Sophie J Lamb QC, Samuel Pape, Laila Hamzi and Eleanor Scogings, Latham & Watkins LLP

This is an extract from the third edition of GAR’s The Guide to Damages in International Arbitration. The whole publication is available here

Bifurcation

The parties and tribunal are often required to consider whether damages should be assessed in the same phase as the merits, or whether the proceedings should be ‘bifurcated’ such that damages should fall to be assessed separately, only after the tribunal has determined the merits of the dispute (as well as admissibility and jurisdiction, if these issues are in dispute). In this section, we consider the circumstances in which it might be appropriate for proceedings to be bifurcated, and if so, how the issues of merits and quantum might be divided between phases.

The practice of bifurcation

All major institutional rules allow for proceedings to be bifurcated, either under the general powers conferred upon the arbitral tribunal or, more infrequently, under specific provisions regarding the bifurcation or phasing of proceedings.

Institutions have shown particular interest in this issue in recent years, viewing bifurcation as a possible procedural tool that can allow for a more efficient and cost-effective dispute resolution process. The revised IBA Rules on the Taking of Evidence in International Arbitration were adopted by the IBA Council on 29 May 2010 (the IBA Rules of Evidence). They advocated a more prescriptive approach to arbitration, inviting the tribunal to promptly identify issues for which a preliminary determination may be appropriate and encouraging the systematic organisation of arbitral proceedings by issues or phases.

In 2012, the ICC Commission on Arbitration in its report, ‘Controlling Time and Costs in Arbitration’ suggested that ‘arbitral tribunals should consider, or the parties could agree on, bifurcating the proceedings when doing so may genuinely be expected to result in a more efficient resolution of the case.’

In a recent issue of its Dispute Resolution Bulletin, the ICC examined its dispute resolution statistics from 2017 and observed that out of a total of 512 awards approved by the ICC Court in 2017, 143 were partial awards. The ICC’s dispute resolution statistics from 2016 showed a similar proportion of partial awards (with 140 partial awards out of a total of 479 awards). The proportion of partial awards (29 per cent in 2016 and 28 per cent in 2017) was higher than in previous years, potentially reflecting an ‘increasing trend to bifurcate proceedings’.

Further, the bifurcation of proceedings, and in particular clarification of the rules on bifurcation, was marked as a topic for consideration in the ongoing amendment of the ICSID Rules.

On 3 August 2018, ICSID released the proposed amendments to its rules, which includes Arbitration Rule 37 and Additional Facility Rule 47, which expressly permit bifurcation and provide that parties must request bifurcation within 30 days of the memorial on the merits or ancillary claim.

Is bifurcation effective in reducing the time and costs of dispute resolution?

The limited studies available on the subject of the effectiveness of bifurcation for these purposes are inconclusive. The major institutions do not tend to release specific data on bifurcation. As one author has observed, the only real way to assess the effect of bifurcation is to examine published awards and to compare cases – bearing in mind the myriad variations in complexity.

Based on this sort of case-by-case analysis in the investor–state context, the author of a study in 2011 found that the time for concluding the (then known) 45 bifurcated ICSID cases was 3.62 years, compared to 3.04 years for non-bifurcated cases. Moreover, of the 19 concluded ICSID Additional Facility cases, for which information was publicly available at the time, 10 bifurcated cases took an average of 3.39 years to reach final award, as compared to 2.96 years for the nine non-bifurcated cases. Of the 10 cases that were bifurcated, eight were split between a jurisdiction and merits phase and two were split between liability and quantum phases.

Given the limited data available, the difficulty in isolating bifurcation as a variable in proceedings in light of all the other parameters in international arbitration, and the fact that studies tend to focus on investment treaty arbitration awards rather than commercial awards (of which there are many more, but most of which are confidential), it is difficult to draw any general conclusions.

In particular cases, tribunals have initially bifurcated the proceedings in the interest of procedural efficiency, or following agreement by the parties, only to later realise that it would have been faster and more cost-effective to proceed to a single award. In other cases, bifurcation of the proceedings between the liability and damages phase has necessarily led to a lengthier timetable. For instance, in the NAFTA case of Mobil Investment and Murphy Oil v. Canada, the damages phase added approximately two additional years to the procedural timetable.

Might bifurcation be appropriate?

