LMH v EGK

the high threshold for establishing a successful s.68 challenge;

the fact that, objectively viewed, there was nothing about the course of the arbitration or Award which was in any way surprising or outside the contemplation of reasonable parties who have agreed to arbitrate their disputes; and

the fact that the one grievance which may have some substance does not fall within any of the limited categories of serious irregularity recognised by s.68.

Ground 1 alleged that the Tribunal failed to deal with the issue of whether the breach of the MSA found by the tribunal caused EGK any loss.

Ground 2 alleged that if and insofar as the Tribunal did deal with that issue, it followed that the Tribunal had failed to give reasons for its Award contrary to s.52(4) of the 1996 Act and/or Article 32(2) of the ICC Rules.

If a tribunal has dealt with an issue in any way, that is sufficient for the purposes of s.68(d) of the 1996 Act, whether the tribunal has dealt with the issue well, badly or indifferently, nor does it matter that the tribunal might have expressed its conclusions at greater length RAV Bahamas Ltd v Therapy Beach Club Inc [2021] UKPC 8, [43].

It is not sufficient to establish that the award does not comply with s.68(2)(h), because it does not “contain the reasons for the award” as required by s.52(4), merely because the reasons are said to be inadequate: UMS Holding Limited v Great Station Properties SA [2017] EWHC 2398 (Comm), [134] and Islamic Republic of Pakistan v Broadsheet LLC [2019] Bus LR 2753, [22]-[23] (although I should record that Mr Trompeter KC reserved the right to challenge the correctness of those decisions, should the opportunity to do so arise in a higher court).

The Tribunal concluded that “the decisive term to be negotiated between the Parties was the revenue-sharing model” (Award, [202]).

LMH had offered the lowest possible margin to EGK from the range communicated to LMH by its appointed expert (Award, [203(3)]).

While Article 3.1 did not have the effect that the parties were not free to refuse to continue the MSA, there was a liability in damages “to the extent that the violation of the specific good faith obligation causes further expectational damages” (Award, [220]) and “the Tribunal finds that this is the case here” (Award, [221]): a clear finding “dealing with” the issue of causation.

LMH’s bad faith deprived EGK “from ensuring a further [MSA] which otherwise it would have been able to obtain” (Award, [222]): another clear finding “dealing with” the issue of causation.

The Tribunal identified various objective factors which were “of guidance” in determining “for which term the Parties may have entered into a new [MSA] if the negotiations had been conducted in good faith” (Award, [269]). The Tribunal concluded that there was “little likelihood” that the parties would have agreed a 20-year term given their mutual hostility (Award, [270]), before arriving at a six-year figure (Award, [275]). The use of objective information on market transactions, adjusted by reference to the fact that the parties’ poor relationship made a long term unlikely, was a perfectly proper (and in any event, appropriately reasoned) means of answering the hypothetical counterfactual issue of what the parties would have done if they had negotiated in good faith.

The Tribunal noted for a second time the central role of revenue-sharing in the negotiations (Award, [276]), and expressly found that a 50:50 revenue share “would be the most probable outcome of negotiations” (Award, [284]). That finding of itself is a further finding on causation (the Tribunal could not have found that good faith negotiations would have led to agreement on a 50:50 revenue share had they not found they would have led to agreement) and the Tribunal gave various reasons for their conclusion, including evidence that this was acceptable to EGK, and fell within an objectively reasonable market range.

The Tribunal therefore expressly found that good faith negotiations would have led to an extension, and expressly explained why they were satisfied that good faith negotiations would have led to a resolution of the key issues of duration and revenue-share, the latter being the principal ground of contention in the negotiations actually conducted.

As Teare J noted at UMS Holding Ltd v Great Station Properties SA [2017] EWHC 2398 (Comm), [28]:

