This HTML version is machine-generated. Always consult the original document.

LSG Building Solutions GbmH and others v. Romania
(ICSID Case No. ARB/18/19)

Dissenting Opinion of Judge O. Thomas Johnson

[Page 1]

1. I dissent from the Majority's decision on the merits. I do so because, in my view, the Majority does not properly apply the fair and equitable treatment standard to the facts of this case. Namely, it does not properly define the specific commitment of Romania that caused Claimants to legitimately expect that they would benefit from the Green Certificate support scheme described in Romanian law for 15 years.

2. How to properly apply the fair and equitable treatment standard in the context of legitimate expectations requires a two-step analysis. The first step is to clarify the possible sources of legitimate expectations. There are two possible sources: specific commitments that bind the State and a general legal or regulatory framework on which the investor relies. The Majority (correctly) describes these sources as follows¹:

"Scholars and tribunals have explained that a State can create two types of potential expectations vis-à-vis foreign investors:

- The first type refers to specific representations made, or assurances given by the State to an investor (or a narrow class of investors or potential investors), to induce foreign investment; these representations or assurances then, undisputedly, become binding upon the State;

- The second type relates to the State's general legislative and regulatory framework; the questions whether the investor relied on the expectation of stability of this framework when deciding to invest, and whether a reform of the framework results in a breach of the investor's regulatory legitimate expectations, are two far more controversial issues."

3. The second step in the analysis is determining which, if either, of the two possible sources of legitimate expectations actually gave rise to legitimate expectations in the present case. The Majority finds in favor of the first source, holding that Romania made a binding specific commitment to Claimants. Thus, the Decision states, at paragraph 1066, that²:

“The Tribunal finds that when it created the GC support scheme, Romania sought to attract investment in RES-E by establishing a clearly defined framework, which permitted investors in the Romanian PV sector to foresee that a PV plant would be legally entitled to receive, for 15 years, certain clearly defined streams of income. More importantly, Romania formalized these essential characteristics in a specific commitment given to the PV Facilities owned by Group A Claimants."

¹ LSG Building Solutions GbmH and others v. Romania (ICSID Case No. ARB/18/19), Decision on Jurisdiction, Liability and Principles of Reparation, dated 11 July 2022 (“Decision"), para. 1033 (citing, inter alia, Doc. CL-22, Masdar Solar & Wind Cooperatief U.A. v. Kingdom of Spain (ICSID Case No. ARB/14/1), Award, dated 16 May 2018 (“Masdar"), paras. 489 et seq.; Doc. RL-172, RWE Innogy GmbH and RWE Innogy Aersa S.A.U. v. Kingdom of Spain (ICSID Case No. ARB/14/34), Decision on Jurisdiction, Liability, and Certain Issues of Quantum, dated 30 December 2019 (“RWE”)), para. 453 (footnotes omitted). ↩
² Decision, para. 1066. (Group A Claimants are those Claimants that made their investments between January 2011 and January 2013, before any changes were made to the GC support scheme that were detrimental to any of the Claimants.) ↩

[Page 2]

4. A few paragraphs later, in paragraph 1073, the Decision describes the “framework” mentioned immediately above as a “regulatory promise" that was³:

"...subsequently formalized in a specific administrative act, performed by ANRE: the issuance of the accreditation certificate for each of the PV Facilities owned by Group A Claimants – a document which acknowledged that the specific PV plant would be entitled to benefit from the Essential Characteristics of the GC support scheme [the “Specific Commitment”]." [Bold and bracketed language in original.]

5. In paragraph 1070, the Decision defines the “Essential Characteristics” that the Majority holds were the subject of Romania's binding Specific Commitment. These were⁴:

"First, the PV plant would be entitled to the sale of electricity on the wholesale electricity market, at the price which would result from offer and demand, without any Government support;

Additionally, (and crucially) the PV plant would be entitled to six GCs for each MWh of electricity produced and delivered to the grid, which could be sold either on the GC market or through GCPAs, at a minimum price which the Law specifically said could not fall below EUR 27/GC indexed annually to European inflation [these traits will be referred to as the "Essential Characteristics" of the GC support scheme]." [Bracketed language and bold in original.]

