Accuracy

Arbitration before the International Centre for Settlement
of Investment Disputes

ICSID Case No. ARB/20/46

Second Expert Report of Accuracy

Confidential

21 September 2022

Transactions

Disputes

Crises

Decisions

[Page 2]

TABLE OF CONTENTS

[Page 3]

[Page 4]

Glossary

[Page 5]

1 Introduction and instructions

A. Factual background

1.1 This report (the “Second Accuracy Report”) has been prepared by Accuracy in connection with the Dispute (ICSID Case No. ARB 20/46) between Lupaka Gold Corp. (Claimant) and the Republic of Peru (Respondent).

1.2 On 1 October 2012, Claimant acquired the Invicta Project, located in the Huaura province in Peru, through its acquisition of IMC. Over a number of years, Claimant obtained relevant authorisations and permits from the Peruvian authorities and completed development works with a view to commencing exploitation of the mine.¹

1.3 On 30 June 2016 (and subsequently amended in 2017 and 2018),² Claimant entered into a loan agreement with PLI in order to fund the development of the Invicta Project through pre-paid gold forward purchase agreements (the PLI Loan).

1.4 On 14 October 2018, the leaders of the Parán community directed men to forcibly evict Claimant and its personnel from the premises of the Invicta Project, and blocked access to the mine. The Blockade continued indefinitely and Claimant was never able to regain access to the site.

1.5 We understand that, at the time of the Blockade, Claimant had materially completed the development of the Invicta Mine and was close to commencing full commercial production. As a result of the Blockade and the alleged failure of the Peruvian authorities to resolve the situation, it was unable to produce gold and generate cash flows to enable it to fulfil its financial obligations to creditors, including to make gold repayments under the PLI Loan.

1.6 Subsequently, on 26 August 2019, Claimant's shares in IMC were seized following the initiation of foreclosure proceedings by Lonely Mountain, which controlled PLI at this date.

1.7 Claimant's case is that Peru's acts and omissions in relation to the Invicta Project breached the FTA, and resulted in the unlawful expropriation of Claimant's investment.³

1.8 Accuracy was instructed by LALIVE, counsel for Claimant in the Dispute, to prepare an expert report assessing the damages sustained by Claimant as a result of Respondent's Alleged Breaches of the FTA as at the Valuation Date of 26 August 2019 (the “First Accuracy Report”). The First Accuracy Report was submitted on 1 October 2021 along with Claimant's Memorial (“Claimant's Memorial”).

¹ The factual background to the Dispute was described in Section 3 of the First Accuracy Report and is summarised above. A more detailed background is set out in in the RfA and Claimant's Memorial. ↩

² Claimant's Memorial, paragraph 42. In the First Accuracy Report, we incorrectly stated that the PLI Loan was initially entered into on 30 June 2017 (First Accuracy Report, paragraph 3.18); this should have read 30 June 2016. ↩

³ Claimant's Memorial, Section 4 ↩

[Page 6]

1.9 Respondent's Counter-Memorial (“Respondent's Counter-Memorial”) was submitted on 24 March 2022. Respondent's comments on quantum are informed by an expert report submitted on the same date by Isabel Santos Kunsman and Alexander Lee of AlixPartners LLP (“AlixPartners”) (the “AlixPartners Report”).

B. Instructions

1.10 We are further instructed by LALIVE to:

a) Review and comment on the AlixPartners Report; and

b) Update our damages assessment in light of the additional information made available since the date of the First Accuracy Report, including an independent mining expert report by Micon International Limited (“Micon”) dated 20 September 2022 (the "Micon Report").

C. Preparation of this report

1.11 The authors of this report are Erik van Duijvenvoorde and Edmond Richards, who authored the First Accuracy Report, and whose CVs were appended therein.

1.12 The disclaimers set out at paragraphs 1.10 to 1.14 of the First Accuracy Report apply equally to this report.

1.13 Information sources which we have relied upon in the preparation of this report are listed in Appendix 1.

1.14 Unless they are redefined in this report, we use the same defined terms as in the First Accuracy Report. The glossary includes all of the newly defined terms used in this report.

1.15 The absence of discussion or rebuttal of any particular topic in the AlixPartners Report should not be interpreted as implicit agreement thereon.

D. Summary of the First Accuracy Report

1.16 In the First Accuracy Report, we were instructed to assess damages incurred by Claimant as a result of Respondent's Alleged Breaches as at 26 August 2019, the Valuation Date, being the effective foreclosure date. We were further instructed to assess damages by reference to the fair market value (FMV) of Claimant's Investment in Peru, in line with Article 812 of the FTA, and to apply pre-award interest at a rate of LIBOR +2% to the Report Date, being 1 October 2021.⁴

⁴ First Accuracy Report, paragraph 2.3 ↩

[Page 7]

1.17 In assessing damages, we also gave consideration to the principle of full reparation, as well as guidance from the Canadian Institute of Mining, Metallurgy & Petroleum on the Valuation of Mineral Properties (CIMVAL).⁵

1.18 We assessed the FMV of Claimant's Investment at the Valuation Date, using an income (DCF) approach, under two distinct scenarios: the 355t/day Scenario and the 590t/day Scenario.⁶

1.19 Our valuation of the 355t/day Scenario relied upon a financial model prepared by SRK, the SRK Model, which set out an initial valuation of the Invicta Project based on the six-year PEA Mine Plan. Assuming production of c.355t/day, the SRK Model computed a post-tax NPV of USD 43.4m.⁷ We made two broad categories of adjustments to the model; first, adjustments to reflect the passage of time between April 2018 and the Valuation Date, and second, adjustments based on our view on certain valuation assumptions. In the former category, the main adjustments related to updated metal price expectations and the updated status of pre-production works, which generally increased SRK's valuation. In the latter category, we applied our own working capital and discount rate assumptions, which generally reduced SRK's valuation. Our resulting valuation of the 355t/day Scenario was USD 44.2m.⁸

1.20 Our valuation of the 590t/day Scenario relied upon a financial model prepared by Red Cloud, a market dealer focused on the junior resource sector, who updated the SRK Model to reflect the prospective purchase of the Mallay Plant, including an estimate of the additional capital expenditure required and assuming an increase in production and processing capacity from c.355t/day to c.590t/day. Based on these updates, the Red Cloud Model computed a post-tax NPV of USD 86.3m.⁹ In addition to applying the adjustments to the SRK Model summarised above, we also adjusted the Red Cloud Model to reflect the average resource grade of the Invicta Project as a whole and increased the discount rate as compared to the 355t/day Scenario to reflect the (limited) additional uncertainty associated with the 590t/day Scenario. These additional adjustments generally decreased Red Cloud's valuation. Our resulting valuation of the 590t/day Scenario was USD 63.6m.¹⁰

1.21 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.¹¹

⁵ First Accuracy Report, paragraph 2.4 ↩

⁶ First Accuracy Report, paragraph 2.7 ↩

⁷ First Accuracy Report, paragraph 2.5 ↩

⁸ First Accuracy Report, paragraph 2.8 ↩

⁹ First Accuracy Report, paragraph 2.6 ↩

¹⁰ First Accuracy Report, paragraph 2.9 ↩

¹¹ First Accuracy Report, paragraph 2.10 ↩

[Page 8]

1.22 To provide further comfort over the reasonableness of our damages assessments using an income approach, we benchmarked our calculations to other indicators of value of the Invicta Project, including (i) Claimant's market capitalisation; (ii) Claimant's sunk costs; (iii) relevant gold mining industry transaction multiples; and (iv) other contemporaneous valuation evidence such as the SRK Model and Red Cloud Model.¹² Our assessments of damages using the income approach were broadly consistent with these benchmarks.¹³

1.23 After applying pre-award interest to the date of the First Accuracy Report, we calculated damages of USD 29.9m and USD 50.5m under the 355t/day Scenario and 590t/day Scenario respectively.

Table 1.1. Summary of damages in the First Accuracy Report

Source: First Accuracy Report, Table 2.1

1.24 We did not give any consideration to the tax consequences of any award and, therefore, if a future award is taxable in the hands of Claimant it may be necessary to apply a tax gross-up to the overall damages figure.¹⁴

E. Summary of the AlixPartners Report

1.25 AlixPartners were asked by Respondent to review the First Accuracy Report and offer their independent expert opinion on whether and to what extent our calculations correctly measured Claimant's damages as a result of Respondent's Alleged Breaches.¹⁵

1.26 Based on our reading and interpretation of the AlixPartners Report, we split their critique of our damages assessment set out in the First Accuracy Report into four sub-sections:

a) AlixPartners' comments on so-called “fundamental flaws" in our analysis;

b) AlixPartners' comments on our damages assessment;

c) AlixPartners' comments on the financing of the Invicta Project; and

d) AlixPartners' comments on the other indicators of value of the Invicta Project.

¹² First Accuracy Report, paragraph 8.1 ↩

¹³ First Accuracy Report, paragraph 2.11 ↩

¹⁴ First Accuracy Report, paragraph 4.52 ↩

¹⁵ AlixPartners Report, paragraph 12 ↩

[Page 9]

AlixPartners' comments on so-called “fundamental flaws” in our analysis

1.27 Firstly, AlixPartners highlight four supposed “fundamental flaws” and risks inherent to the Invicta Project which, if unresolved, would result in the FMV of the project being nil at the Valuation Date.

1.28 To a large extent, these “fundamental flaws” relate to factual matters upon which we were not instructed to comment on in the First Accuracy Report and which fall outside the scope of our expertise. To the extent that we are able to do so, we set out our comments on the "fundamental flaws" in Section 3.

AlixPartners' comments on our damages assessment

1.29 Secondly, AlixPartners provide comments on our damages assessment, based on the FMV of the Invicta Project, set out in the First Accuracy Report. We split their comments into three categories.

Comments on our operational and technical assumptions

1.30 AlixPartners provide a limited number of comments upon the operational and technical assumptions used in our damages assessment, and challenge some of the adjustments we made to the SRK Model and Red Cloud Model in order to estimate the FMV of the Invicta Project under the two different production scenarios:

a) 590t/day Scenario: AlixPartners adjust our FMV calculation to (i) add back cash flows from the Mallay Plant and G&A expenses, as modelled in the Red Cloud Model (which we excluded from our assessment in the First Accuracy Report);¹⁶ and (ii) deduct USD 13m of additional debt required to acquire the Mallay Plant, which AlixPartners argue we fail to account for in our analysis.

b) 355t/day Scenario: AlixPartners propose a revised FMV, in which they reverse our adjustment to remove contingent cash flows related to closure costs.

1.31 In the table below, we set out AlixPartners' revised FMV of the Invicta Project at the Valuation Date (before applying pre-award interest) under the 590t/day Scenario and 355t/day Scenario.

Table 1.2. Revised FMV of the Invicta Project set out in the AlixPartners Report

Source: AlixPartners Report, Figures 21 & 22

¹⁶ First Accuracy Report, paragraph 6.19c) ↩

[Page 10]

Comments on our valuation assumptions

1.32 In addition, AlixPartners provide comments on (i) the discount rate; and (ii) the pre-award interest rate used in the First Accuracy Report, and present sensitivities thereon.¹⁷

Comments on the residual value of IMC's shares in the Actual Situation

1.33 Finally, AlixPartners argue that if the Tribunal finds that Respondent is not liable for PLI's actions, then the value of Claimant's Investment in the Actual Situation could be USD 13m instead of nil (and that damages should therefore be reduced by the same amount).¹⁸

1.34 We set out, and respond to, AlixPartners comments on the assumptions underlying our damages assessment in the First Accuracy Report in further detail in Section 4, and present updates to our damages assessment in Section 6.

AlixPartners' comments on the financing of the Invicta Project

1.35 According to AlixPartners, in the First Accuracy Report, we assumed incorrectly that Claimant would have obtained “friendlier financing terms”¹⁹ from a new investor compared to its existing debt under the PLI Loan and ignored any risks that Claimant would have faced in obtaining additional funds for its planned acquisition of the Mallay Plant under the 590t/day Scenario.

1.36 These comments appear to stem primarily from a misunderstanding of the damages framework we applied in the First Accuracy Report.

1.37 We consider AlixPartners' comments on the financing of the Invicta Project further in Section 5.

AlixPartners' comments on the other indicators of value of the Invicta Project

1.38 Finally, AlixPartners challenge the accuracy, reasonableness, and relevance of our other indicators of value of the Invicta Project and claim that there are both theoretical (conceptual) and technical deficiencies in our analysis.²⁰

1.39 We set out, and respond to, these comments in further detail in Section 7.

F. Overview of the Micon Report

1.40 The Red Cloud Model, which was the starting point for our 590t/day Scenario in the First Accuracy Report, was an updated version of the SRK Model and provided a preliminary estimate of the estimated cash flows that would have been generated under the Mallay Acquisition Plan.

¹⁷ AlixPartners Report, Figures 21, 22, 23 & 24 ↩

¹⁸ AlixPartners Report, paragraph 97 ↩

¹⁹ AlixPartners Report, paragraph 148 ↩

²⁰ AlixPartners Report, Section IX ↩

[Page 11]

1.41 Micon, acting as an independent mining expert, were instructed to:²¹

a) Review the PEA, Red Cloud Model and other contemporaneous technical documents and reports pertaining to the Invicta Project;

b) Adjust the PEA production and cost schedules to reflect the actual situation at the Invicta Mine in October 2018 (i.e. at or around the Blockade Date);

c) Ascertain the validity of assumptions made by Red Cloud, which reflect the acquisition of an offsite processing facility (the Mallay Plant) and which underlie our 590t/day Scenario;

d) Opine on the ability of Claimant to meet its obligations under the existing terms of the PLI Loan and the proposed amendment thereto; and

e) Opine on possible causes of variances between the reported grade of materials mined during pre-production activities and forecast grades.

Micon's updates to the 355t/day Scenario

1.42 Micon have reviewed the PEA Mine Plan underlying the 355t/day Scenario, and adjusted the development and production schedules to reflect the following:²²

a) The actual development works carried out by Claimant prior to the Blockade Date (including applying SRK's assumed percentage of life-of-mine development in each year to the total remaining development requirement);

b) Claimant's need to identify reliable toll millers capable of routinely and efficiently processing 355t/day and to implement closer supervision of toll milling operations; and

c) Restating the schedules into project years running from 1 September to 31 August.

1.43 In order ensure the availability of sufficient toll milling capacity, Micon assume a three-month production ramp-up period in the But-For Situation. Specifically, they have assumed production of 100t/day in November 2018, 200t/day in December 2018, 300t/day in January 2019 and 355t/day thereafter.²³

²¹ Micon Report, paragraph 2 ↩

²² Micon Report, paragraph 83 ↩

²³ Micon Report, paragraph 86 ↩

[Page 12]

Micon's updates to the 590t/day Scenario

1.44 In addition, the Micon Report revisits certain limitations of the Red Cloud Model and makes several adjustments to the 590t/day Scenario to provide more accurate estimates, notably:²⁴

a) Red Cloud assumed that mine development costs would rise pro rata with the increase in production from 355t/day to 590t/day, with no revisions to the mine layout to ensure access to the additional ore. Micon refine this assumption by developing a mine layout to access sufficient material to support the 590t/day Scenario, and include an additional provision for development costs under this scenario;

b) As highlighted in the First Accuracy Report,²⁵ annual average ore grades were assumed to remain the same in the Red Cloud Model (i.e. the 590t/day Scenario) as the PEA Mine Plan (i.e. the 355t/day Scenario), despite the average grades in the PEA Mine Plan being above the average grade for the Invicta Mine as a whole. As noted below, Micon have tested the annual production assumptions and proposed updates to the grade profile of mined resources in the 590t/day Scenario;²⁶

c) Micon increase SRK's (or Red Cloud's) annual provision for employee profit sharing in order to maintain a provision equal to 5.1% of pre-tax operating profit;

d) The unit cost of treating ore at the Mallay Plant used by Red Clouds appears to have omitted electricity costs. Micon have increased annual processing costs accordingly; and

e) The cost of transporting concentrate to the port of Callao was the same in the Red Cloud Model as in the 355t/day Scenario, despite the Mallay Plant being significantly closer to Callao than the assumed location of the toll processing plant in the PEA. Micon have proposed a reduction in corresponding transport costs to account for the relative proximity of the Mallay Plant to Callao.

1.45 Micon performed a detailed review of the Red Cloud Model, and developed an updated mine layout, development schedule and production plan identifying specific materials to be mined in each year under the 590t/day Scenario, in the same way that the PEA identified specific materials to be mined under the 355t/day Scenario.²⁷

1.46 Similar to their review of the 355t/day Scenario, Micon update the 590t/day development schedule to consider actual works carried out by Claimant prior to the Blockade Date, and restate development and production schedules to reflect production years running from 1 September to 31 August.²⁸

²⁴ Micon Report, paragraph 100 ↩

²⁵ First Accuracy Report, paragraphs 6.9-6.10 ↩

²⁶ Micon Report, paragraph 100 (ii) ↩

²⁷ Micon Report, paragraph 102 ↩

²⁸ Micon Report, paragraph 103.N ↩

[Page 13]

1.47 In the 590t/day Scenario, Micon have assumed the same production schedule for Year 1 as the 355t/day Scenario, with a further ramp-up to 590t/day starting in September 2019, following the acquisition of the Mallay Plant. Specifically, Micon assume production of 355t/day between February and August 2019, 450t/day in September 2019, 550t/day in October 2019 and 590t/day thereafter.²⁹

1.48 Pursuant to the amendments they make to the Red Cloud Model, Micon consider that confidence in the production and cost estimates underlying the 590t/day Scenario has been raised to a level comparable to the 2018 PEA prepared by SRK.³⁰

1.49 Our updated assessment of damages under both the 355t/day and 590t/day Scenarios, set out in Section 6, reflects the updates made in the Micon Report.

G. Structure of this report

1.50 The structure of the rest of this report is as follows:

• Section 2: Executive summary
• Section 3: AlixPartners' comments regarding so-called “fundamental flaws"
• Section 4: AlixPartners' comments on our damages assessment
• Section 5: The financing of the Invicta Project in the But-For Situation
• Section 6: Our updated damages assessment
• Section 7: AlixPartners' comments on other indicators of value
• Section 8: Experts' declaration

²⁹ Micon Report, paragraph 103.R ↩

³⁰ Micon Report, paragraph 103.U ↩

[Page 14]

2 Executive summary

A. Our comments on the AlixPartners Report

2.1 A significant portion of the AlixPartners Report is dedicated to the discussion of four matters which AlixPartners present as “fundamental flaws" of the analysis set out in the First Accuracy Report.

2.2 These matters are primarily factual issues. In particular, matters relating to (i) the effectiveness of Peruvian law enforcement authorities in response to the Blockade and subsequent events; and (ii) possible legal requirements under Peruvian law to conclude agreements with local communities, are questions of fact which are outside of the scope of our expertise and which are addressed by Claimant in its pleadings. Furthermore, such matters are not relevant for our assessment of damages under Claimant's case.

2.3 Furthermore, AlixPartners' third assertion, that Claimant would likely have defaulted on the PLI Loan in the But-For Situation, appears to be inconsistent with the evidence submitted and with the analysis set out in the Micon Report.

2.4 The fourth matter raised by AlixPartners, that we supposedly did not appropriately account for financing risk in our assessment of damages, is incorrect and appears to be based on a misunderstanding of our approach in the First Accuracy Report.

2.5 To the extent that AlixPartners provide comments on the assumptions underlying our damages assessment and on other indicators of the value of the Invicta Project, we address these in the corresponding section of our report.

2.6 In the table below, we summarise the key topics discussed by AlixPartners in their report, and indicate where such issues are addressed.

Table 2.1. Summary of topics discussed in the AlixPartners report

Source: AlixPartners Report

B. Our updated assessment of damages

2.7 Micon have produced a report in which they review and update the assumptions underlying both the PEA Mine Plan and the Red Cloud Model, which formed the starting point for our damages assessment under the 355t/day and 590t/day Scenarios, respectively.

[Page 15]

2.8 Micon develop an updated mine layout, development schedule and production plan identifying specific materials to be mined in each year under the 590t/day Scenario, similar to the way in which the PEA identified specific materials to be mined in the 355t/day Scenario. Based on these updates, Micon consider that confidence in the 590t/day Scenario is raised to a comparable level to that of the 355t/day Scenario.

2.9 In this report, we present an updated assessment of damages, updating our production and cost assumptions to align with those of the Micon Report. We also apply a number of additional updates to cost assumptions based on new information and to refine the modelling of certain assumptions from the First Accuracy Report.