“A contention that the tribunal has ignored or failed to have regard to evidence relied upon by one of the parties cannot be the subject matter of an allegation of a serious irregularity within s 68(2)(a) or (d), for several reasons. First, the tribunal’s duty is to decide the essential issues put to it for decision and to give its reasons for doing so. It does not have to deal in its reasons with each point made by a party in relation to those essential issues or refer to all the relevant evidence. Second, the assessment and evaluation of such evidence is a matter exclusively for the tribunal. The court has no role in that regard. Third, where a tribunal in its reasons has not referred to a piece of evidence which one party says is crucial the tribunal may have (i) considered it, but regarded it as not determinative, (ii) considered it, but assessed it as coming from an unreliable source, (iii) considered it, but misunderstood it or (iv) overlooked it. There may be other possibilities. Were the court to seek to determine why the tribunal had not referred to certain evidence it would have to consider the entirety of the evidence which was before the tribunal and which was relevant to the decision under challenge. Such evidence would include not only documentary evidence but also the transcripts of factual and expert evidence. Such an inquiry (in addition to being lengthy, as it certainly would be in the present case) would be an impermissible exercise for the court to undertake because it is the tribunal, not the court, that assesses the evidence adduced by the parties. Further, for the court to decide that the tribunal had overlooked certain evidence the court would have to conclude that the only inference to be drawn from the tribunal’s failure to mention such evidence was that the tribunal had overlooked it. But the tribunal may have had a different view of the importance, relevance or reliability of the evidence from that of the court and so the required inference cannot be drawn. Fourth, s 68 is concerned with due process. Section 68 is not concerned with whether the tribunal has made the ‘right’ finding of fact, any more than it is concerned with whether the tribunal has made the ‘right’ decision in law. The suggestion that it is a serious irregularity to fail to deal with certain evidence ignores that principle. By choosing to resolve disputes by arbitration the parties clothe the tribunal with jurisdiction to make a ‘wrong’ finding of fact.”

As I noted in Rolls-Royce Holdings plc v Goodrich Corporation [2023] EWHC 1637 (Comm), [192], where “the issue to be addressed is a hypothetical one … the weight to be accorded to witness evidence is significantly less than in cases of direct observation of historical facts.”

There is no absolute rule that party must challenge every part of a witness’s evidence in cross-examination if it is to invite a tribunal not to accept that evidence: BPY v MXV [2023] EWHC 82 (Comm), [67]-[68]. I would add this is a fortiori the case when the issue does not relate to a matter of historical fact (what the witness saw, heard or understood at the relevant point), when a challenge will often carry an imputation on the witness’s honesty, but to what is essentially a matter of opinion – what would have happened in a counterfactual scenario had the parties conducted themselves differently.

It was not necessary for EGK to adduce evidence addressing the counterfactual hypothesis of what would have happened had LMH negotiated in good faith.

As set out at [13], the Tribunal’s conclusions as to what would have happened were based on evidence: historic evidence that a 50:50 margin was acceptable to EGK at the time, expert evidence as to the terms on which comparable transactions were being entered into at the relevant time; consideration of what was required for the transaction to be at least break-even for EGK (Award, [203(1)] and EGK’s desire to extend negotiations (Award, [207]).

As to the failure to challenge the evidence of LMH’s witness, I would not accept that the failure to cross-examine a witness as to what the outcome of hypothetical negotiations would have been had their employer acted in good faith involved any form of unfairness in this case. The nature of EGK’s case, and its contention that if LMH had acted in good faith a deal would have been done, was clear from the filings which preceded the service of this statement, and the witness had ample opportunity to offer any observations on that case.

There are two more fundamental problems with this complaint. First, the evidence given by the witness was limited to the fact that a deal with a revenue-sharing proposal of 73:27 or 66:33 in EGK’s favour would not have been acceptable, that “there was no way that the MSA could be prolonged in its current form” (emphasis added) and that, following EGK’s 66:33 offer, “given how far apart we were from each other, particularly on this crucial point, it seemed obvious that we would not manage to come to an agreement on the terms of our collaboration.” However, the Tribunal’s findings were that a very different deal would have been negotiated, with a 50:50 revenue share and a six-year term. The witness had given no evidence on that issue.

Further, the witness gave extensive evidence intended to show that LMH had negotiated in good faith and had not delayed in putting proposals to EGK. However, those contentions were rejected by the Tribunal. The Tribunal clearly rejected the witness’s contention that LMH’s offers were not made on a “take it or leave it” basis” (Award, [203]), her evidence as to the reason for taking the lower end of the range in the expert report was criticised (Award, [203(3)]) and her evidence as to the reasons for delay rejected (Award, [206(1)]). That is a particularly unpromising background for the suggestion that the process was unfair because the witness was not afforded the opportunity to comment on how LMH would have acted had it negotiated in good faith.