6. With one important exception, I agree with this analysis. In enacting its PV incentive scheme, Romania exercised its sovereign discretion to make binding specific commitments to investors for the purpose of attracting a particular type of investment. The important exception to my approval of the Decision's reasoning relates to the scope of Romania's specific commitment concerning the sale of GCs. Romania's commitment was not that GCs could be sold at a minimum price of EUR 27 per GC; it was that they could be sold at whatever price they commanded on the GC market, which price could not fall below EUR 27 per GC nor rise above EUR 55 per GC.⁵

Romania's Specific Commitment

7. I begin by considering the nature of a specific commitment. The Masdar tribunal described Spain's specific commitment at issue in that case as resulting from “a very specific unilateral offer from the State, which an investor would be deemed to have accepted, once it had fulfilled the substantial condition of construction of the plant and the formal condition of registration within the prescribed ‘window.”⁶ This accepted offer, recorded in communications from the State to specific investors, amounted to a

³ Decision, para. 1073. ↩
⁴ Decision, para. 1070 (footnotes omitted). ↩
⁵ The Decision devotes several paragraphs to a discussion of legitimate expectations that can arise from a general regulatory framework. See Decision, paras. 1034-1049. While I find nothing to fault in this discussion, most of it is beside the point where, as is the case here, one sees State conduct that corresponds to the first source of legitimate expectations mentioned above: specific commitments, representations, or assurances. ↩
⁶ Doc. CL-22, Masdar, para. 512. ↩

[Page 3]

specific commitment causing investors to form legitimate expectations that they would be allowed to benefit from the incentives described in the law and in their written communications from the State for the period of time stated in the law and in those communications.⁷

8. There is no meaningful distinction between the Spanish and Romanian renewable-energy incentive schemes. Both made the following offer to potential investors: if you construct solar-power facilities of a certain capacity and connect them to the grid within a specified window of time, you will be permitted to charge tariffs calculated in a specified manner for a specified period of time. Participation in the Romanian scheme required accreditation, which would-be participants in the scheme had to request for their investments. Romania's responses to these requests are clear evidence of a specific commitment to Claimants.

9. The Majority and I are in agreement on this point. After reviewing a particular accreditation certificate, the Majority concludes that these certificates “constitute Specific Commitments, i.e., self-contained administrative acts, specifically addressed to Group A Claimants, which confirm that their PV Facilities met the necessary requirements and were entitled to benefit from the Essential Characteristics of the GC support scheme.”⁸

10. I also agree with the Decision's holding that “pursuant to this Specific Commitment, Group A Claimants, who had already relied on the regulatory promise to make their investments, confirmed their expectation that the Essential Characteristics of the GC scheme, as they existed at the time they made their investment, would remain stable.”⁹ Save for the reference to the so-called “Essential Characteristics," which I will address in the following section, I find this conclusion unimpeachable. The key component of the Decision's reasoning is that Claimants, having received a specific commitment from Romania, legitimately expected that the content of that commitment would remain stable.

The Invented "Essential Characteristics"

11. It is important at the outset to have in mind just what Romania did that has given rise to this case. The Decision accurately describes how a Green Certificate support scheme works at paragraphs 67-70. On the supply side, qualified RES-E generators receive a specified number of Green Certificates per MWh of power produced. On the demand side, the government obligates retail suppliers to procure a certain percentage of their electricity from RES-E in a given period, which translates into an obligation for these suppliers to purchase from RES-E producers a number of GCs per MWh of electricity they sell. The revenue that RES-E producers receive for their GCs is a major source of their income. That revenue is a function of GC supply and demand. Thus, a reduction in the number of GCs per MWh that a producer is entitled to receive will reduce that producer's revenue, other things remaining equal, as will a reduction in the obligation of

⁷ See Doc. CL-22, Masdar, paras. 520-522. ↩
⁸ Decision, para. 1111. ↩
⁹ Decision, para. 1111. ↩

[Page 4]

retail suppliers to purchase GCs. The actions of Romania that are challenged by Claimants adversely affected both GC supply and demand.

