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

INTERNATIONAL CENTRE FOR SETTLEMENT OF INVESTMENT DISPUTES
1818 H STREET, NW | WASHINGTON, DC 20433 | USA
TELEPHONE +1 (202) 458 1534 | FACSIMILE +1 (202) 522 2615
WWW.WORLDBANK.ORG/ICSID

CERTIFICATE

SEVILLA BEHEER B.V. AND OTHERS

V.

KINGDOM OF SPAIN

(ICSID CASE NO. ARB/16/27)

I hereby certify that the attached document is a true copy of the English version of the Tribunal’s Award dated May 22, 2023.

Signature

Meg Kinnear
Secretary-General

Washington, D.C., May 22, 2023

INTERNATIONAL CENTRE FOR SETTLEMENT OF INVESTMENT DISPUTES

In the arbitration proceeding between

Sevilla Beheer B.V. and others
(Claimants)

and

The Kingdom of Spain
(Respondent)

ICSID Case No. ARB/16/27

AWARD

Members of the Tribunal
Dr. Raëd M. Fathallah, President of the Tribunal
Professor Peter D. Cameron, Arbitrator
Professor Attila Tanzi, Arbitrator

Secretary of the Tribunal
Mr. Marco Tulio Montañés-Rumayor

22 May 2023

[Page ii]

This page intentionally left blank

[Page iii]

[Page iv]

REPRESENTATION OF THE PARTIES

[Page v]

FREQUENTLY USED ABBREVIATIONS

[Page vi]

[Page vii]

[Page 1]

I. INTRODUCTION AND PARTIES

1. The claimants in this arbitration are Sevilla Beheer B.V. (“Sevilla Beheer") and Cordoba Beheer B.V. (“Cordoba Beheer”), two private limited liability companies incorporated under the laws of the Netherlands, as well as 57 Spanish companies (the "Claimants").¹ The claimant entities are ultimately controlled by a Dutch national, Mr. Reinier Bouman, the majority owner of the share capital in the Claimants and the CEO of Sevilla Beheer. Mr. Reinier Bouman's daughter owns a minority stake in the Claimants.

2. The respondent in this arbitration is the Kingdom of Spain ("Spain" or the “Respondent").

3. The Claimants and the Respondent are collectively referred to as the “Parties"; the names of their representatives are set out on page iv above.

4. This arbitration concerns a dispute submitted to the International Centre for Settlement of Investment Disputes (“ICSID” or the “Centre”) on the basis of the Energy Charter Treaty (the “ECT”) and the Convention on the Settlement of Investment Disputes between States and Nationals of Other States (the “ICSID Convention”). The ECT entered into force for the Netherlands and Spain on 16 April 1998. The ICSID Convention entered into force on 14 October 1966 for the Netherlands, and on 17 September 1994 for Spain.

5. In this matter, the Claimants allege that the Respondent's legislative and regulatory measures enacted in respect of its renewable energy (“RE”) sector in 2010-2014 (the "Disputed Measures”) and described in detail in Section III.C of the Decision on Jurisdiction, Liability and the Principles of Quantum (the “Decision”), dated 11 February 2022, violated Article 10(1) of the ECT and caused damages to the Claimants' investments in their photovoltaic plants² (the "PV Plants"). The

¹ See Annex I to the Decision on Jurisdiction, Liability and the Principles of Quantum. The 57 Spanish companies will hereinafter be referred to as the "Spanish Project Companies." ↩
² The Claimants have invested in the following PV Plants: Mahora PV, Villar de Cañas PV, Ronda PV, Matapozuelos PV, and Fuentes de Año PV. ↩

[Page 2]

Respondent argues that the Tribunal lacks jurisdiction over the present dispute and that in any event the Disputed Measures did not breach the ECT.

6. The Tribunal refers to Sections III-VII of its Decision for the factual and legal background of the case. The full text of that Decision is hereby made an integral part of this Award.

7. The Operative Part of the Tribunal's Decision reads as follows:

For these reasons, the Tribunal decides as follows:
i. Unanimously, that the Tribunal has jurisdiction over the Parties and the subject-matter of this dispute with the exception that it has no jurisdiction to determine whether the TVPEE breached Spain's obligations under the ECT;
ii. Unanimously, that the Respondent breached Article 10(1) of the ECT to the extent that it applied the New Regime retroactively to the remuneration already received by the Claimants' PV Plants under RD 661/2007 and RD 1578/2008;
iii. By majority, that the Respondent has breached Article 10(1) of the ECT to the extent (if any) that the New Regime does not provide a reasonable return to the Claimants' PV Plants at a rate of 7% after taxes; and
iv. By majority, that all other claims under Article 10(1) of the ECT are dismissed.
The procedure to be followed by the Parties for the purposes of quantifying damages (if any) resulting from the majority's findings on liability and the principles of quantum will be set out by the Tribunal in a separate procedural order.
The issues of damages (if any) and costs are reserved for a Final Award.³

³ Decision, paras. 1080-1082. ↩

[Page 3]

II. PROCEDURAL HISTORY

8. On 17 February 2022, with a view to determining the procedure to be followed by the Parties for the purposes of quantifying damages (if any), and as referred to in the penultimate paragraph of the Decision's Operative Part, the Tribunal issued Procedural Order No. 3 inviting the Parties to submit on or before 21 March 2022, a joint proposal regarding the organization of the next stage of these proceedings (or indicate the differences in their positions in respect of the same).

9. On 19 March 2022, the Parties informed the Tribunal by e-mail that they had reached a partial agreement in respect of the organization of the next stage of these proceedings.

10. On 25 March 2022, having sought and considered the Parties' views regarding the organization of the next stage of these proceedings, the Tribunal issued Procedural Order No. 4 fixing the procedural calendar.

11. On 14 May 2022, the Parties agreed to extend the deadline for the submission of the Joint Expert Report on the Claimants' Internal Rate of Return and compensation for retroactivity (the “Joint Expert Report”) until 20 May 2022.

12. On 20 May 2022, the Parties submitted the Joint Expert Report.

13. On 10 June 2022, the Parties filed their respective submissions addressing the content of the Joint Expert Report.

14. On 29 June 2022, Spain submitted a Request for Reconsideration of the Decision in relation to the Tribunal's findings on jurisdiction (the "Request for Reconsideration”). In its Request for Reconsideration, the Respondent also sought leave to introduce into the record the following legal authorities: (i) Green Power K/S and SCE Solar Don Benito APS v. Kingdom of Spain, SCC Case No. V2016/135, Award, 16 June 2022 and (ii) Standard Chartered Bank (Hong Kong) Limited v. Tanzania Electric Supply Company Limited, ICSID Case No. ARB/10/20, Decision on Annulment, 22 August 2018.

[Page 4]

15. On 7 July 2022, further to the Tribunal's invitation, the Claimants submitted their response to the Respondent's Request for Reconsideration.

16. On 11 August 2022, having considered the Parties' respective positions, the Tribunal rejected the Request for Reconsideration while reserving any decision on costs (the “Decision on the First Request for Reconsideration”).

17. On 30 August 2022, the Tribunal posed to the Parties additional questions regarding certain outstanding quantum issues and invited the Parties to agree on a date for sharing their updated submissions on costs.

18. On 30 September 2022, the Parties submitted their respective answers to the Tribunal's questions.

19. On 14 October 2022, the Parties submitted their comments on the responses of the other Party to the Tribunal's questions.

20. On 11 November 2022, the Parties submitted their updated costs submissions to the Tribunal.

21. On 27 December 2022, the Respondent submitted another Request for Reconsideration of the Decision in relation to the Tribunal's findings on jurisdiction (the "Second Request for Reconsideration"). In its Second Request for Reconsideration, the Respondent also sought leave to introduce into the record the following documents: the judgment by the Svea Court of Appeal, dated 13 December 2022 and the judgment by the Swedish Supreme Court, dated 14 December 2022 (the “Swedish Judgments").⁴

22. On 16 January 2023, further to the Tribunal's invitation, the Claimants submitted their response to the Respondent's Second Request for Reconsideration.

⁴ Svea Court of Appeal, Case No. T 4658-18, Judgment, 13 December 2022 (RL-139); Swedish Supreme Court, Case No. T 1569-19, Judgment, 14 December 2022 (RL-140). ↩

[Page 5]

23. On 30 January 2023, having considered the Parties' respective submissions, the Tribunal rejected the Second Request for Reconsideration and reserved any decision on costs (the “Decision on the Second Request for Reconsideration”).

24. On 10 February 2023, further to the Tribunal's invitation, the Claimants submitted their observations on the issue of the allocation of the arbitration costs and counsel fees incurred because of Spain's Second Request for Reconsideration.

25. On the same date, the Respondent requested an extension of one week to submit its observations on costs. In response, the Tribunal noted that the Respondent's request for an extension of one week was not reasoned and appeared to be excessive in view of the page limit of 2 pages set by the Tribunal. The Tribunal thus invited the Respondent to endeavour to submit its observations as soon as possible and no later than 17 February 2023, noting that no further extensions would be granted.

26. On 15 February 2023, the Respondent filed its Submission on Allocation of Costs Relating to the Respondent's Second Request for Reconsideration.

27. On 6 March 2023, the Tribunal declared the proceeding closed.

[Page 6]

III. REQUESTS FOR RELIEF

28. The Claimants request that the Tribunal:

(a) order that Spain pay the Claimants compensation for a total amount of EUR 6,871,246;
(b) award the Claimants pre-award interest at a rate of 7.58% compounded monthly;
(c) award the Claimants post-award interest, at a rate of 7.58% compounded monthly from the date of the award until full payment thereof;
(d) award the Claimants all costs and expenses of this arbitration on a full-indemnity basis, including all expenses that the Claimants have incurred or will incur in respect of the fees and expenses of the arbitrators, ICSID, legal counsel, experts and consultants; and
(e) order any other relief that the Tribunal may deem just and proper.⁵

29. The Respondent requests the following relief:

i. Declare that the Respondent has not breached Article 10(1) of the ECT to the extent that the New Regime does provide a reasonable return to the Claimants' PV Plants at a rate of 7% after taxes;
ii. Adopt Quadrant Economics' approach when addressing quantum in the Final Award for the reasons expressed in the Joint Report and in this Submission, and quantify the total damages to be paid by the Respondent in [EUR] 4,092,193; and
iii. Order the Claimant to pay all costs and expenses derived from this arbitration, including ICSID administrative expenses, arbitrators' fees and the fees of the legal representatives of the Kingdom of Spain, their experts and advisors, as well as any other cost or expense that has been incurred, all of this including a reasonable rate of interest from the date on which these costs are incurred until the date of their actual payment.⁶

IV. THE REMAINING ISSUES ON LIABILITY AND QUANTUM

30. In accordance with the findings made in the Decision,⁷ the Tribunal will analyze below the remaining issues on liability and quantum.

31. First, the Tribunal will address the issue of compensation due for the retroactive application of the New Regime (i.e. RDL 9/2013 and related subsequent measures

⁵ Claimants’ Quantum Submissions, para. 74. ↩
⁶ Respondent’s Quantum Submissions, para. 55. ↩
⁷ See Operative Part; See also paras. 1019-1021 of the Decision. ↩

[Page 7]

including Law 24/2013, Order IET/1045/2014, and Royal Decree 413/2014⁸) to the remuneration already received by the Claimants' PV Plants under RD 661/2007 and RD 1578/2008 (Section IV.A).

32. Second, the Tribunal will consider whether the Claimants' Internal Rate of Return (the “IRR”) was below the reasonable rate of return, which, pursuant to the Decision, shall be 7% after taxes (Section IV.B).

33. Finally, the Tribunal will address the interplay between the compensation for retroactivity and the compensation to achieve the reasonable return, and fix the amount of damages (Section IV.C).

A. Compensation for Retroactivity

1. The Parties' Positions

a. The Joint Expert Report

34. As to the retroactivity of the New Regime (see paragraphs 903-907 of the Decision), the Experts agree on the methodology and disagree on the calculation of one parameter,⁹ which results in the following compensation according to each Experts' approach:¹⁰

⁸ See Decision, para. 465. ↩
⁹ Joint Expert Report, para. 2. ↩
¹⁰ Joint Expert Report, para. 10, Table 1. ↩

[Page 8]

35. The Experts agree on the following assumptions:

a) [...] as the Claimants' PV Plants all have a different IT code under the New Regime, five in total, it is necessary to calculate this compensation separately for each of the five plants.
b) To use the projected cash flows previously modelled in these proceedings under the actual scenario, with the exception of the actual revision of the regulatory rate of return, which is 7.09%.
c) To use the standard initial investment cost as the initial net asset value [(the "Initial NAV")], which is consistent with Spain's approach under the New Regime, corresponding to the year when each plant was constructed.
d) To calculate a NAV as of July 13, 2013, which is the date that the NAV was calculated for the New Regime.¹¹

36. The Experts also agree to use the following methodology:

a) The first step is to calculate an NAV as of July 13, 2013, given the [Initial Nav] and the pre-tax reasonable rate of return for the years of operation prior to the New Regime, ignoring the actual cash flows achieved during the same period [(the "Adjusted NAV”)].
b) The second step is to use the Adjusted NAV to determine the adjusted investment remuneration parameter [(the "Adjusted Rinv")] from July 13, 2013.
c) The third step is to implement the adjusted Rinv computed in the prior step in the forecasted operations of each plant, to obtain the resulting cash flows generated by each plant.
d) The fourth step is to compute the difference between the cash flows of each plant using the Adjusted Rinv and the actual Rinv.
e) Finally, discount the difference calculated in the prior step using a post-tax WACC of 6.03% to June 20, 2014, to obtain the compensation value.¹²

37. Despite the above, the Experts do not agree on how the Adjusted NAV should be calculated as of 13 July 2013.¹³ The calculation of this parameter requires an assumption of a pre-tax target rate of return between each PV Plants' commissioning date and 13 July 2013. The Experts' respective positions on the relevant pre-tax rate of return are summarized in the below paragraphs.

¹¹ Joint Expert Report, para. 4. Footnotes omitted. ↩
¹² Joint Expert Report, para. 5. Footnotes omitted. ↩
¹³ Joint Expert Report, paras. 7-9. ↩

[Page 9]

i. Compass Lexecon

38. According to the Claimants' Expert, Prof. Pablo T. Spiller from Compass Lexecon, a 10% rate of return should be applied. To calculate this rate of return, Prof. Spiller uses the 7% post-tax determined as the reasonable rate of return as per the Decision and converts it to a pre-tax rate of return by dividing 7% by (1-30%), where 30% is the nominal tax rate prevailing between the PV Plants' commissioning dates and 13 July 2013.¹⁴ Prof. Spiller explains that this approach "is the same approach used by Spain in its most recent calculation of the pre-tax regulatory rate of return under the New Regime, applicable to the regulatory period starting in 2020.”¹⁵

ii. Quadrant Economics

39. Dr. Daniel Flores, the Respondent's Expert from Quadrant Economics (formerly from Econ One), uses a pre-tax rate of return of 7.398%.¹⁶ Dr. Flores notes that this rate is set out in the regulations implementing the New Regime and that both Experts agree to use it in the period from 13 July 2013 through 31 December 2019.¹⁷ In respect of Compass Lexecon's 10% rate, Dr. Flores states that it is a rate that does not appear in any law or decree implementing the New Regime in 2013.¹⁸

b. The Claimants' Quantum Submissions

40. In their Quantum Submissions, the Claimants address the following issues: the operation of the “claw-back” element of the New Regime,¹⁹ the methodology followed by the Experts to quantify the compensation for retroactivity, and the merits of Prof. Spiller's approach in relation to the quantification of the Adjusted NAV during the period between the PV Plants' commissioning dates and 13 July 2013.²⁰

41. As regards the latter, the Claimants argue that Dr. Flores' Adjusted NAV based on the assumption of a reasonable pre-tax return of 7.398% is inappropriate, as it leads to an overestimation of capital recovery as that implemented by Spain through the

¹⁴ Joint Expert Report, paras. 7-8. ↩
¹⁵ Joint Expert Report, para. 8. ↩
¹⁶ Joint Expert Report, para. 9. ↩
¹⁷ Joint Expert Report, para. 9. ↩
¹⁸ Joint Expert Report, para. 9. ↩
¹⁹ Claimants’ Quantum Submissions, paras. 9-16. ↩
²⁰ Claimants’ Quantum Submissions, paras. 17-22. ↩

[Page 10]

New Regime.²¹ The Claimants explain that “[t]he lower the assumed target reasonable rate of return, the higher the assumed capital recovery as of 13 July 2013 and therefore the lower the NAV. [...] By assuming a 7.398% pre-tax reasonable rate of return for pre-2013, the New Regime: (a) purposefully ignores the target rate of return underlying the Original Regime; and (b) overestimates the extent of capital recovery before its entry into force, thus reducing the required level of remuneration going forward to achieve the newly defined ‘target rate of reasonable return'."²²