It can be read as reflecting the expected position under most systems of private law: where no contract would have been concluded even if good faith had been exercised, a reliance measure of loss caused by the bad faith negotiations would be recoverable (e.g. lost management time, delay, lost opportunity etc) whereas if a contract would have been concluded had good faith been exercised, an expectation measure of loss – namely the profits which would have been earned under the contract – is recoverable.

It can be read as meaning that the reliance measure is recoverable unless there was a legal obligation to conclude the contract, as opposed to a legal obligation to negotiate in good faith, even where the exercise of good faith would have led to the conclusion of a contract.

In Weldon Plant Ltd v The Commission for the New Towns [2001] 1 All ER (Comm) 264, [33], His Honour Judge Humphrey Lloyd QC held:

“The latter raises a point that frequently arises: is an arbitral tribunal obliged to confront a party with a proposed finding when it is not one that a party has sought? Obviously the tribunal should inform the parties and invite submissions and further evidence before making an award if the finding is novel and was not part of the cases presented to the arbitral tribunal. On the other hand in many arbitrations, especially those in the construction industry, there are many findings other than those which the parties have invited the tribunal to make. Matters of quantification and valuation frequently lead to the tribunal taking a course which is not that put forward by either party, but which lies somewhere between. ‘Doing the best one can on the material provided’ almost inevitably produces such a result. Provided that the finding is not based on a proposition which the parties have not had an opportunity of dealing with the arbitral tribunal will not be in breach of its duties under s 33 of the 1996 Act nor will its award be liable to challenge under s 68(2)(a) or (d) of that Act if it makes such a finding without giving the parties a chance of dealing with it. In many such cases the tribunal will have been appointed for its expertise so that in addition there would be no obligation to consult the parties. Any other course could defeat the objective of avoiding ‘unnecessary delay and expense’ as provided by s 1(a) of the 1996 Act.”

In Bulfracht (Cyprus) Ltd v Boneset Shipping Co (The Pamphilos) [2002] 2 Lloyd’s Rep 681, 687, Colman J observed:

“It has to be emphasized, however, that the duty to act fairly is quite distinct from the autonomous power of the arbitrators to make findings of fact. Thus, whereas it may normally be contrary to the arbitrator’s duty to fail to give the parties an opportunity to address them on proposed findings of major areas of material primary facts which have not been raised during the hearing or earlier in the arbitral proceedings, it will not usually be necessary to refer back to the parties for further submissions every single inference of fact from the primary facts which arbitrators intend to draw, even if such inferences may not have been previously anticipated in the course of the arbitration. Particularly where there are complex factual issues it may often be impossible to anticipate by the end of the hearing exactly what inferences of fact should be drawn from the findings of primary fact which have been in issue. In such a case the tribunal does not have to refer back its evidential analysis for further submissions. A typical situation is where arbitrators arrive at a conclusion on an issue of expert evidence which differs to some extent from that put forward by either opposing expert. In many cases, such as this, the arbitrators have been appointed because of their professional, legal, commercial or technical experience and the parties take the risk that, in spite of that expertise, errors of fact may be made or invalid inferences drawn without prior warning. It needs to be emphasized that in such cases there is simply no irregularity, serious or otherwise. What has happened is simply an ordinary incident of the arbitral process based on the arbitrator’s power to make findings of fact relevant to the issues between the parties.”

In Cameroon Airlines v Transnet Ltd [2004] EWHC 1829 (Comm), Langley J found that there had been a serious irregularity when the tribunal, faced with “extreme positions adopted by both parties”, had come up with its own measure of the fair value of services, by taking the contract price for the services and deducting commission, when the parties and both experts had conducted the hearing on the agreed basis that the issue was to be determined by reference to the cost of providing the services.

In Republic of Kazakhstan v World Wide Minerals Ltd [2020] EWHC 3068 (Comm), His Honour Judge Pelling KC set aside an award of damages under s.68 of the 1996 Act, when the claimant had presented only a single global damages claim, on the assumption that it would succeed on all of the breaches of the investment treaty it had alleged, but had only succeeded on a limited number of those breaches. The arbitration respondent had submitted that if that was the outcome, the tribunal should not award any damages, and the claimant had submitted that the tribunal should issue a partial award on liability and “come back to the parties on damages”. However, the tribunal had fashioned a damages award of its own devising. His Honour Judge Pelling KC referred to the following passage in Popplewell J’s judgment in Reliance Industries Ltd v The Union of India [2018] EWHC 822, [32]:

The first was the discounting of losses by one third to reflect the fact that the sole damages claim assumed a loss of profits on the basis that the claimant had wrongfully been deprived of a lease which covered both a beach club and a seafood restaurant, whereas it was held that the lease did not extend to the seafood restaurant. This was held to be a serious irregularity because no figures had been put forward to disaggregate the loss of profits claim between the two properties, and “the possibility of there needing to be a disaggregation …. was simply not ‘in play’ and not ‘in the arena’” ([82]).