12. My problem with the Decision comes down to this: the Decision includes within its protected “Essential Characteristics” the supply side of Romania's commitment (six GCs per MWh) but excludes the demand side (the percentage of their electricity that retail suppliers were obligated to obtain from RES-E producers), other than to include the minimum price that RES-E producers were entitled to receive for their GCs. There is simply no reason to treat the supply and demand sides of Romania's GC support scheme differently.

13. That there is no foundation for limiting the scope of Romania's specific commitment to the Majority's “Essential Characteristics” becomes obvious when one even briefly considers the factors cited by the Majority as the reasons why Claimants were entitled to expect Romanian compliance with the “Essential Characteristics.” The Majority cites four reasons: (1) the purpose of Romania's GC support scheme was the promotion of investment in RES-E; (2) the Essential Characteristics were guaranteed by legislation; (3) Claimants relied on Romania's regulatory promise; and (4) Romania certified Claimants' entitlement by issuing accreditation certificates. All of these factors actually demonstrate that the scope of Romania's specific commitment encompassed the entire RES-E incentive, including the GC purchase obligations of retail sellers, and not just the "Essential Characteristics."

Promotion of Investment in RES-E

14. The Majority's first reason is that “Romania approved a regulatory regime, with the specific purpose of promoting investments in RES-E....”¹⁰ This statement is correct. The Majority establishes, on the basis of a survey of documentation including a parliamentary statement of reasons, communications to the European Commission, and Romanian government presentations, that Romania indeed wished to attract investment in RES-E by means of its regulatory regime.¹¹

15. However, the Majority draws an unnatural conclusion from this documentation. It states that¹²:

"Summing up, there is extensive evidence which proves that Romania created the GC support scheme endowed with certain Essential Characteristics to incentivize investment in RES-E in general, and in PV plants in particular. Romania equally represented, through its promotion actions, that the Essential Characteristics of the support scheme would remain stable."

16. The Majority's conclusion is incomplete. It is not just the Essential Characteristics that Romania used to incentivize investment, and it is not just the Essential Characteristics

¹⁰ Decision, para. 1071. ↩
¹¹ Decision, paras. 1074-1084. ↩
¹² Decision, para. 1084. ↩

[Page 5]

that Romania promoted to investors. Romania created the entire GC support scheme to encourage investment in PV plants, and, to the extent Romania represented through its promotion actions that the support scheme would remain stable, it represented that the entire scheme would remain stable, not just some parts of it. All of the evidence that the Majority cites in support of its Essential Characteristics is in fact evidence of this broader conclusion.¹³

Legislative Guarantee

17. The Majority's second reason why Claimants were entitled to expect Romanian compliance with the “Essential Characteristics” is that “The Essential Characteristics were guaranteed by legislation.”¹⁴ The Majority asserts that this legislation “constituted a regulatory promise, made to any future investors in RES-E.¹⁵ The appropriate response to this is: no, Romania's entire PV incentive scheme was “guaranteed by legislation," and its regulatory promise encompassed all elements of that scheme.

18. Once again, the evidence that the Majority lays out does not support its conclusion that Romania's specific commitment to Claimants consisted only of the “Essential Characteristics.” The Majority examines the relevant sections of the legal framework that set the terms of the GC program. In particular, it considers Articles 2(g), 3, 5, 9 and 10 of Law 139/2010,¹⁶ all of which support the Majority's conclusion that the "Essential Characteristics" were “guaranteed by legislation," and none of which supports the conclusion that only the “Essential Characteristics” were guaranteed by legislation.

19. For example, in the section of the discussion titled “The value of GCs,” the Majority reproduces Article 10 of Law 139/2010, which discusses how GCs are priced.¹⁷ The law explicitly stipulates that "the trade value for green certificates...varies between: a) a minimum trade value of 27 EUR/certificate; and b) a maximum trade value of 55 EUR/certificate." From this the Majority concludes that “PV Generators were entitled ... [t]o sell the GCs on the centralized market or through GCPAs, at the guaranteed minimum value of EUR 27/GC indexed annually to European inflation.”¹⁸ The more obvious conclusion to draw from the quoted statutory language is that PV Generators were entitled to sell their GCs for whatever price they could get for them on the centralized market or through GCPAs, within the minimum and maximum prices set by law.