42. Thus, according to the Claimants, Dr. Flores' Adjusted NAV was not sufficiently increased to the point where the “claw-back” of the New Regime would be fully removed.²³

43. The Claimants thus adhere to Prof. Spiller's conversion of the 7% post-tax reasonable rate of return (deemed by the Tribunal to be the target rate of return of the Original Regime) into a pre-tax rate of 10% using a widely applied formula: “dividing the post-tax rate by one minus the nominal tax rate of 30% (7%/(1-30%)=10%).”²⁴

44. The Claimants emphasize that Dr. Flores “does not disagree with this conversion from a post-tax return to a pre-tax return.”²⁵ Rather, his only criticism is that Prof. Spiller's pre-tax return of 10% does not appear in any law or decree implementing the New Regime in 2013.²⁶ This, in the Claimants' view, is a red herring, as Prof. Spiller's 10% rate of return is the pre-tax equivalent of the rate of return underlying the Original (Feed-in Tariff “FiT”-based) Regime (not the New Regime).²⁷ According to the Claimants, “using the Original Regime's rate of return is required precisely to eliminate the second clawback element of the New Regime's clawback (overestimation of capital recovery in the years preceding its entry into force)."²⁸

²¹ Claimants’ Quantum Submissions, para. 19. ↩
²² Claimants’ Quantum Submissions, para. 18. ↩
²³ Claimants’ Quantum Submissions, para. 19. ↩
²⁴ Claimants’ Quantum Submissions, para. 20. ↩
²⁵ Claimants’ Quantum Submissions, para. 21. ↩
²⁶ Claimants’ Quantum Submissions, para. 21. ↩
²⁷ Claimants’ Quantum Submissions, para. 21. ↩
²⁸ Claimants’ Quantum Submissions, para. 21. ↩

[Page 11]

Thus, in the Claimants' view, Dr. Flores' calculation is incomplete and must be rejected.²⁹

c. The Respondent's Quantum Submissions

45. The Respondent argues that the pre-tax rate of return of 7.398% must be used for the calculation of the Adjusted NAV by essentially reiterating the arguments made by Dr. Flores in the Joint Expert Report and summarized at paragraph 39 above.³⁰

d. The Parties' Responses to the Tribunal's Questions

46. The Tribunal put to the Parties the following additional questions concerning the compensation for retroactivity:

1) Based on the Tribunal's findings on liability and the need to quantify and, if applicable, compensate for the 'claw-back' element of the New Regime, and in view of the Claimants' argument that Dr. Flores' approach does not fully eliminate retroactivity [see Claimants' Submissions on Quantum, paras. 18-22], the Respondent and its Expert are invited to explain why should the rate of 7.398% established by the New Regime apply to the period before 13 July 2013 [see Joint Expert Report, para. 9]? Similarly, why, again considering the Tribunal's findings on liability, should the Adjusted NAV be based on a rate indicated in a law or decree implementing the New Regime [see Joint Expert Report, para. 9]?
The Respondent is also invited to comment on the formula and parameters used by Prof. Spiller.
2) The Claimants and their Expert are invited to make comments (if any) regarding Dr. Flores' remark that both Experts agree to use the 7.398% rate of return during the period from July 2013 to December 2019 [see Joint Expert Report, para. 9].³¹

i. The Claimants' Responses to the Tribunal's Questions

47. The Claimants and their Expert, Prof. Spiller, reiterate their arguments on the appropriateness of the Adjusted NAV based on the 7% rate of return post-tax.³²

²⁹ Claimants’ Quantum Submissions, para. 22. ↩
³⁰ Respondent’s Quantum Submissions, para. 7. ↩
³¹ Tribunal’s letter to the Parties, dated 30 August 2022. ↩
³² Claimants’ Response to the Tribunal’s Questions on Quantum, 30 September 2022, paras. 7-9; Prof. Spiller’s Answers to the Tribunal’s Questions on Quantum, 30 September 2022, paras. 2-6. ↩

[Page 12]

48. The Claimants further submit that Dr. Flores' remark that both Experts agree to use the 7.398% rate of return during the period from July 2013 to December 2019 is irrelevant to the issue the Tribunal needs to determine.³³ Specifically, the Claimants argue as follows:

Dr Flores's remark that he and Professor Spiller agree to use the New Regime's 7.398% pre-tax return for the post-July 2013 period has no bearing on this discussion. The only reason that the Experts agree that the 7.398% rate should remain unchanged post-2013 for the isolation of Retroactivity Damages is simply because there is no “claw-back” to isolate after July 2013. The claw-back only operates over pre-2013 revenues, by underestimating the 2013 NAV, therefore reducing future revenues. In other words, the revenues that the PV Plants would receive under the New Regime after July 2013 are determined based on the level of support they had received before July 2013 (which Spain claws-back through the underestimated July 2013 NAV).³⁴

49. Prof. Spiller explains his agreement regarding the use of the 7.398% rate of return for the period after July 2013 as follows:

[...] for the second period, after July 13, 2013, I use the pre-tax rate of return set by the New Regime. This is because I understand that the compensation for retroactivity is meant to eliminate the negative impact on the Claimants' PV Plants resulting from the New Regime taking into account the remuneration obtained by the PV Plants prior to July 2013 when calculating the July 2013 NAV. Therefore, I understand that when calculating the compensation for retroactivity, the 7.398% pre-tax regulatory rate of return implemented by the New Regime after July 2013 is not being disputed. As a result, for computing revenues from July 2013, I agree to use the pre-tax return set by the New Regime, which is 7.398% between July 2013 and December 2019, and 7.09% from January 2020.³⁵

50. In their Comments on Spain's Responses, the Claimants note that Dr. Flores' attempt to justify his approach by reference to BayWa v. Spain should fail, as in contrast with the BayWa tribunal's findings, “this Tribunal has: (a) determined a level of reasonable return prior to July 2013 (namely, 7% post-tax); and (b) never ordered that the New Regime's 7.398% target return be used to calculate the Adjusted NAV.”³⁶

³³ Claimants’ Response to the Tribunal’s Questions on Quantum, 30 September 2022, para. 7. ↩
³⁴ Claimants’ Response to the Tribunal’s Questions on Quantum, 30 September 2022, para. 10. ↩
³⁵ Prof. Spiller’s Answers to the Tribunal’s Questions on Quantum, 30 September 2022, para. 7 (emphasis in the original). ↩
³⁶ Claimants’ Comments on Spain’s Responses to the Tribunal’s Questions on Quantum, 14 October 2022, para. 11. ↩

[Page 13]

51. Prof. Spiller adds that Dr. Flores' interpretation of the Decision, “mixes the mention of the 7.398% pre-tax rate (used to calculate the NAV defined under the New Regime) with the amount that the New Regime was wrong in discounting to calculate the NAV (i.e., 'the past remuneration received by the PV installations that was in excess of the reasonable rate of return').”³⁷ Prof. Spiller has also provided the following figure demonstrating the difference between his and Dr. Flores' approaches (using the Mahora Plant for illustrative purposes):³⁸

Figure 1: Mahora Counterfactual 7% Post-tax Adjusted NAV and 7.398% Pre-tax NAV¹

[Chart showing two lines, one red and one blue, both declining over time from 2007 to 2020. A vertical line marks "Start of New Regime" at 2013. The Y-axis is labeled "NAV €Million/MW" and ranges from 4.0 to 7.0. The X-axis shows years from 2007 to 2020. The blue line is consistently higher than the red line from 2007 to 2013, after which they appear to run parallel. The legend indicates: Blue line is "7% Post-tax (2007-2013)" and "7.398% Pre-tax (2013-2019)". Red line is "7.398% Pre-tax (2007-2013)" and "7.398% Pre-tax (2013-2019)".]

52. Prof. Spiller provides the following commentary: “[f]rom the plant's commissioning date up until 2013, the lines diverge as the Retroactivity applied by Spain (i.e. deducting revenues in excess of 7.398% pre-tax) decreases the net asset value at a faster rate. As the 2013 NAV is lower, Spain's New Regime provides less remuneration accruing to the PV Plants over their remaining regulatory lives (since there is less of initial standard investment remaining to be recovered). Starting from 2013, both scenarios apply the 7.398% pre-tax under the New Regime."³⁹ The

³⁷ Prof. Spiller’s Comments on Dr. Flores’ Answers to the Tribunal’s Questions on Quantum, 14 October 2022, para. 5. ↩
³⁸ Prof. Spiller’s Comments on Dr. Flores’ Answers to the Tribunal’s Questions on Quantum, 14 October 2022, para. 8. ↩
³⁹ Prof. Spiller’s Comments on Dr. Flores’ Answers to the Tribunal’s Questions on Quantum, 14 October 2022, para. 9. ↩

[Page 14]

Claimants thus conclude that “there can be no doubt that 7.398% pre-tax cannot apply for the pre-July 2013 period".⁴⁰

53. The Claimants and Prof. Spiller also disagree with Dr. Flores' and Spain's criticisms of the formula used by Prof. Spiller to calculate the Adjusted NAV. The Claimants and Prof. Spiller note that the absence of any documentary evidence supporting the 10% pre-tax reasonable rate of return is unsurprising given that the sources relied upon in the Decision refer to post-tax rates.⁴¹ The Claimants and Prof. Spiller further note that Spain itself has applied the same formula, for example, to calculate the 2020 update to the regulatory rate of return (7.09%).⁴²

ii. Spain's Responses to the Tribunal's Questions

54. Dr. Flores explains that in his view, the Tribunal did not consider the 7.398% rate from the New Regime to be part of the “claw-back”, as the Decision “stated that the ‘claw-back effect consists in discounting the past remuneration received by the PV installations that was in excess of the reasonable rate of return in the period since their commissioning until July 2013 against the periodically defined rate of return (i.e., 7.398%) to which the installations have become entitled under the New Regime'.”⁴³ Dr. Flores thus notes that the Tribunal did not indicate a rate different from 7.398% to be used prior to 13 July 2013, “let alone the much higher 10% rate of return proposed by Prof. Spiller.”⁴⁴

55. Dr. Flores further notes that he understands that the Tribunal agreed with the BayWa tribunal's analysis related to the retroactivity of the New Regime and should similarly

⁴⁰ Claimants’ Comments on Spain’s Responses to the Tribunal’s Questions on Quantum, 14 October 2022, para. 12. ↩
⁴¹ Claimants’ Comments on Spain’s Responses to the Tribunal’s Questions on Quantum, 14 October 2022, para. 15; Prof. Spiller’s Comments on Dr. Flores’ Answers to the Tribunal’s Questions on Quantum, 14 October 2022, paras. 11-12. ↩
⁴² Claimants’ Comments on Spain’s Responses to the Tribunal’s Questions on Quantum, 14 October 2022, para. 15; Prof. Spiller’s Comments on Dr. Flores’ Answers to the Tribunal’s Questions on Quantum, 14 October 2022, para. 13. ↩
⁴³ Dr. Flores’ Answers to the Tribunal’s Questions on Quantum, 30 September 2022, para. 4. ↩
⁴⁴ Dr. Flores’ Comments on Prof. Spiller’s Answers to the Tribunal’s Questions on Quantum, 14 October 2022, para. 3. ↩

[Page 15]

apply the 7.398% rate to calculate the Adjusted NAV for the purposes of calculating remuneration for Retroactivity.⁴⁵

56. As regards Prof. Spiller's formula for calculating the Adjusted NAV by reference to the 7% reasonable return post-tax, Dr. Flores observes not having “seen any evidence in the documentary record of this Arbitration that would indicate that Claimants expected a reasonable rate of return of 10% pre-tax when they made their investment.”⁴⁶ Moreover, Dr. Flores observes that renewable energy projects in Spain “benefit from numerous tax deductions” and that “the difference between a pre-tax and post-tax rate of return is small.”⁴⁷

2. The Tribunal's Analysis

57. In its Decision, the Tribunal found that “the Respondent breached Article 10(1) of the ECT to the extent that it applied the New Regime retroactively to the remuneration already received by the Claimants' PV Plants under RD 661/2007 and RD 1578/2008".⁴⁸

58. As also noted in the Decision, “[u]nder RD 661/2007 and RD 1578/2008, there was an unlimited possibility to earn returns that could have been in excess of the reasonable rate of return (whether the rate of 7% guaranteed under Law 54/1997 or the 7.398% defined under the New Regime)."⁴⁹

59. To recall, the Claimants have presented the retroactivity element of the New Regime as "[t]he most harmful aspect of Law 24/2013” and in the following terms during the previous phase of these proceedings:

The most harmful aspect of Law 24/2013, however, was that it established that the newly determined ‘reasonable return' set out by RDL 9/2013 was to be calculated 'throughout the regulatory life of the plant'. In practice, this means that whatever payments an installation received in the past in excess of what the Government now considers to be reasonable will have to be discounted from future payments. That has practically the same effect as if the

⁴⁵ Dr. Flores’ Answers to the Tribunal’s Questions on Quantum, 30 September 2022, para. 5. ↩
⁴⁶ Dr. Flores’ Answers to the Tribunal’s Questions on Quantum, 30 September 2022, para. 8. ↩
⁴⁷ Dr. Flores’ Comments on Prof. Spiller’s Answers to the Tribunal’s Questions on Quantum, 14 October 2022, para. 4. ↩
⁴⁸ Decision, paras. 907, 1080(ii). ↩
⁴⁹ Decision, para. 904. ↩

[Page 16]

amounts previously received had to be returned. This aspect of RDL 9/2013 results in ‘maximum degree' retroactivity as it essentially claws back plants' prior earnings. As a result, returns received by the plants many years before the implementation of the New Regime are used to calculate the returns to which they are entitled to now.⁵⁰

60. The Tribunal considered that the operation of the New Regime meant that “future entitlements to subsidies are determined – and decreased – by reference to subsidies that exceeded the statutorily defined rate of return in the past, but that were nevertheless lawfully granted and paid."⁵¹ It agreed with the tribunal in BayWa v. Spain that "[i]t may have been reasonable to take into account, in calculating subsidies going forward, the 7.398% that the Plants were deemed to be entitled to under the Disputed Measures[but] [t]o count against them the amounts previously earned in excess of that threshold was to penalize the Plants for their successful operation during those years."⁵²

61. The Tribunal thus found that “as a matter of principle, this ‘claw back' operation of the New Regime amounts to a violation of Article 10(1) of the ECT,” whilst noting also that "no specific quantification in this respect has been submitted by the Claimants."⁵³

62. In accordance with the Tribunal's directions on quantum, the Parties' Experts have reached an agreement in relation to the methodology to be used in order to calculate the compensation for retroactivity.⁵⁴ Specifically, the Parties' Experts have agreed on the following assumptions:

a) [...] as the Claimants' PV Plants all have a different IT code under the New Regime, five in total, it is necessary to calculate this compensation separately for each of the five plants.
b) To use the projected cash flows previously modelled in these proceedings under the actual scenario, with exception of the actual revision of the regulatory rate of return, which is 7.09%.

⁵⁰ Memorial, para. 223 (footnotes omitted). See also Decision, para. 903. ↩
⁵¹ Decision, para. 905. ↩
⁵² Decision, para. 906, citing BayWa r.e. Renewable Energy GmbH and BayWa r.e. Asset Holding GmbH v. Kingdom of Spain, ICSID Case No. ARB/15/16, 2 December 2019 (RL-125), para. 496. ↩
⁵³ Decision, para. 907. ↩
⁵⁴ See paras. 34-35 above. ↩

[Page 17]

c) To use the standard initial investment cost as the initial net asset value [the Initial NAV], which is consistent with Spain's approach under the New Regime, corresponding to the year when each plant was constructed.
d) To calculate a NAV as of July 13, 2013, which is the date that the NAV was calculated for the New Regime.⁵⁵

63. The Experts have equally agreed to use the following methodology:

a) The first step is to calculate an NAV as of July 13, 2013, given the initial NAV and the pre-tax reasonable rate of return for the years of operation prior to the New Regime, ignoring the actual cash flows achieved during the same period [the Adjusted NAV].
b) The second step is to use the Adjusted NAV to determine the adjusted investment remuneration parameter [the Adjusted Rinv] from July 13, 2013.
c) The third step is to implement the adjusted Rinv computed in the prior step in the forecasted operations of each plant, to obtain the resulting cash flows generated by each plant.
d) The fourth step is to compute the difference between the cash flows of each plant using the Adjusted Rinv and the actual Rinv.
e) Finally, discount the difference calculated in the prior step using a post-tax WACC of 6.03% to June 20, 2014, to obtain the compensation value.⁵⁶

64. As recapitulated at paragraph 36 above, according to the methodology agreed by the Experts, the Adjusted NAV needs to be calculated in view of the initial NAV and “the pre-tax reasonable rate of return for the years of operation prior to the New Regime.”⁵⁷ The Experts however disagreed on the calculation of such a pre-tax reasonable rate of return prior to 2013.

