out. As I have held in rejecting the case of breach of warranty, the claimant did not know that the guardwire had not been reconnected by Initial and that was not something which the claimant ought reasonably to have known. In the circumstances, that is not a matter which was misrepresented or which the claimant ought to have disclosed.

The third element of Risk Improvement Requirement 2006/02 was that the existing movement detectors on the mezzanine should be checked to ensure full coverage, as some of the lights had not come on during Mr Thomas’ walk test. The insurers’ case is that by representing that the Risk Improvement Requirement had been complied with, the brokers were by necessary implication representing that the existing detectors had been checked and did provide coverage. The insurers were highly critical of Mr Cloke’s walk test in December 2006, submitting that the walk test which had in fact been carried out was not one which any reasonable person could have believed was sufficient for the purpose of ensuring that there was full coverage.

Although by his own admission the walk test which Mr Cloke conducted in December 2006 was less thorough than the one he conducted after the theft, I have already indicated, when rejecting the insurers’ allegation of breach of warranty in relation to the angling of the motion detectors, that I accept that Mr Cloke is not to be criticised for failing to notice that some of the detectors were angled upwards. I also accept that during his walk test, the detectors did all illuminate, so he had no reason to suppose they were not providing full coverage. In my judgment, there was no misrepresentation in this respect.

As stated above, the question is whether the misrepresentations made were material, in the sense that they would have influenced the mind of a reasonably prudent underwriter in exercising his underwriting judgment as to whether to remove Endorsement 6 and thereby vary the contract of insurance. Ultimately this is a question for the court, but the court is usually assisted by the opinions of experienced underwriting experts as to the question of materiality.

In the present case, only the claimant’s expert, Mr Allan Guest, was called. In cross-examination, he accepted that:

With high value stock of this kind (cigarettes and tobacco) at the warehouse, a prudent underwriter would expect the stock to be held in a dedicated secure area, with “layered” security in relation to access to the building and then access to the secure area;

A reasonably prudent underwriter would agree with a risk surveyor’s recommendation that such high value stock be kept in a secure cage, with full alarm protection;

The imposition of Risk Improvement Requirements with a timescale attached to them is within the range of actions which a reasonably prudent underwriter would take after a risk survey;

The Risk Improvement Requirements would still have been important to a reasonably prudent underwriter even if there had been some administrative delay in issuing them.[in fact there was not, because, as I have held, the brokers had them in July 2006];

A reasonably prudent underwriter approached on behalf of an insured and asked to lift an exclusion in circumstances where he has not been told whether Risk Improvement Requirements have been complied with, would want to know what steps had been taken to comply and why he was being asked to lift the exclusion. If not given confirmation that the Requirements had been complied with, he would be inclined not to lift the exclusion, but would inquire why the request was being made;

If the insured came back and said we haven’t finished complying with the Requirements but are in the course of doing so, he would be inclined to give the insured the benefit of the doubt and extend the coverage, provided that he was satisfied with the extent of the work the insured had already undertaken;

If the insured said that nothing had been done to comply with the intruder alarm requirements, the reasonably prudent underwriter would certainly take that into account and would be unlikely simply to shrug his shoulders and say it did not matter. One option would be to say the exclusion remained in force until the requirements were complied with. Another would be to extend the time for compliance and yet another would be to impose some co-insurance until they were complied with.

That an important factor in the decision-making process as to whether to lift the endorsement would be the true position in relation to compliance with the Risk Improvement Requirements and if anything was outstanding what it is and how long it was going to take to do.

If the insured had not in fact complied with either of the Risk Improvement Requirements, the prudent underwriter would not simply have given away the exclusion with nothing in return.

That evidence (particularly points (8) and (9)) amounted to acceptance by Mr Guest that the question whether and to what extent the insured had complied with the matters set out in both the Risk Improvement Requirements was material, in the sense that what the true position was about compliance with the Risk Improvement Requirements would have influenced the mind of a reasonably prudent underwriter in determining whether to vary the contract by lifting the exclusion in Endorsement 6.