¹³ For example, the Majority reproduces part of Romania's 2010 NREAP (Decision, para. 1081) and underlines a mention of a minimum price guarantee but neglects to underline the reference to "the green certificate system for electricity,” which must be taken to include the demand mechanism for determining GC prices. Even clearer is the excerpt from an ANRE presentation, which explicitly refers to how GCs would trade "bilaterally or on a centralized market ... at values between 27 EUR/GC and 55 EUR/GC" (Decision, para. 1083). Presumably, the reason the Majority does not emphasize this is because it contradicts its insistence that Romania only specifically committed to a minimum trading value for GCs. ↩
¹⁴ Decision, para. 1085. ↩
¹⁵ Decision, para. 1085. ↩
¹⁶ Doc. C-89. ↩
¹⁷ Decision, para. 1090. ↩
¹⁸ Decision, para. 1091. ↩

[Page 6]

Claimants' Reliance

20. The Majority's third reason is that “Claimants relied on the regulatory promise.”¹⁹ This is also an appropriate consideration, as evidence of what information Claimants relied on when making their investments indicates how Romania's incentive scheme appeared to sophisticated investors. The Majority surveys numerous documents relating to the financing of Claimants' investments, including proposals from Claimants to investors as well as analyses conducted by financers.²⁰ Its conclusion that Claimants expected "that, if they met the necessary requirements, they would benefit from the Essential Characteristics of the scheme"²¹ is once again incomplete. The evidence instead demonstrates that Claimants expected to benefit from the entirety of the scheme, and that neither Claimants nor their financing partners distinguished among different components of the scheme in a way that classified the demand mechanism as less important than other elements.²²

ANRE Accreditations

21. The Majority's fourth reason why Claimants could legitimately expect Romania to maintain the "Essential Characteristics" of the GC support scheme is that Romania certified Claimants' participation in the support scheme by issuing accreditation certificates. The Majority's conclusion in this regard is worth quoting in its entirety²³:

“The ANRE Accreditations constitute Specific Commitments, i.e., self-contained administrative acts, specifically addressed to the Group A Claimants, which confirm that their PV Facilities met the necessary requirements and were entitled to benefit from the Essential Characteristics of the GC support scheme. By issuing these Accreditations to Group A Claimants' PV Facilities, Romania converted its regulatory promise, enshrined in a legal and regulatory framework of general application, into a

¹⁹ Decision, para. 1094. ↩
²⁰ Decision, paras. 1094-1103. ↩
²¹ Decision, para. 1094. ↩
²² An example of the evidence of reliance cited by the Majority is a draft binding term sheet prepared by Raiffeisen Bank, one of the financers of several of Claimants' plants, for Pressburg. The Majority reproduces the following excerpt from that term sheet at paragraph 1096 of the Decision: ↩

According to the Romanian renewable energy system, the producers of electricity from PV power plants receive six green certificates for each MWh delivered into the power grid. The price for each green certificate has to be within a fixed, annually inflation adjusted range (EUR 27 to EUR 55). Additionally, a PPA for the sale of green certificates will be signed with Enel or OMV, in form and substance satisfactory to RBI.

This reflects an accurate understanding of the law. But then the majority summarizes Raiffeisen's position as follows: "Thus, in August 2012, Raiffeisen Bank, a bank prepared to grant financing for the development of RES-E in Romania, included in its term sheets, as one of the factors which mitigated the credit risk, that PV Generators were entitled to receive six GCs for each MWh delivered into the power grid and to sell them through GCPAs, and that the trade value of each of these GCs would not be less than EUR 27/GC, annually adjusted for inflation.” Decision, para. 1097. See Doc. C-123, p. 25 of the PDF (emphasis added). The Majority's summary of Raffeisen's term sheet is obviously incomplete, because Raffeisen referred to the range within which GCs would trade. It did not mention only, or even emphasize, the minimum price.
²³ Decision, para. 1111 (footnote omitted). ↩

[Page 7]

Specific Commitment. And pursuant to this Specific Commitment, Group A Claimants, who had already relied on the regulatory promise to make their investments, confirmed their expectation that the Essential Characteristics of the GC scheme, as they existed at the time they made their investment, would remain stable."