65. This disagreement results in a difference of EUR 804,022 in the calculation of compensation for retroactivity.

66. As summarized at paragraphs 38-39 above, Prof. Spiller converts a 7% post-tax rate into a 10% pre-tax rate of return in the period between the PV Plants' commissioning dates and 13 July 2013, whereas Dr. Flores uses the 7.398% pre-tax rate of return established by the New Regime regulations throughout the same historical period.

⁵⁵ Joint Expert Report, para. 4 (footnotes omitted, emphasis added). ↩
⁵⁶ Joint Expert Report, para. 5. Footnotes omitted. ↩
⁵⁷ Joint Expert Report, para. 5. ↩

[Page 18]

67. Before addressing the Experts' disagreement regarding the calculation of the Adjusted NAV for the purposes of calculating retroactivity compensation, the Tribunal deems it important to clarify from the outset the interaction between its approach regarding the assessment of potential harm (if any) caused by the claw back mechanism, on the one hand, and of any harm caused by a potential failure to provide the 7% reasonable rate of return that the Tribunal has found the Claimants to be entitled to, on the other hand. Indeed, as a matter of principle, the Tribunal considers that these assessments of harm are analytically distinct, whilst being mindful that the analysis of their quantification may overlap. The Tribunal also notes that these differences and potential overlap between these assessments has not been specifically discussed by the Parties during the previous stage of these proceedings, as the impact of retroactivity was not specifically quantified. That being noted, at the quantum stage, the Parties recognize that the quantification of the Claimants' damages will have to take into account the fact that the losses flowing from these respective treaty breaches may, at least in substantial part, overlap. As recognized by the Claimants, compensation for retroactivity, if awarded, will increase the PV Plants' IRRs,⁵⁸ as the additional cash flows they would require to reach a 7% post-tax return would be lower.⁵⁹

68. It is thus clear that the compensation for retroactivity will ultimately have an effect on the assessment of the amount of compensation required to make the Claimants whole. It equally follows from the above that the lower the amount of compensation for retroactivity, the likelier it would become that the 7% reasonable rate of return might not be attained.

69. This can be illustrated by reference to Tables 5 and 6 provided in the Joint Expert Report that set out the compensation to the PV Plants to achieve a 7% post-tax reasonable return after accounting for compensation for retroactivity under different sets of inputs. For example, on the Claimants' case (Table 5), if their claimed amount on retroactivity is retained, the resulting compensation for IRR after retroactivity would be EUR 1,975,031, whereas if the Respondent's retroactivity damages are

⁵⁸ Claimants’ Quantum Submissions, para. 69; Joint Expert Report, para. 48. ↩
⁵⁹ Claimants’ Quantum Submissions, para. 69; Joint Expert Report, para. 51. ↩

[Page 19]

taken as a starting point, the resulting compensation for IRR would be higher as it would amount to EUR 2,664,701 (under the same sets of inputs).⁶⁰

70. This finding, in the Tribunal's view, addresses the Claimants' criticism that Dr. Flores' approach to the calculation of retroactivity damages does not eliminate such retroactivity in full, as it does not account for the Tribunal's findings regarding the legitimately expected reasonable return of 7% post-tax.⁶¹ Indeed, the Tribunal considers that the difference in approach between the Parties' quantum Experts is one of sequence of analysis.

71. In this regard, the Tribunal further notes that the Respondent's approach to the calculation of the Adjusted NAV is consistent with the wording of the Experts' agreed methodology (which refers, as a first step, to calculating the Adjusted NAV on the basis of the initial NAV and “the pre-tax reasonable rate of return for the years of operation prior to the New Regime”), the reasoning adopted in the Decision and the characterization of the operation of the “claw-back” that the Claimants presented in the previous phase of the arbitration.⁶² Indeed, as rightly noted by Dr. Flores, the Tribunal did not specifically condemn the 7.398% rate of return when discussing the retroactive operation of the New Regime.⁶³ The finding of breach in relation to the retroactive operation of the New Regime was made before and independently from the discussion of the Claimants' entitlement to a reasonable return and the level of such a return.⁶⁴

72. The Tribunal therefore concludes that the Adjusted NAV should be calculated in accordance with Dr. Flores' approach.

73. Based on this approach, the harm caused by the retroactive operation of the New Regime is quantified at EUR 4,092,193.

⁶⁰ Joint Expert Report, p. 16, Table 5, rows 1 and 9. ↩
⁶¹ Claimants’ Response to the Tribunal’s Questions on Quantum, 30 September 2022, para. 9. ↩
⁶² See paras. 59-61 above. ↩
⁶³ Dr. Flores’ Answers to the Tribunal’s Questions on Quantum, 30 September 2022, paras. 4-5. ↩
⁶⁴ Decision, paras. 903-905. ↩

[Page 20]

B. The Claimants' IRR

1. The Parties' Positions

a. The Joint Expert Report

74. The Experts agree on the methodology of the IRR calculation and disagree on three parameters.⁶⁵ Specifically, the Experts agree:

a) That the "operational” IRR is to be computed based on the free cash flows to the firm [(the “FCFF”)], from inception through the end of their regulatory lives.
b) To use a date of valuation, and thus information available until, June 20, 2014 (with exception of the actual revision of the regulatory rate of return).
c) To use the revised regulatory rate of return at 7.09% pre-tax rate from 2020 to the end of the Claimants' PV Plants' lives.⁶⁶

75. The Experts however disagree on the following three parameters:

a) The approach to calculating the tax effect of debt;
b) The application of the Decision's instruction regarding operating and maintenance costs;
c) Whether to calculate an IRR separately for each of the PV Plants, or a single IRR for all of the Claimants' plants.⁶⁷

76. The results of the IRR calculation for the Claimants' PV Plants according to each Experts' parameters are as follows:⁶⁸

⁶⁵ Joint Expert Report, para. 2. ↩
⁶⁶ Joint Expert Report, para. 12 (footnotes omitted, emphasis added). ↩
⁶⁷ Joint Expert Report, para. 13. ↩
⁶⁸ Joint Expert Report, para. 14, Table 2. ↩

[Page 21]

77. The Experts' respective positions regarding the three parameters in dispute that yield the difference between 6.29% and 6.95% result in a discrepancy of the amount that must be compensated to the PV Plants to achieve a 7% post-tax IRR:⁶⁹

78. The Experts' respective positions on the three areas of disagreement – tax effect of debt, operating and maintenance (“O&M”) costs, and whether to calculate the IRR on a plant-by-plant basis– are set out below.

i. The Experts' positions on the tax effect of debt

(a) Compass Lexecon

79. Prof. Spiller incorporates the tax effect of debt through the discount rate, the weighted average cost of capital ("WACC").⁷⁰ According to Prof. Spiller, this approach is consistent with his previous reports as well as the financial literature referred to by Dr. Flores, which indicates that adjusting the discount rate “is the most common approach, which is usually implemented via the after-tax weighted-average cost of capital (WACC).”⁷¹

80. This allegedly allows Prof. Spiller to avoid double counting when calculating the FCFF on the basis of an unlevered tax expense (i.e., without the tax effect of debt) –

⁶⁹ Joint Expert Report, para. 45, Table 4. ↩
⁷⁰ Joint Expert Report, para. 17. ↩
⁷¹ Joint Expert Report, para. 17, referring to Richard A. Brealey, Stewart C. Myers, and Franklin Allen, Principles of Corporate Finance, 11th ed, p. 479 (EO-21). ↩

[Page 22]

in contrast with Dr. Flores.⁷² In support of his position, Prof. Spiller also refers to financial literature explaining that the FCFF is computed before the effect of debt.⁷³

81. Furthermore, according to Prof. Spiller, “in order for the resulting IRR to be comparable to the reasonable return, the two measures of return should be computed on a comparable basis.”⁷⁴ In Prof. Spiller's view, the Tribunal's definition of the reasonable return is in a way consistent with Spain's own description, which is: “a return for standard projects amounting to 7% with own resources, before financing and after tax."⁷⁵ Prof. Spiller further explains:

The reference in such definition to "before financing" means before (or absent) debt, and thus absent the tax effect of debt in the cash flows used to compute such return. Dr. Flores claims that such reference to "before financing" does not exclude shareholder loans. Such claim, however, is unsupported. Spain's description of the reasonable rate of return makes no differences between financing with third-party or with shareholders' loans. In fact, if Spain considered that the effect of shareholder loans needs to be accounted for, Spain would have needed to determine corresponding assumptions for standard plants (e.g., principal amounts of shareholder loans, interest rates on shareholder loans). Spain, however, has never taken into account shareholder loans in its determination of the reasonable rate of return or any of the remuneration parameters, either during the FiT regime or in the New Regime.⁷⁶

(b) Quadrant Economics

82. Dr. Flores asserts that the tax effect (or tax benefit) of debt shall be accounted for in the calculation of the PV Plants' cash flows, as interest payments on debt are tax deductible.⁷⁷ By not incorporating the tax benefits of debt, Prof. Spiller, in Dr. Flores' view, understates the actual IRR of the PV Plants.⁷⁸

83. Dr. Flores disagrees that Spain's description of the reasonable return as “a return for standard projects amounting to 7% with own resources, before financing and after

⁷² Joint Expert Report, para. 18. ↩
⁷³ Joint Expert Report, para. 19, referring to Damodaran, A. 2003. Corporate Finance. Theory and Practice, p. 271 (CLEX-22), which calculates the cash flows to the firm based on "EBIT" that stands for “earnings before interest and taxes", with "before interest" meaning absent the impact of debt interest. Such textbook also defines it as "The cash flow to the firms is a pre-debt, after-tax cash flow...". ↩
⁷⁴ Joint Expert Report, para. 20. ↩
⁷⁵ Joint Expert Report, para. 20. See Decision, paras. 168, 197, 985-986. ↩
⁷⁶ Joint Expert Report, para. 20 (footnotes omitted). ↩
⁷⁷ Joint Expert Report, para. 21. ↩
⁷⁸ Joint Expert Report, para. 25. ↩

[Page 23]

tax" implies that the PV Plants would not benefit from the tax effect of debt.⁷⁹ According to Dr. Flores, “before financing" refers to any external debt only and does not exclude, for example, shareholder loans that generate tax-deductible interest.⁸⁰ Dr. Flores refers to Prof. Spiller's own calculations allegedly showing that even after the Claimants' PV Plants had no external debt, they still continued to enjoy the tax benefit of debt because the external debt was recategorized as shareholder loans the interest on which was still available to reduce taxes.⁸¹

ii. The Experts' positions on the operating and maintenance costs

(a) Compass Lexecon

84. Prof. Spiller uses the actual O&M costs incurred by the Claimants' PV Plants, including the re-negotiated O&M contracts from the moment these were renegotiated until the end of the Claimants' PV Plants' lives.⁸² In Prof. Spiller's view, it is consistent with the Tribunal's conclusions that the Experts' task should be “to assess the actual economic impact of the Disputed Measures on the Claimants' PV Plants.”⁸³

85. Dr. Flores' approach, which retroactively reduces the Claimants' PV Plants' operating and maintenance costs by applying the re-negotiated prices from the start of the Claimants' PV Plants' operation (i.e., even before the operating and maintenance agreements were renegotiated) is unacceptable, according to Prof. Spiller, as it is inconsistent with the Tribunal's instruction to calculate the actual impact by using the actual O&M costs.⁸⁴

86. Finally, Prof. Spiller notes that the Claimants' O&M contracts were not inefficient before the re-negotiation: four of the Claimants' PV Plants had operating costs in line with that of the standard facility.⁸⁵

⁷⁹ Joint Expert Report, para. 22. ↩
⁸⁰ Joint Expert Report, paras. 22-23. ↩
⁸¹ Joint Expert Report, para. 24. ↩
⁸² Joint Expert Report, para. 26. ↩
⁸³ Joint Expert Report, para. 26. See Decision, para. 1016. ↩
⁸⁴ Joint Expert Report, para. 27. ↩
⁸⁵ Joint Expert Report, para. 30. ↩

[Page 24]

(b) Quadrant Economics

87. Dr. Flores notes that the Claimants O&M contracts were up to 66% higher than the renegotiated contracts, meaning that the PV Plants were inefficient.⁸⁶ Therefore, Dr. Flores calculated the IRR assuming the renegotiated O&M prices from the moment when the PV Plants began operating. Dr. Flores explains:

Logically, if a PV plant failed to negotiate market-rate O&M contracts and its O&M costs were much higher than normal, then its profitability would be lower than normal and, therefore, its IRR would be lower than reasonable. However, that failure to achieve a reasonable rate of return cannot be attributed to the New Regime, but to the PV plant's own inefficiency.⁸⁷

iii. Whether to calculate the IRR on a plant-by-plant basis

(a) Compass Lexecon

88. Prof. Spiller calculates the compensation on an individual plant basis which is consistent with his previously submitted damages calculations and with the approach the Experts agree to use for the compensation for retroactivity.⁸⁸ Prof. Spiller interprets the Decision as not precluding the compensation (or the IRR) to be computed separately for each PV Plant.⁸⁹ Prof. Spiller criticizes Dr. Flores' combined calculation:

[...] a computation of a single IRR, and a subsequent compensation, for the aggregate of the PV Plants introduces inconsistencies with Spain's regulatory regime. For instance, the PV Plants have different regulatory life end dates, some end in 2037 and others in 2040. A subsequent compensation calculation based on a single IRR extending until the end of the regulatory life of the aggregate of PV Plants (i.e., 2040) would compensate some plants beyond their regulatory useful life. Calculating an IRR and subsequent computation of compensation for each individual plant based on its own regulatory useful life is the closest comparison to Spain's New Regime.⁹⁰

89. Moreover, computation of an individual IRR was, according to Prof. Spiller, supported by other arbitral tribunals (e.g., in Hydro v. Spain).⁹¹

⁸⁶ Joint Expert Report, para. 32. ↩
⁸⁷ Joint Expert Report, para. 35. ↩
⁸⁸ Joint Expert Report, para. 37. ↩
⁸⁹ Joint Expert Report, para. 38. ↩
⁹⁰ Joint Expert Report, para. 39. ↩
⁹¹ Joint Expert Report, para. 40, referring to Hydro Energy 1 S.à.r.l. and Hydroxana Sweden AB v. Kingdom of Spain, ICSID Case No. ARB/15/42, Decision on jurisdiction, liability and directions on quantum, 9 March 2020 (CL-196), paras. 717-726. ↩

[Page 25]

(b) Quadrant Economics

90. Dr. Flores argues that, from an economic perspective, Compass Lexecon's calculation is incorrect as it does not credit the surpluses from some plants against the shortfalls of others.⁹² Since the Claimants' PV Plants were owned by the same entities, it is reasonable to calculate a combined IRR of the PV Plants, according to Dr. Flores.⁹³

b. The Claimants' Quantum Submissions

91. The Claimants provide the following comments in support of Prof. Spiller's approach to the three parameters in dispute (the tax effect of debt, the historical O&M costs and whether to calculate the IRR on a plant-by-plant basis).

i. The tax effect of debt

92. In addition to the arguments set out by Prof. Spiller (see paragraphs 79-81 above), the Claimants assert that the PV Plants' IRRs indeed should not include the tax benefit of debt, as the benchmark reasonable return of 7% does not include such a benefit due to the following:

• [...] none of the "laws and regulations as well as other contemporaneous documents that were publicly available during the Claimants' investment process", which the Tribunal relied on to determine that the reasonable return benchmark is 7% post-tax, provide that the tax benefit of debt should be accounted for.⁹⁴
• On the contrary, these documents make clear that the 7% benchmark is assessed on an all-equity basis with no regard to project financing [...] "Profitability of the type project: calculated on the basis of maintaining an Internal Rate of Return (IRR) [...] a minimum of 7%, with own resources, before financing and after taxes.”⁹⁵

93. The Claimants further refer to the following documents containing similar statements:

• Spain Renewable Energy Plan 2005-2010;⁹⁶

⁹² Joint Expert Report, para. 43. ↩
⁹³ Joint Expert Report, para. 43. ↩
⁹⁴ Claimants’ Quantum Submissions, para. 36. ↩
⁹⁵ Claimants’ Quantum Submissions, para. 36, referring to 2000-2010 Plan for the Promotion of Renewable Energies in Spain, 19 December 1999 (R-67), p. 182 (emphasis added). ↩
⁹⁶ Spain Renewable Energy Plan 2005-2010, 1 August 2005 (R-69)/(C-32), p. 274. ↩