The appropriateness of bifurcating the liability and damages phases of a dispute will depend on the circumstances of the dispute. Significant time and cost efficiencies will result where the quantification of claims is likely to be a time-consuming and complex undertaking, and a finding on liability could ultimately be dispositive of the dispute. For instance, in Methanex v. USA, Methanex sought damages of approximately US$970 million plus interest and costs. The tribunal decided to hear jurisdiction and the merits in a first phase, and to bifurcate quantum. Ultimately, the claims failed in the first phase, averting the need to undertake a complex quantification phase. The tribunal in Suez and Vivendi v. Argentina considered bifurcation of quantum to be appropriate in this case ‘because of the complexity involved in ascertaining damages, a matter extensively argued with widely differing conclusions by each party’.

Bifurcation might also promote the prospects of an early settlement. A finding on liability might encourage the parties to settle, for example, where the scope of the dispute on damages is narrow, or if bifurcation shores up weaknesses in certain aspects of the parties’ arguments as previously articulated.

In addition, bifurcation can sometimes reduce the complexity and scope of the dispute for the purposes of determining quantum, again saving costs and time. For instance, in Suez v. Argentina, the tribunal decided to bifurcate the merits for reasons of judicial economy. The case comprised three consolidated claims arising out of a concession to operate water and waste-water services in Buenos Aires. The tribunal found that Argentina had breached its obligations under three bilateral investment treaties by denying the claimants’ investments fair and equitable treatment. After finding on liability, the tribunal observed that the case was very complex and that the record was extraordinarily voluminous. In addition, each party had engaged financial specialists who had reached vastly different conclusions on damages. In these circumstances the tribunal held that it was appropriate to dedicate a separate phase of the proceedings to damages. In this regard, it was stated that:

the Tribunal by rendering a decision on liability … and thereby defining the scope of its investigation with respect to a determination of damages will be able more efficiently to define the mission of the independent expert that will assist the Tribunal in this determination.

In certain cases, a tribunal’s decision, in the liability phase, on the correct valuation date for assessing damages can result in an important narrowing of issues to be determined during the damages phase. Similarly, cases involving complex questions of causation, or difficult valuation exercises tend to be particularly propitious for bifurcation, though few reported decisions provide reasons for the bifurcation of the damages phase. In Pope & Talbot and SD Myers v. Canada, the investors claimed that the challenged measures affected their ability to export products from Canada. Damages were left to be determined in a standalone phase, presumably on the basis that complex issues of causation were involved.

By contrast, in other disputes, questions of liability and quantum might be so intertwined that bifurcation would result in significant duplication of work, to the extent that the issues can be cogently divided at all. Separating issues of causation, mitigation and remoteness into questions of pure ‘liability’ or ‘damages’ may be somewhat artificial and difficult. In certain cases, the fact that damage has been suffered might even affect a finding on liability such that it is impossible for damages to be considered in isolation. In Electrabel v. Hungary, bifurcation of liability and quantum was found to have been ‘at least unfortunate’. The tribunal, upon the parties’ initiative, had determined to bifurcate the liability and damages phases. In turn, the claim relating to the Energy Charter Treaty’s ‘fair and equitable treatment’ (FET) standard arguably combined issues of both liability and quantum. Because of the bifurcation of the merits, the tribunal decided to defer its decision on both liability and quantum in relation to the FET claim to a later phase of the proceedings. In their submissions in the second phase of the proceedings, the parties agreed that loss was not in fact relevant to the issue of liability under the FET claim. By this stage, however, the claim had already been bifurcated to the second phase, and had given rise to a significant volume of submissions and exchanges by the parties.

Bifurcation of liability and quantum may also not be appropriate in certain cases where there are allegations of expropriation or creeping expropriation. The evidence that is used to determine whether, and to what extent, there has in fact been an expropriation is often also relevant and intertwined with the assessment of whether the claimant has incurred loss.

In other instances, even where there is bifurcation of liability and quantum, it is often the case that there will still be factual witnesses, as well as experts, involved in the quantum stage of the proceedings.

