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

ICSID Case No. ARB/20/46

Lupaka Gold Corp.

CLAIMANT

v.

Republic of Peru

RESPONDENT

SECOND QUANTUM EXPERT REPORT OF
ISABEL SANTOS KUNSMAN AND ALEXANDER LEE

AlixPartners

WASHINGTON DC

25 JANUARY 2023

[Page i]

[Page iv]

Glossary

[Page v]

[Page vi]

[Page 1]

I. Scope of Work and Qualifications

1. As discussed in our previous report dated 24 March 2022 (“First Report”), the Republic of Peru (“Respondent” or “Peru"), through Arnold & Porter (“Counsel”), has asked us to offer our independent expert opinion in relation to the arbitral proceeding (“Arbitration”) that Lupaka Gold Corporation (“Claimant” or “Lupaka") initiated against Peru, under the Free Trade Agreement between Canada and Peru (“FTA”).¹

2. The investment at issue in this arbitration (“Investment”) is Claimant's 100% shareholding in the Invicta gold mine project located in Peru's Huaura province (“Invicta Project” or “Project”). Lupaka held its shareholding in the Project through its wholly owned Canadian subsidiary, Andean American Gold Corp. (“AAG"), and AAG's 99.99% owned Peruvian subsidiary, Invicta Mining Corp. S.A.C. (“IMC").²

3. Claimant claims that alleged acts and omissions ("Measures”) should be considered breaches by Peru of the FTA.³

4. Claimant retained Mr. Erik van Duijvenvoorde and Mr. Edmond Richards of Accuracy (“Accuracy") to quantify the alleged damages to Claimant's Investment resulting from the Measures.⁴ Accuracy presented their opinions in their report dated 1 October 2021 (“Accuracy First Report").

5. In our First Report appended to Respondent's Counter-Memorial, we reviewed the Accuracy First Report and submitted our opinions on whether and to what extent their calculations correctly measure Claimant's alleged damages (measured as the

¹ Our use of defined terms in this report is consistent with those included in the First Report, unless otherwise stated. ↩

² Claimant's Request for Arbitration dated 21 October 2020, ¶ 4; Claimant's Memorial, ¶¶ 22-26. ↩

³ Claimant's Memorial, ¶¶ 266, 312. Consistent with our First Report ¶¶ 92-94, we define the "Measures" as the "[a]lleged acts and omissions the Claimant claims amount to a breach of the FTA, comprising the Authorities' Measures and the Parán Community Measures." Furthermore, the "Authorities' Measures" are defined as the "[a]lleged acts and omissions committed by the Peruvian regional and central state authorities (excluding acts and omissions by Parán Community members allegedly attributable to Peru)", while the "Parán Community Measures" are defined as the "[a]lleged acts and omissions committed by the Parán Community". We understand that in relation to Parán Community Measures, Claimant's position is that the Parán Community's Access Road Protest amounted to a direct expropriation of Lupaka's investment in breach of Peru's obligations under the FTA. We understand that in relation to the Authorities' Measures, Claimant's position is that the following seven alleged acts and omissions amounted to an indirect expropriation of Lupaka's investment in breach of Peru's obligations under the FTA: a) The failure of the police and other state authorities to prevent the occupation of the Project by Parán Community members in June 2018; b) The failure of the police and other state authorities to end the occupation and prevent Parán Community members from damaging Lupaka's property and abusing Lupaka's personnel in June 2018; c) The failure by the police and other state authorities to sanction Parán Community members for the acts on 19 June 2018; d) The failure of the police and other state authorities to prevent the occupation of the Project and the Access Road Protest of the Lacsanga road by Parán Community members on 14 October 2018; e) The ongoing failure (since 14 October 2018) by the police and other state authorities to remove the Access Road Protest, notwithstanding its numerous complaints; f) The failure by the police and other state authorities to sanction Parán Community members for abuse of Claimant's representatives, including on 14 October 2018, 20 March 2019 and 14 May 2019; and, g) The state authorities' alleged support of Parán Community members' actions during the negotiations with Parán Community members following the Access Road Protest. ↩

⁴ Consistent with the First Report, we define the Investment as the "Claimant's 100% shareholding in the Invicta Project". ↩

[Page 2]

fair market value (“FMV") of Claimant's Investment offset by costs Claimant was required to remit to its lender PLI Huaura Holdings L.P. (“PLI") to maintain its Investment because of the Measures).

6. This report is in response to Accuracy's second expert report dated 21 September 2022 ("Accuracy Second Report") which was appended to Claimant's Reply dated 23 September 2022 (“Claimant's Reply”). Claimant also submitted an expert report by Mr. Christopher Jacobs of Micon International Limited (“Micon") dated 21 September 2022 (“Micon Report") containing opinions in relation to the planning, production, and operation of the Invicta Project. This report responding to the Accuracy Second Report and the Micon Report (“Second Report") will be appended to Respondent's Rejoinder Memorial dated 25 January 2023 (“Respondent's Rejoinder").

7. Nothing in our conclusions or opinions in this report is intended to address the Parties' respective legal arguments. This report does not contain any opinions on matters of law that would require legal expertise.

8. We are independent from the Parties, their legal advisors, and the Tribunal. The opinions and analyses contained in this report are ours and represent our considered views in light of our education and experience.

9. This report conforms with the Practice Standards of the Chartered Business Valuators Institute. The relevant Practice Standards for this report are those governing the preparation of expert reports (Practice Standards 310, 320, and 330).⁵

II. Executive Summary

10. In the Accuracy Second Report, Accuracy maintained their Primary Approach, while they adopted Micon's inputs and revised their damages conclusions. Under their Primary Approach, Accuracy calculated alleged damages as the difference between the But-For Scenario and the Actual Scenario using two different production scenarios: the 590t/day scenario and the 355t/day scenario, both of which are offset by the amounts Lupaka would have needed to pay to PLI to maintain its ownership of IMC. Their approach changed in the following principal ways:

a) Accuracy previously assumed that Claimant would default on the PLI Loan Agreement and would be required to pay PLI a settlement payment of US$15.9

⁵ The Practice Standards can be found at https://cbvinstitute.com/members-students/standards-ethics/. ↩

[Page 3]

million to maintain ownership and control of its shares of IMC. In the Accuracy Second Report, Accuracy adopted Micon's revised production schedule with an earlier start date of commercial mining operations allowing Claimant to meet its gold delivery obligations. Therefore, Accuracy modeled financing cash flows (i.e., periodic installment payments by Claimant, rather than one settlement payment) under both production scenarios (590t/day and 355t/day).

b) Accuracy revised their production schedule and cost assumptions under both production scenarios based on inputs received from Micon.

c) Accuracy decreased the discount rate in the 590t/day scenario by reducing their pre-production premium from 6.9% to 3.3%.

d) Accuracy increased the proposed pre-award interest rate from LIBOR+2% to LIBOR+4% or UST+5%.

11. In addition to updating their Primary Approach, Accuracy also revised the analysis of their proposed alternative indicators of value by incorporating some of our comments from our First Report.

12. Accuracy's updated analysis remains an "all or nothing" approach that does not provide the Tribunal with a basis to estimate damages if it determines that some Measures were not FTA breaches, while others were. We also note that Claimant's Reply has continued to rely upon the 590t/day scenario as the only basis for Claimant's damages claim.

A. Accuracy's Quantification of Damages Remains Fundamentally Flawed

13. In our First Report, we raised four fundamental flaws with Accuracy's damages conclusions. The Accuracy Second Report continued to ignore these fundamental flaws. In short, the four fundamental flaws in Accuracy's conclusions are the following:

a) Ongoing social conflict and protests. Accuracy did not directly respond to our point in our First Report that Accuracy's calculations did not take into account that police intervention would not have permanently resolved the conflicts between Claimant and the Parán Community. Accuracy's updated damages were based on the premise that intervention by the police would have been immediately and permanently resolved the Access Road Protest and the broader conflict at no additional cost or delay as of the Valuation Date. Because Accuracy's damages model omitted these likely costs and delays, it did not produce a reliable estimate of damages caused by Respondent's alleged breaches.

[Page 4]

b) SLOs from local communities. Accuracy did not directly respond to the social license risk issues raised in our First Report with respect to the Parán, Lacsanga, Santo Domingo, and Mallay communities and instead dismissed them as “not relevant for our assessment of damages under Claimant's case". In the Accuracy Second Report, Accuracy accelerated their mining production schedule to begin in November 2018 without providing evidence or support that the social license issues would be resolved immediately. Accuracy also did not provide any information as to why the Mallay Community's approval of the Mallay Plant transaction was delayed or if specific social license issues could further impact future dealings between Claimant and the Mallay Community.

As we have demonstrated, obtaining and maintaining the SLO is a critical step for mining projects. Claimant's inability to do so would impair its ability to deliver gold, meet its obligations under the PLI Loan Agreement, and may have further negatively impacted its operations. Not obtaining an SLO is a recognized cause of total project failure that would reduce damages to nil. In any case, the risk associated with obtaining and maintaining SLOs needs to be accounted for and the absence of such accounting is a fundamental flaw in Accuracy's calculations.

c) Defaults on the PLI Loan Agreement Not Caused by Peru. Under the 355t/day scenario, based on the Micon Report, Accuracy accelerated the Project's mining production start date from August 2019 to November 2018 by assuming Claimant would leverage third-party ore processing. Based on this revised schedule, Accuracy also assumed that Claimant would have avoided default under the PLI Loan Agreement. Micon provided no specific basis for their assumption that the failures at the third-party processing facilities would be resolved. Furthermore, we understand from Counsel that Ms. Dufour has concluded that Micon's revised production schedule is not feasible due to missing approvals for both the Invicta Project and for the ore processing plants, which would have prevented production from beginning before July 2020. We also understand from Counsel that Claimant had committed 14 defaults under the PLI Loan Agreement, some of which Claimant has acknowledged were unrelated to any acts or omissions by Peru or by the Parán Community, and that each one of those defaults entitled PLI to foreclose on Claimant's Investment. Accuracy did not account for those contract violations to isolate which damages, if any, should be calculated as the damages caused by Peru. The lack of an accounting for loan defaults not caused by Peru remains a fundamental flaw with Accuracy's analysis that would have resulted in Claimant's damages being reduced to nil.

[Page 5]

Under the 590t/day scenario, Accuracy also assumed that Claimant's lender PLI, controlled by its former parent company Pandion, would have conceded payment extensions to Claimant and lent it more funds for a purchase of the Mallay Plant. This assumption disregarded Claimant's failure to obtain and maintain its SLO, its inability to meet its obligations under the PLI Loan Agreement due to the technical issues at the third-party processing facilities, and its lack of required permits and approvals before officially commencing commercial-scale production, and the events of default on the PLI Loan Agreement committed by Claimant that were unrelated to Peru's conduct. The multiple obstacles to producing marketable minerals discussed above lead us to conclude that it would have been irrational for PLI to provide further extensions and additional funding to Claimant and that this remains a fundamental flaw with Accuracy's analysis that would have resulted in Claimant's damages being reduced to nil.

d) Refinancing Risk. Accuracy assumed that, due to the revised production schedule provided by Micon (which shifted the start date of commercial mining to November 2018, instead of August 2019), Claimant would be able to meet its obligations under the PLI Loan Agreement and thus would not default on the loan and need financing to pay a settlement to PLI. Additionally, under the 590t/day scenario, Accuracy assumed that Claimant could obtain from PLI an additional loan of US$13.0 million for a potential purchase of the Mallay Plant and would thus not need an alternative lender. Due to PLI's continued involvement as Claimant's lender, Accuracy assumed that the Invicta Project faced no refinancing risk. However, this assumption is unjustified as, absent the Measures, Claimant still was likely to default on its obligations under the PLI Loan Agreement (and we understand Claimant indeed did so due to defaults unrelated to conduct by Peru) and Claimant therefore would be required to make a settlement payment to maintain ownership of its shares in IMC. Neither Claimant nor Accuracy have presented any evidence about how the refinancing risk for payment of such settlement to PLI would have been mitigated or avoided in the event of default. Although Mr. Gordon Ellis stated that Claimant raised financing five months before the Valuation Date, his testimony indicated that Claimant's attempt to raise US$1.0 million was undersubscribed and yielded only US$0.7 million, less than 5% of the US$15.9 million that Claimant would need to settle its default on the PLI Loan Agreement.

[Page 6]

B. Accuracy's Damages Under the 590t/day Scenario are Overstated

14. As discussed previously, Accuracy did not adequately address the fundamental flaws. Notwithstanding the fundamental flaws, we assess Accuracy's revised 590t/day scenario including the new and remaining defects in it. To recall, Accuracy's 590t/day scenario assumed that Claimant would have acquired the Mallay Plant, and that Claimant's ownership of this plant would have enabled it to increase its ore production from 355t/day to 590t/day.

15. In its second report, Accuracy assumed that the Project would have started commercial production in November 2018 and that Claimant would have obtained additional funding from PLI when it closed the Mallay Plant transaction in March 2019. However, Accuracy's damages conclusions remain flawed because they:

a) Assumed that the failures at the various third-party processing facilities, that made them unsuitable to process ore from the Invicta Project, would be resolved immediately before the commencement of production without additional cost or time to account for production prior to the acquisition of the Mallay Plant. Although Accuracy, Micon, and Claimant provided insufficient information on how they would resolve the third-party processing problems to enable an estimate of such costs, we note that those additional costs incurred would have the effect of reducing the damages under the 590t/day scenario. Additionally, any delay in obtaining ore processing capacity could lead to a default under the PLI Loan Agreement;

b) Adopted Micon's unrealistic project start date (starting in November 2018 rather than August 2019) that does not consider the time and process required to acquire necessary permits and approvals to begin exploitation of the Project. We understand from Counsel that Respondent's expert in Peruvian mining law, Ms. Miyanou Dufour von Gordon has opined that Claimant could not have begun commercial exploitation of the Invicta Project's resources prior to July 2020;

c) Adopted Micon's unsupported and unauthorized production schedule which extended the life of the Project from seven to ten years;

d) Did not provide documentary evidence or analysis that would support Micon's rationalizations about why the grade of actual gold produced by the Invicta Project was lower than the budgeted grade (and subsequently in Micon's revised production schedule);

[Page 7]

e) Did not consider various additional costs that would be necessary to proceed with production including incremental costs related to obtaining and maintaining the Project's SLO from the Parán Community; and,

f) Understated the discount rate as it did not consider the fundamental flaws above and used a pre-production risk premium that Accuracy revised downward based on only Micon's assurances and one source document that does not support the discount rate that Accuracy selected.

16. Accounting for the above, we have reduced Claimant's maximum damages from US$41.0 million to US$20.5 million, before pre-award interest. We also provided a sensitivity analysis should the Tribunal find it appropriate to increase the discount rate to address the additional risks or costs we have identified.

C. Accuracy's Damages Under the 355t/day Scenario are Overstated

17. As discussed previously, Accuracy did not adequately address the fundamental flaws. Notwithstanding the fundamental flaws, we assess Accuracy's revised 355t/day scenario including the new and remaining defects in it. To recall, the 355t/day scenario was based on Claimant's production plans that did not assume that Claimant would acquire the Mallay Plant.

18. Under the 355t/day scenario, Accuracy adopted Micon's updated production schedule and assumed the failures by the various third-party processors would be resolved immediately before the commencement of production. However, Accuracy and Micon did not explain how these issues would have been resolved or quantify the time and cost to do so. A delay in obtaining ore processing capacity could lead to a default under the PLI Loan Agreement.

19. In addition, Accuracy's 355/day scenario continued to overstate damages because they:

a) Adopted Micon's unrealistic project start date (starting in November 2018 rather than August 2019) that does not consider the time and process required to acquire necessary permits and approvals to begin commercial exploitation of the Project. We understand from Counsel that Ms. Dufour has concluded that Claimant could not have begun exploitation of the Invicta Project's resources prior to July 2020;

b) Did not provide documentary evidence or analysis that would support Micon's rationalizations about why the grade of actual gold produced by the Invicta

[Page 8]

Project was lower than the budgeted grade (and subsequently in Micon's revised production schedule);

c) Did not consider various additional costs that would be necessary to proceed with commercial production including incremental costs related to obtaining and maintaining the Project's SLO from the Parán Community and other affected local communities; and,

d) Understated the discount rate since it did not consider the fundamental flaws discussed above.

20. Accounting for the above, we have reduced Claimant's maximum damages from US$32.1 million to US$22.5 million, before pre-award interest. We also provided a sensitivity analysis should the Tribunal find it appropriate to increase the discount rate to address the additional risks or costs we have identified.

D. Using Alternative Pre-Award Interest Rates Reduces Damages

21. In the Accuracy Second Report, Accuracy revised their selected pre-award interest rate from LIBOR+2% to LIBOR+4% or UST+5%. The total pre-award interest calculated under both damages scenarios increased compared to that in the Accuracy First Report, reflecting Accuracy's higher interest rates and a longer accrual period.