22. I agree that the ANRE Accreditations are indicative of a binding specific commitment to Claimants.²⁴ But I cannot agree with the Majority's conclusion that "pursuant to this Specific Commitment, Group A Claimants ... confirmed their expectation that the Essential Characteristics of the GC scheme, as they existed at the time they made their investment, would remain stable."²⁵ Again the Majority makes an incomplete statement. There is nothing in the ANRE certificates that indicates that Romania's specific commitment consisted only of the “Essential Characteristics.” The certificates refer to the incentive scheme as a whole. What investors "confirmed" in those certificates was their expectation that the entire GC scheme as it existed when they made their investments would remain stable.

23. In summary, all of the documentation reviewed by the Majority fails to support its conclusion that Romania's specific commitment only encompassed those elements of the incentive scheme that the Majority refers to as “Essential Characteristics.” This is plain from Romania's promotional materials, the text of the relevant legislation, evidence of Claimants' and investors' understanding of the program, and the accreditation certificates. But there is an even more fundamental problem with the way in which the Majority defines the scope of the specific commitment.

24. At the heart of the Majority's logic for drawing the boundaries of Romania's specific commitment the way it does is the notion that “GCs are a legal and regulatory construct, which is intrinsically worthless....”²⁶ This is a true statement, even if it is somewhat banal: every instrument or commodity created by regulation is “intrinsically worthless” in the sense that, absent the representations made by a government or some other authority, it would have no value. This does not mean, however, that regulatory constructs such as GCs cannot come to possess real value, or that the promise of their value cannot legitimately induce reliance on the part of would-be participants in government-created markets.

25. The Majority seems to understand this, as it admits that “a PV generator who receives GCs will not find anyone interested in buying these titles, except if the regulation creates demand, e.g., by forcing certain players in the electricity market to purchase a certain

²⁴ The Majority states that Romania “converted" a regulatory promise into a specific commitment by means of the ANRE certificates. This is consistent with my view of the matter, although I think it is more useful to view the "regulatory promise" as a unilateral offer made by Romania to potential investors, which offer could be accepted by performance, i.e., making the sort of investment required by the offer within the timeframe set in the offer. The ANRE accreditation is, I think, best viewed as conclusive evidence that the investor accepted Romania's offer by performance and, therefore, that Romania is now bound by the terms of its offer. In this regard, see Doc. CL-22, Masdar, para. 572. ↩
²⁵ Decision, para. 1111. ↩
²⁶ Decision, para. 1090. ↩

[Page 8]

number of titles.”²⁷ This is exactly right. It is for this reason that Romania created demand by forcing electricity suppliers to purchase GCs in accordance with a mandatory quota. This demand, which is entirely the product of regulation, is what gives GCs value in the context of a certain supply.

26. What does not give the GCs value is the statutory minimum price that the Majority finds essential. The minimum price is just that: a floor beneath which the supply-demand interaction cannot push the price of GCs. The minimum and maximum prices' sole function is to place limits on the movement of the price for GCs as it is determined by the market. There are practical reasons for this: if the price is too low, investors will not invest; if it is too high, the cost of the program becomes unbearable for governments and, in turn, for consumers. The price floor and ceiling work in concert with the demand mechanism to create value at a level acceptable to all parties. A simple thought experiment demonstrates this. If Romania had enacted an incentive scheme providing for a minimum price but no demand mechanism, there would be no guarantee that anyone would ever buy GCs at any price. The minimum price would be akin to a minimum bid requirement imposed by an overly optimistic seller at an auction with no one in the audience. Without something to create demand, the minimum price is an empty figure, and there is no guarantee that any value is realized whatsoever.