Nonetheless, it may in appropriate circumstances be possible to establish ‘a broad division between liability and quantum [and] … to determine the principles on which damages should be awarded, while leaving the pure arithmetical calculations to a second stage’. The tribunal in CME v. Czech Republic  took this approach. It bifurcated the quantum phase upon the parties’ agreement and left only the amount of monetary damages for determination in the quantum phase. An unconventional but emerging method for case management in civil trials in the United States is known as ‘reverse bifurcation’, whereby the issue of damages is heard and determined before liability. This can be a useful tool and one on which those involved in international arbitration may wish to draw, particularly for cases in which there is a serious argument that the claimant has suffered little or no damages. An early determination to this effect may lead to early disposal or settlement of the case. Reverse bifurcation might also alert the parties to weak elements of their respective arguments on liability, again encouraging settlement. Where reverse bifurcation is used in arbitration, care will need to be taken to protect the due process rights of the parties and to ensure that neither party’s case on the merits is pre-judged or otherwise unduly prejudiced.

The procedure in a bifurcated damages phase

Careful consideration needs to be given as to what the procedure within a bifurcated damages phase should look like. The damages phase of proceedings can sometimes be just as long and costly as the merits phase, sometimes unjustifiably. In such instances, bifurcation can serve to complicate and prolong the dispute, rather than assist in its swift and cost-effective resolution. For example, in Suez v. Argentina, the tribunal decided of its own accord at the conclusion of the liability phase, and after the parties had already made detailed submissions on damages (which were later revised), that it would bifurcate the damages phase because of the complexity of ascertaining damages.

One potential pitfall of bifurcation consists in the quantum phase mirroring the liability phase, where this would be disproportionate or otherwise inappropriate. In practice, damages phases will almost invariably involve one or two rounds of written submissions by the parties as well as the submission of written witness statements and expert opinions on quantum (which will often play a central role in the phase). These submissions are often followed by (or take place concurrently with) document production, and an oral hearing. In some cases, such as in CME v. Czech Republic there may be both an ‘evidentiary hearing’ and a ‘final hearing’. An evidentiary hearing gives the parties the opportunity to provide both factual and expert evidence to the tribunal and for counsel to make legal submissions. A final hearing allows the parties to submit their final pleadings prior to the close of the arbitration. The parties are sometimes then called upon to submit post-hearing submissions on damages.

All these steps are not necessarily required in a damages phase, and parties and tribunals would be well served to build the procedure for the quantum phase using a ‘bottom-up’ approach, starting with the bare minimum and considering carefully which features are really necessary and proportionate. Nonetheless, there may well also be cases in which the quantum phase ought rightly, by reason of its significance or complexity, to be lengthier and include more procedural steps than the merits phase.

Documentary evidence, disclosure and confidentiality

In this section, we focus on documentary evidence and particular aspects of disclosure in international arbitration that may be relevant to the assessment of damages. In particular, we draw attention to the treatment of confidential information that may be especially relevant to damages issues.

In international arbitration proceedings, the parties will almost invariably seek to submit the documents on which they wish to rely, either through the submission of standalone indexed bundles, or as exhibits to witness statements or pleadings. In relation to damages, parties will often submit various documents accompanying their expert reports (discussed below). For example, where a discounted cash flow method is used, internal evidence of the business such as documents pertaining to operating cash flow and capital expenditures may accompany an expert report. Similarly, the parties may submit market reports and other data as a comparator for determining growth forecasts and for applying the discount rate.

Where one party may have access to documents relevant to the issue of damages that are not in the possession of the other party, for instance, internal documents identifying whether a party has mitigated or contributed towards its losses, it may be appropriate for the proceedings to include a procedure for documentary disclosure. In other circumstances, document production may be inappropriate.

The IBA Rules of Evidence provide a potential defence to disclosure on ‘grounds of commercial or technical confidentiality that the Arbitral Tribunal determines to be compelling’. This potential defence is often of particular relevance to documents relating to damages. Commercial disputes often arise between competitors (such as between joint venture partners in the extractive sectors), or between parties whose businesses legitimately depend upon securing the confidentiality of certain information (for example, gas price review disputes often involve questions of pricing strategy and internal rates of return that can be of utmost sensitivity in the industry).

The parties invoking confidentiality as a defence to production are required to give reasons for each objection and, in some cases, to indicate the documents, if any, that they would be prepared to produce instead of the documents requested. Generally, tribunals require convincing arguments before they grant confidentiality protection. Such arguments may include that the information has industrial significance and may provide the disclosing party with a competitive advantage, that the information contains data and know-how for the development of the project subject to the dispute, or that the information is protected by a specific confidentiality agreement because it involves third-party proprietary technology.