2.10 We also update our damages analysis to directly model financing cash flows relating to the PLI Loan, to reflect the evidence that, absent the Alleged Breaches, Claimant would likely have acquired the Mallay Plant in early 2019, financed through the receipt of an additional USD 13m of funding from PLI.

2.11 This update results in us calculating Free Cash Flows to Equity (FCFE), representing the cash flows available to equity holders after accounting for debt repayments. As such, we discount these cash flows at the cost of equity of a hypothetical gold mining entity operating in Peru. We also reduce the pre-production premium applied to our discount rate in the 590t/day Scenario to align with that in the 355t/day Scenario, to reflect the comparable level of confidence in both scenarios following Micon's updates.

2.12 We present our estimate of damages at the Valuation Date, after making the above updates, in the table below. For the purposes of this report, we are instructed to apply pre-award interest between the Valuation Date and 31 August 2022 (being a proxy for the date of this report), at alternative rates of (i) LIBOR +4%; and (ii) UST +5%. Our updated damages assessment is detailed in Section 6.

Table 2.2. Summary of damages in the Second Accuracy Report

Source: Appendix 3, sheet "Table 6.13"

2.13 Our assessment of other indicators of value of the Invicta Project, to which we make minor updates in Section 7, provides a relatively narrow range of indicative valuations and appears to confirm the reasonableness of our damages assessments under an income approach.

[Page 16]

2.14 Our indicative assessment based on Claimant's sunk costs provides the lowest valuation benchmark, and is below the FMV of the Invicta Project under both our damages scenarios. Our indicative valuation using a market capitalisation approach is in line with our damages assessment under the 355t/day Scenario, and below our 590t/day Scenario; this appears consistent with the fact that Claimant's intention to purchase the Mallay Plant was not made public, and would therefore not have been reflected in investors' perception of the value of the Invicta Project.

2.15 Meanwhile, the implied equity value multiple of the Invicta Project is below the average multiple of a sample of other recent gold industry transactions (even before taking into consideration the relatively higher grade of the Invicta Project), suggesting that our damages assessment may be conservative.

Figure 2.1. Our updated damages assessments v other indicators of value of the Invicta Project at the Valuation Date

All damages figures in USDm. Transaction multiples in USD/oz.

Source: Appendix 3, tab "Figure 2.1"

[Page 17]

3 AlixPartners' comments regarding so-called “fundamental flaws"

A. Introduction

3.1 AlixPartners claim to have identified four so-called “fundamental flaws” in our analysis which they argue that, if unresolved, would “render IMC's shares in the Invicta Project worthless":³¹

a) According to AlixPartners, intervention by the police would not have resolved the conflict with the Parán Community or the Blockade. Furthermore, the conflict would have remained an obstacle to Claimant's ability to perform its obligations under the PLI Loan;³²

b) Claimant's investment was subject to the risk of not reaching certain agreements, if required under the relevant legislation in Peru, with local communities (i.e. with the Lacsanga, Santo Domingo and Parán Communities and, under the 590t/day Scenario, also with the Mallay community (the “Mallay Community”)). AlixPartners argue that we do not account for such risk, which they refer to as "social license risk";³³

c) AlixPartners further assert that both the actual performance of the Invicta Project prior to the Alleged Breaches and the unavailability of a processing plant with adequate capacity indicate that, even without the Blockade, Claimant was likely to default on the PLI Loan;³⁴ and

d) Finally, AlixPartners allege that our assessment ignored the financing risks associated with Claimant's existing debt and its planned acquisition of the Mallay Plant under the anticipated amendment the PLI Loan.³⁵

3.2 According to AlixPartners, we underestimated or ignored the risks inherent to the Invicta Project which, given the “risky provisions” of the PLI Loan, meant that “any interruption to the operations [...] could result in its termination and the foreclosure proceedings” even without the Alleged Breaches taking place.³⁶

3.3 Ultimately, AlixPartners consider that “the fundamental flaws with Accuracy's damages estimates mean the FMV of the Project could reasonably be nil as of the Valuation Date".³⁷

³¹ AlixPartners Report, paragraph 16; Respondent's Counter-Memorial, paragraph 782 ↩

³² AlixPartners Report, Section V.A. ↩

³³ AlixPartners Report, Section V.B. ↩

³⁴ AlixPartners Report, Section V.C. ↩

³⁵ AlixPartners Report, Section V.D. ↩

³⁶ AlixPartners Report, paragraph 135 ↩

³⁷ AlixPartners Report, paragraph 24 ↩

[Page 18]

B. Our comments on the so-called "fundamental flaws" in our analysis

The so-called “fundamental flaws” primarily relate to questions of fact, and are outside the scope of our expertise

3.4 The so-called “fundamental flaws” identified by AlixPartners primarily relate to questions of fact, which we understand are disputed by the Parties and which the Tribunal will ultimately decide on.

3.5 In particular, matters relating to the effectiveness of Peruvian law enforcement authorities and the requirements that may exist under Peruvian law to conclude agreements with local communities (items (a) and (b) above) are factual issues which fall outside of our scope of expertise and, therefore, we do not opine on them.³⁸ In any case, such issues are not relevant for our assessment of damages under Claimant's case.

3.6 To the extent that items (c) and (d) above are relevant to our scope of expertise as financial experts, we provide our comments below.

AlixPartners' opinion that Claimant was likely to default on the PLI Loan appears to be inconsistent with the evidence submitted

3.7 According to AlixPartners, we “did not address the financing risk associated with Claimant's anticipated amendment to the PLI Loan Agreement”.³⁹

3.8 AlixPartners state that the closing of the third amendment to the PLI Loan (the "Third Amendment to the PLI Loan") and the Mallay purchase agreement (the “Mallay Purchase Agreement")⁴⁰ had to be completed by 15 February 2019. However, AlixPartners do not consider that this could have been achieved as (i) the Mallay Community only approved the transaction in March 2019; (ii) Claimant provided no evidence of alternative financing arrangements or remedies; and (iii) without the Mallay Plant, Claimant faced a high risk of default due to the issues it encountered with third-party processing.⁴¹

³⁸ In the interest of balance, we note that, as at the Valuation Date, Claimant already had in place land surface agreements with two local communities (Santo Domingo and Lacsanga) (First Accuracy Report, paragraph 3.45), in addition to receiving the relevant approvals from the Mallay Community (via Buenaventura) regarding the purchase of the Mallay Plant (Claimant's Memorial, paragraph 94). ↩

³⁹ AlixPartners Report, paragraph 124 ↩

⁴⁰ In the First Accuracy Report, we exhibited an undated draft of the Mallay Purchase Agreement (AC-12) and a draft of the Third Amendment to the PLI Loan dated 26 September (AC-13). We understand that, at the Blockade Date, the latest drafts of the Third Amendment to the PLI Loan (AC-57) and the Mallay Purchase Agreement (AC-58) were dated 5 October 2018; where relevant, we refer to the latest version for the purpose of this report. ↩

⁴¹ AlixPartners Report, paragraph 128 ↩

[Page 19]

3.9 We note that the referenced completion date of 15 February 2019 was an “approximate timeline" based on an anticipated signing date of 15 October 2018,⁴² as the draft Mallay Purchase Agreement provided for a 120-day time period starting from the signing date to complete all conditions precedent to closing.⁴³ However, given that Claimant and the seller of the Mallay Plant (Compañia de Minas Buenaventura S.A.A. (“Buenaventura”) did not ultimately sign the Mallay Purchase Agreement, this time period never started to run.

3.10 One of the conditions precedent of Claimant's acquisition of the Mallay Plant was that Buenaventura obtain the approval of the Mallay Community for the transfer of an easement agreement to Claimant.⁴⁴ We understand 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 2019.⁴⁵

3.11 Notwithstanding the above, whilst the points raised by AlixPartners are, once again, largely factual, they nonetheless appear at odds with the evidence submitted.

Pandion had historically demonstrated significant flexibility regarding the provision of financing to Claimant

3.12 In arguing that Claimant would have likely defaulted on the PLI Loan, AlixPartners ignore the evidence that Pandion (which owned PLI at the Blockade Date) had historically shown significant flexibility regarding the provision of financing to Claimant.

3.13 The provision of additional funding by Pandion for the purchase of the Mallay Plant was already anticipated through the terms of the initial PLI Loan, which granted Pandion a right of first refusal to provide an additional round of financing for an amount between USD 6m and USD 12m⁴⁶ for the acquisition or construction of a processing plant under substantially similar terms.⁴⁷ As we note in Section 5, the effective borrowing rate proposed in the draft Third Amendment to the PLI Loan appears to be similar to the effective borrowing rate under the terms of the prevailing PLI Loan.⁴⁸

⁴² AC-11 Lupaka Gold Corp. Minutes of the board meeting dated 27 September 2018, page 2 ↩

⁴³ AC-58 Draft Mallay Purchase Agreement between Buenaventura and IMC dated 5 October 2018, Section 12.3 ↩

⁴⁴ Claimant's Memorial, paragraph 93; Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraph 36 ↩

⁴⁵ Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraphs 38 & 39 ↩

⁴⁶ Ultimately, Pandion agreed to provide an additional USD 13m, i.e. higher than the amount initially provided for (AC-57 Draft Amendment and Waiver No. 3 to the Second Amended and Restated Pre-paid Forward Gold Purchase Agreement dated 5 October 2018, page 8) ↩

⁴⁷ AC-04 Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017, pages 16 & 17; Section 7(1)(c). "Second Tranche Prepayment" means a prepayment estimated to be an amount between US$6,000,000 and US$12,000,000 (a) the proceeds of which shall be used to fund Seller's acquisition of or construction of a processing plant for the Mine [...] with delivery schedule, volume of ounces and pricing substantially similar to the terms and conditions of this Agreement [...]" ↩

⁴⁸ On an illustrative basis, we calculate the effective interest rate to be 16.1% in the 590t/day Scenario, compared to 17.0% in the 355t/day Scenario. See Section 5. ↩

[Page 20]

3.14 The terms of the draft Third Amendment to the PLI Loan demonstrate that Pandion had agreed in principle in October 2018 to defer the first gold repayment (initially anticipated for December 2018) to a later date. Specifically, under the proposed terms, Claimant would have benefitted from a nine-month grace period between the closing date of the Mallay Plant acquisition and the first gold repayment.⁴⁹ A repayment schedule appended to the draft Third Amendment to the PLI Loan assumed that the first gold repayments would have taken place in September 2019, based on a closing date of December 2018.⁵⁰

3.15 The law firm Mayer Brown, who we understand were acting as Pandion's counsel, noted that the closing date may in fact be later than assumed in the repayment schedule, in which case the first gold repayments would also have been due later:⁵¹

"We note that the delivery schedules to be attached as Schedule P are included for convenience only and are not actually referenced in the operative provisions of the PPF Agreement (even as amended). Instead the operative provisions tie the timing of the gold deliveries to whenever the applicable Effective Date occurs, which may be later than is indicated in Schedule P. As a result, Pandion's position is that the schedules do not need to be updated, but please let us know if you have any additional comments."

3.16 Indeed, we understand from Claimant that Pandion had already demonstrated flexibility on a number of occasions over the course of the Invicta Project including, inter alia:

a) The initial PLI Loan agreement, signed on 30 June 2016,⁵² granted Claimant until the end of 2016 to enter into an agreement with a neighbouring community to use and improve community roads to transport ore from the Invicta Mine. Pandion subsequently extended this deadline by one year, to 31 December 2017.⁵³

b) Pandion amended production forecast and timelines to reflect the actual progress of development works and regulatory processes. For example, in February 2018, pursuant to an amendment to the PLI Loan, Pandion (i) amended the annual production forecast to reflect the additional time to bring the Invicta Project to production, without changing the terms of financing; and (ii) increased Claimant's expenditure allowance and deferred certain expenditures to later production years,⁵⁴ rather than insisting that Claimant comply with the prevailing terms of the PLI Loan at the time.⁵⁵

⁴⁹ AC-57 Draft Amendment and Waiver No. 3 to the Second Amended and Restated Pre-paid Forward Gold Purchase Agreement dated 5 October 2018, page 12 ↩

⁵⁰ AC-57 Draft Amendment and Waiver No. 3 to the Second Amended and Restated Pre-paid Forward Gold Purchase Agreement dated 5 October 2018, page 45 ↩

⁵¹ AC-59 Email from Fernando Ortiz (BVN) to Will Ansley (Lupaka) and Joe Archibald (Pandion) dated 9 October 2018, page 4 ↩

⁵² For the sake of clarity, references to the "PLI Loan" in this report refer to the terms Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017 (AC-04), unless otherwise stated (e.g. "initial PLI Loan agreement", "draft Third Amendment to the PLI Loan", etc.) ↩

⁵³ Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraph 31 ↩

⁵⁴ C-46 Amendment No. 2 to the Second Amended and Restated PPF Agreement dated 6 February 2018 ↩

⁵⁵ Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraph 43 ↩

[Page 21]

c) In December 2008, IMC acquired five mining concessions from a subsidiary of Barrick Gold Corporation in exchange for a royalty which was subsequently assigned to Franco Nevada. In order to obtain the full release of funds under the initial PLI Loan agreement, Claimant had to register a first priority lien over the concessions in favour of PLI Huaura, which meant removing Franco Nevada's existing lien. Pandion granted Claimant a one-year extension to finalise the necessary agreements and conditions, so that it could unlock part of the funds under the PLI Loan before obtaining a complete release from the royalty obligation.⁵⁶

d) Although the terms of the PLI Loan technically required Claimant to have and maintain a long-term mineral offtake agreement in place, Pandion deferred the effective date of this obligation twice, before waiving it indefinitely until Claimant (with Pandion's support) finalised negotiations with offtake bidders.⁵⁷

e) Lastly, Pandion actively continued to support Claimant and tried to find alternative arrangements until Pandion sold its interest to Lonely Mountain, despite Claimant not being able to make its gold repayments. Pandion's behaviour was consistent with the terms of the PLI Loan, which obliged the parties to use good faith efforts to reschedule repayment obligations in case of force majeure before taking steps to terminate the PLI Loan early.⁵⁸

3.17 Therefore, whilst it is true that Claimant's purchase of the Mallay Plant was contingent upon Buenaventura obtaining approval of a transfer agreement from the Mallay Community,⁵⁹ which was not received until March 2019,⁶⁰ the evidence suggests that Pandion would have shown flexibility in relation to the completion of the Third Amendment to the PLI Loan, whilst Buenaventura was obtaining the necessary approvals.

3.18 Indeed, given that the Third Amendment was in the interests of both Claimant and Pandion (as, ceteris paribus, it would have enabled both parties to generate a higher return), it is common sense that there would be a degree of flexibility, particularly given the high frequency of delays to closing on transactions.

Claimant's ability to meet its gold repayment obligations

3.19 Notwithstanding that Pandion would likely have shown flexibility regarding its financing of the Invicta Project (rather than demanding payment of the Early Termination Amount and/or foreclosing upon Claimant's shares in IMC at the earliest possible opportunity, as AlixPartners insinuate),⁶¹ Micon perform an analysis of Claimant's ability to meet its gold repayment obligations both with and without the acquisition of the Mallay Plant.⁶²

⁵⁶ Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraph 44 ↩

⁵⁷ Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraph 45 ↩

⁵⁸ Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraph 48; AC-04 Second Amendment and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017, Section 14(1) ↩

⁵⁹ Claimant's Memorial, paragraph 93 ↩

⁶⁰ C-233 Email from Buenaventura to Lupaka dated 11 March 2019 ↩

⁶¹ AlixPartners Report, paragraph 129 ↩

⁶² Micon Report, Section 6 ↩

[Page 22]

3.20 Based on Micon's updated Year 1 production schedule (which includes a ramp-up in production from November 2018 to reflect Claimant's need to identify reliable toll millers capable of processing 355t/day),⁶³ Micon conclude that, even without the purchase of the Mallay Plant, “but for the blockade that prevented access to and operation of the Invicta mine, [Claimant] 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 [...] and to ship the resulting concentrates to market in time to meet its obligations".⁶⁴

3.21 Ultimately, Micon conclude that “[Claimant's] gold repayment obligations were achievable, both under the 355t/d plan and the 590t/d plan".⁶⁵

3.22 We further note that Claimant had agreed with Buenaventura that it would be able to start processing ore at the Mallay Plant in the interim period, before formally taking over the plant. Indeed, the draft Mallay Purchase Agreement gave Claimant the right, upon signing of the agreement, to process up to 8,000t/month at a rate of 600t/day at the Mallay Plant.⁶⁶ Hence, Claimant considers that it could have processed ore at both Huancapeti (a third-party toll processor) and the Mallay Plant until the conclusion of the acquisition of the Mallay Plant.⁶⁷

AlixPartners' opinion that we ignore financing risks associated with Claimant's debt is based on a misunderstanding of our approach

3.23 AlixPartners further opine that we “ignored this financing risk entirely by making several debt financing assumptions in the But-For Scenarios without providing the logic or evidence to support them".⁶⁸ In particular, AlixPartners allege that we do not account for “difficulties in refinancing”.⁶⁹

3.24 In the First Accuracy Report, we did not assume that Claimant would have obtained new financing under “friendlier” terms, as AlixPartners contend.⁷⁰ AlixPartners' comments on this point appear to be based on a misunderstanding of our application of the FMV framework in the First Accuracy Report, which we address in Section 5.

⁶³ Micon Report, paragraph 83 ↩

⁶⁴ Micon Report, paragraph 125 ↩

⁶⁵ Micon Report, paragraph 133 ↩

⁶⁶ Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraph 86; AC-58 Draft Mallay Purchase Agreement between Buenaventura and IMC dated 5 October 2018, page 39 ↩

⁶⁷ Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraph 89 ↩

⁶⁸ AlixPartners Report, paragraph 124 ↩

⁶⁹ AlixPartners Report, Section V.D. ↩

⁷⁰ AlixPartners Report, paragraph 148 ↩

[Page 23]

C. Consideration of Invicta Project risks in our damages assessment

3.25 In the First Accuracy Report, we adopted a set of prudent assumptions which, in our view, accounted for (either directly or indirectly) a number of the risks that AlixPartners identify in relation to the Invicta Project. Examples of such assumptions include, inter alia, the following:

a) Assumed metal grades for the resources considered within the Red Cloud Model were higher than the average grades for the Invicta Project as a whole.⁷¹ As such, we considered it more prudent to apply the average grade for the Invicta Project as a whole to the 590t/day Scenario. This adjustment resulted in total metal content of 305k oz. AuEq under our Mallay Acquisition Plan (average grade of 6.93g/t) compared to 368k oz. in the Red Cloud Model (average grade of 8.38g/t);⁷²

b) Our valuation under the 590t/day Scenario did not include all of the indicated material reported for the Invicta Project (or even the Atenea deposit) and therefore represented a conservative estimate of its value;⁷³

c) We added a premium to our discount rate to reflect the uncertainty and risk associated with our modelling of the Invicta Project under each production scenario (see further discussion in Section 4);⁷⁴

d) We understand that, by the Blockade Date, Claimant had materially completed development of the Invicta Mine and was close to obtaining final authorisations and commencing full commercial production.⁷⁵ Notwithstanding AlixPartners' comments on Claimant's dealings with local communities⁷⁶ (which are addressed in Claimant's pleadings), from a technical and operational perspective our assumed start date for full commercial production more than ten months after the Blockade Date was therefore conservative;⁷⁷ and

⁷¹ As explained in the First Accuracy Report, paragraph 6.12, the PEA included a cut-off grade sensitivity table (AC-02 SRK Consulting PEA dated 13 April 2018, page 117, Table 78) which indicated a total of 2,024kt of indicated mineralised material at the Invicta Mine at a gold-equivalent cut-off grade of 4.0g/t, corresponding to 449k gold-equivalent oz. of contained metal at an average gold-equivalent grade of 6.90g/t. This was lower than the assumed metal grade of resources considered within the Red Cloud Model (8.38t/g). ↩

⁷² First Accuracy Report, paragraphs 6.9-6.17 ↩

⁷³ First Accuracy Report, paragraph 6.18. We noted that the Mallay Acquisition Plan assumes total production of 1,367kt at a 4.0g/t cut-off grade, equivalent to c.68% (1,367kt/2,024kt = 68%) of total Invicta Mine indicated resources, whereas Atenea vein resources represent c.84% of total Invicta Project resources (calculated on an indicative basis at a 3.0g/t cut-off grade, given that Atenea vein resources were reported at this cut-off grade only). ↩

⁷⁴ First Accuracy Report, paragraph A4.23 ↩

⁷⁵ First Accuracy Report, paragraph 3.49 ↩

⁷⁶ According to AlixPartners, '[i]f the Access Road Protest and broader social conflict would have required additional time to resolve or if beginning production would have required a period longer than 10 months [...] this would at the very least delay the Project's cash flows and thus reduce Claimant's damages" (AlixPartners report, paragraph 117) ↩

⁷⁷ First Accuracy Report, paragraph 5.16 ↩

[Page 24]

e) Given that subsequent actual metal prices exceeded market expectations as at the Valuation Date, our assessment of damages was conservative to the extent that we performed an ex-ante assessment as at the Valuation Date, rather than applying hindsight under an ex-post approach.⁷⁸

⁷⁸ First Accuracy Report, paragraph 5.25 & Table 5.4 ↩

[Page 25]

4 AlixPartners' comments on our damages assessment

A. Introduction

4.1 In this section, we address AlixPartners' comments on:

a) The operational and technical assumptions underlying our assessment of the FMV of the Invicta Project in the First Accuracy Report;

b) The valuation assumptions (discount rate and pre-award interest) used for our damages assessment; and

c) The residual value of Claimant's shares in IMC in the Actual Situation.