The discounting of losses by a further 15% because the figures given in the expert evidence were based on memory and unsupported by any documents. The board held that this was not a serious irregularity because “it should have been obvious to RAV that Mr Ocasio's evidence was problematic and weakened in those respects and it had ample opportunity to address the arbitrator on that matter. In that sense, the issue was ‘in play’ and ‘in the arena’” ([86]).

EGK adduced expert evidence from PwC calculating loss using a discounted cash flow (DCF) methodology on two scenarios: if the existing MSA was continued in its current form for 20 years, or if it was renewed for 20 years on a 33:67 revenue sharing model. It identified inputs into the calculation in the form of (a) an assumed market penetration rate; (b) EGK’s market share; (c) EGK’s average revenue per user or ARPU; and (d) a discount rate.

LMH adduced expert evidence from Mazars attacking the assumptions used by PwC for market penetration, market share, ARPU and the discount rate. LMH’s quantum expert stated “I have not been instructed to provide any alternate calculation of any potential damage as it is for the Claimant to prove the validity and grounds of its alleged damage”. Thus, LMH’s expert did not offer alternative figures for market penetration (stating in cross-examination “I believe that it was not my position to perform the valuation of [EGK’s] claim on behalf of [EGK]”), market share, ARPU (“it was not for me to do that”) or discount rates. He did not offer an alternative growth rate but appears to have been asserting that no growth should be assumed, and he suggested an alternative Compound Annual Growth Rate stick should have been used for ARPU calculation but without providing an alternative figure. He also raised a series of criticisms of the reliability of PwC’s inputs: the reliability of EGK’s audited statements and scepticism as to the reliability of EGK’s financial plan. He noted the evidence of LMH’s telecommunications expert that a 3–5-year duration was closer to industry standards than the 20 years assumed by PwC, but offered no calculation of the impact of that change.

Generally, the criticisms made by LMH’s expert were not the subject of alternative figures, but put forward by way of an attack on the reliability of PwC’s inputs. To the extent that the Tribunal concluded that these criticisms had some force, this was inevitably going to require an impressionistic allowance in any damages calculation.

In EGK’s reply evidence, PwC took the position that the expert evidence served by LMH did not require any adjustment to its figures. In LMH’s rejoinder evidence, Mazars maintained their criticisms of PwC, but without providing alternative inputs or figures, stating that the expert had “not been instructed to provide any alternate calculation of any potential damages as it is for the Claimant to quantify is alleged damage.”

The evidentiary hearing did not see any material change in the evidence before the tribunal. There was no suggestion from either party that, as part of the process of resolving damages issues, the Tribunal might need to revert to the parties or their experts. Indeed, the Tribunal expressly confirmed with the parties that the record was closed. In its post-hearing brief served on 25 August 2021, EGK invited the Tribunal to satisfy itself as to the assumptions used, and observed that it might be necessary for the Tribunal to follow the practice under English law of applying a “broad axe” or “broad brush”, doing its best with the evidence available. LMH served its post-hearing brief on the same date, repeating its earlier submissions that there was no sufficiently reliable estimate of loss.

The proceedings were closed under Article 27 of the ICC Rules on 7 September 2021. The time for the delivery of a Final Award – not an interim Final Award – was fixed for 30 September 2021, although that date was later extended. This was consistent with the fact that neither party had ever suggested that the Tribunal would ever produce anything other than a Final Award, determining all the issues.

First, they adjusted PwC’s calculations to reflect their decision on the duration of the extension.

Second, they made an adjustment to PwC’s figures to reflect their conclusion that the market share would have remained static, rather than grown as PwC had assumed.

However, those figures still involved using PwC figures which assumed a 27:73 revenue share in EGK’s favour. As result, at the third stage, the Tribunal divided this figure by 73, and multiplied it by 50, to reflect its 50:50 revenue share conclusion. That 50:50 figure was itself on the record, being one of the suggestions made by EGK to LMH in the course of negotiations.

