2026 PAW: Affaires d’Etats Vol. 5: Getting Causation Right in Investment Disputes

As part of the 2026 Paris Arbitration Week (“PAW”), Curtis, Mallet-Prevost, Colt & Mosle LLP hosted a webinar titled “Getting Causation Right in Investment Disputes.”  This was the fifth installment in the “Affaires d’Etats” series on Investor-State Dispute Settlement (“ISDS”).

The event featured Sylvie Tabet (Government of Canada), Ali Ssekatawa (Petroleum Authority of Uganda), Paul Key KC (Essex Court Chambers) and Juan Carlos Boué (Curtis, Mallet-Prevost, Colt & Mosle LLP).  The discussion was moderated by Irene Petrelli (Curtis, Mallet-Prevost, Colt & Mosle LLP).

The panel examined how causation intersects with questions of liability and quantum in ISDS and explored how causation has been dealt with in practice, touching upon the question of burden of proof and the evolving theories of loss advanced by claimants, such as loss of opportunity.  The panel also discussed how the failure by the arbitral tribunal to properly address causation impacts post-award remedies.

Practical Implications of Causation

Sylvie Tabet opened the floor by discussing the implications of causation in cases where a project is at an early stage of development and the alleged breach concerns the refusal of a permit or another regulatory authorization by the respondent State.  Tabet highlighted how claimants often plead a combination of different breaches, but present a single, global measure of damages.  In particular, claimants advance a so-called “all or nothing” valuation, often based on a Discounted Cash Flow (“DCF”) analysis, without alternative measures of damages.

Although the burden of proof to establish both causation and quantum should rest with the claimants, claimants’ approach to quantum may create particular difficulties for respondent States, which are faced with the strategic decision of either merely pointing out that claimants failed to fulfil their burden of proof or putting forward an alternative damage valuation. 

Juan Carlos Boué criticized claims for damages relating to unbuilt projects, describing them as being among the most dangerous categories of cases.  According to Boué, tribunals tend to take an abstract approach to the “but-for” analysis in these cases, accepting a long chain of hypothetical events, such as permits being granted, financing materializing and technical obstacles being overcome.  That can result in the counterfactual exercise drifting away from reality and into what was described as a fantasy.

Boué suggested that, in such cases, it is incumbent upon the respondent States to carefully underline the importance of the principle of causation and to resist the temptation to dismiss claimants’ claims as too implausible to warrant detailed engagement. 

Tabet further noted that the problem often lies in the assumptions with which counsel instruct quantum experts.  Where damages models rest on unrealistic assumptions about regulatory approvals, financing, or technical feasibility, the issue is not simply one of arithmetic.  It directly impacts the legal principle of causation as it demonstrates a failure to establish that the losses claimed would have materialized absent the breach.

Although tribunals increasingly appear willing to scrutinize deficient damages cases more closely, Tabet explained that they still react to weak damages evidence not by rejecting the claim outright, but by giving the parties another opportunity to present their valuation, which easily leads to an increase in time and costs.

A related theme addressed by the panel concerned the growing invocation of loss of opportunity claims.  These claims are advanced by claimants in an attempt to soften the standard of proof required.

On this point, Boué stated that claimants routinely invoke the Crystallex v. Venezuela case (“Crystallex”) as the centerpiece of their quantum arguments, citing it for the proposition that once damages are established with certainty, their quantification may be approached with flexibility.  Crystallex is often used as a gateway to speculative quantification once a tribunal accepts that profits were at least conceivable.  If the loss of opportunity argument becomes a mechanism for loosening causation, tribunals risk awarding large sums for scenarios that never had a chance of materializing.

The panel then turned to discussing a different category of cases where the question of causation is of key importance: those involving projects that had been operating for years without ever being profitable.

Ali Ssekatawa cited the Biwater Gauff v. Tanzania case as an example where liability was established, but no damages were awarded as there was no record of profitability.  Ssekatawa also noted how in other cases tribunals took a looser approach to causation and awarded compensation despite the claimant’s inability to show that the project was commercially viable.