4.2 In Section 6, we set out an updated assessment of damages as at 31 August 2022, being a proxy for the date of this report, based on the additional information made available to us since the date of the First Accuracy Report.

B. Comments on our technical and operational assumptions

4.3 In this sub-section, we discuss and respond to AlixPartners' comments on the technical and operational assumptions underlying our damages assessment in the First Accuracy Report, and the adjustments we made to the Red Cloud Model and SRK Model under the 590t/day Scenario and 355t/day Scenario respectively.⁷⁹

Assessment of damages under the 590t/day Scenario

4.4 Although AlixPartners agree that some of our adjustments to the Red Cloud Model “may not be unreasonable”, they also state that our adjustments “do not appear to have documentary support” and “typically have the effect of increasing" the FMV of the Invicta Project.⁸⁰

4.5 To the extent that AlixPartners provide specific critiques of the assumptions we adopted, we address these below.

Metal prices

4.6 Whilst AlixPartners include metal prices within their list of assumptions for which we purportedly did not have documentary support, they do not provide any further comments other than to acknowledge the approach we adopted in the First Accuracy Report, in which we updated metal prices to reflect prevailing market expectations at the Valuation Date using monthly metal futures contract prices from Capital IQ.

⁷⁹ We respond to AlixPartners' comments on the two production scenarios in the same order as they are presented in the AlixPartners Report, i.e. we address their comments on the 590t/day Scenario first. ↩

⁸⁰ AlixPartners Report, paragraph 143 ↩

[Page 26]

4.7 As such, it appears that AlixPartners do not disagree with our approach, which was fully supported and documented.

Ore grades

4.8 As set out in the First Accuracy Report, the Red Cloud Model applied the same grade profile in each year of activity as the PEA Mine Plan, resulting in an assumed average gold-equivalent grade (8.38g/t at a cut-off grade of 4.0g/t) higher than that of the Invicta Project as a whole (6.90g/t at the same cut-off grade). We considered it more prudent to apply the average grade for the Invicta Project as a whole to the Mallay Acquisition Plan, which resulted in a reduced metal content of 305k oz. AuEq compared to that of 368k oz. in the original Red Cloud Model.⁸¹

4.9 According to AlixPartners, this adjustment may be reasonable but there is “insufficient documentation or discussion on the matter" in the First Accuracy Report. Furthermore, we did not "reconcile the 4.85g/t gold grade assumed with the actual gold grade achieved as of October 2018 of 2.25g/t”.⁸²

4.10 AlixPartners do not propose alternative grade assumptions that they consider to be more reasonable.

4.11 The Micon Report sets out an updated ore production schedule for the 590t/day Scenario, identifying sufficient mineral resource to sustain 10 years of mining at a rate of 590t/day.⁸³ Applying Micon's updated annual metal grade assumptions results in an average gold-equivalent grade of 5.48g/t⁸⁴ in our 590t/day Scenario, lower than in the First Accuracy Report and significantly lower than the Red Cloud Model (albeit at a lower cut-off grade of 3.0g/t). In Section 6, we update our assessment of the 590t/day Scenario to reflect the volume and grade assumptions of the Micon Report.

4.12 With respect to the lower gold grade achieved as at October 2018 in comparison to the PEA, Micon provide several possible reasons for this outcome, including:⁸⁵

a) The natural variability of in situ mineralisation;

b) The dilution of in situ mineralisation during mining, resulting in the average grade of development material being typically lower than that of run-of-mine ore generated during commercial production;

c) The segregation of heavy particles enriched in gold during the blasting and loading of broken rock; and

⁸¹ First Accuracy Report, paragraphs 6.9-6.18 ↩

⁸² AlixPartners Report, paragraph 143(b) ↩

⁸³ Micon Report, Table 5.13. Of which c.8 years of steady-state production at 590t/day. ↩

⁸⁴ Appendix 3, sheet "fin_lupaka", cell C8. Table 5.13 of the Micon Report shows a gold-equivalent grade of 5.92g/t, higher than our figure of 5.48g/t. The difference arises because Micon calculate the gold-equivalent grade using metal prices from the original Red Cloud Model, whereas we base our calculation on the updated metal prices adopted in the First Accuracy Report. In any case, the overall gold-equivalent grade calculation is illustrative only; for the purposes of our damages calculation we update the grades of individual metals in each year to align with the Micon Report. Therefore, this does not have any impact upon our damages calculation. ↩

⁸⁵ Micon Report, paragraph 138 ↩

[Page 27]

d) The under-reporting of gold content by third-party toll-millers.

4.13 Micon also note that the lower-than-expected grades were for gold only, whilst grades for silver, copper, lead and zinc were all above forecast.⁸⁶

4.14 Micon ultimately conclude that the onset of full commercial production, the systematic collection of fines⁸⁷ generated during mining and the close supervision of toll-milling operations would have likely enabled Claimant to overcome the gold grade shortfalls in early 2019.⁸⁸ On the final point, we note that the purchase of the Mallay Plant would have brought toll processing in-house, which would have addressed the issue of possible under-reporting of gold content by third-party toll processors.

Unit operating costs

4.15 AlixPartners do not comment on unit operating costs, other than to note that average unit operating costs were marginally lower in the First Accuracy Report (USD 80.93/t) compared to the Red Cloud Model (USD 81.33/t).⁸⁹ As understood by AlixPartners, this was a flow-through adjustment, whereby our comparatively lower zinc metal grade assumption resulted in a lower concentrate transportation cost.⁹⁰

4.16 In our updated assessment of damages in Section 6, we update certain operating cost assumptions, including a number of updates to reflect Micon's revised mine plan assumptions. Overall, these updates result in average unit operating costs increasing by USD 0.11/t⁹¹ in the 355t/day Scenario and USD 13.40/t⁹² in the 590t/day Scenario compared to the First Accuracy Report.⁹³

⁸⁶ Micon Report, paragraph 136 ↩

⁸⁷ Fines are defined as "rock or mineral that has been broken or ground into small particles" (Micon Report, page iv) ↩

⁸⁸ Micon Report, paragraph 149. This supports Claimant's view that, notwithstanding the anticipated acquisition of the Mallay Plant, it could have overcome the issues identified at offsite processing plants such as Huancapeti, Coriland, and San Juan Evangelista by "refining internal procedures to avoid operational errors" and "repairing mechanical failures" without incurring delays or high costs. (Second Witness Statement of Julio Félix Castañeda Mondragón dated 23 September 2022, paragraph 100). ↩

⁸⁹ AlixPartners Report, paragraph 143(c) ↩

⁹⁰ First Accuracy report, Appendix 6, sheet "opex", row 49. ↩

⁹¹ Appendix 4, sheet "opex", cell H62. USD 143.45/t – USD 143.34/t = USD 0.11/t ↩

⁹² Appendix 3, sheet "opex", cell H62. USD 94.33/t – USD 80.93/t = USD 13.40/t ↩

⁹³ We note that Respondent alleges that Claimant defaulted on the PLI Loan for a number of reasons, including deviation from the "Initial Expense Budget" (Respondent's Counter-Memorial, paragraph 717, also alleged by PLI in its Notice of Acceleration (AC-16 Notice of Acceleration dated 2 July 2019, Schedule I(b)(e))). The basis for such an allegation is not clear. Whilst no further substantiation is provided for this claim, we note that total actual operating costs to October 2018 (year-to-date, including G&A costs) were USD 4.4m (AC-10, Invicta Monthly Report dated October 2018, page 10), compared to budgeted direct operating costs and G&A costs totalling USD 13.3m for 2018 (C-46, Amendment No. 2 to the Second Amended and Restated PPF, page 15). ↩

[Page 28]

Capital expenditures

4.17 The SRK Model assumed that Claimant would have incurred pre-production and development costs of USD 4.3m in the first year of production. In the First Accuracy Report, we explained that Claimant actually incurred development costs of CAD 7.3m (c. USD 5.7m) in 2018 and as a result, according to Claimant, very little additional capital expenditure was required at the Blockade Date in order to bring the mine to working order. As such, under both the 590t/day and 355t/day Scenarios, we considered any pre-production and development capital expenditure requirements to be nil at the Valuation Date in the But-For Situation.⁹⁴

4.18 Although AlixPartners agree that this adjustment might be reasonable, it is their understanding that “final inspection and several regulatory authorizations were pending and may have required additional capital expenditures and resulted in further delays".⁹⁵ Whilst this is ultimately a question of fact, this is disputed by Claimant, who asserts that the mine was ready to commence production at the Blockade Date.⁹⁶

4.19 The Micon Report presents an updated capital expenditure schedule in both the 355t/day and 590t/day Scenarios, which we reflect in our updated assessment of damages in Section 6. Compared to our high-level assumption regarding pre-production capital expenditure in the First Accuracy Report, Micon's updates include a more granular estimate of the actual capital expenditure incurred by Claimant prior to the Blockade Date and an estimate of the additional capital expenditure requirements in the 590t/day Scenario based on a specific mine layout.⁹⁷

4.20 Micon's updates represent a more conservative estimate of remaining capital expenditure requirements at the Blockade Date, and result in an increase in life-of-mine capital expenditure of USD 1.3m⁹⁸ in the updated 355t/day Scenario and USD 5.1m⁹⁹ in the 590t/day Scenario compared to the First Accuracy Report.

⁹⁴ First Accuracy Report, paragraphs 5.30-5.31 ↩

⁹⁵ AlixPartners Report, paragraph 143(d) ↩

⁹⁶ Second Witness Statement of Julio Félix Castañeda Mondragón dated 23 September 2022, Section 7 ↩

⁹⁷ Micon Report, paragraph 103.V ↩

⁹⁸ Sum of (i) capital development costs, (ii) infrastructure capex and (iii) annual contributions under the 2015 Mine Closure Plan, equal to USD 8.4m in the First Accuracy Report and USD 9.7m in the Second Accuracy Report. ↩

⁹⁹ Sum of (i) capital development costs, (ii) infrastructure capex, (iii) Mallay Mill purchase price, (iv) Mallay sustaining capital, (v) Mallay closure bond and (vi) annual contributions under the 2015 Mine Closure Plan, equal to USD 37.3m in the First Accuracy Report and USD 42.4m in the Second Accuracy Report. Updated figures include Micon's capital expenditure updates and the addition of a further USD 200k to reflect the cost of an update to the PEA (see Section 6). ↩

[Page 29]

Working capital

4.21 The Red Cloud Model did not consider changes in project working capital in calculating cash flows. For completeness, based on our discussions with Claimant, we included an estimate of changes in project working capital in our own assessment of cash flows:¹⁰⁰

a) Days sales outstanding (DSO) of 30 days for 90% of receivables and 60 days for the remainder, calculated as a percentage of gross sales;

b) Days payables outstanding (DPO) of 30 days, calculated as a percentage of offsite and site operating costs; and

c) Days inventory outstanding (DIO) of nil, as all mined rock is assumed to go straight to toll processing facilities.

4.22 AlixPartners point out that, aside from citing our discussions with Claimant, we failed to provide any “supporting documents that would corroborate these working capital requirements”.¹⁰¹

4.23 AlixPartners do not specify which of our assumptions, if any, they disagree with, and do not provide alternatives.

4.24 In any case, working capital requirements were not considered in either the SRK Model or the Red Cloud Model. Our inclusion of such an assumption, ceteris paribus, reduces damages and therefore reflects a prudent approach.

4.25 We note that our combined DSO and DIO working capital assumptions are consistent with the Micon Report, which states that “Micon is of the opinion that typically one calendar month would elapse between ore mining and receipt of concentrate sales proceeds”.¹⁰²

Cash flows from the Mallay Plant and G&A expenses

4.26 The Red Cloud Model included two hard-coded lines items related to "Mallay Cash Flow" and "Head office G&A", which did not appear in the SRK Model and whose inclusion was unexplained.¹⁰³ In the absence of additional information as to what these cash flows represented, in the First Accuracy Report, we excluded both of these line items from our damages assessment. However, we acknowledged that including these items would reduce our assessment of FMV under the 590t/day Scenario by USD 4.1m.¹⁰⁴

¹⁰⁰ First Accuracy Report, paragraph 6.19(a). ↩

¹⁰¹ AlixPartners Report, paragraph 143(e) ↩

¹⁰² Micon Report, paragraph 119 ↩

¹⁰³ AC-15, sheet "fin_lupaka", rows 364 & 365 ↩

¹⁰⁴ First Accuracy Report, paragraph 6.19(c) and footnote 175 ↩

[Page 30]

4.27 According to AlixPartners, these costs represent Red Cloud's estimate of the “additional G&A costs related to the acquisition of the Mallay Plant” and it would be “self-serving to remove these costs without providing a more substantiated rationale”. As a result, they propose to reverse our adjustment.¹⁰⁵

4.28 The “Head office G&A" line in the Red Cloud Model shows a negative cash flow of USD 1.5m in Y1, and USD 2.0m in each year thereafter. It is unclear how the purchase of the Mallay Plant under the 590t/day Scenario could have resulted in such a significant¹⁰⁶ increase in general and administrative costs compared to the 355t/day Scenario, as AlixPartners' reversal of our adjustment implies.

4.29 Our exclusion of these costs is consistent with an email received from Pandion to Claimant in May 2018, in which Pandion appear to specifically exclude “Buenaventura G&A that did not seem to be related to the mine site (~USD 2m per annum)" from their own contemporaneous Invicta Project cash flow model.¹⁰⁷ Given the similarities in the amounts, it is therefore possible that the “Head office G&A" line in the Red Cloud Model in fact relates to Buenaventura's own head office costs rather than costs that would have been incurred by Claimant upon purchase of the Mallay Plant.

4.30 In the same email, Pandion note that they added “USD 350k to Lupaka's G&A for managing the acquired asset”.¹⁰⁸ Whilst it is not stated exactly what these costs would relate to, for prudency, we include additional G&A costs of USD 350k in each production year in our 590t/day Scenario.¹⁰⁹

4.31 Notwithstanding the above, the draft Mallay Purchase Agreement envisaged that a number of "Mining Concessions” (which we understand to refer to additional mineral plots) would have been acquired by Claimant as part of the Mallay Plant acquisition.¹¹⁰ Whilst we do not have any further detail regarding the “Mallay Cash Flow” line in the Red Cloud Model (totalling USD 5m over two years), it is possible that this line sought to reflect the additional value that Claimant could have extracted from these additional mineral plots. Indeed, in the same email referred to above, Pandion envisage “[processing] material from the Mallay underground mine at 500tpd from October 2018 through June 2019”.¹¹¹

¹⁰⁵ AlixPartners Report, paragraph 143(g) ↩

¹⁰⁶ Ceteris paribus, including these cash flows in our 590t/day Scenario would have reduced the FMV of the Invicta project by c.7% in the First Accuracy Report (USD 4.1m/USD 63.6m = c.7%) (First Accuracy Report, Appendix 6, tab "fin_lupaka", rows 369 & 370. ↩

¹⁰⁷ AC-70, Email from Pandion to Claimant dated 30 May 2018, page 1 ↩

¹⁰⁸ AC-70, Email from Pandion to Claimant dated 30 May 2018, page 1 ↩

¹⁰⁹ Appendix 3, sheet "opex", row 58. We pro rate such costs in Year 1 to reflect the assumed acquisition of the Mallay Plant in March 2019, and again in Year 10 to reflect only a partial year of operation. ↩

¹¹⁰ AC-58 Draft Mallay Purchase Agreement between Buenaventura and IMC dated 5 October 2018, Section 3.1 ↩

¹¹¹ AC-70 Email from Pandion to Claimant dated 30 May 2018, page 1 ↩

[Page 31]

4.32 Our approach is therefore prudent in that it reflects only the economic benefit relating to (i) an increased production rate; and (ii) cost efficiencies resulting from the acquisition of the Mallay Plant, but does not seek to ascribe any further value resulting from the acquisition of additional mineral plots, which could have increased the overall value of the Invicta Project.

Assessment of damages under the 355t/day Scenario

4.33 Whilst AlixPartners “have not identified material flaws” with the adjustments we made to the SRK Model in relation to metal prices, capital expenditures or working capital in the 355t/day Scenario,¹¹² they nevertheless provide a number of comments in respect of each of these items.

4.34 In the First Accuracy Report, our updates to the underlying models in respect of metal prices, capital expenditure and working capital followed the same approach under both the 590t/day and 355t/day Scenarios.¹¹³ As such, our responses to AlixPartners' comments on each of these points under the 355t/day Scenario are the same as those we set out above in relation to the 590t/day Scenario.

4.35 Further to the above, AlixPartners question a minor adjustment we made to closure costs in the First Accuracy Report. We address this below.

Our removal of the contingency on infrastructure capex in the 355t/day Scenario

4.36 In the First Accuracy Report, we removed the 20% contingency on closure costs (not on all infrastructure capex, as AlixPartners appear to suggest)¹¹⁴ given the relative certainty of these amounts under the 355t/day Scenario.¹¹⁵

4.37 AlixPartners believe that our adjustment is “unreasonable” because we “failed to provide any evidence to support” our opinion and, more importantly, it is their understanding that the mine closure plan had not been approved. As a result, they propose to reverse our adjustment.¹¹⁶

4.38 As noted by AlixPartners, this would have only a minor impact on damages; adding back the contingency under the 355t/day production scenario would have reduced our assessment of damages by USD 0.1m in the First Accuracy Report.

¹¹² AlixPartners Report, paragraph 160 ↩

¹¹³ First Accuracy Report, paragraph 6.19(a) ↩

¹¹⁴ AlixPartners Report, paragraph 160(d) ↩

¹¹⁵ First Accuracy Report, footnote 153 ↩

¹¹⁶ AlixPartners Report, paragraph 160(d) ↩

[Page 32]

4.39 We nonetheless disagree with the proposed reversal of our adjustment. The risk premium that we add to the discount rate in the 355t/day Scenario reflects the additional risk resulting from the fact that the Invicta Project had not yet entered into full commercial production, including any risk relating to the obtaining of final approvals from the relevant government authorities. We understand that Claimant was not aware of any reason why the updated mine closure plan would not have been approved;¹¹⁷ as such, including a contingency in addition to the risk premium would, in our view, double count any risk relating to the approval of the mine plan.

4.40 By contrast, as far as we are aware, no updated closure plan had been prepared with the Mallay Acquisition Plan in mind and, therefore, we consider the addition of a contingency in the 590t/day Scenario to be reasonable.