In exceptional circumstances, where the production of a document is objected to, but the tribunal determines that it should not see the document – because of confidentiality concerns – the IBA Rules of Evidence provide that the tribunal may appoint an independent and impartial expert, bound by confidentiality, to decide on the objection. Ultimately, however, determination of the validity of the objection rests with the tribunal.

A recent development that has a bearing on document production in international arbitration is the EU General Data Protection Regulation (GDPR), which came into force in May 2018. The GDPR imposes significant restrictions on the processing of personal data, including when contained in documents that may be relevant to matters in relation to damages. As a general matter, the processing and transferring of personal data during an arbitration must be minimised. This may require pseudonymisation of data so that personal data, which may include, for example, customer information or employee details, is excluded when the relevant document is disclosed. Similarly, the GDPR may in some instances require the redacting of irrelevant personal data.

Treatment of confidential information

In practice, where documents are confidential, tribunals have taken a variety of case-specific measures to protect one party’s information, while taking into account the need to secure the other party’s rights of due process. Issues surrounding the treatment of confidential information are particularly prevalent in gas pricing arbitrations. In these cases tribunals are aware of the particular sensitivities around disclosing, for example, market models, projections and prices, to competitors in the same market. Redaction of documents is a commonly used method, although this will only be helpful where the information to be redacted is not material to the dispute. Where redaction is not appropriate, a tribunal may make protective orders prescribing that access to the document be restricted to certain individuals only. Frequently, these individuals will include members of the tribunal and legal counsel, on the condition that they do not disclose the document to any other person, including the counsel’s client. In these circumstances, however, it may be difficult for legal counsel to assess or test the information if the information is particularly technical and legal counsel are unable to take instructions from their clients in relation to it. The reverse is true when only the tribunal and experts are given access to the document. Accordingly, access to the document is sometimes given to the tribunal, the parties’ counsel and experts, but not the parties. Further, in other cases, the dispute may require that counsel and experts be given additional instructions by the parties themselves, in which case, access may also be given to specified party representatives.

Additional measures used to protect confidentiality include requiring the return or destruction of the documents in addition to restricted access, or restricting access to the visual inspection of the documents rather than circulating them. These measures, among others, together with appropriate technological solutions, are often adopted in cases involving sensitive information, particularly as parties and tribunals increasingly recognise the importance of protecting the confidentiality of sensitive information from cyberthreats.

Witness evidence

Although the building blocks for a party’s case on damages will often be found in hard data and, therefore, documents, the tribunal will often benefit from witness testimony that explains that hard data.

Witnesses of fact

The lion’s share of the evidence on damages is often given by (or at least presented through) the evidence of expert witnesses. However, witnesses of fact can also play an important role in supporting the appropriateness of any valuation or calculation. For example, members of the parties’ finance departments or those who otherwise have first-hand knowledge of the figures and projections on which the claim for damages is based can be of assistance to tribunals, particularly where the source and meaning of the data is in dispute, or at least not self-evident.

Witnesses of fact may also be called upon to give evidence related to issues of causation. For example, a chief executive or other commercial representative may be asked to give evidence on intended business plans, projections and hypothetical scenarios which relate to issues of causation which can have an impact on quantum. Similarly, witnesses of fact may also be required to give evidence related to mitigation including the steps taken to mitigate the loss, and reasons for pursuing particular causes of action or not pursuing others.

Expert witnesses

Expert witnesses provide the tribunal with their opinion on the parties’ cases on damages. They do so on the basis of their particular expertise, rather than their knowledge of particular facts. There are two principal types of experts used for assessing damages: forensic or advising experts, and testifying experts. Forensic or advising experts are often retained where they have had previous involvement in the issues in dispute, and have already generated work product that can form the starting point for the testifying expert’s evidence. The forensic or advising experts may not have a sufficient degree of independence from the party, or may not be experienced in adopting the relevant methodology for the calculation of damages in international arbitration. In contrast, testifying experts will most often be retained by the party’s counsel rather than by the party itself, and will have a greater degree of independence and potentially greater specialist experience to address the subject of their testimony. Notwithstanding bifurcation of the proceedings, parties often consider it useful to engage with experts at an early stage. This can assist the parties with the development of their legal theories and strategies, and help them assess the robustness of their case and expectations as to quantum. Engaging experts at a preliminary stage may also be a necessary requirement if the case is being funded through third-party funding.

