Taimoor Raza Sultan & Rabia Sohail Paracha
Where a public‑interest regulation lawfully remains in force, should investors nonetheless be compensated for the portion of their legitimate expectations that the regulation frustrates, and if so, on what legal and evidentiary basis?
I. From Expropriation to Regulatory FET
Investment arbitration has shifted from classic direct expropriation and formal nationalizations to claims targeting regulatory measures adopted for public purposes such as economic crisis management, public health, or energy transition. These measures typically do not confiscate title or shut down operations; instead, they alter the economic assumptions of the investment by reducing expected returns, modifying tariff schemes, or raising compliance costs, while ownership and control remain with the investor. In such cases, tribunals often accept that regulatory change has caused real economic injury; however, outcomes tend to be binary, i.e., either full compensation following a finding of breach, or complete dismissal on the basis of regulatory legitimacy or the exercise of the State’s police powers.
The current binary approach does not reflect the recurring reality of ‘partial frustration’, where the investment continues to operate but under substantially less favorable conditions than those reasonably relied upon at the time of investment. The issue is not so much with substantive standards, as FET and legitimate expectations provide considerable flexibility, but with the lack of a remedial and evidentiary framework that can accommodate partial harm without either undermining regulatory objectives or entirely withholding compensation.
II. Public‑Interest Regulation and Treaty Breach
Investment treaties and arbitral practice accept that states retain a sovereign right to regulate in the public interest, particularly in areas such as public health and environmental protection. At the same time, tribunals have made clear that the pursuit of legitimate public objectives does not provide a blanket defense against breaches of FET or expropriation clauses; rather, it helps define the limits and application of the state’s obligations without nullifying them.
For instance, LG&E v Argentina (ICSID Case No. ARB/02/1) involved claims by U.S. investors in Argentine gas distribution companies after the measures during 2001-2002 economic crisis, including the Emergency Law, which altered regulatory frameworks like tariff adjustments and peso convertibility. The tribunal found breaches of FET and umbrella clauses but upheld Argentina’s necessity defense for 1st December 2001 to 26th April 2003, exempting financial liability during that period; but still awarded around USD 57.4 million for the breaches extending beyond this period, thus recognizing that not all investor losses were fully immunized by the crisis. Furthermore, Micula v Romania (ICSID Case No. ARB/05/20) concerned Swedish investors who relied on Romanian incentives i.e. tax breaks, customs exemptions to develop a brownfield site in the 1990s. Romania abruptly revoked these incentives in 2005 in compliance with EU accession state-aid rules. The tribunal held that Romania breached the fair and equitable treatment standard by withdrawing investment incentives relied upon by Swedish investors, notwithstanding the public objective pursued by the measure, and awarded damages calibrated to the portion of investments linked to the incentives (RON 376 million plus interest). These cases illustrate a consistent doctrinal principle, i.e., the legitimacy of a regulatory aim does not, in itself, preclude a treaty breach; rather, liability turns on whether the manner, timing, and impact of the measures are compatible with treaty-based investor protection.
The real difficulty arises where tribunals must articulate the legal consequences of regulatory measures that are substantively legitimate, remain in force, and only partially frustrate reasonable expectations. In such situations, current practice often conflates the validity of regulation with immunity from compensation.
III. Partial Frustration as a Recurring Pattern
Partial frustration occurs where a regulatory intervention alters the economic equilibrium of the investment without amounting to substantial deprivation or functional expropriation. The investor continues to own and operate the asset and cash flow persists but profitability declines, expansion becomes unviable, or the compliance costs materially increase compared with those prevailing at the time of the investment. It can be identified in several cases:
- Renewable energy disputes against Spain (e.g., Cube Infrastructure; 9REN) arose from post-2013 reforms that substantially reduced the expected returns of existing renewable investments, including changes to feed-in tariffs. Tribunals generally found breaches of the FET standard due to frustrated expectations of regulatory consistency, while the principal difficulty lay in quantifying compensation through assessments of reasonable rate-of-return shortfalls.
- In Philip Morris v. Uruguay (ICSID Case No. ARB/10/7), Uruguay adopted tobacco control measures requiring graphic health warnings to cover 80% of principal display areas and limiting each brand family to a single product variant (‘single presentation requirement’). These measures eliminated several of Philip Morris’s product variants and reduced brand differentiation and market share, without forcing the cessation of operations. However, unlike Spanish cases, the tribunal rejected claims of breach of fair and equitable treatment and indirect expropriation, holding that the measures constituted a valid, non-discriminatory, and proportionate exercise of the State’s police powers in pursuit of public health objectives.
- Similarly, in Chemtura v. Canada (PCA Case No. 2008-01), the NAFTA tribunal dismissed claims of breach of the minimum standard of treatment and indirect expropriation arising from Canada’s ban on lindane, a pesticide prohibited due to well-documented risks to human health and the environment. The tribunal held that the measure constituted a non-discriminatory and evidence-based exercise of the State’s police powers and did not result in a substantial deprivation of the claimant’s investment, as it affected only a limited segment of the investor’s overall operations. The decision affirms that adverse economic effects alone do not do not transform legitimate regulation into compensable expropriation.