C. Comments on our valuation assumptions

Discount rate (WACC)

4.41 In the First Accuracy Report, we discounted the forecast FCFF of the Invicta Project using an estimate of the cost of capital for companies in the precious metals industry based on data from Professor Damodaran, a leading academic in the field of corporate finance and valuation who publishes discount rate data through his website.¹¹⁸

4.42 Although AlixPartners neither engage with our detailed assumptions nor provide an alternative calculation of the discount rate, they identify the following alleged shortcomings in our approach:

a) We allegedly failed to provide any “calculation, logic, or evidence” to support our pre-production premiums of 3.3% and 6.9% under the 355t/day Scenario and 590t/day Scenario respectively;¹¹⁹

b) We allegedly failed to account for the “fundamental flaws” identified by AlixPartners in relation to the Invicta Project in the project-specific risk premium included in the WACC;¹²⁰ and

c) In the 355t/day Scenario, we allegedly did not adjust the discount rate to address the execution risk associated with the Invicta Project.¹²¹

¹¹⁷ Witness Statement of Julio Felix Castaneda Mondragon dated 1 October 2021, paragraph 21; Second Witness Statement of Julio Félix Castañeda Mondragón dated 23 September 2022, Section 7 ↩

¹¹⁸ First Accuracy Report, paragraph A4.5 ↩

¹¹⁹ AlixPartners Report, paragraphs 144 and 161 ↩

¹²⁰ AlixPartners Report, paragraphs 18, 21 and 161 ↩

¹²¹ AlixPartners Report, paragraph 152 ↩

[Page 33]

4.43 As set out in the First Accuracy Report, the pre-production premiums used in our calculation of the discount rate came from data compiled by Lawrence Devon Smith, an expert in the field of mineral project evaluation, risk assessment and due diligence.¹²² We referenced and exhibited Mr. Smith's findings (Exhibit AC-47) in the First Accuracy Report.

4.44 With respect to our logic, we added pre-production premiums as a prudent adjustment to our estimated discount rates to reflect the different stages of development of the Invicta Project under our two production scenarios:¹²³

a) Under the 355t/day Scenario, we applied a pre-production premium of 3.3% corresponding to the premium typically applied to mineral properties at the feasibility level of study. Although the PEA Mine Plan was at a more advanced stage than some projects at the feasibility stage given that development works were materially completed and financing had been obtained via the PLI Loan, we still considered it appropriate to apply a premium given the lower level of confidence provided by mineral resources in comparison to reserves; and

b) Under the 590t/day Scenario, we applied a pre-production premium of 6.9% corresponding to the premium typically applied to mineral properties at the scoping level of study in order to reflect the higher level of uncertainty attached to the 590t/day Scenario, 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.

4.45 The impact of adding a pre-production premium to our discount rate was, ceteris paribus, a decrease in FMV of USD 3.9m in the 355t/day Scenario, and USD 18.8m in the 590t/day Scenario.¹²⁴

4.46 AlixPartners do not further explain what sort of additional evidence they would expect to see to support our use of pre-production premiums, nor do they seek to calculate an alternative discount rate that they consider to be more appropriate, save for presenting a sensitivity to increase the discount rate under each production scenario by between 1% and 5%.

4.47 Notwithstanding the above, we note the following:

a) Under the FMV approach, the discount rate is intended to reflect the risk that would be faced by a hypothetical investor in the Invicta Project. To the extent that the so-called "fundamental flaws” relate to Claimant-specific operational or financing risks (rather than general legal risks, for example), one would not expect these to be reflected in the discount rate;

¹²² First Accuracy Report, paragraphs A4.22-A4.23; Exhibit AC-47 Lawrence Devon Smith, The RADR Paradox-Discount Rates Risk, & Long Life Projects, 2016 ↩

¹²³ First Accuracy Report, paragraph A4.23 ↩

¹²⁴ First Accuracy Report, Appendices 5 & 6 ↩

[Page 34]

b) According to AlixPartners, Latin America faces specific risks, such as the risk of suspension of mining projects due to community opposition.¹²⁵ As previously discussed, AlixPartners argue that our alleged failure to account for such risk is one of the "fundamental flaws" of our damages assessment. Professor Damodaran's beta calculation for the precious metals industry is based on a sample of 76 sectoral constituent entities; based on our desktop research, approximately 40% of these appear to have operations in Central and/or South America.¹²⁶ Given the high proportion of companies with Latin American operations within Professor Damodaran's sample, this regional risk should therefore be reflected (at least in part) in our discount rate;

c) In addition to the above, we apply a further Peru-specific country risk premium to our discount rate to reflect the additional risk associated with doing business in Peru over a "risk-free" jurisdiction;¹²⁷ and

d) Contrary to AlixPartners' assertion that we did not adjust the discount rate to address the execution risk associated with the Invicta Project in the 355t/day Scenario, the application of a risk premium to our damages assessment specifically aims to capture project-specific risks (such as execution risk) based on, inter alia, the stage of development of the project and the robustness of the underlying assumptions.

4.48 Given the above, AlixPartners' basis for opining that we fail to consider (i) the so-called "fundamental flaws”; and (ii) execution risk in the 355t/day Scenario in our discount rate is unclear.

Pre-award interest rate

Our comments on AlixPartners' alternative pre-award interest rates

4.49 In the First Accuracy Report, we were instructed to apply pre-award interest at a rate of LIBOR + 2%, on a compound basis.¹²⁸

4.50 Ultimately, the question of the correct pre-award interest rate to apply in this case is for the Tribunal. Although there is no consensus on which rate should be applied to calculate pre-award interest and a wide range of rates have been applied in past investment arbitrations, our rate is consistent with the FTA, which stipulates that any award of damages should include interest at a “commercially reasonable rate”.¹²⁹

¹²⁵ AlixPartners Report, paragraph 198 ↩

¹²⁶ AC-69 Damodaran - Precious metals industry constituent companies dated 7 January 2022. Analysis based on January 2022 data, being the earliest available dataset at the time of writing. ↩

¹²⁷ First Accuracy Report, paragraphs A4.15 & A4.16; AC-66 Global Arbitration Review, Country Risk dated 1 February 2021, page 1 ("[...] if generating a cash flow in a certain country implies that the cash flow is riskier, then the discount rate reflects that incremental risk through a higher risk premium: the country risk premium.") ↩

¹²⁸ First Accuracy Report, paragraph 9.13 ↩

¹²⁹ Canada-Peru Free Trade Agreement, paragraph 3 of Article 812 ↩

[Page 35]

4.51 However, given that LIBOR has been “effectively retired as a basis for contracts in the United States and that publication of rates will be discontinued as of 30 June 2023",¹³⁰ AlixPartners propose alternative pre-award interest rates of (i) the 180 day average Secured Overnight Funding Rate (“SOFR") + 2%; and (ii) 1-Year US Treasury Bills ("UST”) + 2%.¹³¹

4.52 Whilst SOFR is being earmarked to replace LIBOR once the latter is discontinued, the two rates are not equivalent. Whilst LIBOR is calculated on the basis of rates that London banks charge each other on the interbank market, and includes seven tenors of forward-looking term rates, SOFR is a backward-looking rate based on observable overnight transactions, reflecting borrowing costs in overnight borrowing collateralised by US Treasury securities.¹³²

4.53 As SOFR is based off overnight transactions it is considered to be nearly risk-free, whilst LIBOR includes a bank credit risk premium. Whilst term SOFR became available during 2021 for one, three and six-month tenors to make it more conducive to replacing LIBOR, the Alternative Reference Rates Committee (“ARRC”) has recommended applying a credit spread adjustment to term SOFR to account for the additional credit risk associated with LIBOR.¹³³

4.54 However, AlixPartners have not applied term SOFR as an alternative to LIBOR, but rather the 180 day historical average. As can be clearly seen in the graph below, using a historical average SOFR results in a significant time lag between movements in LIBOR and SOFR, which has been exacerbated by recent increases in interest rates; as at 1 August 2022, the spread between 1-Year LIBOR and 180 day average SOFR was c. 3%.¹³⁴

¹³⁰ AlixPartners Report, paragraph 23 ↩

¹³¹ AlixPartners Report, paragraph 24; RER-0003-Appendix 4, sheets "US Treasury” & “LIBOR – SOFR" ↩

¹³² AC-60 S&P Global, Understanding SOFR dated 2 March 2021, page 2 ↩

¹³³ AC-61 LexisNexis, Determining Spread Adjustments for SOFR Loans dated 6 March 2022, page 1 ↩

¹³⁴ AC-72 Comparison of 1Y LIBOR, 180-day average SOFR & 1Y UST, page 6 ↩

[Page 36]

Figure 4.1. Comparison of 1-Year LIBOR v SOFR (180-day average) v UST (1-Year) –
August 2021 to August 2022

[Chart showing three lines representing LIBOR, SOFR, and UST from Aug-21 to Jul-22]

LIBOR - USD - 1 Year

SOFR (180 Day Average)

UST (1-Year)

Source: AC-72 Comparison of 1Y LIBOR, 180-day average SOFR & 1Y UST

4.55 AlixPartners' other proposed pre-award interest rate, 1-Year UST, is a risk-free rate, and therefore not a “commercially reasonable rate”, as required by the FTA.¹³⁵ Again, the figure above indicates that a spread would be required to bridge the comparable level of risk reflected in the UST rate and LIBOR.

4.56 In conclusion, 180 day average SOFR is not comparable to LIBOR, and should not be used for calculating pre-award interest. If, alternatively, term SOFR or UST were to be used for calculating pre-award interest, a spread should be added to reflect an equivalent level of risk (and to reflect a “commercially reasonable rate').

4.57 As an illustrative example, at 1 August 2022, the spread between LIBOR and UST was c.0.8%,¹³⁶ i.e. a rate of c.UST +2.8% would be required to reflect the same level of risk as LIBOR +2%. This adjustment would need to be reassessed at the date upon which LIBOR is discontinued, should an award be made subsequent to this date.

4.58 We further note that, given the current inflationary environment, which has impacted countries across the world during the first half of 2022, applying any of the rates proposed by ourselves or by AlixPartners would currently result in a negative interest rate in real terms (see Appendix 6).¹³⁷ To the extent that Claimant could have achieved a higher rate of return on any award received (e.g. by investing the award in another mining project), following the principle of full compensation, a higher pre-award interest rate could be considered appropriate.

¹³⁵ Canada-Peru Free Trade Agreement, paragraph 3 of Article 812 ↩

¹³⁶ AC-72 Comparison of 1Y LIBOR, 180-day average SOFR & 1Y UST, page 6 ↩

¹³⁷ For example, US inflation was c.8.5% in July 2022 (AC-62 The Guardian, US Inflation falls to 8.5% in July but still close to Multi-decade High dated 10 August 2022), compared to an average LIBOR rate of c.3.8% (Appendix 3, sheet "USD 12m LIBOR") ↩

[Page 37]

D. Comments on the residual value of IMC's shares in the Actual Situation

4.59 In the First Accuracy Report, in line with the principle of full reparation, we assumed that damages correspond to the difference between Claimant's economic position in the But-For Situation and its economic position in the Actual Situation. Given Claimant's case that Respondent's Alleged Breaches resulted in the loss of Claimant's entire interest in the Invicta Project, we assumed this value to be nil in the Actual Situation.¹³⁸

4.60 However, according to AlixPartners, were the Tribunal to find that Respondent cannot be held liable for PLI's actions (including the foreclosure upon Claimant's shares in IMC), then the residual value of the shares at the time of foreclosure, appraised “proximate to the Valuation Date” by an independent valuation expert to be USD 13m, should be used as a proxy for the value of Claimant's economic position in the Actual Situation.¹³⁹ In that case, AlixPartners state that our damages assessment would have to be reduced by USD 13m to account for the residual value of IMC's shares in the Actual Situation.

4.61 As acknowledged by AlixPartners, this is a legal matter which we cannot opine on. Claimant's case is that “Lonely Mountain's enforcement against IMC's shares was the direct consequence of Peru's acts and omissions” as it was “publicly known that Claimant would lose its investment if it could not resume mining activities”.¹⁴⁰

4.62 Notwithstanding the above, we note the following:

a) AlixPartners appear to ignore the fact that the value of Claimant's economic wealth in the Actual Situation prior to foreclosure would account for both (i) the value of Claimant's shares in IMC; and (ii) the value of Claimant's liabilities towards PLI. On an illustrative basis, deducting the USD 15.9m claimed by PLI in July 2019¹⁴¹ from a valuation of IMC of USD 13m would suggest that Claimant's economic value in the Actual Situation was, in fact, negative.¹⁴²

b) AlixPartners' position appears inconsistent with their opinion that the "fundamental flaws" could mean that the value of the Invicta Project could be nil in the But-For Situation,¹⁴³ given that these same “flaws” would have applied in both the But-For and the Actual Situation. One would assume that, in light of the prevailing situation at the time, the value in the Actual Situation would be lower by some margin than the one in the But-For Situation.

¹³⁸ First Accuracy Report, paragraph 4.35 ↩

¹³⁹ AlixPartners Report, paragraph 97 ↩

¹⁴⁰ Claimant's Memorial, paragraph 195 ↩

¹⁴¹ AC-17 Notice of enforcement of the Pledge over IMC's shares dated 24 July 2019, page 2 ↩

¹⁴² I.e. USD 13m – USD 15.9m = USD (2.9)m ↩

¹⁴³ AlixPartners Report, paragraph 24 ↩

[Page 38]

c) The Actual Situation is supposed to represent events that actually occurred. We understand that, even if Lonely Mountain had not foreclosed upon IMC shares, Claimant did not have access to the Invicta Mine, which Claimant understands is now being actively exploited by the Parán Community.¹⁴⁴ AlixPartners do not attempt to explain how, in such a scenario, Claimant could have realised any value from the mine.

4.63 Furthermore, we understand from our discussions with Claimant that it never received this valuation, which was ordered by the independent trustee holding the IMC ownership shares in order to value the “IMC ownership shares seized by PLI”.¹⁴⁵ As such, it is not possible to ascertain on what basis the valuation was carried out or which assumptions were used. The only reference to such a valuation we have found is one line in Claimant's 2019 financial statements.¹⁴⁶

4.64 For the reasons set out above, we maintain our assumption that Claimant's economic position is nil (at best) in the Actual Situation.

¹⁴⁴ First Accuracy Report, paragraph 3.48; Claimant's Memorial, paragraph 192 ↩

¹⁴⁵ AC-55, Lupaka Gold Corp. financial statements for the year ended 31 December 2019, Note 11, page 22 ↩

¹⁴⁶ AC-55, Lupaka Gold Corp. financial statements for the year ended 31 December 2019, Note 11, page 22 ↩

[Page 39]

5 The financing of the Invicta Project in the But-For Situation

A. Introduction

5.1 In the First Accuracy Report, we were instructed to assess damages incurred by Claimant as a result of the Alleged Breaches by reference to the FMV of the Invicta Project.¹⁴⁷ We also gave consideration to the principle of full reparation, and accordingly assessed damages as the difference between Claimant's economic wealth in the But-For Situation and Claimant's economic wealth in the Actual Situation.¹⁴⁸

5.2 We assessed the FMV of the Invicta Project using the DCF approach, under which future cash flows are discounted to their present value at the cost of capital.¹⁴⁹ Consistent with the FMV standard, which contemplates a hypothetical transaction between unidentified market participants,¹⁵⁰ we discounted future cash flows at the weighted average cost of capital (WACC) of a hypothetical gold mining entity operating in Peru, adjusted to reflect the status of the Invicta Project.¹⁵¹

5.3 We assessed the value of Claimant's outstanding liabilities to PLI in the But-For Situation to be USD 15.9m in both the 355t/day and 590t/day Scenarios, corresponding to the Early Termination Amount actually claimed by PLI. We deducted this amount from the FMV of the Invicta Project in order to calculate damages at the Valuation Date.¹⁵²

5.4 In this section, we (i) set out and respond to AlixPartners' criticisms of our financing assumptions in the But-For Situation; and (ii) set out our updated financing assumptions.

¹⁴⁷ First Accuracy Report, paragraph 4.36 ↩

¹⁴⁸ First Accuracy Report, paragraphs 4.34-4.37 ↩

¹⁴⁹ First Accuracy Report, paragraphs 4.8-4.12 ↩

¹⁵⁰ The FMV is defined as the price, expressed in cash or equivalents, at which an asset would change hands between a hypothetical buyer and seller, acting at arm's length in an open market, where the parties are knowledgeable, informed, prudent and under no compulsion to transact. See First Accuracy Report, paragraph 4.6. ↩

¹⁵¹ First Accuracy Report, paragraphs 5.47-5.51 & 6.19 d) ↩

¹⁵² First Accuracy Report, paragraphs 7.1-7.7 ↩

[Page 40]

B. AlixPartners' criticisms of our financing assumptions

5.5 AlixPartners set out a number of criticisms in relation to our treatment of the financing of the Invicta Project in the But-For Situation under the 590t/day Scenario, including the following:

a) AlixPartners conclude that the 590t/day Scenario is “subject to high uncertainty and should not be used as the basis for damages” because we did not provide any evidence to support the assumption that “a new investor would have funded the Project with friendlier financing terms”, noting that our “cost of debt is 6.2% whereas the effective interest rate under the PLI Loan Agreement is 10.1%” and that there was “no upside participation under the new loan agreement”;¹⁵³

b) Furthermore, they opine that our assessment of the FMV of the Invicta Project under the 590t/day Scenario is overstated by USD 13m as we failed to account for “Claimant's obligation to pay back the debt that would have been used to fund the acquisition of the Mallay Plant”,¹⁵⁴ i.e. the additional funding that would have been received under the Third Amendment to the PLI; and

c) Finally, AlixPartners argue that we failed to provide any evidence that “Claimant would have a capital arrangement in place for the amount of investment required for the purchase of the Mallay Plant”.¹⁵⁵

5.6 We respond to these criticisms below.

C. Our comments on AlixPartners criticisms

Alleged friendlier financing terms and failure to model Claimant's USD 13m repayment obligation

5.7 The FMV reflects a notional transaction between a hypothetical buyer and seller.¹⁵⁶ As such, Claimant-specific factors or financing should generally be disregarded when performing a valuation under the FMV standard.¹⁵⁷

5.8 Accordingly, in the First Accuracy Report, we assumed in the 590t/day Scenario that a hypothetical investor would have funded the Mallay Acquisition at its own cost of capital, treating the acquisition cost as a project cash outflow in the first year of operations.

5.9 Since we did not assume Claimant would have received additional financing under the PLI Loan (a cash inflow), there was no need to model Claimant's obligation to pay back such additional financing (cash outflows).

¹⁵³ AlixPartners Report, paragraphs 138 & 148 ↩

¹⁵⁴ AlixPartners Report, paragraph 146 ↩

¹⁵⁵ AlixPartners Report, paragraph 148 ↩

¹⁵⁶ First Accuracy Report, paragraph 4.6 ↩

¹⁵⁷ See also AC-21 International Valuation Standards effective 31 January 2020, pages 18-19 ↩

[Page 41]

5.10 In discounting the free cash flows to firm (FCFF) of the Invicta Project, we used a cost of capital of 14.7% (in real terms), reflecting the weighted average cost of equity and debt for a hypothetical market participant looking to operate a mine at a similar development stage as that of the Invicta Project in Peru.¹⁵⁸

5.11 AlixPartners' comparison of the 6.2% cost of debt assumed in our cost of capital calculation to the effective interest rate under the PLI Loan agreement is therefore misleading, as it compares the cost of financing the Invicta Project for a hypothetical market participant to a Claimant-specific cost of debt.

5.12 We agree with AlixPartners that our assumed cost of capital for a hypothetical buyer of the Invicta Project does not include a specific upside participation element.

Alleged failure to evidence that Claimant would have the financing in place required to purchase the Mallay Plant

5.13 AlixPartners criticism appears to stem from a misunderstanding of our approach. In the First Accuracy Report, we did not assume that the Claimant would have terminated the PLI Loan and then renegotiated a new loan with a different lender, as AlixPartners suggest.¹⁵⁹ Rather, as explained above, we assumed that a hypothetical investor would have financed the Mallay Plant acquisition at their own cost of capital. Consistent with the FMV standard, our assumption was not intended to be reflective of Claimant's own financing plans.

5.14 As acknowledged by AlixPartners, in order to fund the Mallay Plant acquisition, “Claimant planned to borrow an additional US$13m from PLI under the Draft Third Amendment to the PLI Loan Agreement”.¹⁶⁰ The evidence in respect of Claimant's plans to finance the Mallay Plant acquisition is therefore clear.

D. Our updated financing assumptions

5.15 There are two possible alternatives for modelling financing in the But-For Situation:

1. Assuming that a prospective purchaser would have acquired the Invicta Project unencumbered by the PLI Loan, and consequently that:
   1. any obligations under the PLI Loan would have been settled by Claimant as part of the transaction; and
   2. any future financing requirements would have been met by the purchaser at its WACC; or
2. Assuming that a prospective purchaser would have acquired the Invicta Project with the PLI Loan in place.

¹⁵⁸ First Accuracy Report, paragraph 6.19(d) ↩

¹⁵⁹ AlixPartners Report, paragraph 138 ↩

¹⁶⁰ AlixPartners Report, paragraph 140 ↩

[Page 42]

5.16 In the First Accuracy Report, our damages models reflected the first alternative. The FCFF were discounted at the WACC of a prospective purchaser.