The Tribunal applied these adjustments to the figures from PwC’s first and second scenarios. Its fourth adjustment was to take the lower of those figures.

However, that still left various criticisms advanced by LMG: the market penetration rate and ARPU being too high, the unreliability of the data used and the discount rate being to low. The Tribunal accepted the validity of these points. These elements were “taken holistically” and a 20% reduction applied to achieve a final figure of €7,590,400.

Although the Award does not expressly say so, experience suggests that the Tribunal are likely to have “sense-checked” that final number against their overall impression of the evidence.

The witness statement of LMH’s solicitor complains that “the Tribunal gave no indication to the parties that it was intending to carry out its own calculation” and “did not offer the parties an opportunity to undertake their own calculations based upon the ‘revised components’ found by the Tribunal”. With respects, that complaint is unrealistic. LMH had chosen not to provide the Tribunal with its own figure or model, nor show the effect of the adjustments for which it was contending for on PwC’s model. Nor was there any suggestion that the Tribunal should do anything other than produce a Final Award determining all issues. In those circumstances, if the Tribunal upheld EGK’s case on breach and causation, then absent the rather unlikely prospect of it finding that no loss had been established, acceptance by the Tribunal of some or all of LMH’s points would require the Tribunal to make its own significant adjustments to the claimed figure.

A respondent who chooses to run a “scorched earth” response to an overstated quantum case necessarily accepts the possibility of the tribunal fashioning a reduced award from the “building blocks” provided.

It was suggested that the Tribunal had failed to follow “the agreed methodology” for calculating loss, namely a DCF. However, these references to an “agreed methodology” must be seen in their context. The extent of that agreement is well-captured in an answer given by LMH’s expert in cross-examination: namely that “in principle” he did not dispute the accuracy or methodology of a DCF calculation (of which, as he noted, there are many kinds). It was a DCF calculation which the Tribunal used (Award, [358]) – making adjustments to the PwC DCF calculation, and applying a general discount to the output which reflects, among other matters, issues which could not have been addressed by adjustments internal to the DCF calculation.

It is the case that the Tribunal identified as one of the reasons why it could not produce a revised DCF calculation the fact that it did not have the underlying spreadsheets for the PwC report (Award, [357]). LMH had sought disclosure of those documents at an earlier stage of the reference, and that application had been refused, for what the Tribunal must have concluded were good and sufficient reasons. The hearing had thereafter been conducted without reference to the spreadsheets, and the record closed. The absence of those spreadsheets was simply an unalterable fact by the time the evidentiary hearing concluded, of which both parties were aware and the implications of which were obvious: the Tribunal would not be plugging numbers into those spreadsheets.

To the extent that complaint is made about the Tribunal’s use of a 20% reduction to address a number of the complaints raised by LMH, the various matters leading to the adjustments were all “in play”; some would admit only of an impressionistic adjustment and LMH had not provided its own calculations of the impact of the others. In these circumstances, as with the 15% reduction for unreliable data in RAV Bahamas, I am satisfied that this fell on the right side of the line.

A calculation of total revenue was made of 160.

Various expenses were deducted including the 27% share of revenue (40) to arrive at a gross profit for the year in question of 100.

Various other adjustments were then made to that figure including tax, depreciation and EGK’s OPEX to arrive at the “free cash flow” of 42 which was discounted to 40.

In making its adjustment to reflect a revenue share of 50:50 rather than 83:27, the Tribunal did not adjust the 40 figure in (ii) to 75, but adjusted the 40 figure in (iii) to 27.

The tribunal can make the correction itself, if the issue falls within the scope of the relevant power.

Even if it does not, it is possible that the arbitral will admit the error, with the result that s.68(2)(i) may be engaged (the categories of irregularity, for this purpose, potentially being wider than those recognised by s.68(2)(a) to (h): CNH Global v PGN Logistics Ltd [2009] EWHC 977 (Comm), [50]-[51]).

If the arbitral tribunal refuse to correct such an error when the relevant power is properly engaged, it may be possible to treat that refusal as constituting, either on its own or in combination with other matters, the serious irregularity required by s.68 (indeed this is the sense of the passage in Merkin and Flannery).

In any event, the tribunal’s response even where it refuses such an application is likely to be informative in the context of any s.68 application (for example the tribunal might explain that the matter complained of reflects a conscious and informed choice).