Reforming Damages in ISDS and the Problem of Full Reparation

The assessment of damages is perhaps the most controversial of all issues covered by UNCITRAL Working Group III (WGIII), whose overall mandate is to identify concerns related to the operation of ISDS and work toward possible reforms. It is also perhaps the most important one. If awards against states did not frequently reach hundreds of thousands of dollars and sometimes even exceed the billion-dollar mark, there would not be such serious concerns about ISDS crippling public finances or preventing public interest regulations, with those related to climate change attracting the most attention. Other issues related to the integrity of the arbitral process would also matter much less.

The reform process on damages has picked up speed over the last year, with the publication of draft proposals by the UNCITRAL Secretariat. This article offers a critical commentary on these ongoing reform efforts. It begins by presenting how arbitral tribunals have come to address the assessment of damages and why this has led to a large increase in amounts awarded. I will then address UNCITRAL’s proposals, the last of which is particularly disappointing since it appears to endorse the approach typical of the arbitral case law and to suggest that the international law principle of full reparation prevents any significant reform in this domain. This is a misconception: a correct understanding of that principle suggests that limits need to be imposed on how tribunals quantify damages since the problem with the current approach is not simply the sheer numbers it leads to, but the fact that investors are receiving in compensation more than legally is their due. I will conclude with some suggestions for WGIII’s future direction, with a view to putting an end to the unjustified inflation of awards and laying the ground for a more legally sound approach.

The mainstream approach to the assessment of damages in ISDS

It is sometimes said that the problem with damages in ISDS is a lack of consistency in how tribunals approach their quantification. This is a mistake. The case law on this issue has, in fact, developed in a remarkably harmonious way. In particular, it has converged around three seemingly technical but highly consequential questions to abandon the approach typically followed by international tribunals until the 90s. These changes have made it possible for investors to obtain significantly inflated awards. 

The first is the idea that the customary international law principle of full reparation, which normally governs the quantification of damages, leaves no role for any equitable factors. It was relatively common for past tribunals to moderate or limit the amounts awarded by discretionally relying on various considerations unrelated to the harm suffered by the claimant, such as the impact on the state’s public finances, the legitimacy of the goal pursued by the state, the investor’s contribution to local development, or its record of reprehensible conduct. Current tribunals, however, tend to consider that the assessment of damages should be a purely factual enquiry, which ought to focus exclusively on ascertaining and measuring, as objectively as possible, the loss suffered by the investor. Arbitrators, therefore, typically reason that equitable considerations are irrelevant to this assessment and that to factor them in would necessarily result in undercompensating the investor, as it would obtain less than the full value of what it objectively lost.

The second major shift relates to how the investor’s loss is valued monetarily. Arbitrators have gradually come to believe that this should be done exactly as it would be by a real-life investor. Thus, they will seek to replicate the valuation methods typically used by firms to make investment decisions, which reduce an asset’s present value to its ability to produce returns in the future. Any amounts historically invested do not really matter: what counts is whether the investment is predicted to generate profits going forward and, if so, how much. The preferred valuation method is, therefore, DCF, which proceeds by first projecting expected cash flows to then establish their present value by applying a discount factor that reflects any associated risk. DCF calculations can yield widely different results, but the method’s popularity in investor decision making has convinced arbitrators of its appropriateness for adjudicating damages. This position is again in contrast to the approach traditionally taken until the 1990s by international tribunals, who viewed DCF as unduly speculative and preferred to rely on amounts historically invested.

The third most consequential shift concerns the calculation of interest to be added to the damages, which normally will be owed for the period since the loss occurred until the date of payment of the award. The choice here is between simple interest (where interest is earned only on the principal) or compound (where interest is periodically added to the principal, which then generates more interest). In practice, the difference between the two can be large. The default rule in international law used to be simple interest because of the significant amounts compounding could generate and because compounding was (and is) often forbidden or restricted under the law of the host state. Over the last three decades, however, arbitrators have generally come to embrace compound interest, disregarding local restrictions. The reason is similar to why they have embraced DCF: compound interest is the norm in modern financial reality, so that should also be the default in ISDS if investors are to be fully compensated. 