22. We recognize there is no consensus regarding the exact appropriate interest rate to calculate pre-award interest. Historically, the most commonly used variable interest rates were based on LIBOR and UST, with a premium of 2%. As LIBOR is being retired, we provide our calculation of the pre-award interest using SOFR+2% and UST+2%.

23. The pre-award interest would be zero if the damages are reduced to nil due to the fundamental flaws. Our conclusion on damages can be summarized as follows:

Figure 1 – Damages Summary Table

[Page 9]

E. There is no Inconsistency between Our Conclusions and the Independent Appraiser's Findings

24. In our First Report, we referred to an independent appraisal performed by PwC that assessed the value of the shares of IMC at US$13.0 million at the time of foreclosure. This amount was the basis of an alternative FMV of the IMC shares in the Actual Scenario (as opposed to Accuracy's assumption that the shares would be valued at nil).

25. Accuracy disagreed with this perspective and maintained that the shares should be valued at nil because Claimant's “economic position" would still be nil net of the settlement payment that it would have made to PLI. However, as noted in this report, Claimant's obligations under the PLI Loan Agreement, or other financing arrangements, do not directly impact the FMV of IMC. A result where Claimant, net of its own debts, would be left with no damages does not contradict the fact that the shares of IMC still had some value (i.e., US$13.0 million) that could be realized by a notional buyer.

26. They also alleged that the appraiser's findings conflict with our view with respect to the fundamental flaws, as we believed that the fundamental flaws could reduce the FMV to nil. We disagree that there is an inconsistency here. Our conclusion is that if the fundamental flaws remained unaddressed or unresolved, the value of the Project would be nil. The independent appraiser may have reached a different conclusion because a notional investor could potentially have addressed those issues more effectively than Claimant did and resolved those issues.

27. Finally, Accuracy contended that regardless, Claimant allegedly had no access to the Invicta Project and therefore the FMV under the Actual Scenario with respect to damages should be nil. However, even if Claimant could not access the Invicta Project before the foreclosure of the PLI Loan Agreement, the Project would still have value to market participants.

28. In conclusion, without the benefit of reviewing the actual independent appraiser's report, it is not possible to reasonably compare our valuation conclusions with those reached by PwC. However, we do not believe there is an inconsistency between our conclusions and those reached by the independent appraiser.

F. Accuracy's Other Indicators of Value are not Relevant

29. Accuracy responded to comments from our First Report and revised their other indicators of value. However, these indicators of value do not demonstrate that

[Page 10]

Accuracy's damages conclusions under the 590t/day and 355t/day scenarios are reasonable:

a) Claimant's market capitalization is an inappropriate benchmark as Accuracy included unjustified adjustments that inflated the value. Correcting Accuracy's calculation reduces Claimant's market capitalization as of the Valuation Date below the range of Accuracy's conclusions and demonstrate that Accuracy's damages are overstated.

b) Claimant's sunk costs is an inappropriate benchmark as sunk costs are not a recommended methodology for valuing a Development Project under CIMVAL. This approach also does not consider the fundamental issues discussed above and does not represent the amount that a willing buyer would pay as of the Valuation Date. Finally, Accuracy's interest rate on the sunk costs inflated their FMV conclusion to an unreasonable degree. Applying an alternative rate based on the pre-award interest rates discussed in this report demonstrate that Accuracy's damages are overstated.

c) The transaction multiples selected by Accuracy are inappropriate benchmarks because they included transactions with multiple producing projects and projects that were located outside Latin America. Selecting a more comparable subset results in lower multiples than the multiples implied by Accuracy's damages conclusions. However, simply adjusting the scope of these transaction multiples does not result in a more reasonable proxy since this analysis does not consider the fundamental flaws discussed above.

d) The NPVs in the SRK and Red Cloud Models are both inappropriate benchmarks as both models do not account for the fundamental flaws discussed above.

III. Summary of Accuracy's Updated Damages Calculation

30. In the Accuracy Second Report, Accuracy continued to apply the discounted cash flow (“DCF”) methodology to calculate the FMV of the Invicta Project (“Primary Approach”). Using the Primary Approach, Accuracy calculated the alleged damages as the difference in i) the alleged FMV of Claimant's Investment absent the impact of the Measures (“But-For Scenario") and ii) nil, the alleged FMV of Claimant's Investment impacted by the Measures ("Actual Scenario"). Using their Primary Approach, Accuracy calculated the value of Claimant's Investment in the But-For Scenario for two different production scenarios: the 590t/day scenario and the

[Page 11]

355t/day scenario, both of which are offset by the amounts Lupaka paid to PLI to maintain its ownership of IMC.⁶

31. However, in the Accuracy Second Report, Accuracy changed their damages calculations and underlying assumptions from the Accuracy First Report in four respects as discussed below.

32. First, Accuracy changed the way that they modelled the offsetting cash outflows that were required for Claimant to meet its obligations under the PLI Loan Agreement. In the Accuracy First Report, Accuracy assumed that Claimant in the But-for Scenario, to prevent foreclosure under the PLI Loan Agreement, would make a settlement payment to PLI of US$15.9 million as of 26 August 2019 (“Valuation Date").⁷

33. In the Accuracy Second Report, however, Accuracy adopted Micon's revised production schedule with an earlier start date of operations. In doing so, Accuracy assumed that Claimant would have been able to meet its gold delivery obligations under the PLI Loan Agreement and thus avoid making a settlement payment of US$15.9 million to PLI on the Valuation Date. In addition, Accuracy modeled the financing cash flows required to operate the mine under both production scenarios and to acquire the Mallay processing plant ("Mallay Plant") in the 590t/day scenario. This change reduced damages by US$12.1 million⁸ and US$12.0 million⁹ for the 590t/day and 355t/day scenarios respectively, on a standalone basis before the pre-award interest.

34. Figure 2 compares Accuracy's quantification of damages for both production scenarios between the two reports.¹⁰

⁶ Accuracy Second Report, ¶ 5.19. ↩

⁷ Accuracy First Report, ¶¶ 7.7 – 7.8. ↩

⁸ Accuracy Second Report, Appendix 3, "Dashboard (ACC)", Cell E59. (1) In "Dashboard (ACC)", set the scenario to "First Accuracy Report" (2) Under "Live inputs", set "Financing cash flows and discount rate" to "Second Accuracy Report" (3) In "Summary", Cell D4, keep the discount of 14.7% under the "First Accuracy Report" scenario. ↩

⁹ Accuracy Second Report, Appendix 4, "Dashboard (ACC)", Cell E59. (1) In "Dashboard (ACC)", set the scenario to "First Accuracy Report" (2) Under "Live inputs", set "Financing cash flows and discount rate" to "Second Accuracy Report" (3) In "Summary", Cell D4, keep the discount of 11.1% under the "First Accuracy Report" scenario. ↩

¹⁰ Accuracy Second Report, Table 6.6 and Table 6.11. ↩

[Page 12]

Figure 2 – Summary of Accuracy's Damage Assessments

35. Second, Accuracy revised the production and cost assumptions of both production scenarios. Notable changes include:¹²

a) For the 590t/day scenario, Accuracy incorporated Micon's extended production schedule of ten years from the previous seven years;

b) For both the 590t/day and 355t/day scenarios, Accuracy incorporated Micon's updated capital expenditure estimates; and,

c) For both scenarios, Accuracy adjusted, in some cases based on Micon's estimates, the Project's operating costs (such as employee profit sharing, G&A costs, direct unit cost of treating ore, etc.).

36. These changes reduced damages under the 590t/day scenario by US$5.2 million¹³ and increased damages under the 355t/day scenario by US$2.5 million¹⁴, on a standalone basis before the pre-award interest.

37. Third, Accuracy decreased the discount rate in the 590t/day scenario. Specifically, they reduced the “pre-production premium”¹⁵ factored in the discount rate in the 590t/day scenario from 6.9% in the Accuracy First Report to 3.3% in the Accuracy Second Report, to reflect the alleged “comparable level of confidence in both scenarios following Micon's updates".¹⁶ This change increased damages under the

¹¹ The difference in the FMV before financing is due to Accuracy's adjustments in the Accuracy Second Report. ↩

¹² Accuracy Second Report, ¶¶ 6.7, 6.19, 6.28, 6.51, 6.55. ↩

¹³ Accuracy Second Report, Appendix 3, "Dashboard (ACC)", Cell E59. (1) In "Dashboard (ACC)", set the scenario to "First Accuracy Report" (2) Under "Live inputs", set "Updates for the Micon Report" and "Accuracy's modelling updates" to "Second Accuracy Report". ↩

¹⁴ Accuracy Second Report, Appendix 4, "Dashboard (ACC)", Cell E57. (1) In "Dashboard (ACC)", set the scenario to "First Accuracy Report" (2) Under "Live inputs", set "Updates for the Micon Report" and "Accuracy's modelling updates" to "Second Accuracy Report". ↩

¹⁵ Risk premium added by Accuracy to account for the uncertainty with the cash flows of a pre-production project based on PEA mine plan. ↩

¹⁶ Accuracy Second Report, ¶ 2.11. ↩

[Page 13]

590t/day scenario by US$9.0 million, on a standalone basis before the pre-award interest.¹⁷

38. Fourth, Accuracy increased the proposed pre-award interest rate from LIBOR+2% in the Accuracy First Report to LIBOR+4% or One-Year U.S. Treasury Bill (“UST”)+5% in the Accuracy Second Report.¹⁸ For each proposed rate, Accuracy continued to use annual compounding.

39. Accuracy's revised damages estimate as of the Valuation Date for the 590t/day scenario decreased by US$6.7 million, equivalent to -14.1% as summarized in Figure 3.¹⁹ Despite reducing the amount of damages that would be accruing pre-award interest, the claim for pre-award interest more than doubled due to the higher interest rate²⁰ and the later valuation date of 31 August 2022.

Figure 3 – Accuracy's Damages for 590t/day Scenario

40. As shown in Figure 4, Accuracy's revised damages estimate as of the Valuation Date for the 355t/day scenario increased by US$3.8 million, equivalent to 13.6% compared to their previous conclusion.²¹ The claim for pre-award interest more than tripled due to the higher interest rate²² and the later valuation date of 31 August 2022.

¹⁷ Accuracy Second Report, Appendix 3, "Dashboard (ACC)", Cell E59. (1) In "Dashboard (ACC)", set the scenario to "First Accuracy Report" (2) In "Summary", update the discount rate to the value in "inputs", Cell C261 (11.1%). ↩

¹⁸ Accuracy Second Report, ¶ 2.12. ↩

¹⁹ Accuracy Second Report, Table 6.6, Table 6.13. ↩

²⁰ Using higher pre-award interest rates, Accuracy increased the pre-award interest for damages under the 590t/day scenario by US$2.7 million (LIBOR+4% vs LIBOR+2%) and US$3.5 million (UST+5% vs LIBOR+2%). Accuracy Second Report, Appendix 3, "Pre-award interest (ACC)", update Cell C12 to 2%. ↩

²¹ Accuracy Second Report, Table 6.11, Table 6.13. ↩

²² Using higher pre-award interest rates, Accuracy increased the pre-award interest for damages under the 590t/day scenario by US$2.1 million (LIBOR+4% vs LIBOR+2%) and US$2.8 million (UST+5% vs LIBOR+2%). Accuracy Second Report, Appendix 4, "Pre-award interest (ACC)", update Cell C12 to 2%. ↩

[Page 14]

Figure 4 – Accuracy's Damages for 355t/day Scenario

41. In addition to updating their Primary Approach, Accuracy also revised the analysis of their proposed alternative indicators of value by incorporating some of our comments from our First Report.

42. Accuracy's updated analysis is still an “all or nothing" approach. Accuracy did not separate and analyze the impact of each individual Measure that Claimant alleges was an FTA breach, nor did Accuracy determine a loss corresponding to each Measure. Should the Tribunal decide that some Measures were not FTA breaches, while others were, Accuracy's approach does not provide a basis to estimate damages. We also note that Claimant's Reply has continued to rely upon the 590t/day scenario as the only basis for Claimant's damages claim.²³

IV. Fundamental Flaws with Accuracy's Quantification of Damages

43. In our First Report, we raised four fundamental flaws in Accuracy's quantification of damages.²⁴ In the following subsections we review and comment on the responses to these fundamental flaws Accuracy provided.

A. Removing the Authorities' Measures Would Not Resolve the Access Road Protest

44. In our First Report, we noted:²⁵

"Intervention by the police appears to have been unlikely to permanently resolve the conflict with the Parán Community or the Access Road Protest, which therefore would have remained an obstacle to Claimant's ability to perform its obligations under the PLI Loan Agreement;"

45. Accuracy did not provide any specific response to this issue in the Accuracy Second Report. We do however note that they have dismissed any potential issues related

²³ Claimant's Reply, ¶ 1062. ↩

²⁴ First Report, ¶ 135. ↩

²⁵ First Report, ¶ 16. ↩

[Page 15]

to Claimant's Social License to Operate (“SLO") as being "not relevant for our assessment of damages under Claimant's case."²⁶

46. Accuracy's damages quantifications under both production scenarios are based on the fundamental assumption that the conflict with the Parán Community would have been immediately and permanently resolved at no additional cost or delay as of the Valuation Date. Since intervention by the police appears to have been unlikely to permanently resolve the conflict with the Parán Community or the protest set up by the Parán Community on 14 October 2018 (“Access Road Protest"),²⁷ Accuracy's damages model omitted likely costs and delays from those issues and did not produce a reliable estimate of the damages caused by Respondent's alleged breaches.

47. We discuss additional aspects of Accuracy's general position regarding the relevance of the SLO in the next section.

B. Accuracy Continued to Ignore the Social License Risk

48. In our First Report, we further noted that:²⁸

"Accuracy failed to consider and incorporate any social license risk in its valuation analysis to account for the continued conflict between IMC and the Parán Community as of the Valuation Date;"

49. As discussed above, in the Accuracy Second Report, Accuracy did not directly respond to the social license issues we raised in our First Report. Instead, Accuracy dismissed these issues as "not relevant for our assessment of damages under Claimant's case."²⁹ Contrary to Accuracy's opinion, and consistent with our position in our First Report, the social license risk the Invicta Project faced does impact Accuracy's analysis and Claimant's damages claim.

50. As we commented in our First Report, Accuracy did not provide any logic, basis or evidence to support the assumption that "Claimant's efforts towards starting production would resume immediately on the Valuation Date under both scenarios and that ore production would begin approximately 10 months after the start of the protest".³⁰ In the Accuracy Second Report, they made an even more aggressive assumption that ore production would begin in November 2018, instead of August

²⁶ Accuracy Second Report, ¶ 3.5. ↩

²⁷ First Report, ¶¶ 105-111. ↩

²⁸ First Report, ¶ 16. ↩

²⁹ Accuracy Second Report, ¶ 3.5. ↩

³⁰ First Report, ¶ 117. ↩

[Page 16]

2019 as assumed in the Accuracy First Report. No evidence was provided to support the assumption that Claimant would obtain and maintain SLOs with all of the affected local communities from November 2018 to the conclusion of Claimant's planned operation of the Invicta Project.

51. Also, as we pointed out in our First Report, and this issue remained in the Accuracy Second Report, Accuracy did not provide any information as to why the approval of the Mallay Plant transaction was delayed or if the delay was due to specific social license issues that could impact future dealings between Claimant and the Mallay Community.³¹

52. Maintaining positive relationships with local communities plays an important role for mining projects. Claimant's failure to obtain and maintain the SLO for the Invicta Project would impact Claimant's ability to extract, transport, process, and deliver gold, meet its obligations under the PLI Loan Agreement, and may have further negatively impacted its operations. Not obtaining an SLO is recognized as a cause of total project failure³² and could reduce damages to nil. Even if this was not the case, the risk associated with Claimant's missing SLO needed to be accounted for, either in the form of costs to obtain and maintain the SLO or a specific premium in the discount rate. Not doing so is a fundamental flaw in Accuracy's damages.

C. Claimant Would Have Defaulted on its Existing and Anticipated Debt Financing Obligations Absent the Measures

53. To fund the development of the Project, on 30 June 2016, Lupaka and PLI, a subsidiary of Pandion Mine Finance L.P. (“Pandion"), executed a definitive Pre-Paid Forward Gold Purchase Agreement.³³ On 2 August 2017, Lupaka and PLI entered into the Second Amended and Restated Pre-Paid Forward Gold Purchase Agreement (“PLI Loan Agreement”).³⁴ Under this agreement, Lupaka received from PLI US$7.0 million, divided in three tranches.³⁵

54. To repay the loaned principal, Lupaka agreed to sell to PLI a set amount of gold each month (“Contract Quantity”) after the grace period for each tranche for which PLI

³¹ First Report, ¶ 118. ↩

³² First Report, ¶ 120. ↩

³³ Witness Statement of Gordon Ellis dated 1 October 2021, ¶ 32. Pre-Paid Forward Gold Purchase Agreement dated 30 June 2016. [C-0044] ↩

³⁴ Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017. [AC-0004] ↩

³⁵ Lupaka presentation Invicta Mining Suite for Difference dated September 2019, p. 4. [AC-0005] ↩

[Page 17]

would pay a discounted price: the market price less US$500/oz (“Sale” or “Sell").³⁶ The total Contract Quantity Lupaka had to Sell was 22,680 oz, over the full term of the PLI Loan Agreement, as displayed in Figure 5 below.³⁷

Figure 5 – Monthly Gold Delivery Obligation Under the PLI Loan Agreement

[Chart showing a line graph with "oz." on the y-axis from 0 to 550 and dates from Nov-18 to Nov-22 on the x-axis. The line starts at 0, jumps to around 200 oz. in Dec-18, then steps up to around 325 oz. in Mar-19, then up to 500 oz. in Jun-19. It stays at 500 oz. until around Jun-22, then steps down to around 325 oz., then down to around 200 oz. before dropping to 0.]