[Page 26]

• Memoria Económica for RD 661/2007;⁹⁷
• ASIF Report: “Hacia una electricidad respetuosa con el medio ambiente” [“Towards environmentally friendly electricity"];⁹⁸
• Universidad Pontificia de Comillas Report, “Solar energy: current status and immediate perspective."⁹⁹

94. According to the Claimants, Dr. Flores recognized “the 7% benchmark return to be an all-equity return that did not account for any debt (or leverage) and its associated tax benefit."¹⁰⁰ The Claimants also refer to the following statement made by Dr. Flores at the Hearing:

So all these documents from "El Sol Puede Ser Suyo", "The Sun Can Be Yours", what they seem to be saying is they're geared towards the investor. They say, "This is the investment that you could realise, depending on how much you leverage yourself". If you get a lot of bank financing, you're going to get higher IRRs; lower bank financing, less. But this has nothing to do with the 7%. The 7% is a project IRR. Very important distinction.¹⁰¹

95. The Claimants explain the rationale behind this approach: “Spain could not possibly have predicted what specific financing arrangements each of the many thousands of individual PV plants in Spain made and, consequently, the extent of the tax shield they could benefit from.”¹⁰²

96. The Claimants further submit that the Respondent itself recognized this in the course of these proceedings¹⁰³ and the Respondent's fact witness, Mr. Montoya, explained the same in RREEF v. Spain:

30. The decision to obtain a loan or credit to undertake the investment is a financial decision based merely on the interests of the promoter. The credit conditions depend therefore on both the decisions taken by the promoter

⁹⁷ Memoria Económica for RD 661/2007 (C-163), p. 14. ↩
⁹⁸ ASIF Report: “Hacia una electricidad respetuosa con el medio ambiente” [“Towards environmentally friendly electricity"], October 2005 (R-246), p. 9. ↩
⁹⁹ Universidad Pontificia de Comillas Report: “Solar energy: current status and immediate perspective”, 2007 (R-272), p. 197. ↩
¹⁰⁰ Claimants’ Quantum Submissions, para. 39, referring to First Econ One Report, para. 302. ↩
¹⁰¹ Hearing Tr., Day 4, 21:24-25 to 22: 1-7; 26: 16-21. ↩
¹⁰² Claimants’ Quantum Submissions, para. 40. ↩
¹⁰³ Claimants’ Quantum Submissions, para. 41, referring to Counter-Memorial, para. 830. ↩

[Page 27]

regarding his own funds and of the agreements reached with other partners or with banking entities.
31. These aspects are unrelated to regulation in renewable energies and it is not possible to incorporate them into the parameters of remuneration, since there are infinite possibilities in the financing of plants that could not be reflected in the parameters of remuneration: the percentages of financed capital are different in each plant, since each partner may have different financing conditions, and such conditions may vary over time or may be altered by a change in the distribution of partners in a particular plant or sale, etc.¹⁰⁴

97. The RREEF v. Spain tribunal thus accepted that no financing should be accounted for in the benchmark return, stating that “[i]t was not the objective of the regulator to take into account the particularities of each installation and of the various modes of financing."¹⁰⁵

98. The Claimants further argue that Dr. Flores' attempt to make a distinction between external and internal (e.g. shareholder loans) financing is misleading and, in any event, in his calculation Dr. Flores accounts for both types of financing, whereas he should have excluded the external one based on his own approach.¹⁰⁶

99. The Claimants emphasize that Dr. Flores' approach is inconsistent with Spain's objective to disregard the individual financial situation of each facility:

By definition, each investor's situation is unique in relation to shareholder loans: some have recourse to such loans, and others do not. For those who do have recourse to such financing, some may rely on it extensively, thereby benefitting from a significant tax shield, and others may use significantly less. Taking into account shareholder loans would therefore introduce a differential treatment between investors that does not exist in the implicit target return set by Spain.¹⁰⁷

100. The Claimants also consider that the alleged distinction between the facilities with external and those with internal financing is plainly unfair.¹⁰⁸

¹⁰⁴ RREEF Infrastructure (G.P.) Limited and RREEF Pan-European Infrastructure Two Lux S.à r.l. v. Kingdom of Spain, ICSID Case No. ARB/13/30, Decision on Responsibility and on the Principles of Quantum, 30 November 2018 (CL-180), para. 529 (emphasis added). ↩
¹⁰⁵ Claimants’ Quantum Submissions, para. 41, referring to RREEF Infrastructure (G.P.) Limited and RREEF Pan-European Infrastructure Two Lux S.à r.l. v. Kingdom of Spain, ICSID Case No. ARB/13/30, Decision on Responsibility and on the Principles of Quantum, 30 November 2018 (CL-180), para. 546. ↩
¹⁰⁶ Claimants’ Quantum Submissions, para. 44. ↩
¹⁰⁷ Claimants’ Quantum Submissions, para. 46. ↩
¹⁰⁸ Claimants’ Quantum Submissions, para. 47. ↩

[Page 28]

ii. Operating and maintenance costs

101. The Claimants submit that the Tribunal's instruction was to use the "actual" O&M costs, which means that Prof. Spiller's approach to use the re-negotiated O&M costs only from the moment when the relevant contracts were renegotiated is the correct one.¹⁰⁹

102. Dr. Flores' approach, according to the Claimants, is not only inconsistent with the Decision, but it is also unsupported and unfair.¹¹⁰ The Claimants argue that RD 661/2007 and RD 1578/2008 were designed to encourage RE installations to produce as much electricity as possible.¹¹¹ For this reason, under the original O&M contracts the contractor's remuneration was based on a percentage of the revenues generated by the plants in order to incentivize the contractor to maximize production, in line with the then-applicable regulations.¹¹²

103. As it was explained by Mr. Bouman during the Hearing:

A: Yes, the idea is correct. But sometimes, in order to produce more, you need to make some maintenance operations. Maintenance can be higher, and the cost is not always compensated by the revenue rise. To give you an example, for instance, let's say the panel cleaning. The solar panels, of course, the cleaner they are, the better the production. So initially we had an agreement for cleaning twice a year. And then we observed that with the regulatory change, we could not recoup the price of that cleaning twice a year, and once a year, it could be enough. So that's one example. If production has primacy, well, we clean twice a year and we have more production. But if you don't have such an incentive for production, then you cannot optimise production in the first place.¹¹³

104. Therefore, it would be unfair, in the Claimants' view, to increase the IRR by refusing to apply the actual O&M costs for the period up to the renegotiation.¹¹⁴

¹⁰⁹ Claimants’ Quantum Submissions, paras. 49-50. ↩
¹¹⁰ Claimants’ Quantum Submissions, paras. 51-54. ↩
¹¹¹ Claimants’ Quantum Submissions, para. 54. See Decision, para. 904. ↩
¹¹² Claimants’ Quantum Submissions, para. 54. ↩
¹¹³ Hearing Tr., Day 2, 204:11-25. ↩
¹¹⁴ Claimants’ Quantum Submissions, para. 56. ↩

[Page 29]

iii. Whether to calculate the IRR on a plant-by-plant basis

105. Similarly to Prof. Spiller, the Claimants note that the Tribunal has not yet decided “as to whether the Parties should compute an IRR for each plant individually or calculate an average IRR across all of the PV Plants.”¹¹⁵

106. The Claimants argue that Dr. Flores' approach is circular and flawed for the following reasons:

• First, the Claimants recall that the 7% post-tax return is a project return (not a shareholder return).¹¹⁶ Therefore, in order to perform a like-for-like comparison, an IRR must also be calculated on a project-by-project (plant-by-plant) basis.¹¹⁷
• Second, according to the Claimants, Dr. Flores' approach is unfair.¹¹⁸ In support of their position, the Claimants refer¹¹⁹ to the reasoning of the Hydro v. Spain tribunal, which rejected Dr. Flores' approach on the following grounds:

  The clear rationale behind the legislation is to consider what constitutes a reasonable rate of return on a plant by plant basis and the Tribunal finds there is no justification for an aggregation approach. The fact that a plant happens to be in a portfolio owned by a particular investor should not impact whether that plant is considered to make a reasonable rate of return. Each plant should be considered on its own, as would have to be the case if each were individually owned by different investors.¹²⁰

• Third, the Claimants note that Dr. Flores' approach is devoid of any economic logic:

  [...] the regulatory life of some of the PV Plants will end in 2037, while others will end in 2040. If the Tribunal were to award compensation based on Dr Flores' aggregated IRR (i.e., by assuming a regulatory lifetime until 2040), it would be awarding those PV Plants whose regulatory life ends in 2037 the cash flows for the years 2037 to 2040 that they would never have received. Conversely, should the Tribunal assume a regulatory lifetime until 2037 across all the PV Plants, those plants whose regulatory lifetime ends in 2040 would be deprived of the cash flows for the years 2037 to 2040, which they would receive in real

¹¹⁵ Claimants’ Quantum Submissions, para. 57. ↩
¹¹⁶ Claimants’ Quantum Submissions, para. 59. ↩
¹¹⁷ Claimants’ Quantum Submissions, para. 59. ↩
¹¹⁸ Claimants’ Quantum Submissions, para. 60. ↩
¹¹⁹ Claimants’ Quantum Submissions, para. 61. ↩
¹²⁰ Hydro Energy 1 S.à.r.l. and Hydroxana Sweden AB v. Kingdom of Spain, ICSID Case No. ARB/15/42, Decision on jurisdiction, liability and directions on quantum, 9 March 2020 (CL-196), para. 724. ↩

[Page 30]

life. This would not be consistent with Spain's New Regime, as Professor Spiller explains.¹²¹

c. The Respondent's Quantum Submissions

107. The Respondent addresses the three areas of disagreement between the Experts (the tax effect of debt, the historical O&M costs and whether to calculate the IRR on a plant-by-plant basis).

i. The tax effect of debt

108. The Respondent supports Dr. Flores' approach accounting for the tax effect of debt when calculating the PV Plants' cash flows for the following reasons:

• First, Dr. Flores' approach avoids overcompensating the Claimants' PV Plants for taxes "they have not paid in real life".¹²² In support of its position, the Respondent refers to the tribunal's reasoning in PV Investors v. Spain that damages in that arbitration “must be computed assuming that the regulator took account of the tax shield and implementing [Quadrant Economics'] estimate of the effective tax rate."¹²³ The PV Investors v. Spain tribunal explained its conclusion as follows:

  Another consideration also plays in favour of taking account of the tax shield. As a matter of law, compensating the Claimants for taxes they have not paid would be contrary to the principle that reparation cannot exceed the harm effectively suffered. In other words, one cannot do better in litigation than in real life.¹²⁴

• Second, according to the Respondent, Prof. Spiller incorrectly claims that it incorporates the tax effect of debt through the discount rate.¹²⁵ This, in the Respondent's view, is inconsistent with the Tribunal's instruction to calculate an "operational" (not an “exit”) IRR.¹²⁶ Prof. Spiller allegedly accepts that the “[o]perational IRR, however, depends only on the cash flows generated by a

¹²¹ Claimants’ Quantum Submissions, para. 62 (footnotes omitted), referring to Joint Expert Report, para. 39. ↩
¹²² Respondent’s Quantum Submissions, para. 16. ↩
¹²³ Respondent’s Quantum Submissions, para. 17, referring to AES Solar and others (PV Investors) v. Spain, PCA Case No. 2012-14, Final Award, 28 February 2020 (RL-129), para. 798. ↩
¹²⁴ AES Solar and others (PV Investors) v. Spain, PCA Case No. 2012-14, Final Award, 28 February 2020 (RL-129), para. 792. ↩
¹²⁵ Respondent’s Quantum Submissions, para. 18. ↩
¹²⁶ Respondent’s Quantum Submissions, para. 18. ↩

[Page 31]

project, and thus it is not subject to the discount rate at the time of exit.”¹²⁷ Thus, when Prof. Spiller claims to have incorporated the tax effect of debt through the discount rate, he, in fact, does the following:

[Prof. Spiller] has applied a discount rate to the additional cash flows required to achieve a 7% post-tax IRR, but it has not applied them to the FCFF used to calculate the operational IRR. For example, as indicated in row 1 of Table 5 of the Joint [Expert] Report, under Compass Lexecon's approach, the PV Plants would require additional cash flows to achieve a 7% post-tax IRR, which, discounted with a discount rate, result in a present value of [EUR] 1,975,031. In other words, as Quadrant Economics notes, Compass Lexecon applies a discount rate to a different stream of cash flows than the FCFF used to compute the operational IRR indicated in Table 2 of the Joint Report.¹²⁸

Third, the Respondent reiterates Dr. Flores' criticism that by excluding the tax effect of debt, Prof. Spiller's IRR calculations understate the actual IRR of the PV Plants.¹²⁹

ii. Operating and Maintenance costs

109. The Respondent argues that Dr. Flores' approach applying the re-negotiated O&M prices to the whole historical period is the correct one, as “a failure to achieve a reasonable rate of return should not be attributed to the PV Plants' own inefficiency."¹³⁰ The Respondent then reiterates the arguments advanced by Dr. Flores in the Joint Expert Report.¹³¹

110. The Respondent further argues that the use of the re-negotiated O&M prices throughout the whole historical period is also consistent with the Tribunal's decision to use standard installation costs as the Claimants' actual initial investment costs seemed to have been above market prices.¹³²

---

¹²⁷ Respondent’s Quantum Submissions, para. 18, referring to Second Compass Lexecon Report, para. 136.a. ↩
¹²⁸ Respondent’s Quantum Submissions, para. 20 (footnotes omitted). ↩
¹²⁹ Respondent’s Quantum Submissions, para. 21. ↩
¹³⁰ Respondent’s Quantum Submissions, para. 25. ↩
¹³¹ Respondent’s Quantum Submissions, paras. 26-27. ↩
¹³² Respondent’s Quantum Submissions, paras. 28-30. ↩

[Page 32]

iii. Whether to calculate the IRR on a plant-by-plant basis

111. The Respondent insists that a single IRR must be calculated for all of the Claimants' PV Plants.¹³³ As regards the Claimants' reliance on Hydro v. Spain, the Respondent notes that the Hydro approach is not unanimous.¹³⁴ For example, in PV Investors v. Spain the tribunal decided that the IRR of the relevant plants was to be calculated based on groups of plants owned by the same group of investors, “rather than for each plant."¹³⁵

d. The Parties' Responses to the Tribunal's Questions

112. The Tribunal presented the Parties with the following questions concerning the IRR calculation:

[...] the Respondent is invited to address the Claimants' criticism that "Dr Flores [...] accounts for both external financing and internal financing in its IRR calculation. Should he (incorrectly) consider that only internal financing should be included in his IRR calculation, he should have removed external financing. He did not do so." (see Claimants' Quantum Submissions, para. 44);
the Claimants are invited to address the Respondent's contention that incorporating the tax effect of debt through the discount rate, as proposed by Prof. Spiller, is incorrect as an operational IRR (in contrast with an “exit IRR") and shall not be discounted [Respondent's Quantum Submissions, para. 18].¹³⁶

i. The Claimants' Responses to the Tribunal's Questions

113. The Claimants note that Dr. Flores' approach, which includes the tax effect of debt into the calculation of the operational IRR, is inconsistent "with his purported agreement that the IRR shall be computed by using the FCFF metric”,¹³⁷ whereas Dr. Spiller calculates the operational IRR by measuring the FCFF accruing to the PV Plants as agreed by the Experts and “in line with standard practice, does not factor the tax benefit of debt into the FCFF computations but applies it, instead, through the

---

¹³³ Respondent’s Quantum Submissions, para. 35. ↩
¹³⁴ Respondent’s Quantum Submissions, para. 35. ↩
¹³⁵ Respondent’s Quantum Submissions, para. 35, referring to AES Solar and others (PV Investors) v. Spain, PCA Case No. 2012-14, Final Award, 28 February 2020 (RL-129), para. 724. ↩
¹³⁶ Tribunal’s letter to the Parties, dated 30 August 2022. ↩
¹³⁷ Claimants’ Response to the Tribunal’s Questions on Quantum, 30 September 2022, para. 19. ↩

[Page 33]

discount rate when discounting the extra FCFF needed for the PV Plants to achieve a 7% post-tax IRR.”¹³⁸

114. The Claimants also argue that Dr. Flores' explanations for his failure to distinguish between external and internal financing are unconvincing.¹³⁹ The Claimants and Prof. Spiller submit that Dr. Flores' approach results in double counting the tax benefit: in the cash flows and in the discount rate.¹⁴⁰

ii. Spain's Responses to the Tribunal's Questions

115. Dr. Flores observes that the Claimants' allegation that he does not remove external financing in his calculation is incorrect, as the tax shield that he includes “does not presume any external financing”.¹⁴¹ Dr. Flores further notes that although he agrees that "the reasonable rate of return was calculated at the project level and should be the same whether or not the project was debt financed", the Claimants erroneously assume that “no external debt" means "no tax shield."¹⁴²