Experts from a variety of disciplines are often instructed to give evidence on damages, for instance, accountants, engineers, economists, financial and valuation analysts, business valuation experts and oil market experts. In some cases, a combination of industry experts and valuation or economic experts may be helpful. For example, in Occidental v. Ecuador,12 experts were appointed, including oil and gas consultants and quantum experts. Similarly, in Exxon Mobil v. Venezuela, the claimants appointed eight experts, including an oil price forecasting expert and oil reserves evaluation experts. Separate quantum experts were also engaged by the claimants to advise on the extent of damages resulting from production and export curtailments. In other instances, expert evidence from environmental specialists or industry technical experts may also be necessary in order to underpin the damages assessment, and may therefore also be requested in conjunction with the quantum report.

Experts are sometimes called upon to report separately to the tribunal, and in other cases to provide an agreed joint statement. For instance, in Occidental v. Ecuador, the experts produced a joint report estimating the fair market value of the investment according to certain agreed parameters. The experts were able to agree to employ a common model and to reach agreement on ‘the most reasonable approaches and methodology for calculating fair market value, consistent with the Tribunal’s instructions’. In turn, the tribunal appeared to accept the experts’ conclusions.

Independence of experts

Experts are most often selected, instructed and remunerated by one party. This gives rise to the question of their independence, and a perception that their testimony may be of limited probative value, despite the fact that their role is not to advocate on behalf of the party instructing them. Expert witnesses must be independent but arbitral rules do not explicitly recognise the existence of a duty as between an expert and the arbitral tribunal. Accordingly, experts must be transparent about the reasons for their conclusions or calculations and should be willing to change their assessment of damages if the underlying facts or assumptions change. Tribunals are often heavily reliant on experts when quantifying damages, and therefore, experts’ credibility is of utmost importance.

Both the IBA Rules of Evidence and the CIArb Protocol require that a party-appointed expert disclose any past or present relationships that may create a conflict of interest with another participant in the arbitration. These types of disclosure regimes are generally accepted in international arbitration as sufficient to address issues of independence and impartiality. Further, the tribunal may sometimes take into account any concerns in this regard in assessing the credibility of, and thereby the weight to be placed on, the evidence of experts.

In contrast, in certain civil law jurisdictions, and under certain institutional rules that are influenced by those civil law jurisdictions, the independence of experts is of such a concern that the default rule is for the tribunal, rather than the parties, to appoint experts. For example, the default position under the Deutsche Institution für Schiedsgerichtsbarkeit (DIS) Rules is that the tribunal appoints the experts unless the parties agree otherwise. In turn, in DIS proceedings, where evidence is given by party-appointed experts, arbitrators sometimes accord less weight to the evidence because of concerns about the experts’ impartiality. We address the issue of tribunal-appointed experts further below.

Witness conferencing

Witness conferencing (colloquially known as ‘hot tubbing’) is an emerging case management tool in international arbitration whereby multiple witnesses are required to give evidence simultaneously, rather than sequentially. The practice involves a simultaneous joint hearing of all witnesses with the aim of addressing divergences in their evidence on the spot. Hearing practices can vary from asking directed questions and asking counter-witnesses to respond, to the use of witness statements as a guide. Such practices give the opposing experts the opportunity to immediately respond to each other’s views in a way that traditional sequential testimony does not allow.

Both the IBA Rules of Evidence and the ICC ‘Arbitration Commission Report on Techniques for Controlling Time and Costs in Arbitration’ (the ICC Report) encourage the use of witness conferencing in appropriate cases. Witness conferencing can potentially enhance efficiency, clarify disputed issues and eliminate weak or irrelevant arguments. Conferencing can also, in principle, draw the tribunal’s attention to the quality and objectivity of an expert’s opinion by offering an immediate contrast between the experts. Further, it may encourage greater tribunal participation in the examination of witnesses than in cross-examination and therefore elicit relevant questioning that may otherwise have been left for post-hearing submissions. However, the practice of witness conferencing is not without its controversies as detractors have argued that conferencing is too interventionist to be used in arbitration and does little to remedy the gap between the opinions of expert witnesses appointed by the parties.