55. Lupaka would be deemed in default of the agreement if it failed to Sell the Contract Quantity of gold according to the agreed schedule or failed to remedy the deficiency within a 15-day grace period.³⁸ Should Lupaka then be deemed to be in default, PLI was able to demand payment of the “Early Termination Amount"³⁹ or enforce against the “Collateral" (i.e., Lupaka's shares in IMC).⁴⁰ The first Sale of the Contract Quantity of gold was scheduled for December 2018, which Claimant never fulfilled. We understand that Claimant claims that this was due to the Measures.

56. In the following subsections we explain that under both production scenarios Claimant was likely to default on its debt obligations regardless of the Measures.

³⁶ Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017, pp. 3-4. [AC-0004]. "Contract Quantity" means a total of 22,680 Ounces of Gold to be sold at a discount as follows: (a) 0 Ounces of Gold for each of the 15 calendar months following the calendar month in which the Gold Prepayment Amount is paid on the First Effective Date [August 2017] and 187 Ounces of Gold for each of the 45 calendar months thereafter; (b) 0 Ounces of Gold for each of the 15 calendar months following the calendar month in which the Gold Prepayment Amount is paid on the Second Effective Date [November 2017] and 139 Ounces of Gold for each of the 45 calendar months thereafter (which, for the avoidance of doubt, shall be in addition to the 187 Ounces of Gold listed in subclause (a) hereof); (c) 0 Ounces of Gold for each of the 15 calendar months following the calendar month in which the Gold Prepayment Amount is paid on the Third Effective Date [February 2018] and 178 Ounces of Gold for each of the 45 calendar months thereafter (which, for the avoidance of doubt, shall be in addition to the 187 and 139 Ounces of Gold listed in subclauses (a) and (b), respectively, hereof); and (d) any Ounces of Gold to be delivered pursuant to Section 7(3) under this Agreement." ↩

³⁷ Accuracy First Report, ¶ 3.21. ↩

³⁸ Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017, p. 52, section 13(1)(a). [AC-0004] ↩

³⁹ Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017, p. 23, section 5(8). [AC-0004] ↩

⁴⁰ Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017, p. 56, section 14(4). [AC-0004] ↩

[Page 18]

i. Claimant's Default of Its Debt Obligations Under the 355t/day Production Scenario

57. Accuracy, based on the Micon Report, moved up the Project's production start date from August 2019 to November 2018 for the 355t/day production scenario. As a result, and as shown in Figure 6 below, according to Accuracy's assumptions Invicta would have been able to produce enough gold to comply with the Contract Quantity requirement under the PLI Loan Agreement.⁴¹

Figure 6 – 355t/day Scenario Production and Gold Delivery

[Chart showing two line graphs. The y-axis is "oz." from 0 to 3,000. The x-axis is dates from Oct-18 to Jan-25. One line, "Gold produced", starts at 0, rises sharply to around 3,000 oz. in early 2019, and stays high with some dips until late 2023, then drops to 0. The second line, "Gold to be delivered...", is a step function starting at 0, jumping to around 200 oz. in Dec-18, then to around 325 oz., then to 500 oz., where it stays until late 2022, then steps down. The "Gold produced" line is consistently above the "Gold to be delivered" line after production starts.]

58. We disagree with Accuracy's proposition that Claimant would have avoided default under the revised 355t/day production scenario because the analysis is based on the flawed assumptions that the Invicta Project could have started production in November 2018 and that the Project could have relied on third-party plants to process the quantities of ore projected under the 355t/day scenario. Specifically, the Micon Report stated:⁴²

"[I]t is Micon's opinion that the gold grade shortfalls in development ore reported by Lupaka were the result of various factors that could have been overcome by early 2019 and, taking into account the previously recorded actual performance by Lupaka, Micon considers that, but for the Blockade that prevented access to and operation of the Invicta mine, Lupaka would otherwise have been able to produce the ore tonnages and grade required to service the PLI facility, to deliver and arrange treatment of this material

⁴¹ Appendix 3, "Monthly Production". ↩

⁴² Micon Report, ¶ 125. ↩

[Page 19]

at the Huancapeti and other third-party toll-treatment plants (e.g., Coriland, Huari), and to ship the resulting concentrates to market in time to meet its obligations as set out above." (emphasis added)

59. With regards to the third-party ore processing, Micon provided no basis for their assumption that third-party ore processing would be available to Claimant as an alternative to the Mallay Plant. Micon ignored the various failures that Claimant's witness, Mr. Castañeda Mondragón, noted in his witness statement. In fact, neither Micon nor Accuracy was clear about which specific facility they assume Claimant would use for ore processing. As we noted in our First Report, Mr. Castañeda Mondragón identified several mechanical failures with the third-party plants:⁴³

"For instance, Coriland, which was the closest to the Site (besides Mallay), lacked a cyanidation treatment option in its tailing facility, which meant potentially losing recoverable gold; San Juan Evangelista also lacked a cyanidation treatment option and had piles of mineral accumulated due to processing commitments which other mining companies; and Huancapeti II needed to postpone works due to unexpected mechanical failures."

60. With regards to the November 2018 start date, we understand from Counsel that Ms. Dufour has concluded that Micon's revised production schedule is not feasible due to missing approvals both for the Invicta Project and for the ore the processing plants, and that these missing approvals would have prevented Claimant from exploiting the Invicta Project before July 2020. We understand that these approvals include:

a) Approval of modifications to the environmental impact assessment (“EIA”);

b) Licenses to use water from sources not contemplated in the 2009 EIA;

c) The Ministry of Energy and Mines of Peru (“MINEM”) authorization to begin commercial exploitation; and,

d) Authorization to purchase and store fuel.

61. Micon did not provide any explanation as to how these technical issues and missing authorizations could be resolved, much less within the three months before the Contract Quantity of gold requirement under the PLI Loan Agreement became due.⁴⁴

⁴³ Witness Statement of Julio Félix Castañeda Mondragón dated 1 October 2021, ¶ 88. ↩

⁴⁴ Micon Report, ¶ 125. ↩

[Page 20]

62. In addition, we understand from Counsel that Claimant had committed 14 defaults under the PLI Loan Agreement, several which were unrelated to any acts or omissions by Peru (or by the Parán Community), and each one entitled PLI to foreclose on Claimant's investment.

63. Therefore, the absence of an accounting for Claimant's defaults on the PLI Loan Agreement due to issues unrelated to conduct by Peru remains a fundamental flaw with Accuracy's analysis that would have resulted in Claimant's damages being reduced to nil.

ii. Claimant's Default of Its Debt Obligations Under the 590t/day Production Scenario

64. As we noted in Section III of our First Report, following Claimant's failure to achieve its production and processing targets, Claimant approached Compañía de Minas Buenaventura S.A.A. (“Buenaventura") to acquire the Mallay Plant in 2018.⁴⁵ A draft agreement dated 21 September 2018 contemplated that Claimant potentially could purchase the Mallay Plant at a price of US$10.4 million.⁴⁶ According to Claimant, a final agreement was planned to be signed on 15 October 2018.⁴⁷ The chart below compares the Contract Quantity of Gold under the Draft Third Amendment to the PLI Loan Agreement with the revised forecast of production under the 590t/day scenario from the Accuracy Second Report:⁴⁸

⁴⁵ First Report, ¶ 79. ↩

⁴⁶ Witness Statement of Gordon Ellis dated 1 October 2021, ¶ 40. Draft Mallay Purchase Agreement between Buenaventura and IMC, 21 September 2018. [C-0048] ↩

⁴⁷ Witness Statement of Gordon Ellis dated 1 October 2021, ¶ 52. Lupaka Board Meeting Minutes, 27 September 2018. [C-0051] ↩

⁴⁸ Appendix 2, "Monthly Production". ↩

[Page 21]

Figure 7 – 590t/day Scenario Production and Gold Delivery

[Chart showing three line graphs. The y-axis is "oz." from 0 to 3,500. The x-axis is dates from Jul-18 to Jan-28. One solid line, "Gold produced", starts at 0, rises sharply to around 2,800 oz. in early 2019, then rises to over 3,000 oz. and stays high with some dips until late 2027, then drops to 0. A dashed line, "Gold to be delivered, starting January 2020 assumed by Accuracy", starts at 0 and rises in steps to around 1,500 oz. A dash-dot line, "Gold to be delivered, starting September 2019 per Draft Third Amendment to the PLI Loan Agreement", also starts at 0 and rises in steps to around 1,500 oz. The "Gold produced" line is consistently above both "Gold to be delivered" lines after production starts.]

65. Accuracy's updated 590t /day scenario is based on the following three assumptions:

a) First, the Invicta Project would have started production in November 2018;

b) Second, Claimant and PLI would have executed the Draft Third Amendment to the PLI Loan Agreement, deferring Claimant's Contract Quantity of gold obligations to January 2020 based on the maximum allowed grace period under the contract; and,

c) Third, the Invicta Project could have relied on third-party processors for the period from November 2018 until at least March 2019 when the Mallay Plant was contemplated to come online.

66. In our First Report, we showed that under the 590t/day scenario Lupaka would have also likely defaulted on the PLI Loan Agreement for three reasons:⁴⁹

a) The potential acquisition of the Mallay Plant was contingent on approval by the Mallay Community, which approved the transaction only in March 2019, after the latest closing date deadline of 15 February 2019;

b) Claimant had no alternative source of financing or other remedies for a failure to close the transaction by the deadline; and,

⁴⁹ First Report, ¶ 128. ↩

[Page 22]

c) The third-party plants were not capable of processing ore from the Invicta Project, putting Lupaka at risk of default its obligation to Sell the Contract Quantity of gold under the existing terms of the PLI Loan Agreement.

67. In their second report, Accuracy stated that our opinion “appears at odds with the [following] evidence submitted":⁵⁰

a) Pandion historically demonstrated flexibility regarding the financing it provided to Claimant;⁵¹ and,

b) Accuracy, referencing Claimant's witness Mr. Gordon Lloyd Ellis, continued that, in "the absence of the Blockade, Claimant and Buenaventura would likely have signed the Mallay Purchase Agreement shortly after the Mallay Community provided its approval of the transfer agreement in March 2019” (emphasis added).⁵²

68. Accuracy's statements above were not a sound basis for calculating damages with respect to the 590t/day scenario for the following reasons.

69. First, Accuracy unreasonably assumed that Pandion and PLI would concede payment extensions to Claimant and lend it US$13.0 million in additional funds, despite Claimant's continued refusal to obtain and maintain its SLO. In reality, rather than wait indefinitely for Claimant to operate the Project, Pandion sold PLI, cut its losses, and ceased its involvement with the Project. Regardless, Accuracy's calculations assume that, absent the Measures, Pandion would have relinquished its contract enforcement rights indefinitely, further deferred Claimant's obligations, and lent more money to Claimant, which seems irrational and inconsistent with Pandion's and PLI's actual decisions and conduct.

70. Second, as discussed in Paragraph 60, with regards to the November 2018 start date, we understand from Counsel that Ms. Dufour's expert opinion is that Micon's revised production schedule is not feasible due to missing approvals both for the Invicta Project and for ore processing plants, and that these missing approvals would have prevented Claimant from exploiting the Invicta Project until July 2020.

71. Third, we understand from Counsel that Claimant had committed 14 defaults under the PLI Loan Agreement, several which were unrelated to any acts or omissions by

⁵⁰ Accuracy Second Report, ¶ 3.11. ↩

⁵¹ Accuracy Second Report, ¶¶ 3.12 – 3.18. ↩

⁵² Accuracy Second Report, ¶ 3.10. ↩

[Page 23]

Peru or by the Parán Community, and all which entitled PLI to foreclose on Claimant's investment.

72. Fourth, although Mr. Gordon Ellis stated that Claimant raised financing five months before the Valuation Date, his testimony indicated that Claimant's attempt to raise US$1.0 million was undersubscribed and yielded only US$0.7 million,⁵³ less than 5% of the US$15.9 million that Claimant would need to settle its default on the PLI Loan Agreement.

73. In conclusion, because multiple obstacles prevented Claimant from commercially exploiting the Invicta Project, it would not have been rational for Pandion or PLI to provide further extensions and an additional US$13.0 million loan to Claimant, and Accuracy should not have assumed otherwise. Therefore, this remains a fundamental flaw with Accuracy's analysis that would have resulted in Claimant's damages being reduced to nil.

D. Accuracy does not Account for Difficulties in Refinancing

74. In our First Report, we noted that:⁵⁴

"Accuracy's But-For Scenarios ignored the actual financing issues that the Project faced. In the But-For Scenarios, Accuracy assumed that on 26 August 2019, the Valuation Date, Claimant paid the early termination fee to settle the PLI Loan Agreement. Without discussing or incorporating any terms of an alternative financing arrangement that would be required to pay for the settlement as well as the Invicta Project's ongoing operations, Accuracy assumed away Claimant's potential refinancing risks asserting that: 'the project would be financed by a hypothetical investor using a mixture of equity and debt.”'

75. In the Accuracy Second Report, however, by adopting Micon's earlier start date of operations, Accuracy assumes that Claimant would have been able to meet its gold delivery obligations under the PLI Loan Agreement and thus avoid making a settlement payment of US$15.9 million to PLI.

76. As we have discussed above, however, absent the Measures, Claimant would have defaulted on the PLI Loan Agreement under both production scenarios due to the unresolved technical issues with the third-party processing plants, missing authorizations preventing operations from starting before July 2020, and the low

⁵³ Second Witness Statement of Gordon Ellis dated 23 September 2022, Annex. ↩

⁵⁴ First Report, ¶ 131. ↩

[Page 24]

likelihood that the Draft Third Amendment to the PLI Loan Agreement would be closed. In addition, we understand from Counsel that Claimant had committed 14 defaults under the PLI Loan Agreement, several which were unrelated to any acts or omissions by Peru or by the Parán Community. Therefore, absent the Measures, Claimant would have had to pay the early termination fee of US$15.9 million to settle the PLI Loan Agreement and prevent losing its shares in IMC. To continue the Invicta Project, Claimant would have needed to obtain new financing to fund a payment of US$15.9 million to PLI, and Claimant also would have needed additional financing if it were to consummate an acquisition of the Mallay Plant.

77. Neither Claimant nor Accuracy have submitted any evidence to show how Claimant would have been able to obtain additional financing to settle the PLI Loan Agreement and under what terms and conditions this financing would be available, including a timeline of the necessary steps to achieve this task, nor did they show how Claimant would have obtained further financing to potentially acquire the Mallay Plant.

78. Lastly, as noted above, although Mr. Gordon Ellis indicated that Claimant raised financing five months before the Valuation Date, his testimony indicated that Claimant's attempt to raise US$1.0 million was undersubscribed and yielded only US$0.665 million,⁵⁵ less than 5% of the US$15.9 million that Claimant would need to settle its default on the PLI Loan Agreement.

E. Conclusion on Fundamental Flaws

79. In conclusion, we do not believe that Accuracy's But-For Scenarios reasonably estimate damages resulting from the Measures because they do not account for the following fundamental flaws:

a) Claimant would not have resolved the issues underlying the Access Road Protest;

b) Claimant would have failed to obtain and maintain its SLO;

c) Claimant was likely to default on its debt obligations regardless of any Measures; and,

d) Claimant did not provide evidence that it could have secured alternative financing arrangements.

80. Based on these fundamental flaws, as shown in the figure below we conclude that Claimant's damages would be nil.

⁵⁵ Second Witness Statement of Gordon Ellis dated 23 September 2022, Annex. ↩

[Page 25]

Figure 8 – Accuracy's Decision Tree for the Invicta Project

[A flowchart showing a decision tree. "Invicta Project" at the top leads to a decision box: "Little or no social conflict without SLO". "No" leads to "Project Failure". "Yes" leads to "Purchase Mallay (590t/day) or No mechanical issues at third-party plants". "No" leads back to the "Little or no social conflict" box. "Yes" leads to "Obtain or update permits". "No" leads to "Project Failure". "Yes" leads to "Meet debt obligations". "No" leads to "Refinance the Project". From "Refinance the Project", "No" leads to "Project Failure" and "Yes" leads back to "Meet debt obligations". From "Meet debt obligations", "Yes" leads to "Project Success".]