However, even without the assistance of such expert evidence, I would still have concluded that the matters misrepresented were material. This is one of those cases where the court can and should adopt the robust approach of Scrutton LJ in the well-known case of Glicksman v Lancashire and General Assurance Company [1925] 2 KB 593 at 609:

I have already indicated, in rejecting Miss Sabben-Clare’s submissions that the representations made about the additional detectors and the guardwire were substantially correct, that in my judgment the misrepresentations implicit in the statement that Risk Improvement Requirement 2006/02 had been “complied with”, that additional detectors had been installed and that there was guardwire on all the walls of the cage, were material misrepresentations, in that they would have influenced the mind of a reasonably prudent underwriter in deciding whether or not to lift Endorsement 6.

In the alternative to the case of misrepresentation, the insurers alleged that the claimant had failed to disclose the true position as regards these matters. Since I have found that there were material misrepresentations, the alternative case of material non-disclosure does not add much to the case of misrepresentation. Nonetheless, it seems to me that even if there were not misrepresentations, there was material non-disclosure by the claimant. The claimant failed to disclose the following facts which were material: (a) that there had not been any additional motion detectors installed on the mezzanine and (b) that the guardwire did not extend around all four walls of the caged area. Either way, materiality is made out.

An important starting point in considering whether or not Mr McAndrew was induced to lift Endorsement 6 by the facts misrepresented or not disclosed is that, quite apart from his own evidence, the contemporaneous correspondence demonstrates that far from being indifferent to the security in place at the warehouse and specifically as to compliance with the Risk Improvement Requirements (as at least Mr White was suggesting in his witness statements), Mr McAndrew was only prepared to lift the Endorsement when satisfied that there had been substantial compliance with the Risk Improvement Requirements.

The insurers were clearly concerned about the state of security at the premises after Mr Player’s survey in May 2006. Mr Sullivan imposed what became Endorsement 6 in his email to Mr White of 2 June 2006 and then sent Mr Thomas to carry out the further survey. After the results of that survey and the Risk Improvement Requirements he recommended were known, these were sent to the brokers in July 2006 and again in December 2006.

Mr McAndrew sent Endorsement 6 to the brokers with the Policy Schedule under cover of his letter of 4 December 2006. When, in reply, the brokers suggested that Endorsement 6 could not be included until they had confirmation from the claimant that the Risk Improvement Requirements had been complied with, Mr McAndrew insisted that the Endorsement remained in place. Then when, on 8 February 2007, Mr White passed on to Mr McAndrew the information from the claimant that a lot of work had been done at the premises and that the cigarettes and tobacco were now in a cage, although Mr McAndrew agreed to organise a resurvey as soon as possible, he still insisted that in the meantime the Endorsement would remain in place.

This was the background to the telephone conversation on 14 February 2007 and the email from Mr Phillips of 20 February 2007, after which Mr McAndrew lifted the endorsement. Given that background and the history of insisting that the Endorsement remained in place, it seems to me that it is inherently unlikely that Mr McAndrew, who struck me as a prudent underwriter, would have simply given the Endorsement away without assurances that the Risk Improvement Requirements had been complied with, and I find that he did not do so.

Miss Sabben-Clare urged me to take a cautious and sceptical approach to Mr McAndrew’s evidence about inducement, on the basis that it would inevitably be self-serving. She relied upon a passage in the judgment of Colman J in North Star Shipping v Sphere Drake Insurance [2005] 2 Lloyd’s Rep 76:

In the present case, I have assessed Mr McAndrew’s evidence by reference to the contemporaneous background to which I have referred, from which it seems to me his evidence does have a logical self-consistency. Furthermore, as Mr Turner pointed out, this caution against accepting too readily what might be self-serving evidence, would apply equally to the evidence given by the brokers, and specifically, Mr White. Whereas Mr McAndrew was a straightforward witness whose evidence tied in with the contemporaneous documents, Mr White’s evidence on the question of inducement, specifically his change of evidence about the telephone conversation on 14 February 2007, was, as I have held, sometimes incredible and, at best, unsatisfactory.