116. Dr. Flores further describes as follows the Claimants' investment process in the PV Plants:

Claimants have used various capital structures to finance and optimize their investment in the PV Plants. Initially, Claimants financed the PV Plants with equity. Once the PV Plants were built, Claimants refinanced the PV Plants with bank loans, while also maintaining shareholder loans that also accrued tax-deductible interest. In 2017, Claimants replaced their bank debt with shareholder debt. Compass Lexecon continues to project tax benefits of debt after Claimants replaced their bank debt with shareholder debt. This means that even with 100% equity, Compass Lexecon still projects a tax shield. Claimants are thus incorrect to argue that the tax shield of debt is only applicable when external debt is used. Compass Lexecon indicates that Claimants were well aware of the tax benefits available to the PV Plants.¹⁴³

---

¹³⁸ Claimants' Response to the Tribunal's Questions on Quantum, 30 September 2022, para. 21. ↩
¹³⁹ Claimants' Comments on Spain's Responses to the Tribunal's Questions on Quantum, 14 October 2022, paras. 22-23. ↩
¹⁴⁰ Claimants' Comments on Spain's Responses to the Tribunal's Questions on Quantum, 14 October 2022, paras. 22-23. ↩
¹⁴¹ Dr. Flores' Answers to the Tribunal's Questions on Quantum, 30 September 2022, para. 12. ↩
¹⁴² Dr. Flores' Answers to the Tribunal's Questions on Quantum, 30 September 2022, para. 13. ↩
¹⁴³ Dr. Flores' Answers to the Tribunal's Questions on Quantum, 30 September 2022, para. 14 (footnotes omitted). ↩

[Page 34]

117. Dr. Flores thus concludes that “Claimants have used tax deductible debt, including shareholder loans, to optimize the profitability of the PV Plants” and that ignoring the tax benefit of debt would result in overcompensation.¹⁴⁴

118. As regards Prof. Spiller's IRR calculations, Dr. Flores notes that by not incorporating the tax shield, he “presents cash flows that are artificially lower than the actual ones, as a result of assuming that the PV Plants could not enjoy any tax benefit of debt and therefore would have to pay higher taxes than they actually paid”.¹⁴⁵ This allegedly results in “post-tax IRR below 7%, creating artificial damages and the artificial need to calculate additional cash flows".¹⁴⁶

2. The Tribunal's Analysis

119. In its Decision, the Tribunal has directed the Parties to use the following parameters for the calculation of the Claimants' IRR:

1. Use an “operational” IRR (as opposed to an “exit” IRR);
2. Assume initial investment costs totalling [EUR] 52.2 million;
3. Use the re-negotiated prices of the operating and maintenance contracts to calculate historical operating costs;
4. Endeavour to agree on a common approach to calculate the tax effect of debt;
5. To apply the actual revision of the regulatory rate of return.¹⁴⁷

120. The Experts have also agreed on a number of additional parameters for the IRR calculation. Specifically, the Experts agree:

1. That the "operational” IRR is to be computed based on the free cash flows to the firm ("FCFF”), from inception through the end of their regulatory lives.
2. To use a date of valuation, and thus information available until, June 20, 2014 (with exception of the actual revision of the regulatory rate of

---

¹⁴⁴ Dr. Flores' Answers to the Tribunal's Questions on Quantum, 30 September 2022, para. 15. ↩
¹⁴⁵ Dr. Flores' Comments on Prof. Spiller's Answers to the Tribunal's Questions on Quantum, 14 October 2022, para. 15. ↩
¹⁴⁶ Dr. Flores' Comments on Prof. Spiller's Answers to the Tribunal's Questions on Quantum, 14 October 2022, para. 15. ↩
¹⁴⁷ Decision, para. 1019. ↩

[Page 35]

return).

3. To use the revised regulatory rate of return at 7.09% pre-tax rate from 2020 to the end of the Claimants' PV Plants' lives.¹⁴⁸

121. The Experts however disagree on the following three parameters:

1. The approach to calculating the tax effect of debt;
2. The application of the Decision's instruction regarding operating and maintenance costs;
3. Whether to calculate an IRR separately for each of the PV Plants, or a single IRR for all of the Claimants' plants.¹⁴⁹

122. The Tribunal will address each of these parameters in the below sections.

a. Tax effect of debt

123. The Parties disagree as to whether the calculation of the PV Plants' cash flows should account for the tax effect (or tax benefit) of debt.¹⁵⁰

124. According to the Claimants and their Expert, Prof. Spiller, such additional accounting (or increase) of the PV Plants' cash-flows is not required, as the 7% post-tax benchmark return does not assume debt financing and does not include such tax benefit.¹⁵¹ Therefore, in order to perform a like-to-like comparison, the Claimants' IRR should not account for any debt financing either. Moreover, the Claimants argue that the tax effect of debt is accounted for by Prof. Spiller through the application of the discount rate,¹⁵² whereas applying the "tax shield" to the PV Plants' cash flows is inconsistent with the chosen FCFF methodology and results in double counting of the tax effect of debt.¹⁵³

125. The Respondent and Dr. Flores argue that by excluding the tax effect of debt, Prof. Spiller understates the actual IRR of the PV Plants. In support of its position,

---

¹⁴⁸ Joint Expert Report, para. 12 (footnotes omitted, emphasis added). ↩
¹⁴⁹ Joint Expert Report, para. 13. ↩
¹⁵⁰ Joint Expert Report, paras. 16-25. ↩
¹⁵¹ Claimants' Quantum Submission, para. 36, referring to 2000-2010 Plan for the Promotion of Renewable Energies in Spain, 19 December 1999 (R-67), p. 182 (Tribunal's translation) (emphasis added). See Decision, para. 985. ↩
¹⁵² Claimants' Quantum Submissions, para. 33; Joint Expert Report, para. 20. ↩
¹⁵³ Claimants' Comments on Spain's Responses to the Tribunal's Questions on Quantum, 14 October 2022, para. 24. ↩

[Page 36]

the Respondent refers to PV Investors v. Spain, where the tribunal decided to apply the so-called “tax shield” noting that “compensating the [c]laimants for taxes they have not paid would be contrary to the principle that reparation cannot exceed the harm effectively suffered.”¹⁵⁴

126. The Tribunal is mindful of the Respondent's concern that the Claimants' PV Plants cash flows should not be artificially lowered for the purposes of the IRR calculation by overestimating the Claimants' taxes.¹⁵⁵ At the same time, the Tribunal needs to make its assessment of the Claimants' IRR and the resulting damages (if any) based on the evidence submitted to it, in an economically sound manner and in accordance with the methodology agreed by the Experts.

127. The first factual element that the Tribunal needs to assess is the Spanish regulatory framework. It appears that the laws and regulations as well as other contemporaneous documents that were publicly available during the Claimants' investment process, which the Tribunal relied on to determine that the reasonable return benchmark was 7% post-tax,¹⁵⁶ provided that the reasonable return benchmarks was to be assessed with no regard to project financing.

128. For example, the 2000-2010 PER provided as follows: “[p]rofitability of the type project: calculated on the basis of maintaining an Internal Rate of Return (IRR) [...] a minimum of 7%, with own resources, before financing and after taxes.”¹⁵⁷ Similarly, the 2005-2010 PER referred to a return on a standard project “calculated on the basis of maintaining an Internal Rate of Return (IRR), measured in legal tender and for each standard project, around 7%, on equity (before financing) and after taxes.”¹⁵⁸ It also seems to have been the perception by the professional associations on the market: the October 2005 ASIF Report mentioned “an internal rate of return on the own equity

---

¹⁵⁴ Respondent's Quantum Submissions, para. 17, referring to AES Solar and others (PV Investors) v. The Kingdom of Spain, PCA Case No. 2012-14, Final Award, 28 February 2020 (RL-129), paras.791-792. ↩
¹⁵⁵ Dr. Flores' Comments on Prof. Spiller's Answers to the Tribunal's Questions on Quantum, 14 October 2022, para. 16. ↩
¹⁵⁶ See Decision, paras. 985-995. ↩
¹⁵⁷ 2000-2010 Plan for the Promotion of Renewable Energies in Spain, 19 December 1999 (R-67), p. 182 (Tribunal's translation) (emphasis added). See Decision, para. 985. ↩
¹⁵⁸ 2005-2010 Renewable Energy Promotion Plan, August 2005 (R-69)/(C-32[SP]), p. 274 (emphasis added). ↩

[Page 37]

invested of between 5 and 7%."¹⁵⁹

129. The content of these documents is uncontested by the Respondent. Moreover, Dr. Flores has acknowledged that "[t]he rate of return benchmark is defined 'with own resources, before financing' to indicate that it is a project-level return and will not be higher or lower depending on the level of external debt used to finance the project [...] the reasonable rate of return was calculated at the project level and should be the same whether or not the project was debt financed.”¹⁶⁰

130. The Tribunal is thus comfortable with the conclusion that the Spanish regulator when setting the reasonable return per project did so without accounting for any particular "financial engineering” that each particular investor might have used to maximize their IRRs.¹⁶¹

131. This conclusion however does not exclude the possibility of applying the tax shield when calculating the Claimants' IRR in this particular case, as the Tribunal's instruction was to assess the economic impact of the disputed measures on the Claimants' PV Plants to see if the New Regime allowed the Claimants to earn the 7% post-tax return which the Tribunal found they were entitled to. If the Claimants benefited from a tax shield (and it appears that they did given that this fact has not been contested by the Claimants¹⁶²), the Tribunal would, in principle, consider it appropriate to factor the effect of such a tax shield into the calculation of the Claimants' IRR in order to ultimately determine the amount of loss suffered by the Claimants (which cannot exceed the harm effectively suffered¹⁶³, even if some amount of damages was mitigated by the investors' own taxation optimization

---

¹⁵⁹ ASIF, Report: “Towards environmentally friendly electricity", October 2005 (R-246), p. 9 (emphasis added), see also, Universidad Pontificia de Comillas Report: “Solar energy: current status and immediate perspective", 2007 (R-272), p. 197. ↩
¹⁶⁰ Dr. Flores' Answers to the Tribunal's Questions on Quantum, 30 September 2022, para. 13. ↩
¹⁶¹ Prof. Spiller's Comments on Dr. Flores' Answers to the Tribunal's Questions on Quantum, 14 October 2022, para. 21. ↩
¹⁶² See Exhibit CLEX I, footnote 83 ("Claimants have informed me that this conversion into participative loans would have also occurred absent the Measures, for tax optimization purposes”). Dr. Flores also noted that he used Prof. Spiller's cash flow calculations to calculate the tax shield: "I rely on the tax effect of debt calculated in the discounted cash flow ("DCF") model in CLEX I and CLEX II, which is based on the actual projected debt payments for the PV Plants". See Joint Expert Report, para. 21. ↩
¹⁶³ See AES Solar and others (PV Investors) v. The Kingdom of Spain, PCA Case No. 2012-14, Final Award, 28 February 2020 (RL-129), paras.791-792 (“compensating the [c]laimants for taxes they have not paid would be contrary to the principle that reparation cannot exceed the harm effectively suffered"). ↩

[Page 38]

measures). In other words, the Tribunal thus does not consider that a tax shield that was used in practice should be disregarded at the stage of loss computation on the ground that the 7% reasonable rate of return was considered to apply irrespective of a PV project's financing structure. The Tribunal's task at this stage of the proceedings is to analyse whether such a return was guaranteed to the Claimants in this specific case. This is the reason why the Tribunal has earlier directed the Parties to perform an actual (not counterfactual) analysis of the performance of the Claimants' PV Plants.¹⁶⁴

132. That being noted, the Tribunal's analysis needs to be performed in accordance with the valuation methodology agreed by the Experts, as the Tribunal has explicitly directed the Parties to “agree on a common approach to calculating the tax effect of debt."¹⁶⁵

133. The Tribunal recalls that Dr. Flores and Prof. Spiller agreed to use the FCFF¹⁶⁶ valuation method.¹⁶⁷ According to financial literature referred to by the Claimants, the FCFF:

is often termed an unlevered cash flow, because it is unaffected by debt payments or the tax benefits flowing from these payments. If you are wondering where the tax benefits from interest payments, which are real cash benefits, show up, it is in the discount rate, when we compute the after-tax cost of debt. If we add this tax benefit as a cash flow to the firm, we will double count the tax benefit.¹⁶⁸

134. The Tribunal observes that other valuation methods such as free cash flows to the equity (the "FCFE”) and adjusted present value (the “APV”) account for the tax effect of debt through cash flows.¹⁶⁹

135. Dr. Flores does not contest these statements as such. His initial response (presented in the Joint Expert Report) in relation to the double-discounting that would necessarily occur if the Tribunal were to follow the Experts' agreement on the application of the

---

¹⁶⁴ Decision, paras. 1002, 1014, 1016. ↩
¹⁶⁵ Decision, para. 1019(iv). ↩
¹⁶⁶ Which stands for “free cash flows to the firm" as defined at para. 74 above. ↩
¹⁶⁷ Joint Expert Report, para.12. ↩
¹⁶⁸ Damodaran A. Applied Corporate Finance (2011), (Exhibit CLEX-364), p. 34 (footnotes omitted). ↩
¹⁶⁹ Exhibit CLEX II, p. 20, Table 2. ↩

[Page 39]

FCFF method whilst following Dr. Flores' approach on the tax shield was as follows:

Prof. Spiller also claims that he uses the projected cash flows excluding the tax effect of debt to avoid “double counting." [...] That is incorrect. He is referring to the discount rate (WACC), which is used to discount the additional cash flows (if any) needed to achieve the target IRR. This is a different steam of cash flows than the FCFF. There is no double counting.¹⁷⁰

136. Prof. Spiller in turn replied that

Dr. Flores claims that the additional cash flows needed to achieve the target IRR are a different stream of cash flows than the FCFF, and thus there is allegedly no double counting in including the tax effect of debt both in the cash flows and in the discount rate (WACC). This is incorrect, the additional cash flows have to be of the same nature as the original FCFFs, as they are the increment in the original FCFF required to achieve the target return. As a consequence, they have to be consistent with the manner the original FCFFs were computed.¹⁷¹

137. Dr. Flores further argued that if the tax benefit is incorporated directly into the cash flows of the PV Plants from the outset of the calculation, the resulting post-tax IRR would not be below 7%, so there would be no need to perform the rest of the analysis that consists in calculating and discounting additional cash flows to allow the Claimants' plants to obtain a reasonable return.¹⁷²

138. The Tribunal finds Dr. Flores' explanations unsatisfactory. First, the Tribunal notes that Dr. Flores has initially argued that there is no double-counting, as the discounting performed by Prof. Spiller applies to a different stream of cash flows.¹⁷³ Dr. Flores then subsequently (in the most recent round of submissions) argued that no discounting is even needed, as the Claimants' IRR would be above 7%, so there would be no need in calculating any additional discounted cash flows.¹⁷⁴ The Tribunal considers these explanations to be inconsistent.

139. Furthermore, even if Dr. Flores' approach could potentially be seen as avoiding inconsistency if only his own IRR calculations were applied (although even based on

---

¹⁷⁰ Joint Expert Report, footnote 30. ↩
¹⁷¹ Joint Expert Report, footnote 23. ↩
¹⁷² Dr. Flores' Comments on Prof. Spiller's Answers to the Tribunal's Questions on Quantum, 14 October 2022, paras. 12-17. ↩
¹⁷³ See Joint Expert Report, footnote 30. ↩
¹⁷⁴ Dr. Flores' Comments on Prof. Spiller's Answers to the Tribunal's Questions on Quantum, 14 October 2022, paras. 12-17. ↩

[Page 40]

his own approach the resulting IRR appears to be slightly below 7%¹⁷⁵), the Tribunal notes that the issue of potential double-counting would remain if one were to use a combination of Dr. Flores' and Prof. Spiller's assumptions (as opposed to using exclusively Dr. Flores' assumptions). For instance, Table 4 of the Joint Expert Report setting out the IRR compensation to the Claimants' PV Plants for different combinations of inputs including the tax effect of debt (without taking into account the compensation for retroactivity) indicates that the Claimants' PV Plants would need some additional annual cash flows to reach the 7% benchmark in all of the presented scenarios. In other words, the discounting (as the final step in the FCFF methodology chosen by the Experts) seems to be unavoidable in any of the presented scenarios.