V. Our Opinion of Accuracy's Updated 590t/day Scenario

81. In their second report, Accuracy revised certain assumptions and calculations under the 590t/day scenario. Notwithstanding the fundamental flaws discussed above in Section IV, we assess Accuracy's revised 590t/day scenario in the following section, including the new and remaining defects in it.

A. Opinions on Accuracy's Changes to the Damage Analysis under the 590t/day Scenario

82. In this subsection, we provide our comments on Accuracy's assumptions related to their damages calculation under the 590t/day scenario.

i. Additional Cost with Respect to the Third-Party Tolling Issues

83. As discussed in Section IV, Accuracy assumed that Claimant would have resolved the Project's third-party ore processing issues by early 2019.⁵⁶ As the Mallay Community did not approve the acquisition of the Mallay Plant until March 2019, Claimant would need to rely on third-party processors for the period from November 2018 until the Mallay Plant was online.

84. However, Accuracy and Micon did not explain how the various third-party ore processing failures would have been resolved during that period or quantify the

⁵⁶ Paragraph 54. ↩

[Page 26]

financial impact of doing so. Instead, Accuracy estimated damages assuming the third-party ore processing problems would be resolved immediately and at no cost, which seems unrealistic. As Accuracy did not provide an estimate of the additional cost that Claimant would have had to pay to resolve the issues, we are not able to quantify the impact. However, any incremental cost would reduce the damages under the 590t/day scenario.

ii. Project Start Date

85. Accuracy adopted Micon's opinion that commercial production would start as early as in November 2018 under their revised 590t/day scenario. This start date is ten months earlier than Accuracy's assumption in their first report of 26 August 2019.⁵⁷ By assuming an earlier project start date, Accuracy increased their damages estimate by approximately US$4.1 million on account of the cash flows occurring earlier and thus discounted less.⁵⁸

86. However, as discussed in Paragraph 6060, Counsel has informed us that Ms. Dufour's expert opinion is that Claimant could not have exploited the Invicta Project's resources prior to July 2020. Therefore, Micon's revised production start date, which Accuracy adopts, is not a reasonable basis for estimating damages. Under Ms. Dufour's estimate, Claimant could not meet its PLI Loan Agreement obligations and would therefore default.

87. Setting aside other flaws and risks and assuming PLI would agree to a delayed gold delivery schedule, revising the production start date to 1 July 2020 reduces damages to US$32.3 million (a reduction of US$8.7 million from US$41.0 million).⁵⁹

iii. Extended Production Schedule

88. The Accuracy Second Report adopted Micon's revised production schedule, which extended the project from seven years per the Red Cloud Model (and the Accuracy First Report) to ten years. Under this longer production schedule, the total quantity of ore mined and processed increased by 32.7% as shown in the Figure 9 below.⁶⁰

⁵⁷ Accuracy Second Report, ¶ 6.11; Accuracy First Report, Footnote 153. ↩

⁵⁸ Appendix 2, "AlixPartners", set the scenario to "Reset to Accuracy" and Cell C4 = "Accuracy First Report". ↩

⁵⁹ Appendix 2, "AlixPartners", set Cell C4 = "July 2020". ↩

⁶⁰ Accuracy Second Report, Appendix 3, "Inputs". ↩

[Page 27]

Figure 9 – 590t/day Scenario Updated Production

[Bar chart showing three bars. "Red Cloud Model" bar is 1,366,916 t. "Accuracy First Report" bar is 1,366,916 t. "Accuracy Second Report" bar is 1,814,171 t, with a label "+32.7%".]

89. We compare the production schedules in Figure 10.⁶¹

Figure 10 – Comparison of Production Schedule

[Line chart comparing two production schedules. The y-axis is "Au Eq oz (000)" from 0 to 220. The x-axis is years from 2018 to 2027. The "Accuracy First Report" line starts in 2019, rises to about 210, stays flat until 2025, then drops to 0. The "Accuracy Second Report" line starts in 2018, rises to about 210 in 2019, stays flat until 2026, then drops to 0. An annotation points to the start of the "Accuracy Second Report" line in 2020, labeled "Jan 2020 - First delivery of gold (Accuracy Second Report)". Arrows indicate "Accuracy First Report, 7-year production" and "Accuracy Second Report, 10-year production".]

90. Accuracy's revised production schedule increased damages under the 590t/day scenario. On a pre-discounting basis, the post-tax cash flows in the last three additional years represent 42.1% of the total post-tax cash flows.⁶² The discounted cash flows of these three years represent 32.1% of the total net present value (“NPV”), which means approximately one third of the FMV in the revised But-For Scenario is from the production schedule extension.⁶³

⁶¹ Accuracy Second Report, Appendix 3, "Inputs". ↩

⁶² Accuracy Second Report, Appendix 3, "Dashboard (ACC)". ↩

⁶³ Accuracy Second Report, Appendix 3, "Summary". ↩

[Page 28]

Figure 11 – Impact of the Extended Production Schedule

[Bar chart showing discounted cash flows per year from 2018 to 2028. A line above the bars shows the cumulative discounted cash flows. The bars show values like 2.3, 3.6, 6.7, 10.3, etc. The cumulative line rises from 2.3 to 41.0. An arrow indicates a "+47.3%" increase in the later years.]

91. Micon justified the three-year production schedule extension as follows:⁶⁴

a) The preliminary economic assessment prepared by SRK Consulting (Canada) Inc. (“SRK”) in early 2018 (“SRK PEA”) mine plan did not consider four mineralized zones and excluded 78% of the mineral resources.

b) The SRK PEA mine plan applied a cut-off grade of 4.0 g/t gold equivalent (“AuEq”), which Micon reduced to 3.0 g/t AuEq. Micon thus recommended mining zones with lower mineral density that, under the SRK PEA, would be left unmined.

c) The SRK PEA mine plan's upper elevation limit of 3,505 meters for mining activities did not represent an upper bound to the presence of mineral resources.

d) The minimum stope width should be 3.0 meters based upon Micon's review of the proposed mining method, rather than 4.0 meters contemplated in the SRK PEA.

92. Micon modified the parameters of the SRK PEA block model in a way that Lupaka, SRK, and Red Cloud elected not to do contemporaneously.⁶⁵ We understand that Micon's extended production schedule was created only for this arbitration and that Claimant has not explained why this higher-value plan for the Project was never presented internally or to any third parties, including Canadian investors and regulators to whom Claimant reported quarterly and to whom it presumably owed

⁶⁴ Micon Report, ¶¶ 76, 78, 79, 103, Table 5.10. ↩

⁶⁵ Micon Report, Table 5.10. ↩

[Page 29]

certain disclosure duties. We also understand that Accuracy, Micon, and Claimant have not reconciled their ten-year mining plan with Claimant's approval for only seven years of production (and at a maximum production rate of 400t/day).⁶⁶ Neither Micon nor Accuracy explained how the Project would have lawfully attained production up to 590t/day and a ten-year production period. Calculating damages based on this production rate and period therefore seems unreasonable.

93. In contrast, Red Cloud's production schedule was created around the time that Claimant was considering acquiring the Mallay Plant and was prepared in the ordinary course of business. The Red Cloud forecast reflects what Red Cloud and Lupaka expected during the negotiations with Buenaventura. The Red Cloud forecast is therefore a more credible basis for estimating damages than the revised production schedule that Micon improvised.

94. Finally, we note that Micon did not extend the production schedule under the 355t/day scenario. Although Micon opined that they think the SRK PEA mine plan also understated the life of the Invicta Project in that scenario, they did not modify the amount of ore mined and processed under the 355t/day scenario (669,813 tonnes).

95. Should the Tribunal find that Micon's adjustment is inappropriate,⁶⁷ Accuracy's damages estimate in the 590t/day scenario should be reduced by 29.6%, from US$41.0 million to US$28.9 million on a standalone basis, as of the Valuation Date.⁶⁸

96. We note that Micon increased the total capital development cost for the 590t/day scenario by US$5.9 million.⁶⁹ We understand that US$2.2 million of this increase is attributable to the development costs that Accuracy excluded in the Accuracy First Report, and which Micon reinstated.⁷⁰ The Accuracy and Micon reports are unclear as to whether the remaining US$3.7 million increase in capital development costs is specifically and only attributable to the production period extension. We also note that Accuracy increased the additional closure costs corresponding to an extended production schedule. Removing the additional income from three additional years of

⁶⁶ MEM Report and Resolution approving the Mining Plan, 11 December 2014, p. 5. [C-0009]; SRK Consulting PEA dated 13 April 2018, p. 145. [AC-0002] ↩

⁶⁷ We are not mine engineering experts and do not purport to provide any testimony or opinion with respect to the technical engineering aspects of the Project. However, we have noted that Micon's approach requires several speculative adjustments and logic leaps. ↩

⁶⁸ Accuracy Second Report, Appendix 3, "Summary". Sum of the discounted cash flows from Year 1 to Year 7, adjusted for the remaining changes in working capital in Y8. ↩

⁶⁹ Micon Report, Table 5.12; Accuracy First Report, Appendix 6, "capex-sum". ↩

⁷⁰ Micon Report, Table 5.12. ↩

[Page 30]

production and eliminating the additional capital development costs and closure costs attributable to the three-year extension, the damages under the 590t/day scenario are reduced to US$30.8 million from US$41.0 million on a standalone basis as of the Valuation Date.⁷¹

Figure 12 – Accuracy's Damages Calculation Under 590t/day Scenario Using Red Cloud's Estimated Capital Expenditure

[Waterfall chart showing a starting bar at 41.0 labeled "Damages, Accuracy Second Report". A downward bar of -12.1 is labeled "Impact of using 7-year production schedule". An upward bar of 1.9 is labeled "Impact of removing the additional capital development cost and closure cost". The final bar is at 30.8, labeled "Damages, removing the additional production years and capital development cost".]

iv. Metal Grades

97. In our First Report, we commented that there was insufficient documentation or analysis to support Accuracy's assumed metal grades. In their second report, Accuracy adopted Micon's modifications to Red Cloud's metal grade assumptions. The metal grades Micon proposed are generally more conservative than those applied in the Red Cloud Model and the Accuracy First Report.⁷²

98. In our First Report, we also observed that the grade of the payable gold that the Invicta Project yielded was less than half of the budgeted grade. Lupaka's report for October 2018 showed that the actual gold ore grade was 2.25g/t, only 46.6% of the budgeted grade of 4.83g/t.⁷³

99. Responding to our observation that the actual gold grade was lower than predicted, Micon considered four possible explanations.

⁷¹ Appendix 2, "AlixPartners", set the scenario to "Reset to Accuracy" and Cell C5 = "ON". ↩

⁷² Accuracy Second Report, Appendix 3, "Inputs". ↩

⁷³ First Report, Figure 12. ↩

[Page 31]

100. First, Micon conjectured that the low grade might be due to the natural variability of mineralization, and that low grade materials might have been presented around the tested elevation levels:⁷⁴

"...material with actual gold grades well below average for the orebody could quite properly have been mined as payable ore based on its gold equivalent grade, owing to the contribution of copper, lead, zinc and silver to the value of the ore."

101. However, Micon did not provide sufficient evidence that would support this statement.

102. Second, Micon suggested that the mix of waste with ore might dilute the average grade during the development stage. We note that Micon did not provide any documentary evidence to support this statement.⁷⁵

103. Third, Micon speculated that the low gold grade might be because of the separation of heavy particles enriched in gold during the blasting of broken rock:⁷⁶

"...Thus, compared to the average grade of the Invicta resource, the fine material was highly enriched in precious metals while being somewhat less enriched in the base metals. Enrichment of the fines in precious metals would necessarily leave the remainder of the ore depleted in precious metals. Fortunately, once recognised, this problem is manageable. The footwall of a stope is washed down once all lump ore has been removed, leaving a heavy sludge that can be collected and transported to the processing plant for separation of the base metal concentrates."

104. This statement was also not supported by any documentary evidence.

105. Fourth, Micon opined on the third-party tolling process:⁷⁷

"The gold grades reported are back-calculated from analysis of the saleable concentrates and mill tailings. However, during processing at third-party toll-treatment mills in the absence of close supervision, it is possible that a portion of the gold content of the ore was separated gravimetrically by the mill operators, leaving only that portion that was intimately mixed with sulphide minerals to be recovered into the base metal concentrates sold by Lupaka."

⁷⁴ Micon Report, ¶¶ 139 – 140. ↩

⁷⁵ Micon Report, ¶ 141. ↩

⁷⁶ Micon Report, ¶ 142. ↩

⁷⁷ Micon Report, ¶ 143. ↩

[Page 32]

106. Again, Micon did not provide any documentary evidence that would support their speculation.

107. Furthermore, we have not identified any contemporaneous data or documents in which Claimant would have raised, observed, or been alerted to any of the issues that Micon listed above. Therefore, the theories that Micon surmised are not a reasonable basis for budgeting a grade 114.7% higher than the actual grade observed.⁷⁸

108. In conclusion, the Micon Report has no analysis to demonstrate that any or all of their hypotheses could reasonably account for the entire difference in the actual and budgeted gold ore grades.

109. According to Micon, the first three issues could be managed and mitigated without incurring additional costs. For the last issue, Micon proposed closer supervision and improved security at third-party processing operations. However, Micon and Accuracy calculated no increase in expenses in the revised 590t/day scenario. Deploying closer supervision and improved security likely would increase Claimant's costs and thus reduce the damages Accuracy calculated.

v. Operating Costs

110. In their revised calculation of damages, Accuracy increased the average unit operating costs by US$13.40/t in the 590t/day scenario.⁷⁹ This change reflected the net changes by Micon and Accuracy to the operating cost inputs and assumptions, including:⁸⁰

a) An increase in employee profit sharing from US$700,000 to US$975,000 per year;

b) An increase of US$5.50/t in the direct unit cost of treating ore to include the cost of electrical energy;

c) A reduction of US$20.00/t in transportation costs, which they attributed to the relative proximity of the Mallay Plant to Callao Port compared to Caraz, where the assumed processing plant was located per the SRK PEA;

d) Additional closure costs to account for the extended mine plan; and,

⁷⁸ Paragraph 98. Calculated as 4.83g/t / 2.25g/t -1. ↩

⁷⁹ Accuracy Second Report, ¶ 4.16. ↩

⁸⁰ Accuracy Second Report, ¶¶ 6.19 – 6.20. ↩

[Page 33]

e) Additional general and administrative costs of US$350,000 per year for management of the Mallay Plant.

111. Accuracy's revised operating costs still do not account for issues we raised in our First Report and discuss below. Accuracy's operating costs are therefore underestimated.

112. First, as discussed above, to make the project viable, Claimant would have to resolve the conflicts with the Parán Community. However, neither Micon nor Accuracy included any costs that Claimant would need to incur to obtain and maintain the Project's SLO from the Parán Community.

113. As summarized in Figure 13 below, we reviewed several contemporaneous documents relating to Claimant's communications with the Parán Community and its commitments to the neighboring communities as a basis to derive a baseline cost for obtaining and maintaining the project's SLO with the Parán Community.

Figure 13 – Summary of Claimant's Agreed Payments to Neighboring Communities

114. Details regarding these costs can be found in Appendix 5.

115. Counsel also asked us to use the average of the following three amounts as a baseline for an estimate of the cost to obtain and maintain the SLO: i) costs Claimant agreed to pay to the Lacsanga Community; ii) costs Claimant agreed to pay to the Santo Domingo De Apache Community; and iii) the amounts that the Parán Community requested. The resulting total costs comprise a one-time payment of US$0.3 million and an annual payment of US$0.3 million for ten years.⁸¹ Applying the discount factor

⁸¹ If these annual payments are for seven years, as would be expected under a seven-year production schedule, then the net cost is US$2.0 million after discounting, on a standalone basis, as of the Valuation Date. ↩

[Page 34]

used in the Accuracy Second Report, the net cost is US$2.4 million, on a standalone basis as of the Valuation Date.

116. As an alternative benchmark, based on the Social Sustainable Solution report dated 20 May 2018,⁸² it appears that the president of the Parán Community intended to negotiate with IMC an annual payment of US$0.6 million.⁸³ Paying this amount annually for ten years would imply a total net cost of US$4.2 million after applying the discounting factors Accuracy used, on a standalone basis as of the Valuation Date.⁸⁴

117. Second, in addition to the one-time and annual payments to the Parán Community, we understand that IMC would have to incur costs to meet regulations and permit requirements. For example, we understand from Counsel that Ms. Dufour has provided several examples of these additional costs: preparing dossiers for permit applications, costs for implementing citizen participation mechanisms to modify the EIA, and costs for dismantling water management systems. We are not currently able to estimate the cost of obtaining and maintaining compliance with these regulations and permits.

118. Third, as mentioned earlier, Accuracy accounted for no incremental costs to supervise and improve security at third-party tolling plants, as Micon proposed to address one of the possible causes that Micon theorized could be responsible for the low gold ore grade.