In such cases, tribunals either (a) treat the injury as falling within ordinary regulatory risk and deny liability, or (b) find a breach but assess compensation using valuation methods developed for cases of complete deprivation, despite the fact that the harm is only partial. The resulting ‘missing middle’ lies not in the recognition of partial harm, which tribunals often accept, but in the failure to provide proportionate compensation for it.
IV. Structural Reasons for Non‑Compensation
There are a few reasons that explain why partial frustration rarely results in partial compensation.
First, the tribunals often treat legitimate expectations in binary terms, i.e., either fully protected when based on explicit commitments or a stable legal framework, or entirely disregarded when they reflect mere hopes or ordinary business risk, leaving little room to recognize expectations that are reasonable but conditional. This encourages tribunals to decide ‘all or nothing’ at the liability stage rather than calibrating the extent of protection.
Second, when tribunals apply proportionality analysis, it is usually limited to deciding whether a breach occurred, balancing the public interest against investor rights, rather than determining how much compensation should be awarded. Once a measure is deemed lawful or proportionate, tribunals often stop the analysis, even if they acknowledge that investors still suffered significant losses.
Third, common valuation methods, like discounted cash flow, are designed to assess an investment as a whole under the assumption of a stable regulatory and income environment. They are not well-suited to measure smaller, partial losses caused by lawful regulatory changes. Because evidence on partial harm can be uncertain, tribunals often choose simple, all-or-nothing solutions, either finding no breach or awarding full compensation as if the investment were entirely lost, rather than attempting a more detailed but complex assessment.
As a result, public interest justifications serve two purposes. They provide a legal basis for the regulation and also give tribunals a practical reason to avoid the difficult task of calculating partial losses. When investors are not compensated, it often reflects procedural and evidentiary limits rather than the absence of actual harm.
V. Towards a Framework for Partial Compensation
A more coherent approach would accept that where a public‑interest regulation remains in force but partially frustrates legitimate expectations in breach of FET, the appropriate remedy is partial compensation adjusted to the scope of the frustrated expectations. This approach preserves the State’s regulatory autonomy while ensuring that a portion of the cost is borne when treaty obligations are breached. There are four conditions which could be required before awarding partial compensation:
- Specific and relied-upon expectations
An investor must demonstrate reliance not on general profit expectations but on specific assurances or a reasonably stable legal environment, capable of creating legitimate expectations at the time of investment, and actually relied upon in structuring the investment.
- Direct causal link
The investor must establish that the alleged losses are directly caused by the regulatory measure, excluding losses attributable to poor management, market volatility, or unrelated policy changes.
- Material but partial economic impact
The economic impact must be significant without reaching the level of substantial deprivation or expropriation, distinguishing it from partial frustration. Indicators include a marked reduction in net present value, curtailed expansion plans, or a notable decline in expected returns, while the investment continues to operate.
- Exclusion of ordinary regulatory risk
Losses that fall within a foreseeable regulatory risk in heavily regulated sectors should remain non‑compensable, consistent with cases emphasizing that investors must anticipate evolving regulation in fields such as energy or environment. Where the change is abrupt, retroactive, or contradicts specific assurances, it may cross the line into partial frustration.
Where these conditions are satisfied, tribunals could employ a simplified indicative formula:
Partial compensation
= (Proven post-measure loss – ordinary regulatory risk) × degree of reliance
The proven post‑measure loss can be estimated by comparing pre‑ and post‑measure cash flows, rates of return, or peer‑based benchmarks, while ordinary regulatory risk reflects a discount for changes that a prudent investor should have anticipated. The degree of reliance factor allows tribunals to reduce compensation where only part of the investment decision can be linked to the frustrated expectation.
Absolute precision is unattainable, but tribunals routinely accept reasonable approximation in quantum, particularly where the respondent’s conduct contributed to uncertainty. The key is to make explicit the assumptions and discounts used, rather than collapsing the analysis into a liability/no‑liability dichotomy.
VI. Doctrinal and Practical Pay‑Off
This remedial approach preserves regulatory autonomy while allocating part of the economic burden of treaty-incompatible measures to the state, leaving the measure itself in effect. It parallels domestic constitutional and administrative practices, where otherwise lawful regulations may give rise to compensation if they impose disproportionate burdens on specific actors, such as in cases of legitimate expectation damages.
Explicitly addressing partial frustration also enhances the legitimacy and predictability of the system. When tribunals acknowledge harm but award nothing, the entire economic cost of regulation is implicitly shifted to the investor, even where the state has contributed to the creation of specific expectations. Partial compensation, grounded in strict evidentiary requirements, signals that treaty rights remain meaningful even in a regulatory state, while states can still pursue public‑interest policies subject to bearing a calibrated financial consequence if they cross the line defined by FET and legitimate expectations.
Seen in this light, the missing middle in investment arbitration is best understood as a remedial and evidentiary gap rather than a deficiency of substantive law. A structured framework for partial compensation, anchored in existing case law on legitimate expectations, necessity, and police powers, offers a legally defensible way to recognize and address partial economic injury without undermining public‑interest regulation.