Ssekatawa then discussed an arbitration against Uganda regarding a concession that had been operated for twenty years without being profitable and in the context of which the claimants had failed to meet the operational targets set in the concession itself.  While the tribunal found liability with respect to the process by which the concession was terminated, it also accepted that the State had reasons to terminate because of the claimants’ non-performance.  Even so, the proceedings moved into a further quantum phase requiring expert evidence and substantial additional cost.

This, according to Ssekatawa, demonstrates how onerous it can be for respondent States to argue lack of causation even where the burden to proving causation formally rests on the claimant.  The tribunal must test whether the assumptions built into the counterfactual scenario are truly attributable to the State’s breach.  But the process is often lengthy and expensive, especially when the case moves into a quantum phase after liability has been established by the tribunal.

The same case against Uganda also involved a claim for the loss of an additional business opportunity that had never crystallized into any contractual right.  The alleged opportunity rested only on preliminary discussions, not on any contractual anchor, and the evidence showed that the opportunity was not viable.  The tribunal approached that claim through a causation lens.  The central question was whether the opportunity would, more likely than not, have materialized but for the alleged wrongful conduct.  That claim was dismissed as being too speculative and not proven.

Post-Award Consequences of Failing to Address Causation

The second part of the discussion turned to post-award remedies, and in particular to the long-running World Wide Minerals v. Kazakhstan saga before the English courts, which was discussed by Paul Key KC.  The case provides an illustration of how a tribunal’s failure to properly address causation can affect an award.

As recounted during the discussion, the claimant had advanced three treaty breaches as a collective package amounting to expropriation.  The tribunal upheld two of the alleged breaches but rejected one, and recognized that Kazakhstan had the right to terminate the underlying contract.

At that point the tribunal faced two options.  It could either accept Kazakhstan’s argument that the claim had been pleaded as an indivisible whole and therefore failed if one constituent breach was dismissed.  Or it could adopt the claimant’s position and issue a liability award only, inviting further submissions on damages.

Instead, the tribunal proceeded directly to damages and developed its own theory: although there had been no expropriation, the two breaches justified recovery of sunk costs.  That approach did not survive judicial scrutiny.  The English High Court held that the tribunal should not have issued a damage award on a theory not advanced by the parties, and that it should instead have gone back to the parties for further submissions.

The matter was sent back to the tribunal.  After further submissions and a hearing, the tribunal again awarded compensation of sunk costs, this time still failing to engage with a critical defense on causation advanced by Kazakhstan, namely that it had the right to terminate the underlying agreement.  Only a single paragraph of the award was devoted to causation and loss.  So the parties went back to the English High Court, which set the award aside a second time.  A newly constituted tribunal has since been appointed to address causation and damages.

As Key observed, the costs incurred over the years may now well exceed the amount in dispute.  The case demonstrates that inadequate treatment of causation is not merely a defect of reasoning internal to the award.  It can lead to prolonged and costly post-award litigation.

Conclusion

In her closing remarks, Tabet emphasized the importance of distinguishing between the legal standard, which applies to the parties’ submissions and is not for quantum experts to discuss, and what counsel is asking the expert to value.

Boué further recommended that respondent States strive to answer everything and not fall into the temptation of cutting corners in their defenses.

Ssekatawa noted that causation is what preserves the distinction between breach and valuation.  It ensures that investor-State arbitration compensates only those losses that were actually caused by the wrongful act, rather than losses attributable to ordinary commercial risk or speculative assumptions about the future.

Key concluded that because arbitrators are generally lawyers rather than accountants, their professional training naturally directs their attention towards obligations and breaches of those obligations, while causation and damages receive less attention and are sometimes treated in a less rigorous manner.  For causation and quantum to receive the attention they deserve, they must be presented in a way that is both understandable and appealing to the tribunal. 

This post is part of Kluwer Arbitration Blog’s coverage of  Paris Arbitration Week 2026.