119. To summarize, Accuracy's operating expenses are underestimated. After adjusting for the additional operating costs required to obtain and maintain the social license from the Parán Community, Accuracy's damages estimate is reduced to US$38.5 million as of the Valuation Date.⁸⁵ As noted above, we are not currently able to quantify the cost of obtaining and maintaining compliance with the regulations and permits, which as Counsel has informed us, were noted by Ms. Dufour. As Accuracy provided no estimates as to the incremental security and supervision costs at the third-party processing plants, and we are unable to locate public information on the cost to impose surveillance and security measures upon third-party processors, we are not able to quantify the financial impact.

⁸² SSS, Weekly Report, Project, 14/05/2018 to 20/05/2018, 14 May 2018 (ENG-SPA), pp. 1-2. [C-0435] ↩

⁸³ PEN 2.0 million. Converted to US$ using the FX rate as of 26 August 2019. Capital IQ, US$PEN Exchange Rates. [AP-0065] ↩

⁸⁴ If these annual payments are for seven years, as would be expected under a seven-year production schedule, then the net cost is US$3.5 million after discounting, on a standalone basis, as of the Valuation Date. ↩

⁸⁵ Appendix 2, "AlixPartners", set Cell C6 = "Scenario 1". ↩

[Page 35]

vi. Capital Expenditures

120. In their second report, Accuracy adopted Micon's revisions to capital expenditures. Micon increased the capital expenditure estimate by US$5.1 million compared to the Accuracy First Report and by US$1.0 million compared to Red Cloud's forecast. As a comparison to these modest increases in capital expenditure (13.6% and 2.4%, respectively), Micon raised the total production volume by 32.7%, as was shown in Figure 10.

121. The three capital expenditure forecasts are presented in Figure 14 below.

Figure 14 – Capital Expenditure Forecasts

122. As explained below, Micon's capital expenditure estimates appear to be underestimated for the following reasons:

123. First, as presented above, Micon reduced infrastructure capital spending by US$1.9 million (US$5.3 million less US$3.4 million), despite extending the production schedule by three years.

124. Micon's annual infrastructure capital spending estimate⁸⁹ appears to be consistent with the SRK PEA, which was based on a 355t/day production rate, as shown in the figure below.⁹⁰ Micon stated no rationale for why infrastructure capital spending,

⁸⁶ Accuracy Second Report, Appendix 3, "capex-sum" (set to "Red Cloud" on "Dashboard (ACC)"). ↩

⁸⁷ Accuracy Second Report, Appendix 3, "fin_lupaka" and Table 6.3 (set to "First Accuracy Report" on "Dashboard (ACC)"). ↩

⁸⁸ Accuracy Second Report, Appendix 3, "fin_lupaka" and "capex-sum" and Table 6.3. ↩

⁸⁹ Accuracy Second Report, Appendix 3, "Micon". ↩

⁹⁰ Accuracy Second Report, Appendix 3, "Micon"; SRK Consulting PEA dated 13 April 2018, Table 102 and Table 103. [AC-0002]; Micon Report, Table 5.8. ↩

[Page 36]

comprised of "site roads and prep", "community infrastructure" and "environmental and closure", under the 590t/day scenario would be the same as in the 355t/day scenario (instead of the estimates used in the Red Cloud Model for the 590t/day scenario⁹¹). We understand that the annual community infrastructure cost in the SRK PEA did not account for any potential costs to provide such infrastructure to the Mallay Community.⁹² Also, Micon discontinued the annual environmental and closure costs in Year 6 (as reflected by the decrease in Year 7), despite extending the production schedule to ten years.⁹³

Figure 15 – Infrastructure Capital by Year

[Chart titled "Figure 15 – Infrastructure Capital by Year" showing three line graphs over time from 2018 (Y1) to 2027 (Y10). The Y-axis is "US$ millions" from 0.0 to 1.4. The X-axis shows years. The legend indicates three scenarios: "SRK PEA, 6-year production (355t/day scenario)", "Accuracy First Report, 7-year production (590t/day scenario)", and "Accuracy Second Report, 10-year production (590t/day scenario)". Data points are labeled on the graph.]

125. Second, we also note that, based on Micon's forecast for ten years of production, the last year Claimant would spend any amount on capital development costs is Year 6, while both SRK and Red Cloud assumed that, for seven years of production, such costs would be incurred through to Year 7 (as shown in Figure 16).⁹⁴ Micon thus assumes that, in the last four years of operation, Claimant would incur no capital development costs, in contrast to the contemporaneous forecasts that expected a such spending throughout the project.

126. Micon commented that “Red Cloud's estimate was very conservative” in the cadence of their projected capital spending (i.e., Micon viewed it as front-loaded in early

⁹¹ Accuracy Second Report, Appendix 3, "Inputs". For comparison purpose, the infrastructure capital in Y1 is not included. ↩

⁹² Accuracy Second Report, Appendix 3, "Micon"; SRK Consulting PEA dated 13 April 2018, Table 102 and Table 103. [AC-0002] ↩

⁹³ Accuracy Second Report, Appendix 3, "Micon"; SRK Consulting PEA dated 13 April 2018, Table 102 and Table 103. [AC-0002] ↩

⁹⁴ Accuracy Second Report, Appendix 3, "Inputs", rows 55-57. To be consistent with Figure 14, the capital development cost in Y1 is not included. ↩

[Page 37]

years). Micon, instead, assumed capital expenditure would spread in a longer period of time, even though they assumed zero capital expenditure after Year 6.⁹⁵ Allocating capital expenditures across a longer period, as Micon did, would decrease the NPV of Project capital expenditures and thus increase the estimated damages.

Figure 16 – Capital Development Cost by Year

[Chart titled "Figure 16 – Capital Development Cost by Year" showing two line graphs over time from 2018 (Y1) to 2027 (Y10). The Y-axis is "US$ millions" from 0.0 to 6.0. The X-axis shows years. The legend indicates two scenarios: "Accuracy First Report, 7-year production" and "Accuracy Second Report, 10-year production". Data points are labeled on the graph.]

127. Third, Counsel has informed us that Ms. Dufour confirms that additional capital expenditure is required for construction of fuel storage facilities, surface water use works, and ground water use works. Any increase in capital expenditure reduces the estimated damages.

128. Based on the discussions above, Micon's capital expenditure estimates lack support and appear to be underestimated.

vii. Financing Cash Flows

129. In the Accuracy Second Report, instead of assuming Claimant would close the PLI Loan Agreement by paying to PLI a lump sum, Accuracy estimated the (periodic) financing cash flows under the loan. Accuracy combined the anticipated cash inflows of US$13.0 million (i.e., the loan disbursement by the lender) for the purchase of the Mallay Plant with the anticipated cash outflows totaling US$32.3 million for delivery to PLI of gold (or its cash equivalent), on an undiscounted basis.⁹⁶ Accuracy also included anticipated cash outflows totaling US$9.9 million that would be due to PLI

⁹⁵ Accuracy Second Report, Appendix 3, "Inputs", row 57. ↩

⁹⁶ Accuracy Second Report, Appendix 3, "Dashboard (ACC)". ↩

[Page 38]

under the PLI Loan Agreement.⁹⁷ The net impact of the financing cash flows is US$17.2 million on a discounted basis.

130. Accuracy adopted Micon's extended production schedule and assumed that Claimant would have met its payment obligations under the PLI Loan Agreement. By making these assumptions, Accuracy reduced Claimant's total costs under the PLI Loan Agreement and increased the estimated damages. As shown below, with the extended production schedule, the net cost of the financing is US$17.2 million, as compared to the total cost of US$28.9 million that we calculated in our First Report (settlement payment of US$15.9 million plus the additional loan of US$13.0 million for the Mallay Plan transaction).

Figure 17 – Impact of the Debt Obligations under the 590t/day Scenario

131. Based on our review of the Accuracy Second Report, these revisions appear to be appropriate only by assuming that Pandion (who would retain control over PLI in the But-For Scenario) would sign the Draft Third Amendment as Accuracy assumed for the 590t/day scenario, and that lawful commercial production at the Invicta Project would have started in November 2018 (e.g., with all permits in place). However, as we have discussed earlier, both of these assumptions are unsupported.

viii. Discount Rate

132. In the Accuracy Second Report, Accuracy discounted the cash flows using the cost of equity ("COE") of a hypothetical gold mining entity operating in Peru, rather than Claimant's WACC, consistent with the Free Cash Flow to Equity (“FCFE”) method.⁹⁹

133. Accuracy maintained the 11.1% cost of equity that they had calculated to be Claimant's cost of equity in the Accuracy First Report by applying the Capital Asset Pricing Model (“CAPM").

⁹⁷ Accuracy Second Report, Appendix 3, "Dashboard (ACC)". ↩

⁹⁸ The net impact of the cash inflow (additional loan) and outflow (gold delivery to PLI and the upside participation). Also impacted by the change in discount rate. Accuracy Second Report, Table 6.6. ↩

⁹⁹ Accuracy Second Report, ¶ 6.37. ↩

[Page 39]

134. In the Accuracy First Report, Accuracy further adjusted the discount rate by adding to the cost of equity of a hypothetical gold mining entity a premium of 6.9%, reflecting "the higher level of uncertainty attached to the 590t/day Scenario under the Mallay Acquisition Plan, which had not been subject to the same level of detailed financial analysis as the 355t/day Scenario and would have required further capital investment and financing”.¹⁰⁰ In the Accuracy Second Report, however, Accuracy reduced this premium to 3.3%, the same premium Accuracy added to the 355t/day scenario, on the basis that “Micon considers the level of accuracy of the revised 590t/d plan to be very similar to that achieved in the 355t/d plan and that both plans are equally reliable."¹⁰¹

135. We disagree with Accuracy's adjustment to the project-specific risk premium. First, the source document Accuracy relied upon (AC-0047) indicates the premium was calculated based on data collected in 1996, 1999 and 2005, at least 14 years prior to the Valuation Date (as shown in the figure below).¹⁰² Second, according to AC-0047, the 3.3% premium is applicable to mining projects that have undergone feasibility studies, whereas the SRK PEA is akin to a Pre-Feasibility Study, for which AC-0047 recommends applying a premium of 5.7%.

¹⁰⁰ Accuracy First Report, ¶ A4.23. ↩

¹⁰¹ Accuracy Second Report, ¶ 6.33. ↩

¹⁰² Lawrence Devon Smith, The RADR Paradox-Discount Rates: Risk, & Long Life Projects, 2016, p. 55. [AC-0047] ↩

[Page 40]

Figure 18 – Screenshot of the Premiums in AC-0047

[Image of a chart titled "Discount Rates - RADR Risk Adjusting for Different Project Stages (Real)". The chart shows two declining lines, one for "Base Metals" and one for "Gold", across project stages: Scoping, PFS, FS, Operating. Below the chart is a table.]

136. There is no dispute that no Feasibility Study was undertaken as evidenced by a news release Claimant published on 28 November 2017:¹⁰³

"The decision to commence potential production at the Invicta Gold Project and the Company's plans for a mining operation as referenced herein (the "Production Decision and Plans") were based on economic models prepared by the Company in conjunction with management's knowledge of the property and the existing estimate of measured, indicated and inferred mineral resources on the property. The Production Decision and Plans were not based on a preliminary economic assessment, a pre-feasibility study or a feasibility study of mineral reserves demonstrating economic and technical viability." (emphasis added).

137. As explained in our First Report, a preliminary economic assessment ("PEA") is subject to a lower confidence level than a Pre-Feasibility Study.¹⁰⁴ Furthermore, the Micon Report and their production schedule do not constitute a Pre-Feasibility Study or Feasibility Study. Therefore, at the very least, the amount of project-specific risk premium should be increased to 5.7%.

¹⁰³ Cision PR News, Lupaka Gold Commences Preliminary Economic Assessment on the Invicta Gold Development Project, Target Q1/18 Release, 28 November 2017. [AP-0066] ↩

¹⁰⁴ First Report, Figure 5. ↩

[Page 41]

138. Additionally, as discussed in Paragraph 92, Claimant obtained from MINEM approval for production up to only 400t/day and for seven years. We understand that Claimant had not modified its approval to allow a 590t/day schedule for 10 years. The uncertainty with respect to the approval of the higher production volume and extended production schedule make the 590t/day scenario risker than the 355t/day scenario. Further, as Counsel has informed us, Ms. Dufour explains that Claimant needed to obtain various (new and revised) permits to commence commercial ore extraction. If damages were deemed to be greater than nil, these factors, among others, also would need be reflected in a project-specific risk premium.

139. As explained in our First Report and Section IV, Accuracy's discount rate explicitly excludes consideration of social license risk, which any reasonable and informed investor would consider.¹⁰⁵ Accuracy attempted to justify their approach by stating that 40% of the companies included in their reference data have operations in Central and/or South America and thus at least part of, if not all, the regional risk should already be considered.¹⁰⁶ Accuracy also contended that the inclusion of Peru's country risk premium reflects "additional risk associated with doing business in Peru over a 'risk-free' jurisdiction”.¹⁰⁷

140. We disagree with Accuracy's explanations. The beta in Accuracy's source data captures the average risk that a generic U.S. mining company faces. Accuracy provided no analysis as to why Claimant's social license risk, which would need to reflect the continuous and intense conflicts that already had materialized with the Parán Community, would be comparable to the risk level that an average U.S. mining company is faces when investing and operating a generic mining project.

141. In addition, as Accuracy explained, the country risk premium compensates investors for the additional risk of doing business in Peru compared to the United States. The country risk premium includes the general political, economic, legal, and other relevant risks that any reasonable investor would incorporate when making investment decisions, no matter what industry or project the investor is considering. Thus, the country risk premium reflects the average risk that an investor faces in Peru. Accuracy did not demonstrate that the project-specific social license risk that

¹⁰⁵ As noted in the Glossary, "Fair Market Value" requires measuring "the (highest) price, expressed in terms of cash equivalents, at which property would change hands between a hypothetical willing and able buyer and a hypothetical willing and able seller, acting at arms-length in an open and unrestricted market, when neither is under compulsion to buy or sell and when both have reasonable knowledge of the relevant facts" (emphasis added). ↩

¹⁰⁶ Accuracy Second Report, ¶ 4.47. ↩

¹⁰⁷ Accuracy Second Report, ¶ 4.47. ↩

[Page 42]

Claimant faced would be fully captured in Peru's general country risk premium. In fact, aware of the unresolved conflicts with the Parán Community, any rational investor would adjust the overall risk premium (to exceed the general country risk premium) when valuing the Project.

Figure 19 – Summary of Accuracy's Considerations of Risks

142. Furthermore, Accuracy's discount rate also does not account for the execution risk the Project faced, such as the uncertainties with respect to the third-party processors and variance in metal grades between actual result and the budget.¹⁰⁸

143. That is, Accuracy accounted for country-specific risks, but failed to account for project-specific risks that were not captured in their cash flow estimates. To summarize, not accounting for the social license risk and execution risk, Accuracy's discount rate was underestimated.

144. As demonstrated above, Accuracy's risk adjustment lacks support. For comparison purposes, we apply the generally applicable premium of 5.7% for projects in the Pre-Feasibility Study phase (according to Accuracy's source, which is based on old data), and we revise the discount rate to 14.6% (from 12.2%). On a standalone basis, this adjustment to the discount rate reduces damages to US$37.4 million (from US$41.0 million).¹⁰⁹

B. Conclusion on the 590t/day Scenario

145. As discussed, by adjusting the following assumptions in Accuracy's 590t/day model, damages are reduced to US$20.5 million:

1. Updating the Project start date to 1 July 2020;
2. Reducing the duration of the production schedule to seven years, consistent with the contemporaneous Red Cloud Model and the period of time that MINEM had approved Project mining.

¹⁰⁸ See Section V.A.i and Section V.A.iv. ↩

¹⁰⁹ Appendix 2, "AlixPartners", set Cell C7 = "ON". ↩

[Page 43]

3. Reducing the additional capital development costs (except the reversal of the cost in Year 1) and closure cost that Micon added to account for the seven (rather than ten) years of production;¹¹⁰
4. Increasing operating expenses to incorporate additional costs to obtain and maintain the SLO from the Parán Community; and,
5. Increasing the discount rate to 14.6% to account for risk factors Accuracy did not incorporate.

146. Our adjustments cover only the issues that we identified as material in the section above and that we were able to quantify. The Tribunal may find it necessary to increase the discount rate further to account for additional risks or costs, such as those we identified above but could not quantify.¹¹¹ As such, we present the following sensitivity analysis indicating the impact on damages of an increase in the discount rate of 1.0% to 5.0%.¹¹²

Figure 20 – Revised Damages and Sensitivity Analysis under the 590t/day Scenario

VI. Our Opinion of Accuracy's Updated 355t/day Scenario

147. In their second report, Accuracy also revised certain assumptions and calculations under the 355t/day scenario. Notwithstanding the fundamental flaws discussed above in Section IV, in this section we analyze Accuracy's revised damages calculations under the 355t/day scenario.