5.17 The available evidence nonetheless indicates that Claimant intended to finance the Mallay Plant acquisition using USD 13m of additional funding received from PLI pursuant to the Third Amendment to the PLI Loan.

5.18 Accordingly, in this report we update our modelling to instead assume that a prospective purchaser would have acquired the Invicta Project with the PLI Loan in place.

5.19 In both the 355t/day and 590t/day Scenarios, we therefore explicitly model the repayments of the PLI Loan and the upside participation. This results in free cash flows to equity (FCFE),¹⁶¹ which we discount back to the Valuation Date at the cost of equity we estimate for a hypothetical market participant looking to operate a mine at a similar development stage. We discuss our updates to the discount rate further in Section 6.

5.20 Under this approach, it is no longer necessary to adjust the resulting valuation for the liabilities that Claimant would have settled with PLI in the But-For Situation.

5.21 We detail our modelling of the PLI Loan in both the 355t/day and 590t/day Scenarios in Appendix 2.

5.22 Ceteris paribus, modelling financing cash flows to calculate FCFE reduces the FMV of the Invicta Project by USD 17.2m in the 590t/day Scenario, and USD 13.7m in the 355t/day Scenario.¹⁶² In comparison, in the First Accuracy Report, our deduction of the Early Termination Amount claimed by PLI from FMV reduced damages at the Valuation Date by USD 15.9m.

5.23 Overall, therefore, directly modelling financing cash flows relating to the PLI Loan has a relatively minor impact compared to our treatment of financing in the First Accuracy Report.

Effective interest rate on the PLI Loan

5.24 Our calculation of the effective interest rate on the PLI Loan of 10.1% in the First Accuracy Report was based on a simplified calculation, given that we did not model the financing of the PLI Loan at that stage. Our simplified calculation assumed that the entirety of the PLI Loan was repaid at the end of the 60-month term, thereby understating the effective interest rate. In the First Accuracy Report, our calculation of the effective interest rate served solely as a proxy for the minimum expected return on Claimant's investment for our sunk costs benchmark; the only impact of this simplification was therefore to understate the indicative value of the Invicta Project based on Claimant's sunk costs.

¹⁶¹ FCFF represent the cash flows available to both shareholders and debt holders, whilst FCFE represent the cash flows available to shareholders after the servicing of external debt. Explicitly modelling financing cash flows accounts for payments made to debt holders (in this instance, Pandion/PLI), and therefore results in a calculation of FCFE. ↩

¹⁶² Figures are the combined impact on damages at the Valuation Date of modelling cash inflows and outflows relating to the PLI Loan, and updating our discount rate from 11.1% to 12.2%, in order to discount FCFE at the cost of equity of a hypothetical gold mining entity operating in Peru. The impact is calculated after first applying other updates to the assumptions used in our damages model; see Table 6.6 and Table 6.11. ↩

[Page 43]

5.25 For the purposes of our updated damages assessment using an FCFE approach we have prepared an amortised loan schedule, which results in a higher effective interest rate of 17.0% for Tranches 1-3 of the PLI Loan.¹⁶³ Whilst this rate is higher than the cost of equity we use for our DCF calculation (12.2%, after applying a 3.3% production premium), the two rates are not directly comparable given the effective interest rate on the PLI Loan is a pre-tax, Claimant-specific rate quoted in nominal terms, whilst the cost of equity is a post-tax shield rate quoted in real terms applicable to a hypothetical gold mining entity in Peru. On an illustrative basis, the post-tax shield effective interest rate on the PLI Loan in real terms would be 9.8%,¹⁶⁴ i.e. lower than our cost of equity.

5.26 Furthermore, the effective rate on the PLI Loan corresponds to Pandion's assessment of Invicta Project risks in June 2016, when it initially made its investment in the project. Following signature of the initial PLI Loan, Claimant commissioned the PEA, carried out extensive pre-production tests and sampling activities, materially completed development works at the Invicta Mine and was in advanced negotiations to purchase the Mallay Plant. Gold prices were also c.16% higher at the Valuation Date than at the date of the initial PLI Loan.¹⁶⁵ Ceteris paribus, each of the factors above would have reduced the overall risk profile of the Invicta Project at the Valuation Date.

5.27 Based on the above factors, the cost of equity we use in order to discount FCFE appears reasonable in comparison to the effective interest rate on the PLI Loan.

¹⁶³ We calculate an effective interest rate of 16.1% (nominal, pre-tax) in our 590t/day Scenario, similar to the equivalent rate on Tranches 1-3 of the PLI Loan. The slight decreases arises from the assumption in the 590t/day Scenario that gold repayments would start in January 2020 rather than September 2019, in accordance with the likely timeframe of the Mallay Plant acquisition and the terms of the Third Amendment to the PLI Loan. ↩

¹⁶⁴ (1 + (17.0% * (1 - 29.5%))) ÷ 1.02 - 1 = 9.8% ↩

¹⁶⁵ USD 1,548.10 as at 26 August 2019 and USD 1,332.90 as at 30 June 2016. 1,548.10 / 1,332.90 = 16.1% (AC-71 COMEX gold prices - June 2016 to August 2019) ↩

[Page 44]

6 Our updated damages assessment

A. Introduction

6.1 In this section, we set out the updates we make to our FMV calculation under both the 590t/day and 355t/day Scenarios, and present the impact on total damages as at 31 August 2022 (being a proxy for the date of this report) after applying pre-award interest.

6.2 In line with the First Accuracy Report, we continue to assess the FMV of the Invicta Project at the Valuation Date using an income-based DCF approach. As explained in Section 5, we update our approach slightly in order to calculate FCFE, after explicitly modelling financing cash flows based on the terms of the PLI Loan.

6.3 Broadly speaking, our updates to our damages assessment can be split into three categories:

1. Updates to reflect Micon's revised production, cost and operational assumptions;
2. Our additional updates to cost assumptions; and
3. Our modelling of financing cash flows.

6.4 Our updated damages models can be found in Appendix 3 (590t/day Scenario) and Appendix 4 (355t/day Scenario) to this report. Our models contain a number of switches and sensitivities, which enable the user to see the impact of updating certain assumptions to our damages assessment, including comparing our updated damages assessment against both the First Accuracy Report and the underlying SRK/Red Cloud Models.¹⁶⁶

Restatement of project years

6.5 In their report, Micon restate production and cost schedules into project years running from September to August; this update is reflected in our updated damages assessment.¹⁶⁷ As such, whilst we present comparisons of our updated assumptions against those of the First Accuracy Report, the figures shown in individual periods may not be directly comparable across the two reports.¹⁶⁸

6.6 For the sake of simplicity, in both the 590t/day and 355t/day Scenarios we assume that all cash flows occurring prior to the Valuation Date (i.e. all Year 1 cash flows) occur on the Valuation Date.

¹⁶⁶ See Appendix 3 and Appendix 4, sheet “Dashboard (ACC)”. Both our updated damages models contain three switches which allow the user to reset to the post-tax NPV of the underlying Red Cloud/SRK Models, First Accuracy Report and Second Accuracy Report. Furthermore, we have also added individual switches which show the impact of stand-alone adjustments, such as including/excluding financing cash flows or applying Micon's updated assumptions, on post-tax NPV under each production scenario. ↩

¹⁶⁷ Technically speaking, each year of our damages calculation ends on 26 August, whilst Micon's schedules end on 31 August; however, we do not consider this minor timing difference to have a material impact on the overall damages calculation. ↩

¹⁶⁸ This is confirmed by Micon, who note that “costs in individual periods are not directly comparable” (Micon Report, paragraph 91) ↩

[Page 45]

B. Updates to our FMV calculation under the 590t/day Scenario

Production plan

Updates to reflect Micon's revised production plan

6.7 We update the Invicta Project production plan in the 590t/day Scenario, in line with the updated production plan set out in the Micon Report.

Ramp-up of production

6.8 In the First Accuracy Report, we adopted the simplified assumption that production would have commenced at the Valuation Date, noting that this represented “a conservative estimate of the date upon which production could have started".¹⁶⁹

6.9 Based on their analysis of available data, Micon consider that, in the But-For Situation (i.e. but for the Blockade and subsequent events), Claimant could have commenced commercial production using third-party toll processors in November 2018, allowing for a three-month ramp-up in production volume to 355t/day in order to ensure the availability of sufficient toll milling capacity.

6.10 Micon then assume that Claimant would have started processing mined ore at the Mallay Plant on or around the Valuation Date (with a further ramp-up to 590t/day from September 2019 onwards). This assumption appears reasonable, given that approval of the relevant transfer agreement by the Mallay Community was obtained by Buenaventura in March 2019,¹⁷⁰ and that Claimant anticipated a further 3 months for completing the registration and the transfer of licences and permits following closing.¹⁷¹

6.11 Micon assume the following ramp-up assumptions:¹⁷²

1. Start date for production in November 2018, starting at 100t/day and ramping up to 355/day by February 2019;
2. Start date for processing of mined ore at the Mallay Plant at or around the Valuation Date, with production ramping up from 355t/day in August 2019 to 590t/day by November 2019.

¹⁶⁹ First Accuracy Report, paragraph 5.16 ↩

¹⁷⁰ C-233 Email from Buenaventura to Lupaka dated 11 March 2019 ↩

¹⁷¹ AC-11 Lupaka Gold Corp. Minutes of the board meeting dated 27 September 2018, page 2 ↩

¹⁷² Micon Report, paragraphs 103.Q & 103.R ↩

[Page 46]

Figure 6.1. Micon's assumed production ramp-up in the 590t/day Scenario

[Chart showing production ramp-up]

Y-Axis: Tons/day (from 100 to 700)

X-Axis: Sep-18, Oct-18, Nov-18, Dec-18, Jan-19, Feb-19, Mar-19, Apr-19, May-19, Jun-19, Jul-19, Aug-19, Sep-19, Oct-19, Nov-19, Dec-19, Jan-20, Feb-20, Mar-20, Apr-20, May-20, Jun-20, Jul-20, Aug-20

Annotation: Valuation Date (between Jul-19 and Aug-19)

Source: Micon Report, paragraphs 103.Q & 103.R

Production volume and grade

6.12 Based on their review, Micon conclude that it is appropriate to apply a gold-equivalent cut-off grade of 3.0g/t in the 590t/day Scenario, which is lower than the cut-off grade of 4.0g/t in the PEA (and also used in the Red Cloud Model).¹⁷³ As explained by Micon, the cut-off grade is the grade at which it is economically viable for mineralised material to be processed.¹⁷⁴

6.13 As a result, Micon assume total production of 1,814kt of mineralised material over 10 years,¹⁷⁵ representing an increase of 447kt from the First Accuracy Report (1,367kt).

6.14 Based on the lower cut-off grade, assumed average metal grades are generally lower in comparison to the First Accuracy Report:

1. Gold grade decreases from 4.85g/t to 4.09g/t;
2. Silver grade decreases from 29.56g/t to 25.97g/t;
3. Copper grade decreases from 0.71% to 0.54%;
4. Lead grade increases from 0.43% to 0.59%;
5. Zinc grade decreases from 0.50% to 0.47%.

¹⁷³ Micon Report, paragraph 103.C ↩

¹⁷⁴ Micon Report, page iv ↩

¹⁷⁵ Of which c.8 full years of steady-state production at 590t/d (Micon Report, paragraph 111) ↩

[Page 47]

6.15 Applying Micon's updated estimate of metal grades in each year results in a lower average gold-equivalent grade of 5.48g/t,¹⁷⁶ compared to 6.40g/t in the First Accuracy Report.¹⁷⁷

Metal prices

6.16 Whilst Micon do not comment upon metal prices in their report, in order to align our damages assessment with Micon's assumed production period of 10 years, we update our metal price assumptions as follows:

1. Year 1 of our damages assessment now pertains to the period prior to the Valuation Date, in line with Micon's revised production plan. For this period only, we apply actual metal spot prices to calculate revenues.¹⁷⁸
2. From Year 2 onward, to the extent available, we apply the same metal futures prices from Capital IQ as in the First Accuracy Report.¹⁷⁹
3. Consistent with our approach in the First Accuracy Report, for periods in which futures prices were not available at the Valuation Date, we apply the CAGR of the relevant metal price across the remainder of the production period.¹⁸⁰ This is a conservative assumption, given that the CAGR for all metals in available periods was negative (i.e. we assume that real metal prices would decrease across the whole production period).

Working capital

6.17 Similarly, in order to align our calculation with Micon's production plan, we extend our calculation of working capital requirements (applying the same assumptions as in the First Accuracy Report) over the whole production period.

Comparison of updated production schedule to the First Accuracy Report

6.18 We set out a comparison of the annual production schedule and average annual gold-equivalent grade between the First Accuracy Report and our updated assessment of damages below.

¹⁷⁶ As noted in footnote 84, Table 5.13 of the Micon Report shows a gold-equivalent grade of 5.92g/t, higher than our figure of 5.48g/t. The difference arises because Micon calculate the gold-equivalent grade using metal prices from the original Red Cloud Model, rather than our updated metal prices in the First Accuracy Report. This calculation is illustrative only and does not have any impact on our damages assessment. ↩

¹⁷⁷ The 6.40g/t figure in the First Accuracy Report was calculated after applying our adjustments for both (i) cut-off grade and (ii) updated metal prices (First Accuracy Report, footnote 171) ↩

¹⁷⁸ To calculate Year 1 revenues, we apply the average spot prices for each metal between November 2018 and August 2019. This is a prudent approach, to the extent that Micon's ramp-up assumptions result in higher production at the end of Year 1, when spot prices were higher. ↩

¹⁷⁹ First Accuracy Report, paragraph 5.22 ↩

¹⁸⁰ First Accuracy Report, paragraph 5.23 ↩

[Page 48]

Table 6.1. 590t/day Scenario – Production plan – First Accuracy Report v Second Accuracy Report¹⁸¹

Source: Appendix 3, sheet “Table 6.1”

Operating costs

Updates to reflect Micon's revised operating cost assumptions

6.19 For the most part, Micon retain the operating cost inputs and assumptions used in the Red Cloud Model.¹⁸² However, they make the following specific updates:

1. Updating SRK's provision of USD 700k p.a. for employee profit sharing,¹⁸³ so that total life-of-mine profit sharing remains the same as a percentage of pre-tax operating profits (5.1%).¹⁸⁴ On this basis, Micon estimate a provision of USD 900k p.a.; we re-estimate this figure based on our own assessment of pre-tax operating profits and increase the provision to USD 975k p.a.¹⁸⁵
2. Increase in the direct unit cost of treating ore at the Mallay Plant from USD 20.0/t to USD 25.5/t to include the cost of electrical energy, which was previously omitted.¹⁸⁶
3. Reduction in the cost of delivering concentrate to the Callao Port from USD 51.35/t to USD 31.35/t, to account for the relative proximity of the Mallay Plant to Callao compared to Caraz, the assumed location of the processing plant in the PEA (which was not updated by Red Cloud).¹⁸⁷

¹⁸¹ In the Second Accuracy Report, our Year 1 production volumes reflect Micon's revised production plan, which restates production years to run from September to August and, based on an assumed start date of November 2018, results in 10 months of production in Year 1 (with a ramp-up in the first three months). In the First Accuracy Report, our Year 1 production volume was based on the assumptions of the PEA, which was effective from 13 April 2018 (AC-02 SRK Consulting PEA dated 13 April 2018, page ii) and stated on a calendar year basis (i.e. Year 1 in our model equated to 2018 in the PEA, which appears to account for c.8.5 months of production). As a result, total Year 1 production volume is slightly higher in our updated damages assessment than in the First Accuracy Report despite the ramp-up applied by Micon. ↩

¹⁸² Micon Report, paragraph 112. Micon conclude that “the Red Cloud model made reasonable simplifying assumptions and their first-order approximations do not deviate substantially from Micon's more accurate results”. ↩

¹⁸³ AC-29 SRK Model, sheet “opex”, cell C40 ↩

¹⁸⁴ Micon Report, paragraph 100iii) ↩

¹⁸⁵ Appendix 3, sheet “opex”, row 40. Provision applied pro rata in line with production days in each year. Our calculation is based on a higher pre-tax operating profit figure than Micon (principally as a result of the fact that Micon do not consider metal price updates in their analysis), resulting in a higher provision. ↩

¹⁸⁶ Micon Report, paragraph 100iv) ↩

¹⁸⁷ Micon Report, paragraph 100v) ↩

[Page 49]

4. Correction of a formula error in the underlying SRK Model,¹⁸⁸ which excluded copper and lead concentrate tonnages from the unit rate for concentrate transport resulting.
5. Notwithstanding the above changes, consistent with their production forecast (which assumes that Claimant would process mined ore at third party processors in Year 1), Micon apply the same unit cost rates for processing and concentrate transport as in the 355t/day Scenario in Year 1.¹⁸⁹

Our additional updates to operating cost assumptions

6.20 In addition to the above updates based on the Micon Report, we make the following further updates to operating cost assumptions:

1. In the First Accuracy Report, we directly modelled closure costs based on our understanding of the latest Mine Closure Plan, conservatively assuming that closure costs would be front-loaded and ultimately paid off three years before the end of the production period.¹⁹⁰ In their report, Micon do not make any adjustment to closure costs as a result of their extended production plan. For prudency, we extend our calculation of closure costs, assuming that additional costs would be incurred on a pro rata basis (c. USD 100k p.a.) across the extended mine plan;¹⁹¹
2. We correct for what appears to be an incorrect formula in the underlying SRK Model, which understated the zinc treatment charge by USD 0.1m in the First Accuracy Report;¹⁹² and
3. As explained in Section 4, we add additional G&A costs of USD 350k p.a.¹⁹³ in relation to the management of the Mallay Plant.

Comparison of updated operating cost assumptions to the First Accuracy Report

6.21 The above updates result in an increase unit operating costs per tonne milled from USD 80.93/t in the First Accuracy Report¹⁹⁴ to USD 94.3/t.¹⁹⁵ In conjunction with the increased total production volume compared to the Red Cloud Model, this leads to an increase in total life-of-mine operating costs from USD 110.6m in the First Accuracy Report to USD 171.1m in our updated damages assessment.

¹⁸⁸ AC-15 Red Cloud Model, sheet “opex”, row 8 ↩

¹⁸⁹ Micon Report, paragraph 100vi). Specifically, in Year 1 (i) processing/owner supervision costs are calculated using a unit cost of USD 35.0/t as compared to USD 25.5/t thereafter; (ii) transportation and stockpile maintenance costs are calculated using a unit cost of USD 50.2/t as compared to USD 18.0/t thereafter; and (iii) selling costs are calculated using a unit cost of USD 51.35/t as compared to USD 31.35/t thereafter. ↩

¹⁹⁰ First Accuracy Report, paragraph 5.32 ↩

¹⁹¹ Appendix 3, sheet “Closure Costs (ACC)”. In line with our assumption in the First Accuracy Report, we assume that closure costs would be front-loaded and paid off three years before the end of the production period. ↩

¹⁹² First Accuracy Report, Appendix 6, sheet “fin_lupaka”, cell E229. The formula for the Zinc treatment charge referenced the gold refining charge (cell E147) rather than the net treatment charge (cell E193). ↩

¹⁹³ On a pro rata basis. ↩

¹⁹⁴ And USD 81.33/t in the Red Cloud Model (AC-15, Red Cloud Model, sheet “fin_lupaka”, row 254) ↩

¹⁹⁵ Appendix 3, sheet “opex”, row 63 ↩

[Page 50]

6.22 A comparison of operating costs in each production year between th First Accuracy Report and our updated damages assessment is set out below.

Table 6.2. 590t/day Scenario – Operating costs – First Accuracy Report v Second Accuracy Report

Source: Appendix 3, sheet “Tables 6.2-6.3”

Capital expenditure

Updates to reflect Micon's revised capital expenditure assumptions

6.23 As noted by Micon, the capital cost of development in the Red Cloud Model was pro-rated from SRK's estimates, maintaining the same average cost of development per tonne.¹⁹⁶ In the First Accuracy Report, we updated the capital expenditure assumptions of the Red Cloud Model to exclude pre-production and development capex of USD 4.3m in Year 1, on the basis that “construction works, including the new access road, [were] materially completed by September 2018".¹⁹⁷

6.24 Micon have reviewed the development plan in the Red Cloud Model and estimated annual capital expenditure based on a specific three-dimensional mine layout, development sequencing and analysis of annual development required.¹⁹⁸ In doing so, Micon have (i) taken into account the actual metreage of development ore mined prior to the Blockade Date;¹⁹⁹ and (ii) extended the annual provision for sustaining capital expenditure to cover the entire production period.²⁰⁰ We have applied Micon's revised development plan in our updated damages assessment.