140. The problem of double-counting of the tax effect of debt thus remains unresolved.¹⁷⁶ The Tribunal also notes that the Respondent has not submitted any alternative calculations based on a different methodology. On the contrary, the Experts have agreed on the application of the FCFF method, which the Tribunal is now bound to follow.

141. For the above reasons, the Tribunal opts for the Claimants' approach to the tax treatment of debt.

b. Operating and maintenance costs

142. In its Decision, the Tribunal reached the following conclusion in relation to the O&M costs of the Claimants' PV Plants:

Since the Tribunal's task here is to assess the actual economic impact of the Disputed Measures on the Claimants' PV Plants, it seems appropriate to use the actual prices of the operating and maintenance agreements. Therefore, the adjustment that resulted from the re-negotiation must be applied.¹⁷⁷

143. The Parties disagree on the application of this instruction. According to the Claimants' Expert, Prof. Spiller, the re-negotiated prices should be applied only from the actual dates of these re-negotiations, whereas the Respondent's Expert, Dr. Flores, applies

---

¹⁷⁵ See Table 2 of the Joint Expert Report reproduced at para. 76 above; Joint Expert Report, para. 14. ↩
¹⁷⁶ Claimants' Response to the Tribunal's Questions on Quantum, 30 September 2022, para. 20. ↩
¹⁷⁷ Decision, para. 1016. ↩

[Page 41]

the re-negotiated prices “retrospectively” (and instead of the actual, historical O&M costs) from the PV Plants' commissioning dates.¹⁷⁸

144. The Respondent's and Dr. Flores' argument here is that the Tribunal should only use the costs that are reasonable (as it did for the initial investment costs), as the Claimants' actual O&M costs were “excessively high".¹⁷⁹

145. The Tribunal is not convinced that the analogy that the Respondent is trying to draw between the initial investment costs and O&M is correct. The Claimants' initial investment costs were not subject to any adjustment and seemed above average regardless of the assumptions underlying the New Regime, whereas the O&M costs were corrected by the Claimants once the regime was changed to provide less support to the PV Plants and in this sense the Claimants' approach was not an inefficient or unreasonable one. Retrospectively lowering the O&M prices, as proposed by Dr. Flores, does not seem fair in the circumstances and is, indeed, inconsistent with the Tribunal's instruction to assess the actual economic impact of the Disputed Measures on the PV Plants.

146. Therefore, it seems appropriate to implement the Tribunal's direction in accordance with the Claimants' parameters for the O&M costs.

c. Whether to calculate the IRR on a plant-by-plant basis

147. In its Decision, the Tribunal observed that neither Party has calculated an IRR separately for each PV Plant (i.e. on a “plant-by-plant" basis).¹⁸⁰ This comment reflected the Tribunal's understanding of the content of the Parties' positions as presented in the Decision. The Tribunal thereafter set out its instructions to the Parties for the purpose of preparing their joint IRR calculation (paragraph 1019 of the Decision). The issue of whether the IRR should be calculated on a plant-by-plant basis or in aggregate was thus left to be considered and agreed upon between the Parties and their Experts, in accordance with the Tribunal's final instruction inviting the Parties “to agree on any other outstanding issues that may arise in the course of the

---

¹⁷⁸ Joint Expert Report, paras. 26-35. ↩
¹⁷⁹ Respondent's Quantum Submission, paras. 28-30. ↩
¹⁸⁰ Decision, para. 996. ↩

[Page 42]

preparation of the Joint IRR Calculation".¹⁸¹ It is on this basis that the Tribunal conducts the below analysis.

148. As a preliminary point, the Tribunal notes that the Claimants in these proceedings include Sevilla Beheer B.V., Cordoba Beheer B.V., as well as the 57 Spanish Project Companies.¹⁸² This means that certain Claimants are claiming in a capacity of shareholder whilst the majority of the Claimants have direct rights relating to the operation of the PV installations. Each of the Spanish Project Companies therefore has a direct right to receive the compensation in relation to its plants' returns (if any) resulting from these proceedings. The Tribunal however considers it unnecessary to calculate the IRR at the level of the Spanish Project Companies (which would require calculating 57 individual IRRs) and finds the aggregation at the project level for each plant to be sufficient for the below reasons.

149. As determined at paragraphs 983-995 of the Decision, the reasonable rate of return of 7% was guaranteed by the Respondent as a project return, not a shareholder return. Therefore, in order to carry out a like-to-like comparison, the Tribunal deems it appropriate to calculate the PV Plants' IRR on a plant-by-plant basis, in accordance with the Claimants' and Prof. Spiller's approach.

150. The Tribunal notes that this approach was equally supported by the Hydro v. Spain and RREEF v. Spain tribunals.¹⁸³ Specifically, the Hydro v. Spain tribunal ruled as follows:

The clear rationale behind the legislation is to consider what constitutes a reasonable rate of return on a plant by plant basis and the Tribunal finds there is no justification for an aggregation approach. The fact that a plant happens to be in a portfolio owned by a particular investor should not impact whether that plant is considered to make a reasonable rate of return. Each plant should be considered on its own, as would have to be the case if each were

---

¹⁸¹ Decision, para. 1020. ↩
¹⁸² Decision, paras. 108-109. ↩
¹⁸³ Hydro Energy 1 S.à r.l. and Hydroxana Sweden AB v. Kingdom of Spain, ICSID Case No. ARB/15/42, Decision on Jurisdiction, Liability and Directions on Quantum, 9 March 2020 (CL-196), para. 724; RREEF Infrastructure (G.P.) Limited and RREEF Pan-European Infrastructure Two Lux S.à r.l. v. Kingdom of Spain, ICSID Case No. ARB/13/30, Decision on Responsibility and on the Principles of Quantum, 30 November 2018 (CL-180), para. 545. ↩

[Page 43]

individually owned by different investors.¹⁸⁴

151. The Tribunal finds that Spain's reference to PV Investors¹⁸⁵ to be inapposite, as in that case the decision to calculate the harm “per investor” rather than for each plant was agreed between the experts.¹⁸⁶

152. Consequently, the Claimants' parameter shall be applied and the IRR shall be calculated on a plant-by-plant basis.

d. Conclusion on the Claimants' IRR

153. Therefore, on the basis of the above parameters, the Claimants' operational IRR is determined by the Tribunal as constituting 6.29% with the following breakdown per project: Mahora – 5.66%; Villar de Cañas – 6.24%; Ronda – 6.32%; Matapozuelos – 7.13%; Fuentes de Año – 7.22%.

154. In view of this finding and before deciding on the total amount of damages, the Tribunal needs to address the Respondent's argument that the Claimants' IRR of 6.95% or even of 6.29% is “within the margin of error of a 7% post-tax rate of return", as “if instead of basing their analysis in the Joint Report on cash flow models already in the record of the Arbitration, the Experts had sought to do a more realistic modelling of the PV Plants' cash flows, including lower tax payments (and, therefore, higher cash flows) they would plausibly have arrived at a conclusion that the IRR of the Claimants' PV Plants was in fact not below 7% post-tax”.¹⁸⁷ The Respondent thus submits that it had not failed to offer a reasonable return to the Claimants' PV Plants.¹⁸⁸

155. In the Tribunal's view, the Respondent's request belongs to the discussion on liability. Spain is effectively asking the Tribunal to revise its finding regarding the parameters

---

¹⁸⁴ Hydro Energy 1 S.à r.l. and Hydroxana Sweden AB v. Kingdom of Spain, ICSID Case No. ARB/15/42, Decision on Jurisdiction, Liability and Directions on Quantum, 9 March 2020 (CL-196), para. 724. ↩
¹⁸⁵ Respondent's Quantum Submission, para. 35. ↩
¹⁸⁶ See AES Solar and others (PV Investors) v. The Kingdom of Spain, PCA Case No. 2012-14, Final Award, 28 February 2020 (RL-129), para. 727 ("... in light of the experts' agreement and the absence of objection by the Parties, the Tribunal will establish the harm allegedly suffered by Claimant, that is for each of the 19 Claimant entities..."). ↩
¹⁸⁷ Respondent's Quantum Submissions, para. 40. ↩
¹⁸⁸ Respondent's Quantum Submissions, para. 43. ↩

[Page 44]

for deciding whether there was a breach of the reasonable return guarantee by saying that an IRR within the range of 7% would be sufficient to establish that no violation occurred. The Tribunal does not agree to do so. First, the Tribunal considers its analysis regarding the 7% rate to be fully supported by the evidentiary record in this case, as discussed at paragraphs 983-995 of the Decision. Second, the 7% after-tax rate has also been used by Spain's Expert, Dr. Flores, as the relevant benchmark.¹⁸⁹ Third, the amount of compensation is in any case proportional to the profitability of the Claimants' PV Plants. Therefore, a very small difference between the IRR and the 7% rate will result in compensation, regardless of the fact that its amount may be comparatively low. Finally, the Tribunal cannot make a finding that there was no breach or that some amount of compensation (even a very small one) is not due simply based on the allegation that the models used in the present proceedings are not perfectly realistic, as these are the models to which both Parties and their Experts agreed. Moreover, it cannot be established with certainty that a “more realistic modelling” mentioned by Spain would not in fact result in a greater gap between the Claimants' IRR and the 7% benchmark. In view of these considerations, the Tribunal rejects the argument that an IRR slightly below 7% should not result in an obligation to compensate the Claimants for the harm suffered as a result of the Disputed Measures.

156. The Tribunal will now turn to calculating the total amount of damages due in view of its findings regarding the compensation for retroactivity and the Claimants' IRRs.

C. Interplay between the compensation to achieve the reasonable return and the compensation for retroactivity

1. The Parties' Positions

a. The Joint Expert Report

157. The Experts disagree on how compensation based on the joint IRR calculation would interact with compensation for retroactivity.¹⁹⁰

---

¹⁸⁹ See Second Econ One Report, 28 September 2018, Figure 7. ↩
¹⁹⁰ Joint Expert Report, para. 2. ↩

[Page 45]

i. Compass Lexecon

158. According to Prof. Spiller, the Tribunal's Decision “has not provided any instructions as to whether retroactivity compensation should be accounted for in the Joint IRR calculation."¹⁹¹ Therefore, Prof. Spiller provided two alternatives: (i) IRR calculation independent of the compensation for retroactivity and (ii) IRR calculation accounting for the compensation for retroactivity.

159. In order to compute the IRR compensation accounting for the compensation for retroactivity, Prof. Spiller uses the following approach:

I first include the compensation for Retroactivity by incorporating into the revenues of the PV Plants the respective Adjusted Rinv values [...].¹⁹²

I then calculate the IRR of the PV Plants with such adjusted revenues [...].¹⁹³

Next, I determine the additional yearly payment needed to achieve a 7% post- tax return [...]¹⁹⁴.

160. The results of Prof. Spiller's calculations are demonstrated in the below table:¹⁹⁵

Retroactivity Compensation (1)	Tax Effect of Debt (2)	O&M Costs (3)	Combined PV Plants or Plant by Plant (4)	Compensation for IRR after Retroactivity (€) (5)
1. Compass Lexecon	Compass Lexecon	Compass Lexecon	Compass Lexecon	1,975,031
2. Compass Lexecon	Compass Lexecon	Compass Lexecon	Quadrant	456,703
3. Compass Lexecon	Compass Lexecon	Quadrant	Compass Lexecon	1,577,306
4. Compass Lexecon	Compass Lexecon	Quadrant	Quadrant	0
5. Compass Lexecon	Quadrant	Compass Lexecon	Compass Lexecon	0
6. Compass Lexecon	Quadrant	Compass Lexecon	Quadrant	0
7. Compass Lexecon	Quadrant	Quadrant	Compass Lexecon	0
8. Compass Lexecon	Quadrant	Quadrant	Quadrant	0
9. Quadrant	Compass Lexecon	Compass Lexecon	Compass Lexecon	2,664,701
10. Quadrant	Compass Lexecon	Compass Lexecon	Quadrant	1,272,643
11. Quadrant	Compass Lexecon	Quadrant	Compass Lexecon	2,257,642
12. Quadrant	Compass Lexecon	Quadrant	Quadrant	4,177
13. Quadrant	Quadrant	Compass Lexecon	Compass Lexecon	174,642
14. Quadrant	Quadrant	Compass Lexecon	Quadrant	0
15. Quadrant	Quadrant	Quadrant	Compass Lexecon	0
16. Quadrant	Quadrant	Quadrant	Quadrant	0

---

¹⁹¹ Joint Expert Report, para. 47. ↩
¹⁹² See Joint Expert Report, para. 49, footnote 56 (“This is equivalent to taking the additional yearly cash flows computed for the compensation for Retroactivity, and adding them to the cash flows of the PV Plants.") ↩
¹⁹³ Joint Expert Report, para. 49. ↩
¹⁹⁴ Joint Expert Report, para. 49. ↩
¹⁹⁵ Joint Expert Report, para. 49, Table 5: Compensation to the PV Plants To Achieve a 7% Post-Tax IRR, Accounting for Compensation for Retroactivity (with approach used by Compass Lexecon), Under Different Sets of Inputs on Which the Experts Disagree. ↩

[Page 46]

161. The compensation presented in the above table corresponds exclusively to the compensation to achieve a 7% post-tax IRR, and not to the compensation for retroactivity.

ii. Quadrant Economics

162. Dr. Flores argues that any compensation for retroactivity will increase the IRR of the PV Plants.¹⁹⁶ Therefore, Dr. Flores adds his calculation of the compensation for retroactivity (EUR 4,092,193) to the cash flows of the PV Plants, as demonstrated in the below table:¹⁹⁷

Compensation for Retroactivity (1)	Tax Effect of Debt (2)	O&M Costs (3)	Combined PV Plants or Plant by Plant (4)	Joint IRR Compensation Before Retroactivity (€) (5)	Compensation for Retroactivity (€) (6)	Compensation for IRR after Retroactivity (€) (7)
1.	Compass	Compass	Compass	Compass	5,703,356	4,896,215	2,017,532
2.	Compass	Compass	Compass	Quadrant	5,384,366	4,896,215	488,151
3.	Compass	Compass	Quadrant	Compass	5,272,311	4,896,215	1,586,486
4.	Compass	Compass	Quadrant	Quadrant	4,054,635	4,896,215	0
5.	Compass	Quadrant	Compass	Compass	3,169,568	4,896,215	86,044
6.	Compass	Quadrant	Compass	Quadrant	1,751,625	4,896,215	0
7.	Compass	Quadrant	Quadrant	Compass	2,733,328	4,896,215	0
8.	Compass	Quadrant	Quadrant	Quadrant	395,438	4,896,215	0
9.	Quadrant	Compass	Compass	Compass	5,703,356	4,092,193	2,701,442
10.	Quadrant	Compass	Compass	Quadrant	5,384,366	4,092,193	1,292,173
11.	Quadrant	Compass	Quadrant	Compass	5,272,311	4,092,193	2,270,396
12.	Quadrant	Compass	Quadrant	Quadrant	4,054,635	4,092,193	0
13.	Quadrant	Quadrant	Compass	Compass	3,169,568	4,092,193	459,548
14.	Quadrant	Quadrant	Compass	Quadrant	1,751,625	4,092,193	0
15.	Quadrant	Quadrant	Quadrant	Compass	2,733,328	4,092,193	0
16.	Quadrant	Quadrant	Quadrant	Quadrant	395,438	4,092,193	0

---

¹⁹⁶ Joint Expert Report, para. 51. ↩
¹⁹⁷ Joint Expert Report, paras. 51-52, Table 6: Compensation to the PV Plants To Achieve a 7% Post-Tax IRR, Accounting for Compensation for Retroactivity (with approach used by Quadrant Economics), Under Different Sets of Inputs on Which the Experts Disagree. ↩

[Page 47]

b. The Claimants' Quantum Submissions

163. In their Quantum Submissions, the Claimants state that taken in isolation, the Claimants' retroactivity and “reasonable return" damages are, respectively, EUR 4,896,215 and EUR 5,703,356.¹⁹⁸

164. The Claimants note that although the Tribunal has not given any instructions as to the interplay between these two heads of damages, Prof. Spiller nevertheless presented an approach accounting for both retroactivity and reasonable return damages.¹⁹⁹

165. The Claimants further note the Experts' agreement that the compensation for retroactivity if awarded, will increase the PV Plants' IRRs and the cash flows they would require to reach a 7% post-tax return would be lower.²⁰⁰

166. The Claimants endorse Prof. Spiller's approach to accounting for the retroactivity compensation, which includes the following steps:

• the incorporation of the retroactivity damages into the revenues of the PV Plants;
• the calculation of the IRR of the PV Plants using the adjusted revenues;
• the calculation of the additional yearly payment needed to achieve a 7% post- tax return; and
• the discounting of the additional annual cash-flows to 20 June 2014 at a 6.06% WACC.²⁰¹

167. Dr. Flores, according to the Claimants, adopts a slightly different approach by simply subtracting the "retroactivity damages" from the "reasonable return damages".²⁰²

168. In the Claimants' view, Prof. Spiller's approach is more precise and thus preferable despite the fact that it yields lower damages in all scenarios but one. According to Prof. Spiller's calculations, the Claimants shall be awarded EUR 6,871,246

---

¹⁹⁸ Claimants' Quantum Submissions, para. 67. ↩
¹⁹⁹ Claimants' Quantum Submissions, para. 68. ↩
²⁰⁰ Claimants' Quantum Submissions, para. 69. ↩
²⁰¹ Claimants' Quantum Submissions, para. 70. ↩
²⁰² Claimants' Quantum Submissions, para. 70. ↩

[Page 48]

(retroactivity damages of EUR 4,896,215 + reasonable return damages of EUR 1,975,031 as adjusted in view of the compensation for retroactivity).²⁰³

c. The Respondent's Quantum Submissions

169. The Respondent's primary case is that only compensation for retroactivity is due, as the Claimants' IRR was at 6.95% (on Dr. Flores' case), which is only slightly below 7% and should not give rise to compensation.²⁰⁴ According to Spain, an IRR of 6.95% is within the margin of error and could plausibly yield an even higher (7%) result, had the Experts sought to do a more realistic modelling of the PV Plants' cash flows, including lower tax payments in view of the available tax benefits.²⁰⁵ In this connection, the Respondent argues:

[...] based on the calculations presented by the Experts, there is an eminently reasonable chance that the true IRR of the Claimants' PV Plants, taking into account their effective tax rate in real life, has in fact met the reasonable rate of return of 7% post-tax. In these circumstances, awarding compensation based on a formulaic shortfall between a calculated IRR and 7% would be fundamentally unfair and would imply overcompensating Claimants for a harm they have not suffered, hence breaching the principle of full reparation.²⁰⁶

170. The Respondent also emphasizes that even on the Claimants' own case (an IRR of 6.29%) their investments suffered nothing akin to the destruction of value.²⁰⁷

171. In view of these considerations, the Respondent requests that the Tribunal find that "Spain has not breached its obligations under the ECT to offer a reasonable rate of return to the PV Plants, and therefore, that no compensation be awarded under this head of damages.”²⁰⁸

172. Therefore, the Respondent's primary case is that the Claimants are only entitled to retroactivity damages of EUR 4,092,193 (Dr. Flores' valuation).