A. Opinions on Accuracy's Changes to the Damages Analysis under the 355 t/day Scenario

148. In this subsection, we provide our comments to Accuracy's assumptions related to their damages calculation under the 355t/day scenario.

¹¹⁰ Paragraph 96. ↩

¹¹¹ Paragraphs 84, 109, 119 and 127. ↩

¹¹² Appendix 2, "AlixPartners". ↩

[Page 44]

i. Additional Cost with Respect to the Third-Party Tolling Issues

149. As discussed in Section IV, Accuracy did not address Claimant's inability to secure a viable third-party ore processing facility. Instead, Accuracy adopted Micon's extended production schedule and assumed that Claimant would have resolved the Project's third-party tolling issues by early 2019.¹¹³

150. However, Accuracy and Micon did not explain how the various third-party ore processing failures would have been resolved during that period or quantify the financial impact of doing so. Instead, Accuracy estimated damages assuming the third-party ore processing problems would be resolved immediately and at no cost, which seems unrealistic. As Accuracy did not provide an estimate of the additional cost required to solve the issues, we are not able to quantify the impact. However, any incremental cost would reduce the damages under the 355t/day scenario.

ii. Project Start Date

151. Accuracy adopted Micon's opinion that commercial production would start as early as in November 2018 under their revised 355t/day scenario. This start date is ten months earlier than Accuracy's assumption in their first report of 26 August 2019.¹¹⁴ By assuming an earlier project start date, Accuracy increased their damages estimate by approximately US$2.8 million on account of the cash flows occurring earlier and thus discounted less.¹¹⁵

152. As discussed in Paragraph 60, we have been informed by Counsel that Ms. Dufour's expert opinion is that IMC would need to obtain the water use license for the execution of the mining activities of the Project and the possible start date was July 2020. Therefore, Micon's revised production start date, which Accuracy adopts, is not a reasonable basis for estimating damages. Under Mr. Dufour's estimate, Claimant could not meet its PLI Loan Agreement obligations and would therefore default.

153. Setting aside other flaws and risks and assuming PLI would agree to a delayed gold delivery schedule, revising the production start date to July 2020 reduces damages to US$26.0 million (a reduction of US$6.1 million from US$32.1 million).¹¹⁶

¹¹³ Micon Report, ¶ 125. ↩

¹¹⁴ Accuracy Second Report, ¶ 6.43; Accuracy First Report, Footnote 153. ↩

¹¹⁵ Appendix 3, "AlixPartners", set Cell C4 = "Accuracy First Report". ↩

¹¹⁶ Appendix 3, "AlixPartners", set Cell C4 = "July 2020". ↩

[Page 45]

iii. Production Schedule

154. In the Accuracy Second Report, Accuracy adopted Micon's revised production schedule under the 355t/day scenario, which includes a three-month ramp-up period starting from the updated project start date of November 2018 (instead of August 2019 assumed in Accuracy First Report).¹¹⁷ We note that Micon retained the production years and total production volume that was assumed in the SRK Model.¹¹⁸

iv. Metal Grades

155. Similar to the changes to the production schedule, Accuracy adopted Micon's updated metal grades in each production year. The average gold-equivalent grade was kept the same between the two Accuracy reports.¹¹⁹

156. As discussed in Section V.A.iv, Micon proposed four possible explanations for the lower metal grade that Claimant actually achieved compared to the budgeted grade, but, as explained above, those four theories are not supported by documentary evidence.

157. As discussed in Paragraph 109, one of Micon's theories for mitigating the metal grade deficiency required closer supervision of third-party tolling facilities. However, Micon and Accuracy calculated no increase in expenses in the revised 355t/day scenario. Deploying closer supervision and improved security likely would increase Claimant's costs and thus reduce the damages Accuracy calculated.

v. Operating Costs

158. In the Accuracy Second Report, Accuracy corrected an error in their operating cost calculations from their first report, which increased the unit operating cost by US$0.11/t.¹²⁰ Accuracy's other changes included:¹²¹

1. Updating the provision for employee profit sharing to be consistent with the assumptions used in the 590t/day scenario, which increased the annual operating cost by US$90 thousand; and,

¹¹⁷ Accuracy Second Report, ¶ 6.43. ↩

¹¹⁸ Accuracy Second Report, Appendix 4, "Inputs". ↩

¹¹⁹ Accuracy Second Report, ¶ 6.45. ↩

¹²⁰ Accuracy Second Report, ¶ 4.16 and Appendix 4, "opex". ↩

¹²¹ Accuracy Second Report, ¶ 6.49. ↩

[Page 46]

2. Correcting an error in the underlying SRK Model that mistakenly excluded copper and lead concentrate tonnages from the calculation of the concentrate transportation cost.

159. However, as discussed in Section V.A.v, Accuracy's operating cost estimate did not to account for the following issues:

1. Any costs required to obtain and maintain the SLO with the Parán Community;
2. Any costs in relation to meet the regulations and permit requirements; and,
3. Any costs required to conduct close supervision and improved security as Micon recommended to improve and maintain the gold grade.

160. In Figure 13, we provide our estimates for the costs related to obtaining and maintaining the SLO with the Parán Community under two scenarios. Applying Accuracy's discounting factors in the 355t/day scenario, the net impacts to damages are US$1.9 million and US$3.1 million under scenarios 1 and 2, respectively.¹²²

161. As explained in Paragraph 117, neither Accuracy nor Micon have provided estimates for the cost of obtaining and maintaining compliance with the regulations and permits that, as Counsel has informed us, Ms. Dufour has considered, nor have they quantified the incremental cost of increased security at the third-party processing plants. We have also not been able to quantify financial impact of these costs.

vi. Capital Expenditure

162. In the Second Accuracy Report, Accuracy adopted Micon's updated capital expenditure estimate. We note that Micon reduced the SRK PEA's capital expenditure estimate by US$3.0 million, as in Micon's opinion these costs related to site roads, underground development for mine access, and stope preparation Invicta carried out prior to October 2018 and should be excluded.¹²³

163. As discussed in Paragraph 127, Counsel has informed us that Ms. Dufour points out that additional capital expenditure would be required for the construction of fuel storage facilities, surface water use works, and ground water use works. Any increase in capital expenditure will reduce the damages.

¹²² Appendix 3, "AlixPartners", set Cell C5 = "Scenario 1" or "Scenario 2". ↩

¹²³ Micon Report, ¶ 93. ↩

[Page 47]

vii. Financing Cash Flows

164. To be consistent with the FCFE method, Accuracy deducted cash flows that were required by Claimant to meet its obligations under the PLI Loan Agreement. Accuracy calculated the anticipated cash outflows for the delivery of gold (or its cash equivalent) and for the PLI Loan Agreement upside participation clause.

165. Consistent with our findings under the 590t/day scenario discussed in Paragraph 131, Accuracy's calculation appears to be appropriate if Lupaka would be able to commence production in November 2018 and meet its obligations under the PLI Loan Agreement. However, this assumption is not well supported due to the issues with the third-party tolling plants and the permitting issues that, as Counsel has informed us, have been raised by Ms. Dufour which would also impact the production start date.

viii. Discount Rate

166. Consistent with our adjustment to the discount rate for the 590t/day scenario discussed in Section V.A.viii, we believe that an increase in Accuracy's "pre-production premium” to 5.7% is warranted. This is consistent with projects that have reached a Pre-Feasibility Study stage. The overall COE discount rate has increased to 14.6% (from 12.2%). As a result of this change, damages are reduced to US$30.7 million (from US$32.1 million) on a standalone basis.¹²⁴ However, as shown in Figure 19, even incorporating this increase, Accuracy's discount rate omitted consideration of social license risk and execution risk.

B. Conclusion on the 355t/day Scenario

167. As discussed above, after adjusting the following assumptions, Accuracy's damages under the 355t/day scenario are reduced to US$22.5 million (from US$32.1 million) on account of the following adjustments:

1. Updating the project start date to 1 July 2020;
2. Adding the costs that were required to obtain and maintain the SLO with the Parán Community; and,
3. Increasing the discount rate to 14.6% to incorporate the risks that Accuracy did not consider.

¹²⁴ Appendix 3, "AlixPartners", set Cell C6 = "ON". ↩

[Page 48]

168. We further present the sensitivity analysis as follows in case the Tribunal finds it necessary to increase the discount rate to account for the additional costs and risks that Accuracy did not consider and we are unable to quantify.

Figure 21 Revised Damages and Sensitivity Analysis under the 355t/day Scenario

VII. Our Opinion on the Residual Value

169. Accuracy indicated that they calculated Claimant's alleged damages as the difference between:¹²⁵

"Claimant's economic position in the But-For Situation and its economic position in the Actual Situation."

170. Accuracy premised their damages calculation on the value of Claimant's economic position in the Actual Scenario being nil since Claimant lost all its shares in the Invicta Project due to PLI's foreclosure action.¹²⁶

171. In our First Report, we noted that if the Tribunal agrees with Peru's position that PLI's foreclosure action on 26 August 2019 is not attributable to Peru, then the value of IMC's shares at the time of the foreclosure would be relevant to calculating Claimant's economic position in the Actual Scenario. An independent appraiser, PwC, assessed the equity value of the shares of IMC at US$13.0 million close to the foreclosure date (which is also the Valuation Date).¹²⁷ Therefore, the residual value of the IMC shares in the Actual Scenario is approximately US$13.0 million instead of nil. Accordingly, if PLI's foreclosure action was not attributable to Respondent, Claimant's damages would be reduced by a further US$13.0 million to account for IMC's residual value.

172. Accuracy disagrees with our opinion noting in the Accuracy Second Report that:

1. In the Actual Scenario, Claimant's "economic position" prior to the foreclosure should be the value of Claimant's shares in IMC of US$13.0 million, less the value of Claimant's liability to PLI, which Claimant approximated as PLI's settlement

¹²⁵ Accuracy Report, ¶ 4.34. ↩

¹²⁶ Accuracy Report, ¶ 4.35. ↩

¹²⁷ First Report, ¶ 97. ↩

[Page 49]

claim of US$15.9 million in July 2019. Therefore, Claimant's “economic value" in the Actual Scenario was negative.¹²⁸

2. Our opinion that the value of the Invicta Project could be nil due to the unaddressed fundamental flaws (especially those discussed again in Section IV) contradicts the appraised value of IMC's shares (i.e., US$13.0 million) in the Actual Scenario.¹²⁹
3. The Actual Scenario should represent the events that actually occurred. Accuracy stated that "even if Lonely Mountain had not foreclosed upon IMC shares, Claimant did not have access to the Invicta Mine", which they equate to the shares having no value under the Actual Scenario.¹³⁰

173. We disagree with Accuracy's opinions above in three respects.

174. First, as stated in the Accuracy Second Report, Accuracy's damages calculation was based on Claimant's overall economic position (i.e., the amount Claimant would have retained net of its obligations to its creditors), and not the value of Claimant's Investment (i.e., the value of its shares in IMC):¹³¹

"Consistent with the principle of full reparation, we offset damages owed to Claimant by the value of the debts that would have been settled with PLI in the But-For Situation, which were claimed (and settled through foreclosure) in the Actual Situation and amounted to USD 15.9m."

175. Claimant and not IMC held the obligation under the PLI Loan Agreement, would have defaulted on this obligation as discussed in Section IV, and would have needed to pay a settlement of US$15.9 million to avoid foreclosure on the IMC shares. Since Claimant, rather than IMC, held the obligation the residual value of the IMC shares should not be adjusted for financing obligations of the parent company.

176. Second, our opinion that the FMV of the Invicta Project could be nil is not inconsistent with the value the independent appraiser determined. Our conclusion assumes that, if the fundamental flaws, including the social license risk, financing risk, and execution risk, remained unaddressed and unresolved, the FMV of the Project would

¹²⁸ Accuracy Second Report, ¶¶ 4.59-4.62. ↩

¹²⁹ Accuracy Second Report, ¶¶ 4.59-4.62. ↩

¹³⁰ Accuracy Second Report, ¶¶ 4.59-4.62. ↩

¹³¹ Accuracy Second Report, ¶ 1.21. Here, Accuracy is referring to its approach from the Accuracy First Report. ↩

[Page 50]

be nil. However, the independent appraiser would have likely excluded the financing risk and foreclosure risk if a notional investor did address and resolve those issues.

177. Third, we disagree with Accuracy's statement that Claimant's inability to access the Invicta Project implied that the shares had no value. Under the Primary Approach that Accuracy advocated in both of their reports, “future cash flows are discounted to their present value at the cost of capital," based on a "hypothetical transaction between unidentified market participants.”¹³² Even if Claimant could not access the Invicta Project before the foreclosure of the PLI Loan Agreement, the available future cash flows from the Invicta Project still would have value to market participants.

178. Lonely Mountain Resources S.A.C. (“Lonely Mountain") was one such market participant: it purchased PLI from Pandion when Claimant already was in debt distress and requiring waivers of its defaults under the PLI Loan Agreement. As a rational investor acquiring a non-performing loan, Lonely Mountain would have valued PLI based on the FMV of the collateral (i.e., the Invicta Project), not based on Claimant's credit profile at the time.

179. Accuracy claims that Claimant never received the independent appraiser's valuation.¹³³ It is not possible to reasonably compare our conclusions with those of the appraiser without carefully examining the differences in the underlying assumptions between our analysis and the appraiser's assessment. We understand that Respondent requested the appraisal report, and that Claimant has not produced it, which prevents us (and Accuracy) from evaluating the reasonableness of the assumptions in the appraisal report and comparing them to our assumptions.

VIII. Our Opinion on the Calculation of Pre-Award Interest

180. In the Accuracy Second Report, Accuracy recognized that "[u]ltimately, the question of the correct pre-award interest rate to apply in this case is for the Tribunal" and that a "wide range of rates" have been used in the past cases.¹³⁴ We agree with Accuracy on those two points.

¹³² Accuracy Second Report, ¶ 5.2. ↩

¹³³ Accuracy Second Report, ¶ 4.63. ↩

¹³⁴ Accuracy Second Report, ¶ 4.50; First Report, ¶ 168. ↩

[Page 51]

181. With partial support from Accuracy,¹³⁵ Claimant alleged that the current high inflation environment would encourage the Respondent to delay payment of any damages that might become due:¹³⁶

"Indeed, in the current inflationary context, a respondent State would then have every incentive to delay payment of an award because an interest rate of UST+2% or SOFR+2% (or even LIBOR+2%) would mean that, contrary to the principles justifying awarding interest, the value of the amount owed to the Claimant would decrease over time in real terms. In other words, delaying payment or enforcement of an award would make the respondent State better off economically since the award would be worth less with the passage of time and the interest awarded for late payment would not be sufficient to offset the decrease in value of the principal."

182. We disagree with the statement above for the following three reasons.

183. First, the indexed rates that we have referenced (i.e., UST, SOFR, and LIBOR) are all variable, market-determined interest rates that respond to fluctuations in the inflationary and macroeconomic environment.

184. Second, the high inflation Claimant referenced is not expected to last into 2023. The United States Federal Reserve projected in December 2022, that the inflation rate is expected to decrease to 3.1% in 2023 and 2.5% in 2024.¹³⁷

185. Third, Claimant's statement ignores the mismatch in the currency in which Claimant has requested damages (US$) and the currency in which Peru collects revenues (PEN). This statement ignored any potential issues around the foreign exchange rate and the fact that Peru's income is primarily generated in local currency. As shown in Figure 22 below, Peru's local currency has been generally depreciating against US dollars.¹³⁸ This means that from Peru's perspective, when Peru's Sol is depreciating against the dollar the “value of the amount owed" would increase rather than "decrease". That is, Peru would have to pay a higher amount of local currency to obtain and pay the same amount of US dollars.

¹³⁵ Accuracy Second Report, ¶¶ 4.58, 6.64-6.65, and Appendix 6. Accuracy opined that the rates proposed by both Accuracy and AlixPartners would currently result in a negative real interest rate. ↩

¹³⁶ Claimant's Reply, ¶ 1054. ↩

¹³⁷ Federal Reserve, Summary of Projections, December 2022, Table 1. [AP-0067] ↩

¹³⁸ Capital IQ, US$PEN Exchange Rates [AP-0065]. From the Valuation Date of 26 August 2019 to the 31 August 2022 (the ending date for Accuracy's pre-award interest calculation), the Peruvian Sol deprecated by 12.0%. This equals to an annualized rate of 3.82%, which is higher than the average SOFR+2% (2.52%) or UST+2% (2.76%) in this period. See Appendix 2, "Pre-award interest". ↩

[Page 52]

Figure 22 – The Trend of US Dollar per Peruvian Sol

[Line chart showing the fluctuation of the US Dollar per Peruvian Sol from August 2018 to December 2022. The Y-axis ranges from 0.00 to 0.35. The X-axis shows dates at two-month intervals.]

186. The only reasonable situation that would incentivize Respondent to delay a payment due in foreign currency is when the local currency appreciates at a rate higher than the pre-award interest rate (as shown below), which is inconsistent with the historical trend.