Pre-hearing meeting of experts

A pre-hearing meeting of expert witnesses to identify areas of agreement and areas in dispute is a method of case management, available under the IBA Rules of Evidence and endorsed in the ICC Report. The aim of these meetings is for expert witnesses to meet (in the absence of counsel) to identify issues in common and refine the issues in dispute, which a tribunal needs to deal with. The CIArb Protocol sets out detailed guidance on pre-hearing meetings between experts, stipulating that they ought to hold a conference to: identify issues on which their opinions will be sought; identify tests or analyses to be conducted; and if possible, to agree upon those issues, and tests, and the manner in which such tests shall be conducted.

Where experts are able to identify common findings, the parties and tribunal are likely to accept those findings, enabling them to focus on the issues in dispute. In this way, pre-hearing meetings of experts can directly reduce costs and delays. However, the practice may prove less fruitful in circumstances where one, or both, of the experts take adversarial positions in an attempt to advocate on behalf of their appointing party. The outcome may be an increase in costs, without the associated benefit of clarifying the issues in agreement. In this respect, where a pre-hearing meeting is being considered, it may be worth requiring that the meeting take place before the experts have submitted their opinions.

Tribunal proactivity

Tribunal proactivity in relation to damages can take several forms, from active case management and the suggestion that the parties be required to articulate their theory of damages early on, to departing from party-proposed approaches to assessing quantum. The latter practice is not without controversy as it may give rise to concerns in relation to the parties’ due process rights. In this section, we consider some of these proactive procedures available to tribunals and, where relevant, associated due process considerations.

Proactive case management

In the majority of cases, issues relating to damages are left to be addressed towards the end of the proceedings. Problematically, where damages have not been considered early on in the proceedings, critical points of law or fact may arise which could have been more efficiently dealt with earlier in the proceedings. Issues such as causation, remoteness of loss, mitigation and valuation dates are particularly susceptible to overlap between questions of liability and damages. Additionally, there is increased risk that evidence pertaining to damages presented near the end of the case may be based on previously unarticulated theories of compensation. By contrast, a late examination of the damages phase may eclipse valuable opportunities to narrow the scope of damages issues, or to exclude improbable damages claims, and therefore, to bring about a timely and more cost-effective resolution of the dispute.

Accordingly, tribunals may be well-served by giving at least some consideration to damages issues early on. The International Institute for Conflict Prevention & Resolution has provided detailed guidance in its Damages Protocol on how to raise damages issues with the parties early on. In particular, the Damages Protocol proposes that arbitrators address damages in or around the initial conference with the parties, and that the arbitrators require the parties to articulate their theories of compensation and their defences.

Taking such measures can assist in reducing the scope of the dispute and in pre-empting complex or problematic issues that are likely to result in delay and increased cost later in the proceedings.

Independent tribunal initiatives

There is some disagreement on the extent to which a tribunal may utilise a procedure in international arbitration that is different from the procedures proposed by the parties. In relation to the issue of damages, each party may have a set idea of the procedure that would best allow it to present its case. The tribunal may have a very different view from the parties as to the most appropriate procedure. In this section, we examine the question of tribunal-appointed experts and tribunal-led assessments of quantum.

Tribunal-appointed experts

In a minority of cases in international arbitration, tribunals appoint their own experts, either because they deem any experts already appointed by the parties to be insufficient to allow the tribunal to fulfil its mandate, or because the parties themselves have requested the tribunal to do so.

To a common law lawyer, this approach may seem at odds with the principle of party autonomy and the adversarial litigation process. However, arbitrators from a civil law background may expect that the expert will be selected by, and responsive to, the tribunal rather than the parties, in the interest of the expert’s independence, as explained above.

A number of institutional rules (including the UNCITRAL Arbitration Rules and the LCIA Rules), and some BITs, even make express and specific reference to the possible use of tribunal-appointed experts, after consultation with the parties, though none of the rules requires the use of tribunal-appointed experts. The expert, if appointed, is to report to the tribunal on specific issues, determined by the tribunal. Under the US Model BIT, the tribunal’s discretion is more limited. It can appoint an expert upon the request of a disputing party, or on its own initiative, unless the parties disagree, to address ‘any factual issue concerning environmental, health, safety, or other scientific matters raised by a disputing party in a proceeding’.