Overall, these three developments are the most important factors driving the increase in amounts awarded, as investors are able to claim very large amounts in expected profitability, while equitable judgment is not allowed any moderating role. A well-known example is Tethyan Copper v Pakistan, where a mining conglomerate that was denied a mining licence was awarded more than USD 4 billion in compensation (calculated through DCF) and close to a further USD 2 billion in interest, with the tribunal considering it irrelevant that the investor had only invested around USD 150 million or that the final award amounted to 2% of the debt-stricken country’s GDP. 

The UNCITRAL Secretariat’s changing proposals

In July 2023, the UNCITRAL Secretariat finally produced a set of draft provisions for discussion. These included a lengthy one on the assessment of damages. The provision was in many ways defective, but it had the merit of acknowledging the problem of damages inflation and of wanting to do something about it. Indeed, it contained three major proposals, addressing each of the aspects of the case law that I have just outlined. Two were radical: damages were capped at amounts actually invested, and compound interest was banned altogether. Moreover, it required arbitrators to consider the investor’s compliance with human rights standards, which is an example of equitable consideration. All of these proposals went decidedly against current arbitral practice and would have had a significant impact if adopted. It would likely be the end of the “mega-awards” that have made headlines over recent years.

The draft provision on damages was, however, immediately criticized, most of all for running counter to the sacrosanct principle of “full reparation.” The Secretariat accordingly produced a heavily revised version, which was only published this July. This new version goes in a starkly different direction. It can no longer be said to reflect the goal of containing excessive awards, which was meant to be the driving rationale of the reform process. Compound interest is again allowed, any reference to human rights compliance has disappeared, and sunk costs have ceased to serve as a cap. Claims based on the loss of future profits are made conditional on not being “inherently speculative,” but this is an idea that current tribunals already admit. 

All in all, the latest proposal contains no significant attempt at curbing award inflation and appears to be a codification of the current arbitral case law—the very case law that is the root of the problem that the WGIII is meant to correct. The Secretariat has also promised to draft a set of additional “guidelines” to orient how the provision is applied, but it is doubtful that this could be enough to resolve the issues with the case law. In any case, if the draft proposal were to be maintained as it is, it would mean that the WGIII may actually make things worse since the status quo will have been preserved but now will enjoy the WGIII’s blessing and, therefore, an added layer of authority. 

Understanding full reparation 

WGIII is wrong to think that its hands are tied by the principle of full reparation. The fact that it seems to think so comes from a rigid and reductive understanding of that principle, which is unfortunately widespread among arbitrators. 

The remedy of damages, whether under international or domestic law, never protects against every negative consequence of unlawful conduct but only against losses that the law considers to be compensable or that it is the responsibility of the defendant to prevent. Defining the perimeter of those protected losses is a problem of legislative drafting and legal interpretation rather than one of pure fact, but it is also not one that the principle of full reparation can help with. That principle requires that compensable harm be fully compensated but does not tell us what harm actually qualifies as compensable. 

What characterizes the approach often followed by ISDS tribunals is the fact that they are defining what counts as compensable harm, and therefore the extent of states’ responsibility vis-à-vis investors, by reference to the financial norms and expertise that govern investment decision making, rather than on the basis of applicable legal sources. Indeed, the widespread recourse to DCF valuation and compound interest has meant that the loss of financial returns expected by investors is now generally seen as a type of compensable harm. As a result, current tribunals tend to treat investors as legally entitled to that projected profitability, even where this is specifically excluded under the applicable law of the host state (e.g., through a ban on compound interest).

What this also means is that investors are being routinely overcompensated. Indeed, they are receiving more than they were legally entitled to prior to the state’s breach—in violation, therefore, of the principle of full reparation. Thus, the current approach to damages is not simply problematic due to the sheer numbers awarded: it is also fundamentally wrong as a matter of law. 