The Majority misapprehends the functioning of the GC support scheme and, in so doing, confuses issues going damages with issues going to liability

27. The Majority attempts to address my concerns under the heading “Ancillary characteristics of the regulatory regime,” beginning at paragraph 1115. In that paragraph the Majority reveals that it misapprehends what is at issue. According to the Majority, Claimants “submit that the legal framework was designed in such a way that allows investors [to] reliably predict GC supply and demand, which in turn allowed them to predict that GCs would be trading at or near their ceiling value.” Claimants may well assert that Romania's GC support scheme allowed them to “reliably predict GC supply and demand." Whether that assertion has merit, however, is a matter that belongs in the consideration of damages; it has no bearing on the question of liability. Whether Claimants could legitimately expect the statutory demand mechanism (and not just the six GCs per MWh supply mechanism) to remain in place, however, goes directly to liability.

28. Continuing with this line of thought, the Majority asserts (at paragraph 1123) that “[t]here is no support in Romania's legal and regulatory framework for the proposition that the GC support scheme guaranteed a defined level of either supply ... or demand ... for GCs, which in turn would allow investors to predict that GCs would be trading at or near the maximum value.” I am not sure that Claimants ever assert this proposition. That the Majority nonetheless devotes many paragraphs to refuting it reveals that the Majority does not understand how GC support schemes are supposed to work, for the functionality,

²⁷ Decision, para. 1090. ↩

[Page 9]

and indeed attractiveness, of such schemes is predicated on their not guaranteeing a particular level of GC supply or demand.

29. What makes a Green Certificate support program different from a fixed feed-in tariff (FiT) or a feed-in premium (FiP) is its ability to self-correct for changes in the supply of RES-E or the demand for electricity. When there is more RES-E investment than anticipated, the supply of GCs increases with the increase of RES-E power generation, which causes the price of GCs to decrease (all else remaining equal), which serves to reduce the investment incentive, which serves to discourage new investment going forward. Similarly, when electricity consumption is less than expected, the demand for GCs decreases (demand being set as a percentage of total electricity consumption), thus lowering the price of GCs, thus again reducing the investment incentive. The opposite occurs when either RES-E investment drops or electricity consumption increases.²⁸ Thus, it is of the essence of a GC support scheme that the scheme allow levels of GC supply and demand to fluctuate because the whole point of such a scheme is to let changes in supply and demand encourage or discourage additional RES-E investment as circumstances change.

30. It also is of the essence of such schemes that investors have certainty as to the parameters that will determine the supply of and demand for GCs. Thus, a potential investor will rely both on the elements of a GC support scheme going to supply (the number of GCs per MWh) and the elements going to demand (the percentage of total consumption to be met from RES-E). Romania's Disputed Measures significantly reduced both the supply of and the demand for GCs, allegedly to the great detriment of Claimants. The notion that Romania's specific commitment went only to the supply side of the equation ignores the essential nature of a GC support scheme.

31. The Romanian demand mechanism is set forth in Article 4 of Law 139/2010²⁹ and in Article 10 of ANRE Order 45/2011,³⁰ which are described more or less accurately at paragraphs 1129-1146 of the Decision. The workings of the demand mechanism come down to this: By statute, Romania set Annual Mandatory Quotas (for the amount of electricity produced from renewable sources which benefit from the incentive scheme) that increased gradually to 20 percent of total electricity consumption by 2020. ANRE set the Annual Acquisition Quotas for electricity suppliers as the ratio of the expected supply of GCs for the coming year to the estimated total consumption of electricity for the same period of time. “If the estimated Acquisition Quota was lower than the Annual Mandatory Quota (i.e., the maximum amount of RES-E that could benefit from the GC scheme, as defined in Law 139/2010), then the Acquisition Quota was confirmed. But if

²⁸ See C-I, para. 95. See also Roques I, paras. 4.42-4.46. “The efficiency of a GC scheme therefore critically depends on the predictability of the GC demand defined ex ante by the RES target and the ability of investors to form a view on the evolution of new RES projects cost evolution." (Roques I, para. 4.43); Jones I, para. 3.55. ↩
²⁹ Doc. C-89. ↩
³⁰ Doc. C-101. ↩

[Page 10]

it was higher, the Acquisition Quota would be reduced to the percentage of the Annual Mandatory Quota set out in the Law.”³¹