6.25 To reflect the tax impact of Micon's updated capital expenditure assumptions, we update the depreciation calculation in our model in order to ensure that depreciation, and therefore taxable income, are aligned with our capital expenditure assumptions.²⁰¹

¹⁹⁶ Micon Report, paragraph 103.V ↩

¹⁹⁷ First Accuracy Report, paragraphs 5.30-5.31 ↩

¹⁹⁸ Micon Report, paragraph 103.V ↩

¹⁹⁹ Micon Report, paragraph 103.V ↩

²⁰⁰ Micon Report, paragraph 103.P ↩

²⁰¹ For expediency, we calculate depreciation until the end of year 10 only; as depreciation is a non-cash item, any further depreciation beyond the end of this period this does not have an impact on our DCF. In practice, Claimant may have been able to extract additional value from the sale of capital assets with a positive residual value at the end of the production period; however, we do not account for this in our damages assessment. ↩

[Page 51]

Our additional updates to capital expenditure assumptions

6.26 Micon accept as “reasonable and appropriate" Red Cloud's estimates for the purchase and modification of the Mallay Plant, as well as the associated cost of a closure bond.²⁰² However, Micon opine that, had the Mallay Plant acquisition gone ahead, Claimant would likely have commissioned an independent update to the PEA reflecting the expanded operation of the Invicta Project, at a cost of USD 0.1m to USD 0.2m.²⁰³

6.27 As a prudent assumption, we include an additional cost of USD 0.2m in Year 1 of our damages assessment, to cover professional fees or auxiliary costs associated with the purchase of the Mallay Plant.²⁰⁴

Comparison of updated capital expenditure assumptions to the First Accuracy Report

6.28 As a result of the above updates, total capital expenditure in the 590t/day Scenario in our updated damages assessment (including costs associated with the purchase of the Mallay Plant) is USD 42.4m, compared to USD 37.3m in the First Accuracy Report.

Table 6.3. 590t/day Scenario – Capital expenditure – First Accuracy Report v Second Accuracy Report

Source: Appendix 3, sheet “Tables 6.2-6.3”

Modelling of financing cash flows

6.29 As described in Section 5, for the purposes of our updated damages assessment, we explicitly model financing cash flows relating to the PLI Loan within our DCF calculation. As such, no further adjustment is required to our FMV calculation to reflect the value of debts to be settled with PLI in the But-For Situation.

6.30 In order to accurately model the tax impact of PLI Loan repayments, we deduct the annual financing cost (i.e. the value of gold repayments that correspond to interest on the PLI Loan) and upside participation from taxable income.²⁰⁵

²⁰² Micon Report, paragraph 103.O; AC-15 Red Cloud Model, sheet “fin_lupaka”, rows 263-265 ↩

²⁰³ Micon Report, paragraph 94 ↩

²⁰⁴ Appendix 3, sheet “capex-sum”, cell F10 ↩

²⁰⁵ Appendix 3, sheet “fin_lupaka”, rows 362 & 363. In the First Accuracy Report, we applied Claimant's carried forward tax losses in Peru of USD 3.2m at the end of 2018, which could have been used to offset any tax liabilities in the But-For Situation, to the tax calculation in our damages models (First Accuracy Report, paragraph 5.45). In order not to double count the impact of any financing costs already included within Claimant's tax losses at the end of 2018, we deduct the interest expense from taxable income from January 2019 onwards only. ↩

[Page 52]

6.31 As demonstrated in Table 6.6 below, ceteris paribus, directly modelling financing cash flows reduces FMV by USD 17.2m, slightly higher than the USD 15.9m adjustment we made in the First Accuracy Report to account for debts to be settled with PLI.

Discount rate

6.32 We make two updates to the discount rate used for the purposes of our updated damages assessment.

Update to our pre-production risk premium based on the Micon Report

6.33 Having reviewed and revised the 590t/day cost forecasts and production schedule, Micon “considers the level of accuracy of the revised 590 t/d plan to be very similar to that achieved in the 355 t/d plan and that both plans are equally reliable".²⁰⁶

6.34 As such, the level of uncertainty attached to the 590t/day Scenario is reduced, and Micon consider the use of the same discount rate in the 355t/day and 590t/day Scenarios to be appropriate.²⁰⁷

6.35 Accordingly, given that we now align our production and cost assumptions with those of the Micon Report, we reduce the pre-production risk premium applied to our discount rate in the 590t/day Scenario to 3.3%, in line with the 355t/day Scenario.

Our application of a cost of equity to reflect the calculation of FCFE

6.36 As explained in Section 5, directly modelling financing cash flows (as we now do) results in a calculation of FCFE (which represent cash flows available to equity holders only), rather than FCFF (which represent cash flows available to both debt and equity holders).

6.37 We therefore now discount cash flows in each year using the cost of equity of a hypothetical gold mining entity operating in Peru, rather than its weighted average cost of capital (WACC).

²⁰⁶ Micon Report, paragraph 114 ↩

²⁰⁷ Micon Report, paragraph 115 ↩

[Page 53]

Our updated discount rate calculation

6.38 As a result of the two updates above, we now apply a discount rate of 12.2% to our DCF calculation in both the 590t/day and 355t/day scenarios.

Table 6.4. Our updated discount rate calculation in the 590t/day and 355t/day Scenarios

Source: Appendix 3, sheet “Table 6.4”

Our updated assessment of FMV under the 590t/day Scenario

6.39 After applying the updates described above to our DCF calculation, we calculate the FMV of the Invicta Project at the Valuation Date in the 590t/day Scenario to be USD 41.0m.

Table 6.5. 590t/day Scenario – Updated damages assessment – annual cash flows

Source: Appendix 3, sheet “Table 6.5”

6.40 As we (i) explicitly model financing cash flows; and (ii) assume Claimant's economic wealth to be nil in the Actual Situation, damages are now equal to the FMV of the Invicta Project at the Valuation Date.

6.41 Overall, our updates result in a decrease in damages of USD 6.7m at the Valuation Date in the 590t/day Scenario in comparison to the First Accuracy Report.

Table 6.6. 590t/day Scenario – Updated damages at the Valuation Date – Comparison to the First Accuracy Report

Source: First Accuracy Report, Appendix 6; Appendix 3, sheet “Table 6.6”

[Page 54]

C. Updates to our FMV calculation under the 355t/day Scenario

Production plan

Updates to reflect Micon's revised production plan

Ramp-up of production

6.42 We update the Invicta Project production plan in the 355t/day Scenario, in line with the updated production plan set out in Micon Report.

6.43 Consistent with the approach taken in the 590t/day Scenario, in order to ensure the availability of adequate toll processing capacity, Micon assume a start date for production in November 2018, starting at 100t/day and ramping up to steady state production of 355t/day (using third party toll processors) by February 2019.²⁰⁸

Figure 6.2. Micon's assumed production ramp-up in the 355t/day Scenario

[Chart showing production ramp-up]

Y-Axis: Tons/day (from 100 to 700)

X-Axis: Sep-18, Oct-18, Nov-18, Dec-18, Jan-19, Feb-19, Mar-19, Apr-19, May-19, Jun-19, Jul-19, Aug-19, Sep-19, Oct-19, Nov-19, Dec-19, Jan-20, Feb-20, Mar-20, Apr-20, May-20, Jun-20, Jul-20, Aug-20

Annotation: Valuation Date (between Jul-19 and Aug-19)

Source: Micon Report, paragraph 83iii)

Production volume and grade

6.44 Micon do not make any updates to production volumes in the 355t/day Scenario, assuming the same overall production volume of 670kt²⁰⁹ as in the SRK Model (as reflected in the First Accuracy Report).

6.45 Similarly, they assume the same average gold-equivalent grade of 8.58g/t,²¹⁰ with minor variances in metal grades in individual years resulting from the restatement of production years to run from September to August in each year.

²⁰⁸ Micon Report, paragraph 83iii) ↩

²⁰⁹ Micon Report, Table 5.6 ↩

²¹⁰ Micon Report, Table 5.6 ↩

[Page 55]

Table 6.7. 355t/day Scenario – Production plan – First Accuracy Report v Second Accuracy Report²¹¹

Source: Appendix 4, sheet “Table 6.7”

Metal prices

6.46 Consistent with our approach in the 590t/day Scenario, given that Year 1 of our damages assessment now pertains to the period prior to the Valuation Date, we apply actual metal spot prices to calculate revenues in this period.²¹²

Operating costs

Updates to reflect Micon's revised operating cost assumptions

6.47 Micon consider the PEA operating cost estimates (as presented in the SRK Model and the First Accuracy Report) to be “reasonable and appropriate to the proposed scale of operation".²¹³

6.48 They further note that the transport cost assumptions in the PEA were conservative, given that there were several available toll milling plants which were closer to the Invicta Project than Caraz, the assumed location of the toll treatment plant in the PEA.²¹⁴

²¹¹ As noted in footnote 181, the slight differences in production volumes in the Second Accuracy Report compared to the First Accuracy Report are the result of Micon's assumed start date for production in November 2018 and their restatement of production years from September to August, resulting in 10 months' production in Year 1. Overall production volumes do not change in the 355t/day Scenario. ↩

²¹² Consistent with our 590t/day Scenario, we apply the average spot prices for each metal between November 2018 and August 2019. ↩

²¹³ Micon Report, paragraph 87 ↩

²¹⁴ Micon Report, paragraph 88 ↩

[Page 56]

6.49 Micon make the following specific changes to operating cost assumptions (also reflected in their updates to the 590t/day Scenario), which we adopt in our updated assessment of damages:

1. Updating SRK's provision for employee profit sharing.²¹⁵ Consistent with our approach in the 590t/day Scenario, we re-estimate the annual provision so that total life-of-mine profit sharing remains the same as a percentage of pre-tax operating profits as in the SRK Model (5.1%). On this basis, we increase the provision from USD 700k p.a. to USD 790k p.a.;²¹⁶ and
2. Correction of a formula error in the underlying SRK Model, which excluded copper and lead concentrate tonnages from the unit rate for concentrate transport.²¹⁷

Our additional update to operating cost assumptions

6.50 Consistent with our 590t/day Scenario, we correct for what appears to be an incorrect formula in the underlying SRK Model, which understated the zinc treatment charge by USD 0.1m in the First Accuracy Report.²¹⁸

Comparison of updated operating cost assumptions to the First Accuracy Report

6.51 After applying the above updates, unit operating costs are USD 143.5 per tonne milled,²¹⁹ a marginal increase from USD 143.3/t in the First Accuracy Report.²²⁰ Total operating costs in our updated damages assessment are USD 96.0m, in line with the First Accuracy Report.

Table 6.8. 355t/day Scenario – Operating costs – First Accuracy Report v Second Accuracy Report

Source: Appendix 4, sheet “Tables 6.8-6.9”

²¹⁵ Micon Report, paragraph 100iii) ↩

²¹⁶ Appendix 4, sheet “opex”, row 40. Provision applied pro rata in line with production days in each year. Our calculation is based on a higher pre-tax operating profit figure than Micon (principally as a result of the fact that Micon do not consider metal price updates in their analysis), resulting in a higher provision. ↩

²¹⁷ AC-29 SRK Model, sheet “opex”, row 8 ↩

²¹⁸ First Accuracy Report, Appendix 5, sheet “fin_lupaka”, cell E229. The formula for the Zinc treatment charge referenced the gold refining charge (cell E147) rather than the net treatment charge (cell E193). ↩

²¹⁹ Appendix 4, sheet “opex”, cell H62 ↩

²²⁰ First Accuracy Report, Appendix 5, sheet “opex”, row 62 ↩

[Page 57]

Capital expenditure

Updates to reflect Micon's revised capital expenditure assumptions

6.52 Micon update the capital expenditure assumptions of the PEA to exclude their estimate of expenditure on site roads and underground development for mine access and stope preparation actually incurred by Claimant prior to the Blockade Date. This results in a reduction to life-of-mine capital expenditure of c.USD 3m compared to the PEA.

6.53 Micon's update, which we adopt in our updated damages assessment, is more conservative than the approach we took in the First Accuracy Report, whereby we excluded all pre-production and development capital expenditure (totalling USD 4.3m) in Year 1.

6.54 Consistent with our updates to the 590t/day Scenario, we update the depreciation calculation in our model in order to ensure that depreciation, and therefore taxable income, are aligned with our capital expenditure assumptions.

Comparison of updated capital expenditure assumptions to the First Accuracy Report

6.55 As a result of Micon's updates, capital expenditure in our updated damages assessment (USD 9.7m) is higher than that assumed in the First Accuracy Report (USD 8.4m).

Table 6.9. 355t/day Scenario – Capital expenditure in the First Accuracy Report v Second Accuracy Report

Source: Appendix 4, sheet “Tables 6.8-6.9”

Discount rate

6.56 Consistent with the 590t/day Scenario, we discount FCFE at a rate of 12.2%, being the cost of equity of a hypothetical gold mining entity operating in Peru, inclusive of a pre-production premium of 3.3%.

Modelling of financing cash flows

6.57 Consistent with the 590t/day Scenario, we explicitly model financing cash flows relating to the PLI Loan within our DCF calculation.

[Page 58]

6.58 Again, to ensure the correct modelling of the tax impact, we deduct the annual financing cost and upside participation from taxable income.²²¹

6.59 As shown in Table 6.11 below, ceteris paribus, directly modelling financing cash flows increases FMV by USD 13.7m, slightly lower than the USD 15.9m adjustment we made to FMV in the First Accuracy Report to account for debts to be settled with PLI in the But-For Situation.

Our updated assessment of FMV under the 355/day Scenario

6.60 After applying the updates described above to our DCF calculation, we calculate the FMV of the Invicta Project at the Valuation Date in the 355t/day Scenario to be USD 32.1m.

Table 6.10. 355t/day Scenario – Updated damages assessment – annual cash flows

Source: Appendix 4, sheet “Table 6.10”

6.61 As explained above, under our approach, damages at the Valuation Date are equivalent to the FMV of the Invicta Project.

6.62 Overall, our updates result in an increase in damages of USD 3.8m at the Valuation Date in the 355t/day Scenario in comparison to the First Accuracy Report.

Table 6.11. 355t/day Scenario – Updated damages at the Valuation Date – Comparison to the First Accuracy Report

Source: First Accuracy Report, Appendix 5; Appendix 4, sheet “Table 6.11”

²²¹ Appendix 4, sheet “fin_lupaka”, rows 299 & 300. As in our 590t/day Scenario, given we already model the impact of carried forward tax losses until the end of 2018, we deduct the financing cost from taxable income from January 2019 onwards. ↩

[Page 59]

D. Summary of our updated assessment of damages

Summary of damages at the Valuation Date

6.63 Following the updates to our damages assessment described above, we estimate damages at the Valuation Date to be:

1. USD 41.0m in the 590t/day Scenario, a decrease of USD 6.7m in comparison to the First Accuracy Report; and
2. USD 32.1m in the 355t/day Scenario, an increase of USD 3.8m in comparison to the First Accuracy Report.

Table 6.12. Summary of damages at the Valuation Date – Second Accuracy Report v First Accuracy Report

Note: FAR stands for First Accuracy Report and SAR stands for Second Accuracy Report

Source: Appendix 3, sheet “Table 6.12”

Application of pre-award interest to 31 August 2022

6.64 In light of current high inflation rates, we are instructed to calculate pre-award interest from the Valuation Date to 31 August 2022, being a proxy for the date of this report, using alternative interest rates of:

1. LIBOR +4%; or
2. UST +5%.

6.65 In Appendix 6 we demonstrate how, from an economic perspective, applying the two rates above between the Valuation Date and 31 August 2022 results in pre-award interest broadly in line with inflation over the same period (in other words, applying these rates results in Claimant effectively “breaking-even” over the period in real terms). Conversely, applying LIBOR +2%, as we did in the First Accuracy Report, results in negative pre-award interest in real terms.

6.66 After applying pre-award interest at LIBOR + 4%, we estimate damages to be USD 47.6m and USD 37.3m at 31 August 2022 under the 590t/day Scenario and 355t/day Scenario respectively.

6.67 Alternatively, applying pre-award interest at UST + 5% would result in damages at 31 August 2022 of USD 48.5m under the 590t/day Scenario, or USD 38.0m under the 355t/day Scenario.

[Page 60]

Table 6.13. Updated assessment of damages as at 31 August 2022

Source: Appendix 3, sheet “Table 6.13”

6.68 Our calculation of pre-award interest would need to be updated in order to reflect the date of any actual award.

[Page 61]

7 AlixPartners' comments on other indicators of value

7.1 In the First Accuracy Report, we benchmarked our damages assessments at the Valuation Date to other indicators of value of the Invicta Project, including (i) Claimant's market capitalisation; (ii) Claimant's sunk costs; (iii) relevant gold mining industry transaction multiples; and (iv) other contemporaneous valuation evidence such as the SRK Model and Red Cloud Model.²²²

7.2 AlixPartners opine that these indicators are “either inaccurate or do not demonstrate the reasonableness of Accuracy's quantification of damages under the Primary Approach”.²²³

7.3 In this section, we discuss and respond to the arguments presented in the AlixPartners Report. Our updated calculations are detailed in Appendix 5.

A. Claimant's market capitalisation

First Accuracy Report

7.4 As an illustration of the value of Claimant's Investment but for Respondent's Alleged Breaches, we assessed Claimant's hypothetical market capitalisation at the Valuation Date as follows:²²⁴

1. We applied the percentage movement in the GDXJ between 25 October 2018 (the date of Claimant's public announcement in relation to the Blockade) and the Valuation Date (26 August 2019) to Claimant's market capitalisation of CAD 21.4m on 25 October 2018; and
2. Given that Claimant's Investment represented the entirety of its interest in the Invicta Project, we added a control premium to reflect the additional consideration that a hypothetical investor would pay over a marketable minority equity value in order to own a controlling interest in a company or project.

7.5 Based on this market capitalisation approach, we derived an indicative value of USD 33.4m at the Valuation Date.²²⁵ We did not make any further adjustment for Claimant's obligation to pay the Early Termination Amount as this was already reflected in the market capitalisation.²²⁶

²²² First Accuracy Report, paragraph 8.1 ↩

²²³ AlixPartners Report, paragraph 175 ↩

²²⁴ First Accuracy Report, paragraph 8.13 ↩

²²⁵ First Accuracy Report, paragraph 8.22 ↩

²²⁶ First Accuracy Report, paragraph 8.21 ↩

[Page 62]

The AlixPartners Report

7.6 AlixPartners consider that, whilst Claimant's market capitalisation may be relevant to valuing Claimant directly, “it is likely that the result does not represent the FMV of the Invicta Project”, and that our approach inflates the value of IMC.²²⁷

7.7 AlixPartners set out a number of arguments to support their position:²²⁸

1. CIMVAL guidance states that market capitalisation is more applicable to the valuation of single property asset junior companies than to properties;
2. According to AlixPartners, our indicative valuation is based on the assumption that a 1% increase in the VanEck Vectors Junior Gold Miners index (GDXJ) would lead to a 1% change in Claimant's market capitalisation which, AlixPartners argue, is simplistic and inaccurate because market indices are composites of various individual stocks and reflect systematic changes in the industry being tracked rather than the risk of an individual stock;
3. AlixPartners claim that Claimant's historical share price has not tracked the GDXJ Index reliably, which would in turn demonstrate that our analysis, which focuses on one particular date, would be unreliable;
4. AlixPartners estimate that regression analysis shows that the coefficient between the daily change of Claimant's share price and that of the GDXJ is 0.3406 with an R² of 1.8%;
5. AlixPartners opine that the addition of a control premium artificially inflates the indicated value in ways not attributable to the economic value of the investment; and
6. Finally, AlixPartners consider that Claimant's share price might be distorted due to liquidity issues.

Our comments

Our market capitalisation approach was an illustrative assessment only, and may be conservative for a variety of reasons

7.8 AlixPartners fail to recognise that, from March 2018 onward,²²⁹ Claimant was in fact a single property asset junior company, given that it no longer held any interests in the Crucero and Josnitoro projects and the Invicta Project was materially its only asset.²³⁰ Claimant's value was therefore intrinsically linked to the value of the Invicta Project; as such, the use of a market approach based on Claimant's market capitalisation as an indicator of the value of the Invicta Project is supported by CIMVAL guidance.