First, consider the compensability of the loss of profits. From a legal point of view, the award of damages for this specific loss assumes that an actual right to the profits in question exists (instead of a mere financial expectation). Whether this is the case will normally depend on the law of the host state. Imagine, for instance, that the investor is wronged in the process of trying to obtain a licence to operate a profitable activity such as mining, but the licence in question had not yet been obtained. A right to the profits expected from operating the mine may not, therefore, have yet consolidated. This may be made clear by the consequences that the local legal regime attaches to the wrongful denial of such a licence, for instance by limiting damages to costs incurred. The loss of future profits, in other words, is not here recognized as a form of compensable harm. Where this is the case, the use of DCF valuation is not legally justified, even if the investor’s expectation of profit was reasonably certain.

Second, consider the quantification of damages for the loss of profits. Where a right to earn profits is legally recognized, it does not follow that this right extends to whatever income streams happen to be projected. Indeed, international investment law recognizes states’ right to regulate corporate profits, most of all by increasing taxes. This means that any projection of profits is almost always legally precarious rather than held in ownership. Relying on those projections when quantifying damages will, therefore, exceed full reparation. For this reason, investors who do enjoy a right to certain profits should usually receive compensation for their loss based on no more than a reasonable rate of return on their investment.

Finally, it is also wrong to assume that equitable judgment plays no role in defining the extent of the losses for which the investor is entitled to full reparation. The scope of rights, including the right to compensation, is also sensitive to wide-ranging principles and legitimate countervailing interests. The European Court of Human Rights recognizes this most clearly, as it states that the calculation of compensation for expropriation must strike a “fair balance” between the interests of the expropriated individual and those of the broader public. Thus, the more sensitive the interests at stake, the more discretion it will allow in reaching the appropriate sum. There is no reason to exclude the possibility of similar balancing in international investment law, particularly where it concerns principles that international law recognizes as legitimate or even obligatory. Thus, for instance, the compensation for the expropriation of an oil concession should not be calculated in the same way if it is motivated by purely confiscatory reasons than if it is the indirect result of environmental legislation aimed at limiting fossil fuel extraction in line with obligations under international law. To exclude the possibility of such balancing, as tribunals tend to do, leads again to overcompensating claimants since it unreasonably assumes that the definition of investors’ rights is done entirely in isolation from any other interest or legal principle, however compelling it might be. 

Summary of takeaways

What, therefore, is the way forward for ISDS reform if it is to take seriously the problem of exorbitant awards? It may be useful to summarize the key takeaways, which I hope may inform the revision of the Secretariat’s latest proposal, as well as the drafting of any additional guidelines that may follow.

• Applying the principle of full reparation cannot be reduced to a purely factual enquiry: it necessarily involves defining what harm is legally compensable, a key issue that WGIII should not neglect. 
• The perimeter of compensable harm should reflect the extent of investors’ legal entitlements and states’ responsibilities under the applicable legal sources. It should not be defined exclusively by reference to the financial expertise that guides investor decision making, as this leads to overcompensating investors.
• Damages for the loss of profits must only be awarded where the investor’s right to those profits is legally consolidated. This should normally be established by reference to the host state’s legal system, which arbitrators tend to disregard when assessing damages. 
• Where a right to profits is indeed recognized under the applicable law, damages for the loss of profits should be limited to a reasonable rate of return on the investment unless the claimant can also establish that a specific level of profitability was legally guaranteed.
• Simple interest should be the default rule, and compounding should be reserved for exceptional cases where it is allowed under the host state’s law, and the claimant can provide concrete proof that it would have obtained compound interest but for the state’s breach.
• The assessment of damages should factor in equitable considerations, particularly those derived from principles recognized as binding commitments under international legal practice, such as those related to climate change or human rights.  

Author

Toni Marzal is Professor of Law at the University of Glasgow