32. The Majority makes much of the fact that the Annual Mandatory Quota is described as the maximum amount of RES-E that could benefit from the GC support scheme.³² This is true, but it only illustrates the important effect of Romania's lowering that maximum. The Annual Acquisition Quota was set initially as a simple ratio of the expected supply of GCs and the expected total electricity consumption for the coming year.³³ So, for example, if the expected supply of GCs represented 1,500 MWh of electricity, the expected total consumption of electricity was 10,000 MWh, and the Annual Mandatory Quota was 20 percent, an Annual Acquisition Quota of the GCs representing all RES-E production would be confirmed. (In this example the electricity represented by the GCs would be 15 percent of total consumption, well below the 20 percent quota.) If, however, the Annual Mandatory Quota was reduced to 10 percent, the Annual Acquisition Quota would be reduced to the 10 percent cap. In the first case, GC supply and demand would be in balance, with the Acquisition Quota set to equal the supply of GCs, and GCs could be expected to trade at prices above the minimum. In the second case, there would be an oversupply of GCs, causing them to trade at or near the minimum price. In short, the Annual Mandatory Quota matters, and it matters quite a lot. Claimants' expectation that this component of the incentive scheme would remain stable is therefore not only legitimate with respect to Romania's express representations – it is a logically necessary expectation to have should one believe that Romania intended the GC scheme to function the way GC schemes are generally intended to function.

Conclusion

33. It is unfortunate for our purposes that the Romanian incentive scheme is as complicated as it is because these complications make it more difficult to understand what Romania has done with its challenged measures and to understand just how the majority has erred. So, I will close with a simpler hypothetical.

34. Romania's commitment to Claimants may fairly be analogized to an oil-production investment law in which the State offers potential investors, in exchange for making specified investments in oil exploration and production, the right to sell for their own account all oil they produce over the life of their investments but allocates the risk of variations in the price of oil by guaranteeing investors that they always will receive at least $X per barrel of oil produced, with the State making up any shortfall in price, and never will receive more than $2X per barrel of oil, with the State keeping any excess over that price. If the State in this example changed its investment law such that investors were allowed to sell only half of their oil production, with the State selling the other half, but left investors' rights otherwise unchanged, one may presume that the Majority would find this to be a violation of the State's FET obligation but would find that obligation to extend only to the guaranteed minimum price.

³¹ Decision, para. 1145 (footnote omitted). ↩
³² Decision, paras. 1133-1134. ↩
³³ See Decision, paras. 1144-1145; Doc. C-101, Art. 10. ↩

[Page 11]

35. This example, I think, makes obvious the fundamental error in the Majority's analysis. In my example, it is plain that any investor contemplating an oil-production investment under this hypothetical law would base its decision to invest, in part, on its expectation of where, between the permitted minimum and maximum prices, the price of oil would fall over the course of the investment's life. The investment certainly would be less attractive if the investor thought the market price of oil would remain at or below the minimum, and even less attractive if the investor believed that the State would never permit it to sell the oil for more than the minimum price whatever the market price might be. In this example what the investor was “buying” from the State with its investment was the chance to earn between $X and $2X for every barrel of oil produced. The State effectively took from the investor its chance to make that revenue on the half of its production that it could no longer sell under the revised law. Whether the price of oil could reasonably have been expected to be any greater than the minimum price would be a question for the damages phase. The violation of FET would be the State's violation of its specific commitment to allow the investor to sell its oil at a price between $X and $2X per barrel.

36. This, of course, is exactly the situation with which we are presented. There is no evidence in this case, nor even any argument by Respondent, that Claimants – or any PV investor – would have invested in the Romanian PV sector if that investor had understood Romania's commitment to be that the investor would receive the minimum price, and only the minimum price, for its GCs. All the evidence instead suggests that Claimants invested for the chance to earn a price per GC between the minimum and maximum trading values set by statute. Yet the Majority characterizes Romania's commitment in a way that is belied by this evidence. That characterization is strained and plainly wrong. It may be that the evidence before us does not support a conclusion that the price of Claimants' GCs could reasonably have been expected to exceed the minimum price at any time in the 15-year span of Romania's specific commitment. But that is a question that goes to damages, not to liability.