In certain cases, concerns may be raised that the tribunal-appointed experts may become the de facto arbitrator. Accordingly, arbitrators must be careful not to delegate their decision-making responsibilities to the experts they appoint, unless the parties agree to this. While the English courts have not found that tribunal-appointed experts constitute an improper delegation of the tribunal’s powers, the IBA Rules of Evidence have sought to address these concerns by recommending that parties examine the correspondence between the tribunal and the expert, and that the parties should be able to question the expert at the evidential hearing. In a similar concern for transparency (and perhaps a recognition that the expert will ultimately be paid for by the parties), English law, for example, prevents arbitrators from meeting privately with the tribunal-appointed experts regardless of the content of the discussions.

Where an expert is to be appointed by the tribunal, it would normally be appropriate for the parties to be called upon to provide input on the identity of the expert. For example, the parties may each be asked to draw up a list of proposed experts and then for the tribunal to select any overlapping candidates to act as tribunal-appointed experts. Similarly, it would normally be appropriate for a tribunal to provide the parties with an opportunity to comment on the tribunal-appointed expert’s report.

Tribunal-appointed experts can also be useful where the tribunal does not have the benefit of evidence from both parties. This was the case in National Grid v. Argentina, where the tribunal did not have a damages valuation from Argentina. The tribunal-appointed expert did not use its own valuation model but did observe that the claimant’s expert report contained ‘manifest errors’. The tribunal ultimately accepted a number of its own expert’s recommendations on discounted cash flow (DCF) valuation inputs. Similarly, in a recent high-value dispute, the tribunal opted to appoint its own experts even after hearing evidence from the party-appointed experts, noting that those experts ‘were effectively shooting at different targets’ and the tribunal was not comfortable determining the issues on the basis of the parties’ experts’ evidence.

A tribunal may also deem it appropriate to appoint its own expert in circumstances where the case is complex and there are vastly different conclusions reached by the party-appointed experts. In a 2014 survey conducted by the International Bar Association Subcommittee, the majority of responses considered tribunals as reasonably well-equipped to quantify damages. There was also strong support for the use of tribunal-appointed damages experts.

Tribunal-led assessments of damages

In a number of cases, the tribunal may wish to depart from the parties’ assessment of damages and adopt what then becomes, to varying degrees, its own methodology.

In some cases, tribunals may be entirely dissatisfied with the models used by the parties to quantify their damages. For instance, both experts may have failed to consider relevant factors or may have departed from a well-accepted principle of valuation without any justification. In these circumstances, tribunals sometimes choose to adopt their own methodologies and depart from what has been put forward by the parties. By way of example, in SCB v. TANESCO, the tribunal found that the experts had not taken the correct approach to calculating a new tariff. This ultimately led to the tribunal rejecting both parties’ proposed internal rates of return. Instead, the tribunal made its own estimate of the internal rate of return.

Tribunal-led assessments of damages may also be used as a strategy to avoid the ‘ships passing in the night syndrome’ in which the experts fail to engage with each other because, for example, they chose to adopt fundamentally different methodologies. Where the party-appointed experts are not aligned on the models to be used for assessing quantum, tribunals are often much more proactive and interactive.

For instance, in Copper Mesa Mining v. Ecuador, having considered the differences between the party-appointed experts’ valuations, the tribunal dismissed several valuation methodologies in favour of a historical cost approach that reflected the claimant’s alternative valuation method of proven expenditure.

In other cases, tribunals opt to change the calculation methodology and then request that the experts provide recalculated forecasts. In cases where income-based forecasts are presented on the basis of a DCF valuation, tribunals sometimes request to be provided with the party-appointed experts’ financial model. The tribunal can then input any permutation of figures or assumption drawn from the party-appointed experts’ respective forecasts to arrive at a recalculated forecast of its own. This is an interesting approach and one that the relevant experts may resist because of risks associated with disclosing proprietary information contained in their models.

PwC’s international arbitration damages research paper refers to a trend in recent awards of tribunals requesting alternative calculations from experts that are dictated by the tribunal’s own assumptions and fact patterns. For example, in Flemingo v. Poland, the tribunal requested that the parties’ expert provide it with calculations based on alternative scenarios.

Alternatively, a tribunal might simply want to test the effect of disputed assumptions by adjusting model inputs and ‘testing the sensitivity of the important components of the calculations.’