---

²⁰³ Claimants' Quantum Submissions, para. 72. ↩
²⁰⁴ Respondent's Quantum Submissions, paras. 37-38. ↩
²⁰⁵ Respondent's Quantum Submissions, paras. 38-41. ↩
²⁰⁶ Respondent's Quantum Submissions, para. 41. ↩
²⁰⁷ Respondent's Quantum Submissions, para. 42. ↩
²⁰⁸ Respondent's Quantum Submissions, para. 43. ↩

[Page 49]

173. Alternatively, the Respondent argues that any compensation based on the IRR of the Claimants' PV Plants needs to account for any compensation for retroactivity.²⁰⁹

174. According to the Respondent, Prof Spiller's approach on how to compute compensation relating to the IRR of the PV Plants taking into account compensation for retroactivity “produces non-sensical results”.²¹⁰ The Respondent's criticisms are as follows:

[...] as indicated in Table 4 of the Joint Report, if one were to accept all the inputs proposed by Compass Lexecon, the PV Plants would require an additional [EUR] 5,703,356 to achieve a 7% post tax IRR, before taking into account compensation for Retroactivity. However, if the PV Plants were to receive [EUR] 4,896,215 in compensation for Retroactivity, as calculated by Compass Lexecon, then the PV Plants only require [EUR] 807,141 to achieve a 7% post-tax IRR. Despite this, as indicated in row 1 of Table 5 of the Joint Report, Compass Lexecon's approach to computing compensation relating to the IRR of the PV Plants taking into account compensation for Retroactivity more than doubles the [EUR] 807,141 required amount, to [EUR] 1,975,031.55 What this means is that if the PV Plants were to receive the [EUR] 1,975,031 proposed by Compass Lexecon on top of [EUR] 4,896,215, the PV Plants would end up with a post-tax IRR higher than 7% [...].²¹¹

175. In view of the above, the Respondent endorses the approach followed by Dr. Flores requiring the addition of the compensation for retroactivity to the cash flows of the PV Plants, which reduces the compensation required to obtain a 7% post- tax IRR.²¹² According to Dr. Flores' calculations, if all his assumptions are accepted, no additional compensation would be required to achieve a 7% return.²¹³

2. The Tribunal's Analysis

176. According to the Tribunal's Decision, “the Respondent breached Article 10(1) of the ECT to the extent that it applied the New Regime retroactively to the remuneration already received by the Claimants' PV Plants under RD 661/2007 and RD 1578/2008".²¹⁴

---

²⁰⁹ Respondent's Quantum Submissions, para. 49. ↩
²¹⁰ Respondent's Quantum Submissions, para. 52. ↩
²¹¹ Respondent's Quantum Submissions, para. 52 (footnotes omitted, emphasis in the original). ↩
²¹² Respondent's Quantum Submissions, para. 53. ↩
²¹³ Respondent's Quantum Submissions, para. 53. ↩
²¹⁴ Decision, para. 1080. ↩

[Page 50]

177. In addition, the Tribunal ruled (in accordance with its findings in respect of the Claimants' legitimate expectations claim) that there would also be a breach of Article 10(1) of the ECT, to the extent that the Claimants' IRR was below 7% post-tax, as a result of the adoption of the Disputed Measures.²¹⁵ Such a breach would also result in an obligation to compensate the Claimants.

178. The Tribunal has also dismissed the Respondent's argument that the Claimants' IRR in this case is “within the margin of error of a 7% post-tax rate of return” and should not result in a finding of breach.²¹⁶

179. As noted by the Claimants, the Experts agree that the compensation for retroactivity, if awarded, will increase the PV Plants' IRRs²¹⁷ and “the cash flows they would require to reach a 7% post-tax return would be lower."²¹⁸

180. The Tribunal thus reiterates its earlier finding (see paragraphs 67-69 above) that any compensation to achieve a 7% IRR should incorporate an increase in the PV Plants' cash flows resulting from the compensation for retroactivity.

181. As regards the method for incorporating the compensation for retroactivity into the IRR damages, the Tribunal considers the Claimants' approach to be the appropriate one. To recall, Prof. Spiller first includes the compensation for retroactivity by incorporating into the PV Plants' revenues the respective Adjusted Rinv values (see paragraphs 35-38 above). He then calculates the IRR of the PV Plants with these adjusted revenues following the same method that the Experts agreed to use to calculate the IRR without accounting for retroactivity (see paragraph 74 above). Prof. Spiller then determines the additional yearly payment needed to achieve a return of 7% post-tax and discounts these additional cash flows to 20 June 2014 at 6.06% (the WACC).²¹⁹ Although Dr. Flores' approach is simpler to implement (he subtracts the compensation for retroactivity from the IRR damages²²⁰), the Tribunal finds Prof.

---

²¹⁵ Decision, para. 1080. ↩
²¹⁶ See paras. 154-156 above. ↩
²¹⁷ Claimants' Quantum Submissions, para. 69; Joint Expert Report, para. 51. ↩
²¹⁸ Claimants' Quantum Submissions, para. 69. ↩
²¹⁹ Joint Expert Report, paras. 41 and 49; Claimants' Quantum Submissions, para. 70. ↩
²²⁰ Joint Expert Report, para. 51. ↩

[Page 51]

Spiller's method to be more precise, as it simultaneously accounts for both findings made by the Tribunal in a single calculation.

182. The final amount to be awarded is thus EUR 6,756,894 computed as a sum of EUR 4,092,193 (compensation for retroactivity alone) and EUR 2,664,701 (compensation for IRR after retroactivity following Prof. Spiller's approach).

V. INTEREST

A. The Parties' Positions

1. The Claimants

183. The Claimants submit that it is “well established that interest forms an integral part of any award of compensation, the aim of which is to achieve ‘full reparation' and to re- establish the situation which would have existed had the illegal acts not been committed"²²¹ and request that the Tribunal award both pre- and post-award interest at a rate of 7.58% (compounded monthly).²²²

184. As regards pre-award interest, the Claimants argue that the applicable rate is a commercial rate established on a market basis.²²³ The Claimants and their Expert consider that such a commercial rate established on a market basis is the Claimants' cost of equity, i.e. 7.58%.²²⁴

185. Prof. Spiller explains that the Claimants' cost of equity is the appropriate approach in this case as 95% of the estimated damages correspond to losses to the Claimants' equity stake and less than 5% correspond to losses to the Claimants' debt stake.²²⁵

186. With respect to Dr. Flores' alternative proposal to use a short-term risk-free rate such as, for example, the six-month or one-year EURIBOR, Prof. Spiller argues that a risk- free rate is not appropriate as the Claimants' damages could not be converted into a

---

²²¹ Memorial, paras. 479-480, referring to Article 38 of the ILC Articles (CL-41). ↩
²²² Claimants' Quantum Submissions, para. 73. ↩
²²³ Memorial, para. 478, referring, by analogy, to Article 13 (Expropriation) of the ECT; Reply, para. 749. ↩
²²⁴ Claimants' Reply, para. 750; First Compass Lexecon Report, para. 105. ↩
²²⁵ First Compass Lexecon Report, para. 105. ↩

[Page 52]

certain, non-contentious and legally binding promise of payment at a certain date by the defendant, or that the defendant was perceived to be risk-free.²²⁶

187. Moreover, Prof. Spiller notes that EURIBOR has been negative for most of the period under analysis and that Dr. Flores' proposal “results in the unreasonable outcome that Claimants would be effectively paying Spain interest on the amounts owed to them."²²⁷ The Claimants add that an interbank interest rate, such as EURIBOR, is insufficient on its own. To use such a rate, one would need to add a margin on top of it.²²⁸

188. Finally, the Claimants and their Expert provide two further criticisms of Dr. Flores' position:

Following the introduction of the Disputed Measures, the Claimants were deprived of receiving additional dividends. The opportunity cost in respect thereof is the Claimants' cost of equity. Furthermore, given the uncertainty associated with any payment of damages at the date of valuation, the Claimants would be unable to raise funds in order to cover the cash shortfall generated by the Disputed Measures at the risk-free rate.²²⁹

189. As regards Vestey v. Venezuela referred to by Dr. Flores, Prof. Spiller notes that the present case is different as it is not concerned with an expropriation scenario, where the investor is no longer doing business in the host State.²³⁰

190. As regards post-award interest, given its dual purpose of ensuring prompt compliance and preventing unjust enrichment, the Claimants submit that a rate higher than 7.58% remains apposite.²³¹

---

²²⁶ Second Compass Lexecon Report, para. 89. See also Claimants' Comments on Spain's Responses to the Tribunal's Questions on Quantum, 14 October 2022, para. 29. ↩
²²⁷ Second Compass Lexecon Report, para. 89. ↩
²²⁸ Reply, para. 754, referring to I Marboe, Calculation of Compensation and Damages in International Investment Law (Oxford University Press), 2009 (CL-40), para. 6.144. ↩
²²⁹ Reply, para. 755, referring to Second Compass Lexecon Report, para. 90. ↩
²³⁰ Second Compass Lexecon Report, para. 87. ↩
²³¹ Reply, paras. 756-758, referring to Bernhard von Pezold and others v Republic of Zimbabwe, ICSID Case No. ARB/10/15, Award, 28 July 2015 (CL-103), para. 943; Gold Reserve Inc. v Bolivarian Republic of Venezuela, ICSID Case No. ARB(AF)/09/1, Award, 22 September 2014 (CL-139), para. 856. ↩

[Page 53]

2. The Respondent

191. Dr. Flores disagrees with Prof. Spiller regarding the use of the Claimants' cost of equity as an interest rate:

Claimants should not be compensated for risks they do not bear. Any compensation amount that the Tribunal could eventually award to Claimants is not subject to the ex ante risks that are captured in the cost of equity. Consequently, the cost of equity ought not to be used as an interest rate.²³²

192. Dr. Flores further explains:

Had Claimants received additional funds in 2013, for example, they would face the decision of what to do with those funds. By applying the cost of equity to this past amount, Compass Lexecon is assuming that any proceeds received from this Arbitration would have been reinvested in equally risky endeavors. Had Claimants done so, those risks may have materialized in negative returns. But Claimants were never exposed to those risks, and thus it would be wholly inappropriate to compensate Claimants for risks they never faced.²³³

193. Therefore, according to Spain, using a short-term, risk-free rate, such as the six-month EURIBOR, compounded semi-annually, would be appropriate in the present case.²³⁴

194. Finally, Dr. Flores notes that Prof. Spiller's approach “is far outside the mainstream – the cost of capital, which is equivalent to the cost of equity in the absence of debt, was used to calculate interest on damages in only 2% of cases."²³⁵

B. The Tribunal's Analysis

195. The Tribunal considers that the payment of interest is required in the present case in order to ensure the full reparation of the damage caused by Spain to the Claimants.

196. The Tribunal notes that the ECT does not address the question of interest on compensation awarded for breaches of Article 10(1) of the ECT. The Claimants suggest that the applicable rate is a commercial rate established on a market basis,

---

²³² First Econ One Report, para. 337. ↩
²³³ Second Econ One Report, para. 262. ↩
²³⁴ Spain's Responses to the Tribunal's Questions on Quantum, 30 September 2022, para. 3; Second Econ One Report, para. 260. ↩
²³⁵ Second Econ One Report, para. 266, referring to PwC, 2015 – International Arbitration Damages Research, Closing the Gap between Claimants and Respondents, 2015 (EO-174), p. 9. ↩

[Page 54]

referring to Article 13 of the ECT (Expropriation).²³⁶ The Claimants however provide no explanation as to why Article 13 of the ECT should apply by analogy in the present case, where compensation is being awarded for a time-adjusted value of lost future cash flows.

197. As regards the Claimants' proposal to use the cost of equity (7.58%), the Tribunal is of the view that the Claimants have not substantiated that they actually faced the risk they allege needs to be compensated by the suggested rate. The Tribunal therefore agrees with the Respondent that the Claimants should not be compensated for risks they did not prove to have borne.²³⁷

198. In this regard, the Tribunal notes its agreement with the findings of the tribunal in Hydro v. Spain, which observed:

Although being part of reparation of damages, interest is not simply an additional element in the overall calculation of compensation, which would increase the principal sum and would duplicate calculation of damages. Therefore, it cannot be considered as compensation for potential "additional profits”, determined by the cost of equity, as suggested by Professor Spiller's quantum expert opinion. Rather, the payment of interest compensates the injured person for the fact that for a certain period of time the principal sum is not at its disposal, and that it is temporarily deprived of liquidity. Interest does not compensate for the loss of money but for the loss of the use of money, a “secondary element, subordinated to the principal amount of the claim.”²³⁸

199. The Claimants have also argued that “even if Spain were correct that the interest rate should be determined according to Spain's borrowing costs [...], those costs are best reflected by the yield on the Spanish 10-year bond, not the EURIBOR [rate]."²³⁹ However, the amounts awarded to the Claimants have not been subject to the borrower default risks encapsulated by this rate.

---

²³⁶ Memorial, para. 478, referring, by analogy, to Article 13 (Expropriation) of the ECT; Reply, para. 749. ↩
²³⁷ Spain Responses to the Tribunal's Questions on Quantum, 30 September 2022, para. 3; Second Econ One Report, para. 260. ↩
²³⁸ Hydro Energy 1 S.à r.l. and Hydroxana Sweden AB v. Kingdom of Spain, ICSID Case No. ARB/15/42, Award, 5 August 2020, para. 143 (footnotes omitted). The use of the cost of equity as an interest rate was also rejected by the RWE v. Spain tribunal at RWE Innogy GmbH and RWE Innogy Aersa S.A.U. v. Kingdom of Spain, ICSID Case No. ARB/14/34, Award, 18 December 2020, para. 132 ("the Tribunal is mindful that interest rates have been historically low in the period from June 2014, and does not consider it appropriate to award a rate as high as 7.61% as derived by the Claimants as the opportunity cost of equity"). ↩
²³⁹ Claimants' Post-Hearing Brief, paras. 223-224. ↩

[Page 55]

200. The Tribunal thus agrees with the Respondent that in the present case interest should be based on a short-term risk-free rate.