Figure 23 – Illustration of the Interaction between Foreign Exchange Rate and the Pre-award Interest Rate

187. In the figure below, we update our calculation of pre-award interest under the rates of SOFR+2%¹³⁹ and UST+2% as of 31 August 2022 for the 590t/day and 355t/day scenarios.¹⁴⁰

¹³⁹ In our First Report, we used the 180-day average SOFR rate. In this report, we update to the SOFR rate sourced from the Federal Reserve Bank of New York as the basis for our calculation. ↩

¹⁴⁰ Appendix 2, "AlixPartners" and "Pre-award Interest". ↩

[Page 53]

Figure 24 – Revised Damages plus Pre-award Interest

188. The detailed calculations can be found in Appendix 2 and Appendix 3 for the 590t/day and 355t/day scenarios respectively.

IX. Our Opinion on Accuracy's Other Indicators of Value

189. As in the Accuracy First Report, Accuracy also proposed four additional analyses as indicators of the value of Claimant's Investment and as benchmarks for assessing the reasonableness of their damage assessments using the Primary Approach:

1. Market capitalization approach;
2. Sunk costs approach;
3. Transaction multiples approach; and,
4. Valuation results per the SRK Model and the Red Cloud Model.

190. Although Accuracy modified these analyses in their second report based on our comments, we are still of the view that these indicators are either inaccurate or do not demonstrate the reasonableness of Accuracy's quantification of damages under the Primary Approach.

A. Accuracy's Market Capitalization Approach Remains Problematic

191. In the Accuracy First Report, Accuracy estimated Lupaka's market capitalization on the Valuation Date as a benchmark to damages. Accuracy took the market capitalization as of 25 October 2018, the date on which Claimant made the public announcement concerning the Invicta Project and before the market reflected this information in Lupaka's share price.¹⁴¹ Accuracy then applied the daily percentage changes in the VanEck Vectors Junior Gold Miners ETF (“GDXJ") between 25 October 2018 to 26 August 2019 to Lupaka's actual market capitalization on 25 October 2018

¹⁴¹ Accuracy First Report, ¶ 8.5. ↩

[Page 54]

to calculate Lupaka's implied market capitalization on 26 August 2019. Finally, Accuracy added a control premium of 43.2% to the implied market capitalization.¹⁴²

192. In the Accuracy Second Report, Accuracy continued to use the approach from their first report. Accuracy provided three values (US$33.4 million, US$32.8 million and US$27.8 million) and concluded that the value of the Invicta Project would fall into this range from US$27.8 million to US$33.4 million. The US$33.4 million was sourced to the Accuracy First Report,¹⁴³ whereas the US$32.8 million and US$27.8 million were calculated with the following adjustments incorporated respectively:

1. Regarding the US$32.8 million, Accuracy accepted our position from our First Report that Claimant's stock price may have been distorted due to liquidity issues, and that the market capitalization on a specific date is not always realizable.¹⁴⁴ Therefore, Accuracy modified their approach to use the average of Lupaka's market capitalization from 13 September 2018 to 25 October 2018 (i.e., a 30-day trailing average) as the starting point,¹⁴⁵ instead of using Claimant's actual market capitalization as of 25 October 2018 as shown in Figure 25.
2. Regarding the US$27.8 million, Accuracy acknowledged our comment in our First Report that the implied market capitalization under Accuracy's approach was always lower than the actual market capitalization after 28 March 2018.¹⁴⁶ Therefore, Accuracy adjusted Lupaka's implied market capitalization by an "underperformance rate” of 16.9% between Claimant's share price and the GDXJ as of 25 October 2018.¹⁴⁷ Figure 25 below summarizes Accuracy's updated calculations.¹⁴⁸

¹⁴² Accuracy First Report, Appendix 7, "Market Cap". ↩

¹⁴³ Accuracy First Report, ¶ 8.18. ↩

¹⁴⁴ First Report, ¶ 186. ↩

¹⁴⁵ Accuracy Second Report, Appendix 5, "Share price v GDXJ". ↩

¹⁴⁶ First Report, ¶ 183. ↩

¹⁴⁷ Accuracy Second Report, Appendix 5, "Share price v GDXJ". ↩

¹⁴⁸ Accuracy Second Report, Appendix 5. ↩

[Page 55]

Figure 25 – Damages Calculated Under Accuracy's Updated Market Capitalization Approach

193. In summary, we disagree with Accuracy's result from their original approach in the Accuracy First Report (US$33.4 million) for the following reasons:

1. Accuracy relied on daily percentage changes in the GDXJ which demonstrated a weak relationship with Lupaka's share price;¹⁵³
2. Accuracy's control premium was not applicable;¹⁵⁴
3. Accuracy did not account for the underperformance of Lupaka relative to the GDXJ after 28 March 2018; and,¹⁵⁵
4. Accuracy did not adjust this number for the issue that Lupaka's share price may be distorted due to low trading volume.¹⁵⁶

¹⁴⁹ Accuracy First Report, Appendix 7, "Market Cap". ↩

¹⁵⁰ Accuracy's first alternative method in the Accuracy Second Report was to use Lupaka's 30-day trailing average market capitalization as the starting point. ↩

¹⁵¹ Accuracy's second alternative method in the Accuracy Second Report was to adjust the value in their first report by an underperformance rate ↩

¹⁵² US$16,021,120 is the 30-day trailing average of Lupaka's actual market capitalization from 13 September 2018 to 25 October 2018. ↩

¹⁵³ Paragraph 196. ↩

¹⁵⁴ Paragraph 201. ↩

¹⁵⁵ First Report, ¶ 183. ↩

¹⁵⁶ First Report, ¶ 186. ↩

[Page 56]

194. In summary, we disagree with Accuracy's result from the first alternative method in the Accuracy Second Report (US$32.8 million) for the following reasons:

1. Accuracy relied on daily percentage changes in the GDXJ which demonstrated a weak relationship with Lupaka's share price; and,¹⁵⁷
2. Accuracy's control premium was not applicable.¹⁵⁸

195. In summary, we disagree with Accuracy's result from the second alternative method in the Accuracy Second Report (US$27.8 million) for the following reasons:

1. Accuracy relied on daily percentage changes in the GDXJ which demonstrated a weak relationship with Lupaka's share price;¹⁵⁹
2. Accuracy's control premium was not applicable; and,¹⁶⁰
3. Accuracy's underperformance rate was understated.¹⁶¹

196. First, Accuracy based their calculations on an inappropriate dataset of daily percentage changes of the GDXJ. In our First Report, we performed a regression analysis on the daily data between returns of Claimant's stock and returns of the GDXJ from 1 January 2013 to 25 October 2018. Our analysis showed a beta coefficient of 0.34 and an R-squared of 1.8%, signifying weak and unreliable predictions in the response variable (i.e., daily percentage changes of Lupaka's stock price).¹⁶² As a market index, the GDXJ reflected general systematic risks, but did not capture the Claimant-specific risks discussed in Section IV, such as social license, execution, regulatory, and financing risks.

197. We understand that the regression analysis included in the Accuracy Second Report suggested a stronger relationship between returns of Claimant's stock price and returns of the GDXJ, with a beta of 0.88 and a R-squared of 17.7%.¹⁶³ However, Accuracy's regression analysis was based on the monthly data from 25 October 2013 to 25 October 2018, which was less volatile than the daily data that Accuracy applied for their implied market capitalization calculations. A regression analysis based on the daily data between 25 October 2013 to 25 October 2018 shows a low beta

¹⁵⁷ Paragraph 196. ↩

¹⁵⁸ Paragraph 201. ↩

¹⁵⁹ Paragraph 196. ↩

¹⁶⁰ Paragraph 201. ↩

¹⁶¹ Paragraph 209. ↩

¹⁶² First Report, Appendix 5, "Regression". ↩

¹⁶³ Accuracy Second Report, Appendix 5, "Monthly Beta". ↩

[Page 57]

coefficient of 0.15 and an R-squared of 0.2%.¹⁶⁴ The variance of the monthly data ranged from -18.7% to 27.0%, whereas that of the daily data ranged from -42.7% to 68.4%.¹⁶⁵ Figure 26 below demonstrates the differences between monthly data and daily data.

Figure 26 – Correlation Between Returns of Lupaka's Stock Price and Returns of the GDXJ from 25 October 2013 to 25 October 2018

198. Although Accuracy appeared to prove a stronger relationship between Claimant's stock price and the GDXJ on a monthly basis, they still applied the daily percentage changes of the GDXJ, rather than the monthly percentage changes, to calculate Lupaka's implied market capitalization. This mismatch and the weaker relationship indicated by the regression analysis on daily data lead us to conclude that this approach remains unjustified.

199. Moreover, Accuracy applied the percentage changes in the GDXJ market capitalization on Claimant's market capitalization in a 1:1 ratio, thus assuming a beta coefficient of 1 (i.e., ΔP_Lupaka=ΔP_GDXJ). In theory, all else equal, a beta coefficient of 0.88 means that for 1 unit increase in the GDXJ, Claimant's stock price would increase by 0.88 unit rather than the 1 unit that Accuracy assumed.

200. Applying the monthly percentage changes in the GDXJ from October 2018 to August 2019 and integrating the beta coefficient of 0.88 reduce Lupaka's implied market capitalization to US$22.6 million or US$23.0 million on a standalone basis, as shown in Figure 27.¹⁶⁶

¹⁶⁴ Appendix 4, "Daily Beta". ↩

¹⁶⁵ Appendix 4, "Share price v GDXJ (Monthly)" and "Share price v GDX (Daily)". ↩

¹⁶⁶ Appendix 4, "Implied Market Cap". ↩

[Page 58]

Figure 27 – Revised Calculation of Lupaka's Implied Market Capitalization Using Monthly Data Only

201. Second, Accuracy's 43.2% control premium inflated Claimant's implied value. Accuracy stated that adding a control premium was "consistent with both common valuation practice and observable market evidence”.¹⁶⁷ However, as we will discuss below, this approach did not actually follow the common practice or market evidence.

202. Accuracy oversimplified the concept of control premiums. The controlling interest itself has little inherent value but may be valuable if the exercise of control enhances the expected economic benefits and/or reduces the enterprise's risk for a hypothetical investor.¹⁶⁸ Therefore, a persuasive FMV calculation should be supported by a clear explanation of the benefits that are expected to be earned from control.¹⁶⁹ The value of control derives from and relies on a market participant's perspective on future growth in cash flows or reduced risks.

203. Due to the Invicta Project's high social license, execution and financing risks, its expected economic benefits are difficult to be reliably measured, reducing an

¹⁶⁷ Accuracy Second Report, ¶ 7.20. ↩

¹⁶⁸ The Appraisal Foundation, "The Measurement and Application of Market Participant Acquisition Premiums", 6 September 2017, pp. 9-12. [AP-0068] ↩

¹⁶⁹ The Appraisal Foundation, "The Measurement and Application of Market Participant Acquisition Premiums", 6 September 2017, pp. 13-17. [AP-0068] ↩

[Page 59]

investor's willingness to pay a control premium. As a result, Accuracy's method of applying historical premiums without clearly explaining what the value of control represents in this case inflates this indicator of value.

204. Moreover, some finance experts have proposed that the expected value of control is already reflected in a firm's share price. For example:¹⁷⁰

"In 2001, Pratt further clarified his position in the first edition of Business Valuation Discounts and Premiums. Pratt quoted an article by Mark Lee and then wrote that "given the current state of debate, one must be extremely cautious about applying a control premium to public market values to determine a control level of value." Similarly, Mercer's 2004 book agreed with Nath's 1990 article and explained that "unless there are cash flow-driven differences between the enterprise's financial control value and its marketable minority value, there will be no (or very little) minority interest discount.

Several professors of corporate law have questioned the assumption that most market prices include an IMD [implicit minority discount]. Professor Richard Booth wrote in 2001 that "it is not necessarily the case that actual market price is always less than fair market price." Professor William Carney argued in 2003 against assuming that market prices of most publicly traded shares include a significant IMD, concluding, "Even if all values, both present and potential, are valued in the market price for the firm's shares, one would not expect to find a discernible control premium in a widely held firm that is well managed and appears to offer little probability of a transfer of control."

The 2007 Hamermesh and Wachter article argued that premiums paid in acquisitions are not justification for assuming that market prices include IMD, stating, "[N]ot a single piece of financial or empirical scholarship affirms the core premise of the IMD - that public company shares systematically trade at a substantial discount to the net present value of the corporation."

205. Professor Aswath Damodaran of New York University's Stern School of Business noted that the stock price of every publicly traded firm includes an expected value of control, reflecting both the likelihood of changes in management and the value of the

¹⁷⁰ Fishman, Jay E., et al., "Standards of Value_Theory and Applications", 2013, pp. 165-166. [AP-0069] ↩

[Page 60]

changes.¹⁷¹ Although the extent to which an expected value of control is built into a stock price varies across firms and industries, applying an additional premium calls for a careful analysis on the underlying expectations of a stock price, especially when a firm like Claimant experienced ongoing SLO issues.

206. Accuracy also misused the reference premium data cited in the Accuracy Second Report. Accuracy's control premium relied upon a MergerStat report dated Q1'2020 and an RSM study from 2017.¹⁷² In the MergerStat report, the 10-year mining sector average control premium of 50.5% that Accuracy used included data from the metal mining industry, as well as the oil and gas extraction industry.¹⁷³ The control premiums of the metal mining industry alone averaged 34.9% in 2019.¹⁷⁴

207. In the RSM study, the average 20-day pre-announcement premium of 35.8% that Accuracy adopted fell 16.5% below the average 2-day pre-bid control premium of 29.9%, which reflected better the expectations of hypothetical buyers in case of a high level of bid speculation.¹⁷⁵

208. Therefore, the two reference premiums that Accuracy used overstated the 43.2% control premium compared to the more relevant reference points included in the same source documents. Setting those issues aside, we maintain our belief that a control premium should not be added to the calculation of Claimant's implied market capitalization. Removing the control premium would reduce Lupaka's adjusted market capitalization to US$22.9 million or US$19.4 million on a standalone basis, as shown in Figure 28.

¹⁷¹ Damodaran, Aswath, "The Value of Control_Implications for Control Premia", pp. 44-50. [AP-0070] ↩

¹⁷² Accuracy First Report, ¶ 8.16. ↩

¹⁷³ Mergerstat, Control Premium Study, 1st Quarter 2020, p. 8. [AC-0035] ↩

¹⁷⁴ Mergerstat, Control Premium Study, 1st Quarter 2020, p. 27. [AC-0035] ↩

¹⁷⁵ RSM, Control Premium Study 2017, p. 9. [AC-0036] ↩

[Page 61]

Figure 28 – Revised Calculation of Lupaka's Implied Market Capitalization By Removing Control Premium Only

209. Third, Accuracy's adjustment factor of 16.9% understated Lupaka's underperformance relative to the GDXJ from 28 March 2018 to 25 October 2018. During this period, Claimant's actual market capitalization compared to the GDXJ changed at a rate ranging from 0.6% to -35.9%, as displayed in Figure 29.¹⁷⁶

¹⁷⁶ Appendix 4, "Share price v GDXJ (Daily)". ↩

[Page 62]

Figure 29 - Lupaka's Underperformance Rate between 28 March 2018 to 25 October 2018

[Line chart showing Lupaka's underperformance rate from March 2018 to September 2018. The Y-axis ranges from -40.0% to 5.0%. The X-axis shows dates at one-month intervals.]

210. The average and median underperformance rates were 20.8% and 21.0%, each above the 16.9% rate that Accuracy used in their calculation.¹⁷⁷ Using the average percentage decrease of 20.8% would reduce Accuracy's adjusted market capitalization to US$26.5 million on a standalone basis,¹⁷⁸ as shown in Figure 30.

Figure 30 – Revised Calculation of Lupaka's Market Capitalization Modifying Underperformance Rate Only

¹⁷⁷ Appendix 4, "Share price v GDXJ (Daily)". ↩

¹⁷⁸ Accuracy only applied an "underperformance rate" in the second alternative method of Accuracy Second Report. ↩

[Page 63]

211. In summary, revising Accuracy's approach by applying the monthly percentage changes in the GDXJ, removing the irrelevant control premium, and using the average underperformance rate reduces the adjusted market capitalization to US$22.6 million or US$18.2 million,¹⁷⁹ as shown in Figure 31.

Figure 31 Revised Calculations of Lupaka's Implied Market Capitalization

212. In conclusion, Accuracy's methodology of calculating Claimant's implied market capitalization was not generally accepted valuation practice, and the resulting range from US$27.8 million to US$33.4 million was overstated. Therefore, Accuracy's market capitalization approach does not support Accuracy's Project FMV conclusion.

B. Accuracy's Sunk Costs Approach Is Irrelevant

213. In the Accuracy Second Report, Accuracy reiterated their belief that Claimant's sunk costs was a relevant benchmark and should represent the floor of the damages estimate. Accuracy attributed the exploration expenditures expensed prior to 31 July 2017 and those capitalized between 1 August 2017 and 26 August 2019 to the total amount of sunk costs without adjustments,¹⁸⁰ as detailed in Figure 32.