I am quite satisfied that Mr McAndrew did not agree to lift the Endorsement until he had been assured by the brokers that both Risk Improvement Requirements had been complied with (save to the extent that there was no ceiling on the cage). There are two answers to Miss Sabben-Clare’s suggestion that, by 20 February 2007, all Mr McAndrew was interested in was the cage. First, the answer he gave in evidence that by that stage he had had confirmation that the requirements in 2006/02 had been complied with so the only issue outstanding was the nature of the cage. Second his evidence that he considered the risk improvements were all interlinking, which I have quoted at paragraph 109 above. It was quite clear from that evidence, that he only agreed to delete Endorsement 6 even though the cage did not have a ceiling, because he had been assured that the alarm system had been extended in accordance with Risk Improvement Requirement 2006/02. I regard that as compelling evidence that Mr McAndrew was indeed induced to agree to lift Endorsement 6 by the brokers’ representations in the phone conversation on 14 February 2007 and in the email from Mr Phillips on 20 February 2007, that Risk Improvement Requirement 2006/02 had been complied with.

As for Miss Sabben-Clare’s submission that the respects in which the claimant had not complied with 2006/02 were so narrow that they would not have made a difference to Mr McAndrew’s decision to lift the Endorsement, in a very real sense that submission is defeated by the materiality of the respects in which the claimant had not complied with 2006/02, namely the absence of any additional movement detectors in the caged area to provide comprehensive coverage and the fact that there was no guardwire on the northern or eastern walls of the cage, both of which I have concluded were material to the risk. Furthermore, far from being narrow respects in which the claimant had not complied with the Risk Improvement Requirement, the true position was that the claimant had not really complied with it at all, save for extending the guardwire along the western weldmesh wall and carrying out a walk test from which it appeared the lights were illuminating on the existing detectors.

In all the circumstances, I consider that, if the claimant had disclosed to Mr McAndrew on 20 February 2007 the true position as regards its limited compliance with 2006/02, one of two results is likely to have ensued, either of which is sufficient to establish inducement. Mr McAndrew might well have continued to insist, without any further discussion, that the Endorsement remained in place until the security at the premises was improved in accordance with the Risk Improvement Requirements. On the other hand, he might have decided to discuss the matter with Mr Thomas. It is most unlikely that the concerns which Mr Thomas had about security at the premises would simply have evaporated, on being told of the limited work in fact carried out. In all probability, after such further discussion and clarification, Mr McAndrew would have still refused to lift the endorsement until the Risk Improvement Requirements were complied with.

That position is not saved, so far as the claimant is concerned, by the late emergence of the infra-red beams covering the ground floor, not relied upon during the trial and the extent and efficacy of which remain unclear. Of course, it was never put to Mr McAndrew in cross-examination that he would have been content to accept guardwire along only two sides of the cage because of infra-red beams protecting the ground floor, but I have little doubt that his response is likely to have been (whether without consulting Mr Thomas or after such consultation) that he still required the Risk Improvement Requirements to be complied with before he would lift the Endorsement.

It is just possible (but in my view unlikely) that, after consulting Mr Thomas, Mr McAndrew might have been prepared to relax the terms of the Endorsement, but on the strict understanding that the claimant complied with the Risk Improvement Requirements within a short specified time frame or on the basis of further security measures such as security guards on the premises outside business hours being taken, until the Risk Improvement Requirements had been complied with.

What in my judgment would not have happened, if Mr McAndrew had been told the true position on 20 February 2007, is that he would simply have lifted the Endorsement without imposing terms. Accordingly, I am satisfied that Mr McAndrew was induced by material misrepresentation or non-disclosure to lift the Endorsement and vary the contract of insurance.

In the circumstances, the defences of collusion and breach of warranty fail. However, the insurers are entitled to avoid or rescind the variation and, accordingly, Endorsement 6 remained in place at the time of the theft. It necessarily follows that, since the cigarettes and tobacco were stolen from the mezzanine floor, there was no theft cover in place in respect of them. The claimant’s claim for an indemnity under the Policy therefore fails.