201. In his First Expert Report, Dr. Flores stated that calculating interest using the six- month or one-year EURIBOR as an approximation for the risk-free rate “would be consistent with economic theory and practice."²⁴⁰ In response to Dr. Flores, Prof. Spiller noted that “since EURIBOR has been negative for most of the period under analysis, Econ One's proposal results in the unreasonable outcome that Claimants would be effectively paying Spain interest on the amounts owed to them."²⁴¹ The Claimants then referenced authorities that argue in favor of increasing the relevant EURIBOR rate by an additional margin.²⁴² However, neither the Claimants nor their Expert have specified which margin, if any, should be applied in addition to the EURIBOR in order to reach an interest rate they would consider appropriate.

202. The Tribunal agrees with the Claimants that using EURIBOR on its own, as proposed by Spain, would be insufficient, because EURIBOR is a benchmark rate applicable to loans on the inter-bank market. The Tribunal specifically notes that Spain has not contested the position that a short-term risk-free rate might need to be supplemented by an additional margin.²⁴³ Moreover, the possibility of adding a margin to a standard rate has been widely endorsed by investment tribunals.²⁴⁴ The Tribunal therefore considers that the inclusion of an additional margin to the standard rate is appropriate in the case at hand.

---

²⁴⁰ First Econ One Report, para. 344; Spain's Responses to the Tribunal's Questions on Quantum, 30 September 2022, para. 3. See also Second Econ One Report, para. 30; Rejoinder, para. 1427. ↩
²⁴¹ Second Compass Lexecon Report, para. 89. ↩
²⁴² Reply, para. 754, referring to I Marboe, Calculation of Compensation and Damages in International Investment Law (Oxford University Press), 2009 (CL-40), para. 6.144. ↩
²⁴³ Spain's Responses to the Tribunal's Questions on Quantum, 30 September 2022, paras. 3-5; Second Econ One Report, paras. 259-266; Respondent's Post-Hearing Brief, paras. 179-186. ↩
²⁴⁴ See, for example, OI European Group B.V. v. Bolivarian Republic of Venezuela, ICSID Case No. ARB/11/25, Award, 10 March 2015, paras. 942-944; Georg Gavrilovic and Gavrilovic d.o.o. v. Republic of Croatia, ICSID Case No. ARB/12/39, Award, 26 July 2018, para. 1295; Dominicana Renovables, S.L. v. The Dominican Republic, ICC Case No. 23364/JPA, Final Award, 20 January 2021, para. 966; Lion Mexico Consolidated L.P. v. United Mexican States, ICSID Case No. ARB(AF)/15/2, Award, 20 September 2021, para. 919. See also I Marboe, Calculation of Compensation and Damages in International Investment Law (Oxford University Press), 2009 (CL- 40), para. 6.144 ("the interbank interest rate in its 'generic' form would not be sufficient. As is usual to add a surcharge on the interbank rates for commercial loans to customers, this should also be done with regard to the interest on an amount of damages"). ↩

[Page 56]

203. For the above reasons, the Tribunal thus orders that the Respondent shall pay interest on the principal sum outstanding under this Award (EUR 6,756,894) from 20 June 2014 until the date of payment at the rate of EURIBOR plus 1 % established and compounded semi-annually. The Tribunal does not deem it necessary to distinguish between pre- and post-award interest in the circumstances, as the duty to pay interest does not pursue any punitive objective.²⁴⁵

VI. COSTS

A. The Parties' Positions

1. The Claimants

204. The Claimants request that the Tribunal grant an award pursuant to Article 61(2) of the ICSID Convention ordering that Spain bear the costs of this arbitration, as well as the Claimants' costs for legal representation, in the amount of EUR 4,537,712.81.²⁴⁶

205. The Claimants argue that the Tribunal has broad and unfettered discretion with respect to the allocation of costs.²⁴⁷ The Claimants also observe that "the Tribunal has confirmed in its decision that it had jurisdiction over the dispute notwithstanding Spain's jurisdictional challenges and found that Spain had breached the ECT and caused harm to the Claimants."²⁴⁸

206. As regards specifically the costs incurred as a result of Spain's Second Request for Reconsideration, the Claimants argue that the Respondent shall bear the entirety of the arbitration costs and counsel fees based on two criteria: the relative success (or lack thereof) of a party's claim and its procedural conduct.²⁴⁹ In this connection, the Claimants observe that: (i) the Tribunal rejected Spain's Second Request for Reconsideration, (ii) which was submitted in disregard of the stringent threshold set in the Tribunal's previous decision “in a transparent attempt to stall and delay the issuance of the Final Award” and (iii) that Spain's Second Request for

---

²⁴⁵ See Hydro Energy 1 S.à r.l. and Hydroxana Sweden AB v. Kingdom of Spain, ICSID Case No. ARB/15/42, Award, 5 August 2020, para. 144. ↩
²⁴⁶ Claimants' Statement of Costs, paras. 2, 17-20; Claimants' Letter, dated 10 February 2023. ↩
²⁴⁷ Claimants' Updated Statement of Costs, para. 17. ↩
²⁴⁸ Claimants' Updated Statement of Costs, para. 17. ↩
²⁴⁹ Claimants' Letter dated 10 February 2023. ↩

[Page 57]

Reconsideration was based on the evidence submitted in violation of the agreed procedural rules, as recognized by the Tribunal.²⁵⁰

207. The Claimants' costs incurred in connection with these proceedings consist of the following categories:

1. Legal costs and related disbursements totaling EUR 2,676,236.76:
   • Allen & Overy's legal fees of EUR 2,345,483.49 + EUR 15,000.00²⁵¹;
   • Dikeos Abogados' legal fees of EUR 25,531.50;
   • Helm Legal Services' legal fees of EUR 116,327.30;
   • Stibbe's legal fees of EUR 33,072.00;
   • De Breij Evers Boon's legal fees of EUR 12,910.00;
   • Disbursements invoiced through Allen & Overy LLP of EUR 127,912.47;
2. Expert fees and disbursements totaling EUR 1,063,619.41;
3. Travel and accommodation expenses directly incurred by the Claimants for hearing and pre-hearing meetings totaling EUR 13,624.68;
4. Claimants' payments to ICSID totaling EUR 784,231.96²⁵².

208. Finally, the Claimants submit that they should not be liable for any of the Respondent's costs.²⁵³

---

²⁵⁰ Claimants' Letter dated 10 February 2023. ↩
²⁵¹ Incurred in responding to Spain's Second Request for Reconsideration. See Claimants' Letter, dated 10 February 2023. ↩
²⁵² USD 875,000. See Claimants' Updated Statement of Costs, Appendix 1. ↩
²⁵³ Claimants' Updated Statement of Costs, para. 19. ↩

[Page 58]

2. The Respondent

209. The Respondent requests that the Tribunal order pursuant to Article 61(2) of the ICSID Convention that the Claimants bear the costs of this arbitration, as well as the Respondent's costs for legal representation in the amount of EUR 1,925,061 plus a compounded reasonable rate of interest from the date on which these costs are incurred until the date of their actual payment.²⁵⁴

210. The Respondent's costs incurred in connection with these proceedings consist of the following categories:

1. Legal fees totaling EUR 315,250²⁵⁵;
2. Expert fees totaling EUR 744,150;
3. Translations, courier services, editing services, and travelling expenses totaling EUR 81,258.17;
4. Respondent's payments to ICSID totaling EUR 784,375.83.

211. In support of its position that all the costs should be borne by the Claimants, Spain alleges that tribunals typically take into account the extent to which a party has succeeded on its claims and arguments.²⁵⁶ The Respondent further argues that it has shown in these proceedings that it has respected its obligations under Article 10 of the ECT and therefore Spain “should have never been charged with the burden and the costs of defending itself through this arbitration proceeding”.²⁵⁷

212. As regards the Second Request for Reconsideration, the Respondent argues that the Tribunal shall exercise its broad discretion to make an award in Spain's favor and in any case, the Respondent should never be ordered to bear the Claimants' arbitration or representation costs.²⁵⁸ The Respondent argues that its Second Reconsideration

---

²⁵⁴ Respondent's Updated Statement of Costs, paras. 1, 5. ↩
²⁵⁵ No additional costs were claimed in relation to the Second Request for Reconsideration. See Submission on Allocation of Costs Relating to the Respondent's Second Request for Reconsideration, 15 February 2023, para. 6. ↩
²⁵⁶ Respondent's Updated Statement of Costs, paras. 4-5. ↩
²⁵⁷ Respondent's Updated Statement of Costs, paras. 4-5. ↩
²⁵⁸ Submission on Allocation of Costs Relating to the Respondent's Second Request for Reconsideration, 15 February 2023, para. 9. ↩

[Page 59]

Request "is solidly founded and has been prompted by two recent landmark Judgments rendered by the Swedish Courts on issues that directly relate to jurisdictional matters discussed in the present arbitration".²⁵⁹ Furthermore, Spain argues that its conduct in relation to the filing of the Second Request for Reconsideration has been adequate, as it submitted said Second Request “as soon as possible, just a few days after the rendering of the Swedish Judgements.”²⁶⁰

B. The Tribunal's Analysis

213. The costs of these proceedings are as follows:

Dr. Raëd M. Fathallah	USD 468,538.50
Prof. Attila Tanzi	USD 243,458.93
Prof. Peter D. Cameron	USD 259,500.00
Prof. Rudolf Dolzer	USD 82,074.69
ICSID's administrative fees	USD 284,000.00
Other direct expenses (estimated)261	USD 144,011.80
TOTAL	USD1,481,583.92

214. These costs have been paid out of the advances made by the Parties. The remaining balance shall be reimbursed by ICSID to the Parties in the same proportions as each Party paid its advances to ICSID.

215. The rules governing arbitral costs are set out in Article 61(2) of the ICSID Convention:

In the case of arbitration proceedings the Tribunal shall, except as the parties otherwise agree, assess the expenses incurred by the parties in connection with

---

²⁵⁹ Submission on Allocation of Costs Relating to the Respondent's Second Request for Reconsideration, 15 February 2023, para. 4. ↩
²⁶⁰ Submission on Allocation of Costs Relating to the Respondent's Second Request for Reconsideration, 15 February 2023, para. 5. ↩
²⁶¹ This amount includes expenses related to meetings, stenographic and translation services. It excludes expenses related with courier services of this Award (courier, printing, among others). ↩

[Page 60]

the proceedings, and shall decide how and by whom those expenses, the fees and expenses of the members of the Tribunal and the charges for the use of the facilities of the Centre shall be paid. Such decision shall form part of the award.

216. In accordance with the above provision and as agreed by the Parties,²⁶² the Tribunal has broad discretion in relation to the allocation of arbitral costs. This is unchanged by the ECT, which is silent on the matter.

217. Rule 28 of the 2006 ICSID Arbitration Rules (which apply to the present arbitration) equally confirms the Tribunal's broad discretion:

Rule 28 Cost of Proceeding

1. Without prejudice to the final decision on the payment of the cost of the proceeding, the Tribunal may, unless otherwise agreed by the parties, decide:

1. at any stage of the proceeding, the portion which each party shall pay, pursuant to Administrative and Financial Regulation 14, of the fees and expenses of the Tribunal and the charges for the use of the facilities of the Centre;
2. with respect to any part of the proceeding, that the related costs (as determined by the Secretary-General) shall be borne entirely or in a particular share by one of the parties.

2. Promptly after the closure of the proceeding, each party shall submit to the Tribunal a statement of costs reasonably incurred or borne by it in the proceeding and the Secretary-General shall submit to the Tribunal an account of all amounts paid by each party to the Centre and of all costs incurred by the Centre for the proceeding. The Tribunal may, before the award has been rendered, request the parties and the Secretary-General to provide additional information concerning the cost of the proceeding.

218. Although the applicable procedural framework does not provide much guidance on allocating costs, ICSID tribunals often take into account such factors as the outcome of the proceedings, the conduct of the parties as well as the reasonableness of the costs claimed. Both Parties have invoked the outcome of the proceedings as one of the relevant factors²⁶³ whereas the Respondent has also mentioned the criterion of reasonableness.²⁶⁴

---

²⁶² See Claimants' Updated Statement of Costs, para. 3; Respondent's Updated Statement of Costs, para. 2. ↩
²⁶³ Claimants' Updated Statement of Costs, para. 17; Respondent's Updated Statement of Costs, para. 4. ↩
²⁶⁴ Respondent's Updated Statement of Costs, para. 7. ↩

[Page 61]

219. As regards the outcome of this arbitration, it shall be recalled that the Tribunal asserted its jurisdiction over these proceedings notwithstanding the Respondent's Intra-EU Objection and found Spain to be in breach of its obligations under Article 10(1) of the ECT, but only to the extent that the New Regime was unlawfully retroactive and potentially capable of lowering the support guaranteed to the Claimants' PV plants below 7% post-tax (i.e. below the reasonable level). At the same time, the total amount of damages awarded by the Tribunal is EUR 6,756,894, whereas the total damages initially claimed by the Claimants amounted to EUR 38.5 million. Spain has also succeeded on its jurisdictional objection regarding the ECT's taxation carve-out as applied to the TVPEE (i.e. the 7% tax on the value of the production of electricity established by Law 15/2012) as well as on the claims concerning an alleged lack of transparency and a violation of the umbrella clause.²⁶⁵ The Tribunal thus concludes that both Parties had partial success in pursuing their respective claims.

220. Furthermore, the Tribunal is generally satisfied with the manner in which these proceedings have been conducted by the Parties, save for one instance – the Respondent's Second Request for Reconsideration.²⁶⁶ The fact that it was ultimately rejected by the Tribunal in and of itself does not require that either arbitration or representation costs be borne by the Respondent. It is rather the manner in which the Second Request for Reconsideration has been presented by Spain that leads the Tribunal to the conclusion that the Respondent shall bear all the costs incurred by the Claimants as a result of this additional procedural development. In particular, the Tribunal finds it difficult to understand why Spain has not attempted to engage with the findings made by the Tribunal in its Decision on the First Request for Reconsideration, including the inapplicability and irrelevance of the Swedish law governing arbitration proceeding seated in Sweden, an EU Member State, to these proceedings.²⁶⁷ The absence of any substantiation in relation to this critical issue (or any mentioning of any other new facts or considerations pertaining to the Intra-EU Objection) leads the Tribunal to question the seriousness of the Respondent's Second

---

²⁶⁵ Decision, paras. 943, 958. ↩
²⁶⁶ See para. 23 above. ↩
²⁶⁷ Decision on the First Request for Reconsideration, para. 30. ↩

[Page 62]

Request for Reconsideration as well as its own confidence about the likelihood of succeeding at having the Tribunal's findings on jurisdiction revisited as a result thereof. In the Tribunal's view, the substance of the Respondent's Second Request for Reconsideration was more akin to a simple request to introduce additional documents (the Swedish Judgments) into the record accompanied by a detailed description of the content of these documents. Requests for reconsideration shall not be used for such purposes due to the obvious procedural inefficiencies and delays they entail. Overall, the Respondent's procedural conduct increased the duration and costs of these proceedings.

221. In light of the above findings, the Tribunal decides that it is fair to order that each Party shall bear its legal representation costs and that the Parties shall share in equal parts the costs of these proceedings, except for the representation and arbitration costs incurred as a result of the Second Request for Reconsideration, which shall be borne by the Respondent alone. These costs include the Claimants' representation costs of EUR 15,000, as well as the Tribunal's and ICSID's costs of USD 15,000.

[Page 63]

VII. AWARD

222. For these reasons, the Tribunal unanimously decides as follows:

1. The Respondent shall pay to the Claimants a sum of EUR 6,756,894 as compensation for the damages resulting from its violations of the ECT as determined above and in the Tribunal's Decision;
2. The Respondent shall pay interest on the sum awarded above from 20 June 2014 to the date of payment in full of all sums due pursuant to this Award at a rate of EURIBOR plus 1 %, compounded semi-annually;
3. The costs of these proceedings shall be shared equally between the Parties, except for the costs incurred as a result of the Respondent's Second Request for Reconsideration in the amount of USD 15,000 that shall be borne by the Respondent;
4. Each Party shall bear its own legal representation costs, except for the Claimants' legal costs incurred as a result of the Respondent's Second Request for Reconsideration in the amount of EUR 15,000 that shall be borne by the Respondent;
5. All other claims and requests are dismissed.

[Page 64]

Signature Professor Peter D. Cameron Arbitrator Date: 22 May 2023

Professor Attila Tanzi Arbitrator

Dr. Raëd M. Fathallah
President of the Tribunal

[Page 65]

Professor Peter D. Cameron Arbitrator

Signature Professor Attila Tanzi Arbitrator Date: 22 May 2023

Dr. Raëd M. Fathallah
President of the Tribunal

[Page 66]

Professor Peter D. Cameron Arbitrator

Professor Attila Tanzi Arbitrator

Signature

Dr. Raëd M. Fathallah
President of the Tribunal
Date: 22 May 2023