In all such types of cases, the tribunal might ultimately find that its mandate is best served through adopting a more interventionist approach. For example, in CMS v. Argentina, the tribunal built its own model with the assistance of its experts. It then ‘tested a number of scenarios by changing different variables . . . [focusing] on the most important determinants of value (as well as the main sources of uncertainty)’. Ultimately, the tribunal was able to modify the assumptions underpinning the claimant’s expert’s calculation of loss. Similarly, in Enron v. Argentina, the tribunal-appointed expert assisted the tribunal in addressing the parties’ competing positions. Notably, the tribunal found its expert’s approach to be ‘more balanced and realistic’ than that of the party-appointed experts. Accordingly, the tribunal accepted its own expert’s recommendations on several inputs into the DCF valuation that was used.

Due process

Where tribunals use these sorts of proactive measures, they ought to be particularly mindful of their obligation to ensure that the parties are given the opportunity to be heard. Where due process rights are violated, the relevant award may be liable to annulment. Due process requires that parties have a reasonable opportunity to be informed about, and comment on, the bases for a tribunal’s decision on damages. Accordingly, if after hearing the parties as to the appropriate methodology for calculating damages, a tribunal decides to take a third approach, or to appoint its own expert, the tribunal may at that point find it appropriate to communicate its proposed course of action to the parties and invite their comments or further submissions.

Unconventional procedures

In this final section, we consider two unconventional procedures for determining damages, which certain parties may find of interest, and which tribunals might wish to propose to the parties in appropriate cases.

The first of these procedures is known as ‘baseball arbitration’, a term that originates in the resolution of salary disputes in Major League Baseball in the US. In baseball arbitration, rather than award the actual quantum of damages, the tribunal is mandated to decide between two fixed offers made by the opposing parties. There are two variants of baseball arbitration. ‘Day baseball’ arbitration occurs when the tribunal is aware of the offers submitted by the parties and then chooses between them. By contrast, in ‘night baseball’ arbitration, the submitted offers are kept confidential as between the parties until the tribunal has rendered the award on liability and damages. At this point, the offers are disclosed to the tribunal and the offer closest to the damages awarded is awarded.

A key advantage of baseball arbitration is that it may incentivise the parties to make rational offers, or even discounted offers as compared with what their positions on quantum might otherwise be. The method can therefore help parties to settle. Finally, baseball arbitration can accelerate the deliberation process and reduce costs because the tribunal may, for day baseball at least, not be required to include as much detail in the award on damages.

The procedure, of course, has its limits and is likely not to be suitable unless both parties accept that there is a negative financial consequence and there is some common ground between the parties on the parameters for measuring the loss. Some commentators have questioned whether baseball arbitration constitutes ‘arbitration’ at all given that the procedure significantly curtails the tribunal’s discretionary powers – a concern that reflects the tension between party autonomy and the tribunal’s adjudicative role. Finally, critics of baseball arbitration argue that limiting the tribunal’s remedial powers can result in ‘arbitrary and unintended’ outcomes and ‘raises enforceability issues in some jurisdictions’.

The second procedure that parties and tribunals may wish to consider is known as ‘high-low arbitration’. Parties can include a high-low clause in their arbitration agreement, or agree to such a clause on an ad hoc basis when a dispute arises, in order to set a floor and ceiling for an award of damages. The tribunal is not necessarily informed of the high-low agreement between the parties. In these circumstances, if quantum assessed by the tribunal exceeds the maximum amount specified by the parties, then the award is fixed at that maximum amount, and, conversely, upon a finding of liability, the quantum cannot be below the minimum. Like baseball arbitration, high-low arbitration has attracted some criticism and may not be suitable for most cases. In particular, it has been suggested that high-low arbitration favours risk-takers and invites parties to gamble, as a ‘“moderately-extreme” position seems to offer the best risk-benefit ratio’.

Conclusion

In this chapter, we have outlined some of the procedural issues relevant to the determination of damages in international arbitration. This is certainly an area in which users of arbitration can benefit from counsel and arbitrators striving to devise the optimal procedure in each case, thinking creatively and drawing on past experience where necessary. Early tribunal proactivity in this area may be particularly helpful in cases where the parties themselves (or their counsel) are insufficiently focused on damages issues at the outset of the dispute, and can, at least in some cases, lead to significant savings of time and costs.

Subscribe here for related content, breaking news and market analysis from Global Arbitration Review.

Covering commercial and investment arbitration, Global Arbitration Review is the unofficial broadcaster for members of the international arbitration community, keeping them feeling up to date and informed.