²²⁷ AlixPartners Report, paragraph 176 ↩

²²⁸ AlixPartners Report, paragraphs 176-186 ↩

²²⁹ Following the lapse in Claimant's option in the Josnitoro Project (First Accuracy Report, paragraph 3.17) ↩

²³⁰ First Accuracy Report, paragraph 3.13 ↩

[Page 63]

7.9 This said, the market capitalisation approach was used as an illustrative indicator only; we acknowledged that market capitalisation reflects investors' perception of Claimant's own capacity to develop the Invicta Project, rather than that of a hypothetical investor, and therefore does not strictly follow the FMV standard.²³¹

7.10 We also set out several reasons why our figure may, in fact, be conservative. Inter alia, (i) the Invicta Project would likely have imminently entered into full commercial production in the But-For Situation, and Claimant's share price may have subsequently outperformed the market; and (ii) investors may not have access to all information regarding the advanced stage of development of the Invicta Project, nor Claimant's future plans regarding the prospective acquisition of the Mallay Plant, both of which could have increased Claimant's share price if known by the wider market.²³²

Movements in the GDXJ are a reasonable proxy for movements in Claimant's market capitalisation in the But-For Situation

7.11 Notwithstanding the above, we disagree with AlixPartners' suggestion that our use of movements in the GDXJ as a proxy for movements in Claimant's market capitalisation beyond the Blockade Date in the But-For Situation is unreliable.²³³

7.12 As support for their position, AlixPartners rely on a regression analysis of Claimant's share price on the GDXJ index.²³⁴ Regression analysis is a statistical technique used to estimate the relationship between variables; which can be described by statistical measures including (i) the beta coefficient; and (ii) R-squared:

1. The beta coefficient indicates how sensitive the dependent variable (in this instance, Claimant's share price) is to the independent variable (the GDXJ index);²³⁵ the higher the beta, the more sensitive it is. For example, if the beta coefficient is 1.2, the dependent variable's change will be 120% of that of the independent variable; and
2. The R-squared indicates the proportion of the change in the dependent variable that can be attributed to the independent variable.²³⁶ The higher the R-squared, the more that the variance in the independent variable explains the variance in the dependent variable.

²³¹ First Accuracy Report, paragraph 8.23c) ↩

²³² First Accuracy Report, paragraphs 8.23d) & 8.23e) ↩

²³³ AlixPartners report, paragraph 179 ↩

²³⁴ We note that, on 11 August 2022, Claimant consolidated all of its issued and outstanding common shares on the basis of 1 post-consolidated share for every 10 pre-consolidated shares. Claimant's stock price is therefore 10 times greater following the share consolidation compared to pre-consolidation, whilst its share volume is 10 times lower. Whilst any references to Claimant's daily share price and/or share volume in the First Accuracy Report and AlixPartners Report are pre-consolidation and therefore technically outdated, the net impact on market capitalisation is nil; this therefore has no impact on our analysis. (AC-63 Lupaka Gold Corp. Lupaka Gold to Consolidate its Common Shares dated 11 August 2022). ↩

²³⁵ AC-64 Chatterjee & Hadi, Regression Analysis by Example dated 2012, page 67 ↩

²³⁶ AC-64 Chatterjee & Hadi, Regression Analysis by Example dated 2012, page 46 ↩

[Page 64]

7.13 AlixPartners argue that “selecting a single day and applying the index percentage change is incorrect as the variance between [Claimant's] stock price and the GDXJ ranges from -26.6% to 39.1%".²³⁷ Furthermore, based on historical daily trading data, AlixPartners calculate a beta coefficient of 0.3406, and an R-squared of 1.8%, which they imply suggests a weak relationship between Claimant's share price and the GDXJ.²³⁸

7.14 Clearly, an individual stock will not exactly mirror a stock index, and one would expect variances on a daily basis. This is especially the case for a stock with a relatively low trading volume such as Claimant's;²³⁹ indeed, studies have historically shown that beta estimates are biased downwards for stocks which are traded infrequently, and that the bias is greater for high-frequency data.²⁴⁰ Given that the purpose of our analysis is to estimate expected returns over more than a ten-month period,²⁴¹ not daily returns, we consider it inappropriate to calculate the beta coefficient and R-squared based on daily trading data (as AlixPartners have done), and instead we recalculate these two coefficients using monthly data.

7.15 Running a regression analysis on monthly, rather than daily, trading data over a five-year period gives a beta coefficient of 0.88,²⁴² and an R-squared of 18%,²⁴³ which suggests a stronger relationship between Claimant's returns and the GDXJ than that given by AlixPartners based on daily data.

7.16 As such, whilst the relationship is (as one would expect) not perfect, we nonetheless consider movements in the GDXJ (of which Claimant was a constituent) over the long-term to be a reasonable proxy for movements in Claimant's market capitalisation in the But-For Situation.

7.17 Finally, AlixPartners note that Claimant's market capitalisation underperformed the GDXJ between 28 March 2018 (when Claimant's option for the Josnitoro Project was terminated) and the Blockade Date.²⁴⁴ On an illustrative basis, adjusting our assessment of the value of the Invicta Project by Claimant's underperformance of c.17% during this period would decrease the value from USD 33.4m to USD 27.8m.²⁴⁵

²³⁷ AlixPartners report, paragraph 179 ↩

²³⁸ AlixPartners report, paragraph 184 ↩

²³⁹ AlixPartners Report, paragraph 186; AP-0032-ENG ↩

²⁴⁰ AC-65 Armitage and Brzeszczynski, Heteroscedasticity and interval effects in estimating beta dated 18 July 2011, page 5. “It is known from previous research that thin or infrequent trading in shares causes severe downward bias in OLS beta estimates, and that the bias is worse in higher-frequency data. While we confirm the impact of thin trading, we find in addition that the difference between betas estimated from daily and monthly data is strongly related to daily volatility” ↩

²⁴¹ I.e. between the Blockade Date and the Valuation Date ↩

²⁴² Appendix 5, sheet “Monthly beta”, cell N10 ↩

²⁴³ Appendix 5, sheet “Monthly beta”, cell T10 ↩

²⁴⁴ AlixPartners Report, paragraph 183 ↩

²⁴⁵ Appendix 5, sheet “Share price v GDXJ”, row 151. To calculate the performance of Claimant's share price relative to the GDXJ between 28 March 2018 and the Blockade Date, we index movements in both indicators to a base of 100 at 28 March 2018. Based on this index, Claimant's share price at the Blockade Date was 74.5, whilst the GDXJ was 89.6. Therefore, Claimant underperformed the GDXJ by c.17% (1- (74.5/89.6)) = 16.9%. ↩

[Page 65]

The application of a control premium to minority shareholdings is a well-established concept

7.18 The application of a control premium to the value of a minority shareholding when valuing a controlling stake in a company or project is a well-established concept. Whilst there is some debate amongst academics as to how such a premium should be quantified, in practice, a common approach is to apply a premium derived from studies which compare observed prices paid for a controlling share in publicly traded assets to the price prior to the transaction.²⁴⁶

7.19 We adopted a similar approach in the First Accuracy Report, ultimately taking an average premium of 43.2%, which was below the average premium evidenced in the mining industry in the decade to 2019.²⁴⁷

7.20 Clearly, the exact premium that is paid for control may differ from one transaction to another, and may be influenced by a variety of factors. However, for the purposes of our illustrative assessment of the value of Claimant's Investment using a market capitalisation approach, we consider that our approach is reasonable to the extent that it is consistent with both (i) common valuation practice; and (ii) observable market evidence.

AlixPartners' liquidity analysis is of limited relevance

7.21 AlixPartners point to Claimant's low trading volumes in August 2019 as evidence that Claimant's market capitalisation may not be relevant to the value of the Invicta Project.²⁴⁸

7.22 Claimant's market capitalisation in August 2019 is indeed of limited relevance to the value of the Invicta Project, not because of low trading volumes but rather since by this time Claimant had not had access to the mine for 10 months and PLI had already terminated the PLI Loan agreement.

7.23 Our benchmark is based on Claimant's market capitalisation on 25 October 2018, immediately prior to its announcement of the Blockade, indexed by the movement in the GDXJ between this date and the Valuation Date.²⁴⁹

7.24 To mitigate the risk that the relatively low trading volume of Claimant's shares would distort the share price on any given day, we re-perform our market capitalisation analysis by instead using Claimant's 30-day²⁵⁰ trailing average market capitalisation at 25 October 2018 as the starting point for our calculation. This results in an illustrative indicator of the value of Claimant's Investment at the Valuation Date of USD 32.8m, after applying a control premium (i.e. a decrease of 0.6m compared to our calculation in the First Accuracy Report).²⁵¹

²⁴⁶ AC-67 Global Arbitration Review, M&A and Shareholder Arbitrations dated 1 February 2021, page 7 ↩

²⁴⁷ First Accuracy Report, paragraphs 8.16-8.17 ↩

²⁴⁸ AlixPartners Report, paragraph 186 ↩

²⁴⁹ First Accuracy Report, paragraphs 8.5-8.6 & 8.14 ↩

²⁵⁰ Based on the previous 30 trading days. ↩

²⁵¹ Appendix 5, sheet “SAR - Market cap”, cell D4 ↩

[Page 66]

Summary of outcomes of our market capitalisation approach

7.25 In the First Accuracy Report, we obtained an illustrative valuation of the Invicta Project of USD 33.4m under a market capitalisation approach, after (i) applying the percentage movement in the GDXJ between 25 October 2018 and the Valuation Date to Claimant's market capitalisation on 25 October 2018; and (ii) applying a control premium.²⁵²

7.26 In this report, we calculate two alternative illustrative valuations using a market capitalisation approach:²⁵³

1. Using Claimant's 30-day trailing average market capitalisation at 25 October 2018 as the starting point for our calculation, a valuation of USD 32.8m; or, alternatively
2. Adjusting our valuation in the First Accuracy Report to reflect the 17% underperformance of Claimant's share price relative to the GDXJ between 28 March 2018 and the Blockade Date, a valuation of USD 27.8m.

7.27 Our market valuation approach therefore yields an illustrative valuation of the Invicta Project between USD 27.8m and USD 33.4m.

7.28 This range is in line with our estimate of the FMV of the Invicta Project in the 355t/day Scenario, and below that in the 590t/day Scenario. This is consistent with our explanation in the First Accuracy Report that our market capitalisation approach may be conservative in that, inter alia, Claimant's share price would not have reflected the additional value of the Invicta Project associated with the Mallay Acquisition Plan, which had not been publicly announced.

B. Claimant's sunk costs

First Accuracy Report

7.29 In the First Accuracy Report, we benchmarked our damages assessments against Claimant's sunk costs as follows:²⁵⁴

1. We identified costs incurred by Claimant in relation to its investment in the Invicta Project between the Acquisition Date and the Valuation Date, which were subsequently lost as a result of Respondent's Alleged Breaches; and
2. We applied interest from the date upon which sunk costs were incurred to the Valuation Date as a proxy for the expected minimum level of return upon those investments.

²⁵² First Accuracy Report, paragraph 8.22. See also Appendix 5, sheet “FAR - Market cap” ↩

²⁵³ Appendix 5, sheet “SAR - Market cap”; Appendix 5, sheet “Share price v GDXJ” ↩

²⁵⁴ First Accuracy Report, paragraph 8.24 ↩

[Page 67]

7.30 Based on the above, we assessed the value of Claimant's sunk costs at the Valuation Date to be USD 39.5m.²⁵⁵ After deducting the Early Termination Amount of USD 15.9m claimed by PLI in July 2019, our benchmark amounted to USD 23.6m.²⁵⁶

The AlixPartners Report

7.31 AlixPartners highlight the following alleged deficiencies in our calculation of sunk costs:²⁵⁷

1. According to International Financial Reporting Standard (IFRS) 6, only expenditures recognised as exploration and evaluation assets, to the extent that they can be associated with finding specific mineral sources, should be measured at cost on the financial statements. In contrast, expenditures related to the development of mineral resources should not be recognised as exploration and evaluation assets. Therefore, given that these costs were reclassified as “mineral property under development” by Claimant from August 2017 onwards, we should not have “categorized the entirety as sunk costs as presented on the financial statements, without consideration for any adjustments". Furthermore, we did not account for the foreign exchange adjustments included in Claimant's financial statements;
2. Some sunk costs “may no longer be contributing to the overall value of the Project and may not represent what a buyer would pay for the assets as of the Valuation Date". For example, expenditures that “date from more than five years prior the valuation date are usually excluded under the cost approach”;
3. The interest rate we used to bring forward historical spending to the Valuation Date, based on an estimate of the annual effective interest rate on the PLI Loan,²⁵⁸ “does not represent growth in the value of the Project's historical costs going forward” as it could be “abnormally” high due to the absence of a feasibility study for the Invicta Project. Furthermore, the rate is allegedly inconsistent with the cost of debt used in our estimate of WACC for a hypothetical gold mining entity operating in Peru. AlixPartners believe that it would be more appropriate to recognise Claimant's historical costs “incurred at their actual dollar value, rather than with a grossed-up rate of return that may not represent the actual return on these expenditures” or, as an alternative to bringing forward these expenses to the Valuation Date, to apply interest at the same rate as pre-award interest;
4. CIMVAL does not recommend the cost approach to value a Development Property; and
5. Sunk costs “only represent the amount of money that Claimant had spent in relation to the acquisition and exploration of the mine” and do not “account for the uncertainty and risk arising from the social license and execution risks” which would be considered by a rational investor looking to acquire the Invicta Project.

²⁵⁵ First Accuracy Report, paragraph 8.33 ↩

²⁵⁶ First Accuracy Report, paragraph 8.34 ↩

²⁵⁷ AlixPartners Report, paragraphs 187-193 ↩

²⁵⁸ First Accuracy Report, paragraph 8.31 ↩

[Page 68]

Our comments

Sunk costs are not an appropriate primary approach for damages in this case, but remain a relevant benchmark

7.32 As noted in the First Accuracy Report, a cost-based approach excludes future income generation and growth and does not capture the value of intangible assets. When applied to a business or asset which is a going concern, such an approach is therefore typically used to derive a floor for a valuation.²⁵⁹

7.33 Notwithstanding any discussion as to the appropriate classification of the Invicta Project,²⁶⁰ we agree with AlixPartners that sunk costs is not an appropriate primary approach for the assessment of damages in this case; however, we consider it provides a relevant benchmark against the low end of our damages scenarios.²⁶¹

AlixPartners' argument that some costs should be excluded from our assessment is unsubstantiated

7.34 CIMVAL sets out that the cost approach is “based on the principle of contribution to value [...] where exploration expenditures are analysed for their contribution to the exploration potential of the Mineral Property".²⁶²

7.35 As explained in the First Accuracy Report, exploration expenditures in relation to the Invicta Project were expensed via the income statement until receipt of Tranche 1 of the PLI Loan in July 2017, and capitalised thereafter.²⁶³

7.36 There is therefore a difference between our assessment of sunk costs at 31 December 2018 and the value of the “Mineral property under development” asset on Claimant's balance sheet at the same date.²⁶⁴ The former includes exploration expenditures expensed prior to July 2017, which would not have been capitalised on the balance sheet. It is also unclear what the relevance of their comment on foreign exchange variances is, given that we convert CAD amounts to USD at the prevailing spot rate for the purposes of our sunk costs assessment.²⁶⁵

²⁵⁹ First Accuracy Report, paragraph 9.8 ↩

²⁶⁰ In the First Accuracy Report, we considered that the Invicta Project “was at a more advanced stage than a typical Mineral Resource Property” (First Accuracy Report, paragraph 4.31) however we did not seek to categorise the project further. CIMVAL guidance itself acknowledges that boundaries between categories of mineral properties may not be clear-cut, and that “it may be difficult to classify some Mineral Properties so they fit in only one specific category” (AC-22, CIMVAL Code for the Valuation of Mineral Properties dated 29 November 2019, Section 3.3.3.). ↩

²⁶¹ Logically, our damages benchmark based on sunk costs incurred of USD 23.6m was being the damages value generated by the income approach under both the 355t/day and 590t/day Scenarios. ↩

²⁶² AC-22, CIMVAL Code for the Valuation of Mineral Properties dated 29 November 2019, page 17 ↩

²⁶³ First Accuracy Report, paragraph 8.27 ↩

²⁶⁴ AlixPartners Report, paragraph 189 ↩

²⁶⁵ First Accuracy Report, Appendix 8, sheet “Sunk Costs” ↩

[Page 69]

7.37 Accounting convention aside, AlixPartners do not seek to justify why they believe that those exploration expenditures that were expensed prior to July 2017 (or any other expenses incurred since) would not have contributed to the value of the Invicta Project, or what adjustment they would propose to our calculation.

7.38 Whilst it is outside of our expertise to assess which historical sunk costs would have contributed to the value of the Invicta Project at the Valuation Date, we note that our indicator of value based on Claimant's sunk costs implies a damages figure at the low end of the range of our damages assessments and other indicators of value.²⁶⁶

The effective rate on the PLI Loan is a Claimant-specific rate, and an appropriate proxy for the minimum expected return on the Invicta Project

7.39 With respect to our application of interest to bring Claimant's sunk costs forward to the Valuation Date, we apply a Claimant-specific rate (i.e. the effective rate of interest on the PLI Loan) to the investments made by Claimant, on the basis that it would have wished to (at the very least) service its debts.

7.40 We therefore consider the annual effective interest rate on the PLI Loan to be an appropriate proxy for the expected minimum return on Claimant's investments in the Invicta Project.

7.41 By contrast, AlixPartners' comparison of this rate to the 6.2% cost of debt used for our WACC calculation²⁶⁷ is not relevant in this context, given that the latter relates to the average cost of debt for companies in the precious metals industry, not Claimant. Similarly, AlixPartners' alternatives based on their proposed pre-award interest rate (SOFR +2% or UST +2%) are risk-free rates plus a spread, and do not reflect a Claimant-specific rate.

Our updated assessment of Claimant's sunk costs at the Valuation Date

7.42 As noted in Section 5, our calculation of the effective interest rate on the PLI Loan in the First Accuracy Report was based on a simplified calculation which understated the effective interest rate.

7.43 We now explicitly model financing cash flows in our updated damages assessment, and estimate the actual (nominal) effective interest on Tranches 1-3 of the PLI Loan to be 17.0%²⁶⁸ on a pre-tax basis, or 12.0%²⁶⁹ on a post-tax basis.

7.44 On an illustrative basis, updating our assessment of Claimant's sunk costs by applying the post-tax effective interest rate of 12.0% to sunk costs incurred in each year (as a proxy for a rate of return) would result in a value of sunk costs of USD 43.1m at the Valuation Date.²⁷⁰

²⁶⁶ First Accuracy Report, paragraph 9.2 ↩

²⁶⁷ AlixPartners Report, paragraph 191 ↩

²⁶⁸ Appendix 4, sheet “PLI Loan (ACC)”, cell C28 ↩

²⁶⁹ Appendix 4, sheet “PLI Loan (ACC)”, cell D47 ↩

²⁷⁰ Appendix 5, sheet “Claimant's sunk costs” ↩

[Page 70]

7.45 Based on nominal sunk costs of USD 24.8m,²⁷¹ our assessment of USD 43.1m at the Valuation Date, after applying an illustrative rate of return, implies a cost-to-value multiple of 1.74x.²⁷² This appears conservative in comparison to mining industry guidance, which recommends a typical “Prospectivity Enhancement Multiple" of 3x²⁷³ for exploration properties where indicated resources have been defined but a pre-feasibility study has not recently been completed.²⁷⁴

7.46 Consistent with our approach in the First Accuracy Report, deducting the Early Termination Amount of USD 15.9m claimed by PLI in July 2019 therefore provides a benchmark for the value of the Invicta Project of USD 27.2m.

7.47 This implied valuation is below our damages assessment in both the 590t/day and 355t/day Scenarios. This outcome appears reasonable, given that a rational investor would wish to invest in a project with a return greater than (rather than equal to) their cost of borrowing.

C. Transaction multiples

First Accuracy Report

7.48 In the First Accuracy Report, we also benchmarked the multiple implied by our assessment of the FMV of the Invicta Project to other recent gold industry transactions.