¹⁷⁹ Appendix 4, "Summary_Market Cap". ↩

¹⁸⁰ Accuracy First Report, Appendix 8, "Sunk Costs". Accuracy Second Report, Appendix 5, "Claimant's sunk costs". ↩

[Page 64]

Figure 32 - Accuracy's Calculation of Sunk Costs

214. On the grounds that the sunk costs approach reflected Claimant's wish to "(at the very least) service its debts”,¹⁸¹ Accuracy then applied interest rate factors to each year's actual sunk costs. In their first report, Accuracy used the annual interest rate of 10.1% from the PLI Loan Agreement, based on a total amount of interest of US$7.0 million paid on the principal of US$11.3 million over 60 months.¹⁸² In their second report, Accuracy switched to a higher rate of 12.0%, which was estimated based on the financing cash flows in the 355t/day scenario.¹⁸³

215. We disagree that historical exploration expenditures of the Invicta Project would be a proxy for its value in this case.

216. First, according to the Special Committee of the Canadian Institute of Mining, Metallurgy & Petroleum on the Valuation of Mineral Properties (“CIMVAL”) Standards, the cost approach is not appropriate for the valuation of development properties.¹⁸⁴ According to CIMVAL's definition, a Development Property is a Mineral Property that:¹⁸⁵

"a Mineral Property that contains Mineral Reserves and/or Mineral Resources and for which economic viability has been demonstrated by a Feasibility Study or Pre-Feasibility Study and includes a Mineral Property that has a Current positive Feasibility Study or Pre-Feasibility Study but that is not yet in production."

217. As discussed in our First Report, although the Project was not subject to a pre-feasibility study or feasibility study, we understand that it could be classified as a Development Property since it contained mineral resources, obtained a preliminary

¹⁸¹ Accuracy First Report, ¶ 8.31. ↩

¹⁸² Accuracy First Report, ¶ 8.31, Footnote 203. ↩

¹⁸³ Accuracy Second Report, ¶ 7.43. ↩

¹⁸⁴ CIMVAL Standards (2019), p. 16. [AP-0007] ↩

¹⁸⁵ CIMVAL Standards (2019), p. 32. [AP-0007] ↩

[Page 65]

economic assessment and was proceeding towards commercialization.¹⁸⁶ Accuracy also acknowledged that the Invicta Project was "at a more advanced stage" than a typical mineral resource property,¹⁸⁷ which indicated that the Project would fall into the category of development properties, as shown in Figure 33.¹⁸⁸ However, Accuracy did not explain how the cost approach was a relevant indicator of value.

Figure 33 – Valuation Approaches per CIMVAL Standards

218. Furthermore, Claimant's sunk costs were not an appropriate estimate of costs that contributed to the Project's value, as those costs captured only the past exploration expenditures incurred up to the Valuation Date but could not account for the uncertainty and risks that Claimant faced (i.e., the fundamental flaws discussed in Section IV).

219. Second, Accuracy's application of interest rate factors to sunk costs inflated the value of sunk costs inappropriately. These interest rate factors ranged from 3.8% to 118.9% and increased the actual dollar amount of sunk costs by 73.6%, approximately doubling the actual dollar amount of sunk costs in some instances.¹⁸⁹ To calculate their interest rate factors, Accuracy chose the higher rate (12.0%) of the two effective interest rates in the 355t/day scenario (12.0% on a post-tax basis) and the 590t/day scenario (11.4% on a post-tax basis), without any explanation on why or how one rate should be preferred over the other.

220. Although Accuracy ignored Claimant-specific risks in their other damage calculations, they chose to use a Claimant-specific interest rate instead of a market interest rate to bring forward the value of Claimant's sunk costs.

221. As we explained in our First Report, recognizing Lupaka's historical costs at their actual dollar value, rather than with a cumulative rate of return, would be more appropriate. To bring these amounts forward to the Valuation Date, applying a pre-

¹⁸⁶ First Report, ¶ 102. ↩

¹⁸⁷ Accuracy First Report, ¶ 4.31. ↩

¹⁸⁸ CIMVAL Standards (2019), p. 16. [AP-0007] ↩

¹⁸⁹ Accuracy Second Report, Appendix 5, "Claimant's sunk costs". ↩

[Page 66]

award interest rate of UST+2% or SOFR+2% should be sufficient. Since the Federal Reserve Bank of New York began the publication of SOFR on 2 April 2018,¹⁹⁰ we used UST+2% to illustrate a revised amount of interest on sunk costs,¹⁹¹ as shown in Figure 34. This method reduced the interest from US$18.3 million to US$3.7 million.

Figure 34 – Revised Accuracy's Interest on Sunk Costs Using UST+2%

222. In conclusion, Accuracy's sunk costs approach is not a relevant or reasonable benchmark of the FMV of the Project.

C. Accuracy's Market Transactions Are Not Comparable

223. In the Accuracy First Report, Accuracy identified 26 recent gold industry corporate transactions over US$1.0 million and calculated indexed valuation multiples using implied enterprise value of the acquired companies. Accuracy compared these multiples with the implied valuation multiples in the 355t/day and 590t/day scenarios to benchmark the reasonableness of the Primary Approach. The Accuracy Second Report retained the same set of transactions and updated the analysis to compute multiples based on equity value, rather than enterprise value.¹⁹² Figure 35 below summarizes Accuracy's transactions multiples based on implied equity value in their second report.¹⁹³

¹⁹⁰ The Federal Reserve Bank of New York, "Additional Information about SOFR". [AP-0071] ↩

¹⁹¹ Appendix 4, "UST+2%". ↩

¹⁹² Accuracy Second Report, ¶¶ 7.55-7.56. ↩

¹⁹³ Accuracy Second Report, ¶¶ 7.57, 7.65. ↩

[Page 67]

Figure 35 - Accuracy's Updated Transactions Multiples

224. However, this approach is still unreliable, as Accuracy did not address the following issues that we raised in our First Report.

225. First, Accuracy ignored the comparability issues between the 26 transactions and the Project due to the variability in their production stages and geography. We noted that 18 of the 26 transactions had more than one producing mine. Only eight of the 26 transactions were at the development or exploration stage, the stage to which the Invicta Project was closest. Figure 36 shows the indexed multiples based on the equity value of these two subsets.¹⁹⁴

Figure 36 -Transaction Multiples of Illustrative Subsets

226. In the Accuracy Second Report, Accuracy contended that the valuation multiples that resulted from excluding transactions with producing mines were lower.¹⁹⁵ However, Accuracy did not explain whether or why those lower multiples would be inappropriate (except to note they were lower). Companies with more than one producing mine benefited from more diversified risks and more predictable cash flows. Given that Claimant had no producing mines, and that the Invicta Project faced various project-specific risks, using transactions with more than one producing mine as benchmarks inflated Claimant's indicated value. Figure 37 details the number of

¹⁹⁴ Appendix 4, "Revised Market Transactions". ↩

¹⁹⁵ Accuracy Second Report, ¶ 7.58. ↩

[Page 68]

producing mines the target companies in Accuracy's comparable transactions owned.¹⁹⁶

Figure 37 - Number of Producing Mines of Accuracy's Comparable Transactions

227. In response to Accuracy, we examined the target companies of these 18 transactions. For example, one of the target companies, Goldcorp Inc. announced its transaction on 14 January 2019, with an indexed multiple of US$225.7/oz based on equity value.¹⁹⁷ As a subsidiary of Newmont Corporation, it had eight producing mines in five countries and generated US$2,984.9 million in revenue as of 31 December 2018.¹⁹⁸ Goldcorp Inc.'s debt was considered investment grade by Standard and Poor's, indicating low risks and positive prospects.¹⁹⁹

¹⁹⁶ Appendix 4, "Revised Market Transactions". ↩

¹⁹⁷ Accuracy Second Report, Appendix 5, "Table 7.1". ↩

¹⁹⁸ Capital IQ, Goldcorp Inc_PrivateCompany, p. 1. [AP-0072] The revenue of Goldcorp Inc. was CA$4,072.9 million as of 31 December 2018. Appendix 4, "CAD per USD". The CAD/USD exchange rate was 1.3645 on 31 December 2018. CA$4,072.9 million / 1.3645 = US$2,984.9 million. ↩

¹⁹⁹ Capital IQ, Goldcorp Inc_PrivateCompany, pp. 5-6. [AP-0072] ↩

[Page 69]

228. Another target company, Lake Shore Gold Corp., announced its transaction on 8 February 2016, with an index multiple of US$1,041.7/oz based on equity value.²⁰⁰ It owned two producing mines in Canada and had generated revenues of US$195.6 million as of 31 December 2015.²⁰¹ Therefore, the 18 transactions were not comparable to the Invicta Project and thus should be excluded.

229. Although no two companies or transactions are exactly the same, a valuation multiples analysis assumes that the companies or transactions used should have similar underlying assets, size, cash flows, etc. Misusing companies or transactions that are not comparable to the subject creates unjustified biases on its value.²⁰²

230. Second, Accuracy did not analyze the relative risks faced by the projects acquired in these 26 transactions. Looking at the eight target companies with only development or exploration projects first, we noted that Romarco Minerals Inc. which had the highest indexed multiple of US$451.7/oz was located in South Carolina, a region described as a “low-risk jurisdiction”.²⁰³ Removing Romarco Minerals Inc. from the eight companies reduced the mean indexed multiple from US$133.0/oz down to US$87.4/oz.²⁰⁴

231. Risks and opportunities in the mining industry vary among jurisdictions. Indices such as the Investment Attractiveness Index rank mining risks by country.²⁰⁵ Contrary to Accuracy's allegation, we did not assume that countries across Latin America were exposed to have “exact same level of risk".²⁰⁶ The fact is that most countries in this region face persistent conflicts due to "insufficient consultation with affected communities".²⁰⁷ Therefore, it is not reasonable to include mining companies or transactions without taking geographic characteristics into consideration.

232. We also noted that in their second report, Accuracy identified that three out of the eight transactions (i.e., two transactions of Rio Novo Gold Inc. and one transaction of Bison Gold Resources Inc.) had pre-production projects that appeared to be "at

²⁰⁰ Accuracy Second Report, Appendix 5, "Table 7.1". ↩

²⁰¹ Capital IQ, Lake Shore Gold Corp_PrivateCompany, p. 1. [AP-0073] The revenue of Lake Shore Gold Corp. was CA$271.4 million as of 31 December 2015. Appendix 4, "CAD per USD". The CAD/USD exchange rate was 1.3874 on 31 December 2015. CA$271.4 million / 1.3874 = US$195.6 million. ↩

²⁰² Damodaran, Aswath, "An Introduction to Valuation", pp. 24-25. [AP-0074] ↩

²⁰³ Financial Post, "OceanaGold Corp buys Romarco Inc for $856 million, gains low-cost mine in low-risk area", 30 July 2015. [AP-0038] ↩

²⁰⁴ Appendix 4, "Summary_Market Transactions". ↩

²⁰⁵ Fraser Institute, "Annual Survey of Mining Companies", 2021. [AP-0075] ↩

²⁰⁶ Accuracy Second Report, ¶ 7.62. ↩

²⁰⁷ Purdy, Caitlin and Castillo, Rodrigo, "The Future of Mining in Latin America", July 2022, pp. 7-8. [AP-0076] ↩

[Page 70]

significantly earlier stage of development" than the Invicta Project.²⁰⁸ Accuracy calculated the mean and median indexed multiples of US$190.4/oz and US$212.8/oz by excluding only these three transactions. However, we believe that the 18 transactions with more than one producing mine and the Romarco Minerals Inc. transaction should also be removed.

233. Figure 38 illustrates the impact of revising the scope of Accuracy's existing list of comparable transactions without adjusting for the risks that Claimant faced. As shown below, Accuracy's post-tax NPV of US$34.3 million and US$46.9 million from the 355t/day scenario and the 590t/day scenario are both higher than those of our three alternative scenarios.

Figure 38 - Claimant's Implied Equity Value Based on Adjusted Market Transactions

²⁰⁸ Accuracy Second Report, ¶ 7.61. ↩

²⁰⁹ Accuracy Second Report, Appendix 5, "Table 7.2". ↩

²¹⁰ Accuracy Second Report, Appendix 5, "Table 7.2". ↩

²¹¹ Accuracy Second Report, Appendix 5, "Table 7.1". ↩

²¹² Accuracy Second Report, Appendix 5, "Table 7.1". ↩

²¹³ Appendix 4, "Revised Market Transactions". ↩

²¹⁴ Appendix 4, "Revised Market Transactions". ↩

[Page 71]

234. Regardless of the above, neither the full list of 26 transactions nor the smaller subsets in the illustrative calculations above adequately demonstrate that Accuracy's damages are reasonable. This approach does not account for any of Claimant-specific risks (see Section IV) and when making adjustments to improve the comparability of the transactions, few relevant data points remain. In conclusion, Accuracy's market transaction analysis is unreliable and does not indicate the FMV of the Project.

D. The SRK Model and Red Cloud Model

235. Accuracy's fourth indicator of value was the NPVs from the SRK Model and the Red Cloud Model. Accuracy claimed that because these NPVs were in line with or higher than their calculations for the 590t/day scenario and the 355t/day scenario, their damage assessments using the Primary Approach were reasonable. However, these two models are not indicative of Claimant's damages.

236. First, an issued that we raised in our First Report, the SRK Model and the Red Cloud Model did not consider the same fundamental flaws that Accuracy ignores in their own analysis (see Section IV). Neither model took social license, regulation, execution, and financing risks into account. Any of these risks could undermine Claimant's ability to continue operations or generate positive cash flows.

237. Specifically, the SRK Model does not account for the third-party ore processing issues discussed in Section V.A.i. Whether or when these issues could be resolved would impact Claimant's production schedule and increase its risk of default.

238. Second, Accuracy states that the SRK Model under the SRK PEA's mine plan was sufficient for estimating mineral resources. In Accuracy's But-For Scenario, Claimant could have "materially completed all development work” and “obtained third-party financing via the PLI Loan"; as a result, Accuracy claims that Claimant would not have needed further any feasibility studies.²¹⁶ However, Accuracy again ignored the fundamental flaws (see Section IV) and the uncertainty that the SRK PEA itself acknowledged:

²¹⁵ Appendix 4, "Revised Market Transactions". ↩

²¹⁶ Accuracy Second Report, ¶ 7.70. ↩

[Page 72]

"This preliminary economic assessment is, however, not adequate to confirm the economics of the study. A preliminary-feasibility study, or feasibility study, as defined in Canadian Securities Administrators National Instrument 43-101, containing mineral reserve estimates is required for this purpose.”²¹⁷

239. Third, Accuracy believed that applying a higher premium to the discount rate in the 590t/day scenario could help to offset the unreliability of the Red Cloud Model. As discussed in Section V.A.viii, Accuracy's selected discount rate was not adequate to capture all of the relevant risks that the Project was exposed to.

240. In conclusion, the NPVs in the SRK Model and the Red Cloud Model are not relevant indicators of the FMV of the Project.

Χ. Expert Declaration

241. We declare that:

1. We understand that our duty in giving evidence in this Arbitration is to assist the Tribunal in deciding issues in respect of which expert evidence is adduced. We have complied with, and will continue to comply with, that duty.
2. We understand that our expert report is to be objective and impartial and that it is to include everything we consider relevant to the expert opinions expressed.
3. We confirm that this is our own independent, objective unbiased opinion which has not been influenced by the pressures of the dispute resolution process or by any party to the arbitration.
4. We confirm that all matters upon which we have expressed an opinion is within our area of expertise.
5. We confirm that we have referred to all matters which we regard as relevant to the opinion We have expressed and have drawn to the attention of the Tribunal all matters, of which we are aware, that might adversely affect our opinion.
6. We confirm that, at the time of providing our written opinion, we consider it to be complete and accurate and constitute our true, professional opinion.

²¹⁷ SRK Consulting PEA dated 13 April 2018, p. xi. [AC-0002] ↩

[Page 73]

7. We confirm that if, subsequently, we consider this opinion requires any correction, modification, or qualification, we will notify the Parties and the Tribunal forthwith.
8. We confirm that we have made clear which facts and matters referred to in this report are within our knowledge and which are not. Those that are within our knowledge, we confirm to be true.

Signature

Isabel Santos Kunsman
Partner & Managing Director
25 JANUARY 2023

Signature

Alexander Lee
Director
25 JANUARY 2023

[Page 74]

Appendix 1 - Scope of Review

[Page 75]

[Page 76]

Appendix 2 - Adjustments to Accuracy's 590t/day Scenario

[Transmitted in native format]

[Page 77]

Appendix 3 - Adjustments to Accuracy's 355t/day Scenario

[Transmitted in native format]

[Page 78]

Appendix 4 - Adjustments to Other Indicators of Value

[Transmitted in native format]

[Page 79]

Appendix 5 - Estimate of the Cost for Social License

[Transmitted in native format]