7.49 Using Capital IQ, we identified transactions with a value over USD 1m for a majority stake (>50%) in entities operating in the gold industry which closed within the five-year period prior to the Valuation Date. Given the lack of data to perform a benchmarking exercise against transaction multiples for projects with only reported resources, we retained transactions with reported proven and probable gold reserves in either the year of the announcement date or, alternatively, the preceding year.²⁷⁵

7.50 We obtained median and mean multiples of USD 229.5/oz. and USD 357.1/oz. across our sample of 26 transactions which satisfied the above criteria.²⁷⁶

²⁷¹ Appendix 5, sheet “Claimant's sunk costs”, cell C15 ↩

²⁷² USD 43.1m ÷ USD 24.8m = 1.74 ↩

²⁷³ For illustrative purposes, applying a multiple of 3x to nominal sunk costs of USD 24.8m would imply a valuation of USD 58.5m, after deducting the Early Termination Amount claimed by PLI in July 2019 (i.e. (USD 24.8m x 3) – USD 15.9m = USD 58.5m) ↩

²⁷⁴ AC-68 Australian Institute of Minerals Valuers & Appraisers (AIMVA), Valuation and Appraisal of Mineral Projects dated 28 August 2014 ↩

²⁷⁵ First Accuracy Report, paragraph 8.40 ↩

²⁷⁶ First Accuracy Report, paragraphs 8.41 and 8.43 ↩

[Page 71]

7.51 We then benchmarked the multiples provided by our sample to our valuation of the Invicta Project. Given that over half of the transactions in our sample related to entities/projects that had already commenced production at the transaction date, in order to ensure comparability, we removed the pre-production premium from the discount rate used in our DCF valuation. This gave an illustrative post-tax NPV of USD 48.1m under the 355t/day Scenario and USD 82.3m under the 590t/day Scenario.²⁷⁷

7.52 Our adjusted DCF valuations produced implied multiples of USD 260.1/oz. and USD 270.4/oz. under the 355t/day and 590t/day Scenarios respectively, which lie between the median and mean multiples for our sample of gold industry transactions.²⁷⁸

7.53 Furthermore, our analysis of gold industry transactions demonstrated a positive correlation between gold grade and the gold reserves multiple of a given project. Almost all of the transactions in our sample were for properties with a significantly lower grade than the Invicta Project; as such, benchmarking our implied multiple against the mean or median of the entire dataset was conservative.²⁷⁹

The AlixPartners Report

7.54 AlixPartners opine that our analysis of transaction multiples suffers from the following shortcomings:²⁸⁰

1. We allegedly did not analyse the microeconomic factors, such as country-specific risks and development stages, which drive the value of the comparable companies and allegedly failed to consider whether these factors are similar to those driving the value of Claimant. After excluding (i) 18 transactions for companies which had at least one producing mine; and (ii) Romarco Minerals Inc., which has one developing mine in the “low-risk jurisdiction” of South Carolina from the sample, AlixPartners propose significantly lower mean and median transaction multiples of USD 89.46/oz. and USD 66.00/oz. respectively;
2. Our transaction multiples range from USD 7.59/oz. to USD 1,187.94/oz. with a “large” standard deviation of USD 352.42/oz.; and
3. Latin America faces specific risks that other geographies may not face, including the frequency with which mining projects in that region are suspended due to community opposition compared to any other factor.

²⁷⁷ First Accuracy Report, paragraph 8.51 ↩

²⁷⁸ First Accuracy Report, paragraph 8.53 ↩

²⁷⁹ First Accuracy Report, paragraphs 8.44-8.47 ↩

²⁸⁰ AlixPartners Report, paragraphs 194-200 ↩

[Page 72]

Our comments

Updates to our gold industry transaction multiples calculation

7.55 In the First Accuracy Report, our multiples calculation was based upon the implied enterprise value of the target company (i.e. the total value of the company, including both equity and debt) in recent gold industry transactions.²⁸¹ Consistent with this approach, we compared the USD/oz. multiple implied by our sample of transactions to the multiple implied by our calculation of the FMV of the Invicta Project before adjusting for the value of debts that would have been settled with PLI in the But-For Situation.

7.56 For the purposes of our updated damages assessment, we now directly model financing cash flows relating to the PLI Loan in order to calculate FCFE, resulting in an FMV which is equivalent to damages at the Valuation Date (with no further adjustment for debts to PLI required). For consistency, we amend our multiples analysis to benchmark the implied multiple of our updated FMV calculation to the equity value (i.e. the value of cash flows available to the shareholders of a company) of our retained sample of gold industry transactions.

7.57 We now obtain median and mean multiples, based on implied equity value, of USD 219.1/oz. and USD 368.0/oz. respectively (compared to USD 229.5/oz. and USD 357.1/oz in the First Accuracy Report, based on enterprise value) as shown in the table below.²⁸²

²⁸¹ First Accuracy Report, paragraphs 4.17 & 8.35 ↩

²⁸² For the transaction involving Atlantic Gold Corporation, announced on 14 May 2019, no implied equity value was provided on Capital IQ. Therefore, for simplicity, we assume that the implied equity value would be the same as the transaction value of USD 595m. ↩

[Page 73]

Table 7.1. Transaction multiples based on implied equity value

Source: Appendix 5, sheet “Table 7.1”

Our response to AlixPartners' comments on our transaction multiples approach

7.58 In order to calculate a revised transaction multiple, AlixPartners exclude all transactions with a producing mine and retain a small sample of only eight transactions involving companies with exploration or development projects only. These transactions have, on average, lower multiples.

7.59 Given the heterogeneity of the mining industry, it is unrealistic to expect to find fully comparable transactions. AlixPartners allege that we do not analyse the microeconomic factors driving the value of the transactions in our sample, but do not seek to do so themselves when providing a revised calculation.

7.60 We note that the Invicta Mine was, in fact, in production at the Blockade Date, albeit not full commercial production. It was also historically subject to a feasibility study and optimised feasibility study carried out in 2009 and 2010 respectively by the Lokhorst Group,²⁸³ which incorporated an estimate of mineral reserves; these studies were subsequently superseded by later reports, including the PEA. At the Blockade Date, Claimant was on the verge of commencing full commercial production at the Invicta Mine and, in our view, it is reasonable to consider that the Invicta Project was at a more advanced stage than a typical property with exploration or development projects.

²⁸³ Claimant's Memorial, paragraph 32; AC-02 SRK Consulting PEA dated 13 April 2018, pages 24-25 & 28-29 ↩

[Page 74]

7.61 With respect to the eight transactions retained by AlixPartners:

1. As demonstrated in the First Accuracy Report,²⁸⁴ our sample shows a positive correlation between gold grade and the gold reserve multiple of a project. However, the average gold-equivalent reserve grades of the eight transactions range from 0.87g/t to 1.74g/t, significantly lower than the grade of the Invicta Project.
2. Whilst we do not have extensive information regarding the projects in question, some of these appear to be at a significantly earlier stage of development than the Invicta Project. For example:
   1. Two of the eight transactions retained by AlixPartners relate to the purchase of shares in Rio Novo Gold Inc. in July 2016 and December 2017. The latest financial statements of this company at each date note that “the Company is currently in the process of exploring the exploration property interests under government granted permits and has not yet determined whether these interests contain reserves that are economically recoverable”.²⁸⁵
   2. Similarly, the June 2017 financial statements of Bison Gold Resources Inc. state that “the properties in which Bison currently has an interest are in the exploration stage".²⁸⁶
3. Were we to exclude the above three transactions, the remaining five non-producing transactions involving non-producing mines have median and mean transaction multiples of USD 212.8/oz. and USD 190.4/oz. respectively,²⁸⁷ which is much closer to the implied multiple suggested by our updated FMV assessment (see above), especially when rationalised against (i) the much lower grade of those properties compared to the Invicta Project; and (ii) the advanced stage of development of the Invicta Project compared to an average project in the exploration or development stage.²⁸⁸

7.62 With respect to AlixPartners' comments on geographical risk, we note that:

1. “Latin America” is a broad geographical area; it is not reasonable to assume that the exact same level of risk is encountered across the entirety of this region (or that similar risks to those in Peru would not be encountered in other developing nations outside of South America); and

²⁸⁴ First Accuracy Report, Figure 8.3; paragraph 8.45 ↩

²⁸⁵ AP-0057-ENG & AP-0060-ENG ↩

²⁸⁶ AP-0058-ENG ↩

²⁸⁷ Appendix 5, sheet “Table 7.1”, rows 36 & 37 ↩

²⁸⁸ First Accuracy Report, paragraph 8.43 ↩

[Page 75]

2. Whilst there are not enough data points to make a meaningful comparison, we identify one transaction in our sample which relates to a company whose operations appear to be wholly based in Peru (Rio Alto Mining Limited; February 2015), which provides a transaction multiple of USD 451.7/oz., significantly above that implied by our damages assessment.²⁸⁹

Updated implied multiple of the Invicta Project

7.63 As we did in the First Accuracy Report, we compare the multiple of recent gold industry transactions to the multiple of the Invicta Project implied by our updated damages assessment. Consistent with our previous approach, for comparability, we remove the pre-production premium from the discount rate used in our DCF valuation, which results in an updated, illustrative discount rate of 8.9% for the purposes of calculating the implied multiple of the Invicta Project.²⁹⁰

7.64 We set out below the implied equity value multiples for both the 355t/day and 590t/day Scenarios by dividing our updated FMV calculations at the Valuation Date (as set out in Section 6), by total gold-equivalent produced ounces in each production scenario.

Table 7.2. Implied equity value multiples (USD/oz.) as per our updated FMV of the Invicta Project²⁹¹

Source: Appendix 5, sheet “Table 7.2”

7.65 We calculate implied equity value multiples of USD 202.3/oz. for the Invicta Project under the 355t/day Scenario and USD 146.7/oz. under the 590t/day Scenario, both of which are:

1. Lower than the median and mean equity value multiples of USD 219.1/oz. and USD 368.0/oz. respectively across our full sample of gold industry transactions;

²⁸⁹ Appendix 5, sheet “Table 7.1”, row 31 ↩

²⁹⁰ 8.9% being the real cost of equity shown in Table 6.4, before adding a pre-production premium, i.e. 12.2% - 3.3% = 8.9%. This step is a purely illustrative step in order to compare our valuation on a like-for-like basis to industry transactions. ↩

²⁹¹ In the First Accuracy Report, our calculation of Invicta Project gold-equivalent resources in the 355t/day Scenario was based on an average gold-equivalent grade of 8.58g/t as calculated in the PEA and SRK Model, and in the 590t/day Scenario on our adjusted gold-equivalent grade of 6.93g/t. For consistency, the calculation of gold-equivalent resources should have reflected our updates to metal prices in the First Accuracy Report, which resulted in an average gold-equivalent grade of 7.81 in our 355t/day Scenario (First Accuracy Report, Appendix 5, sheet “fin_lupaka”, row 8) and 6.40 in our 590t/day Scenario (First Accuracy Report, Appendix 6, sheet “fin_lupaka”, row 8). Correcting these calculations would have resulted in implied multiples of USD 285.82/oz. and USD 292.96 under the 355t/day and 590t/day Scenarios respectively. Given that these multiples still lie between the median and mean multiples obtained from our sample of transactions used in the First Accuracy Report, our conclusions do not change. Our updated calculation above accurately reflects the impact of subsequent updates made to our damages assumptions on gold-equivalent grade and resources. ↩

[Page 76]

2. Broadly consistent with the median and mean transaction multiples of USD 212.8/oz. and USD 190.4/oz. respectively for the five non-producing transactions referenced above (after excluding those which appear to be at a particularly early stage of development); and
3. Significantly lower than the multiple of USD 451.7/oz. for the only transaction in our sample which relates to a company whose operations appear to be wholly based in Peru.

7.66 In our view, whilst we agree that the multiples approach is subject to a number of limitations in this case, comparison to other gold industry transactions supports the reasonableness of our damages assessment under an income approach.

D. Other contemporaneous valuation evidence

First Accuracy Report

7.67 Finally, we benchmarked our assessment of the FMV of the Invicta Project to other contemporaneous valuation exercises and found that it was (i) in line with the post-tax NPV computed by the SRK Model under the PEA Mine Plan (USD 43.4m, compared to our calculation of USD 44.2m under the 355t/day Scenario); and (ii) significantly lower than the post-tax NPV computed by the Red Cloud Model under the Mallay Acquisition Plan (USD 86.3m, compared to our calculation of USD 63.6m under the 590t/day Scenario).²⁹²

The AlixPartners Report

7.68 AlixPartners believe that the NPVs computed by the SRK Model and Red Cloud Model are “too speculative” and not indicative of damages for the following reasons:²⁹³

1. The SRK Model and Red Cloud Model were based on the PEA, which itself acknowledged that it is subject to higher risk than a pre-feasibility or feasibility study. Furthermore, there was no certainty that the PEA Mine Plan would be realised;
2. The SRK Model and Red Cloud Model did not consider the risk of not reaching agreements with the local population, as may have been required under the relevant legislation, or the financial risks presented by the PLI Loan;
3. The SRK Model, which assumed a third-party processing plant would be employed, did not incorporate the fact that all the third-party processing plants identified by Claimant had failed the tests conducted in the pre-production phase; and
4. The Red Cloud Model, as we acknowledged in the First Accuracy Report, was not as reliable as the SRK Model as some inputs were not subject to the same level of technical and financial analysis as the PEA.

²⁹² First Accuracy Report, paragraph 8.55 ↩

²⁹³ AlixPartners Report, paragraphs 201-206 ↩

[Page 77]

Our comments

7.69 The SRK Model was prepared for the purposes of the PEA, an independent technical report issued by SRK in 2018, and therefore represents the best available contemporaneous source of data to use as a starting point for our damages calculation. AlixPartners do not propose an alternative source of data which they consider to be more appropriate for the purposes of a damages assessment.

7.70 Micon explain that the geology of the Invicta Mine meant it may not have been cost-effective to drill from the surface at sufficiently close intervals to demonstrate continuity in the mineralisation that would warrant a pre-feasibility study. Instead, Claimant gained access to mineralisation and demonstrated continuity of mineralised structures through exploratory underground development, a process which “greatly reduces a key geological risk”.²⁹⁴ Furthermore, in our experience, one of the primary reasons to obtain a feasibility study is for a mine operator to obtain financing. Following issue of the PEA (which indicated that significant further capital investment was not required), Claimant did not see a need to commission a further feasibility study given that it had both (i) obtained third-party financing via the PLI Loan; and (ii) materially completed all development works by the Blockade Date.²⁹⁵

7.71 In the First Accuracy Report, we specifically accounted for the additional risk associated with a project at an earlier stage of development than an operating mine by applying a pre-production premium to our discount rate.²⁹⁶

7.72 As AlixPartners note, the Red Cloud Model was not subject to the same technical and financial analysis as the PEA; we reflected the additional risk associated with the 590t/day Scenario by applying a higher premium to our discount rate than in the 355t/day Scenario. In our updated damages assessment in this report, we align the discount rate in each production scenario to reflect the higher level of reliability (and lower risk) attached to the 590t/day Scenario, following the additional analysis performed by Micon.

7.73 Points (b) and (c) raised by AlixPartners above are primarily factual issues, which are outside of the scope of our expertise. To the extent possible, we address AlixPartners' comments on financing risks in the previous sections of this report. We note that, even ignoring the financing obtained through the PLI Loan, Claimant had a track record of raising funds in relation to the Invicta Project through a mixture of (i) private placements, (ii) bridge loans; and (iii) conversions of shares to debt, and expected that its ability to raise funds would have further improved after commencing full commercial production.²⁹⁷

²⁹⁴ Micon Report, paragraph 38 ↩

²⁹⁵ Witness Statement of Julio Félix Castañeda Mondragón dated 1 October 2021, paragraph 26 ↩

²⁹⁶ First Accuracy Report, Table A4.1 ↩

²⁹⁷ Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraph 101 and Annex ↩

[Page 78]

8 Experts' declaration

8.1 We understand that our duty in giving evidence in this arbitration is to assist the arbitral tribunal to decide the issues in respect of which expert evidence is adduced. We have complied with, and will continue to comply with, that duty.

8.2 We confirm that this is our own, impartial, objective, unbiased opinion which has not been influenced by the pressures of the dispute resolution process or by any party to the arbitration.

8.3 We confirm that all matters upon which we have expressed an opinion are within our area of expertise. We have indicated in this report where we have relied upon the opinions of experts in other fields.

8.4 We confirm that we have referred to all matters which we regard as relevant to the opinions we have expressed and have drawn to the attention of the arbitral tribunal all matters, of which we are aware, which might adversely affect our opinion.

8.5 We confirm that, at the time of providing this written opinion, we consider it to be complete and accurate and that it constitutes our true, professional opinion.

8.6 We confirm the attribution of the entirety of the report to each author.

We confirm that if, subsequently, we consider this opinion requires any correction, modification or qualification we will notify the parties to this arbitration and the arbitral tribunal forthwith.

[Page 79]

Appendix 1 Sources of information used in this report

Table A1.1. Documents provided by Counsel

Table A1.2. Accuracy exhibits prepared for the First Accuracy Report

Table A1.3. Additional documents we have researched or prepared

[Page 80]

[Page 81]

Appendix 2 Modelling of financing in the 355t/day and 590t/day Scenarios

A2.1 As explained in Section 5, for the purposes of our updated damages assessment, we directly model financing cash flows in relation to the PLI Loan in order to calculate FCFE in each year of our model.

A2.2 Our estimates of financing cash flows comprise two main elements in each production scenario:

1. Principal and interest; and
2. Upside participation.

A2.3 We summarise our calculation in each production scenario in further detail below. Our detailed calculations can be found in Appendix 3 and Appendix 4.²⁹⁸

A. Modelling of the PLI Loan in the 355t/day Scenario

Principal and interest

A2.4 As set out in the First Accuracy Report, Claimant entered into the PLI Loan in order to fund the continued development of the Invicta Project through pre-paid gold forward purchase agreements.²⁹⁹

A2.5 Under this agreement, Claimant received total gross proceeds of USD 7.0m in three tranches; Tranche 1 received in August 2017, Tranche 2 received in November 2017 and Tranche 3 received in February 2018.³⁰⁰ For each tranche, following a grace period of 15 months, Claimant was to make gold repayments over a further 45 months, equivalent to a discount of USD 500/oz. on a total quantity of 22,680 oz. of gold.³⁰¹

A2.6 Accordingly, repayment of Tranche 1 was due from December 2018, Tranche 2 from March 2019 and Tranche 3 from June 2019.³⁰²

A2.7 In line with the terms of the PLI Loan, we model loan repayments as follows:³⁰³

1. Tranche 1 – 187 oz./month x 45 months = 8,415 oz. x USD 500/oz. = USD 4.2m (monthly payments starting December 2018);

²⁹⁸ Tabs "PLI Loan (ACC)" and "Upside Participation (ACC) in Appendix 3 and Appendix 4 for the 590t/day and 355t/day Scenarios respectively. ↩

²⁹⁹ First Accuracy Report, paragraph 3.18 ↩

³⁰⁰ First Accuracy Report, paragraph 3.19 ↩

³⁰¹ First Accuracy Report, paragraphs 3.19-3.21 ↩

³⁰² First Accuracy Report, paragraph 3.21 ↩

³⁰³ AC-04 Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017. The repayment schedule is defined under "Contract Quantity" on pages 3-4 and USD 500/oz. is defined under "Gold Price Discount" on page 6 ↩

[Page 82]

2. Tranche 2 – 139 oz./month x 45 months = 6,255 oz. x USD 500/oz. = USD 3.1m (monthly payments starting March 2019); and
3. Tranche 3 – 178 oz./month x 45 months = 8,010 oz. x USD 500/oz. = USD 4.0m (monthly payments starting March 2020).

A2.8 In total, we model repayments of USD 11.3m, of which (i) USD 7.0m in principal; (ii) and USD 4.3m in interest, at an effective interest rate of 17.0%, over the full term of the PLI Loan.

Upside participation

A2.9 Under the terms of the PLI Loan, PLI would have benefitted from upside participation, whereby Claimant and PLI would have shared the incremental profits arising if metal prices rose above pre-determined prices (the “Base Spot Price”). PLI’s share of the upside was 30%, calculated on a monthly basis based on monthly payable production volumes and prices as determined pursuant to applicable mineral offtake agreements:³⁰⁴

“The Seller shall pay an amount equal to the greater of zero and the sum of the amount determined for each Covered Metal (for the avoidance of doubt, which may be a positive or negative number) as follows: the product of (a) 30% of the Monthly Payable Production of such Covered Metal and (b) an amount equal to the price as determined pursuant to the Mineral Offtake Agreement for such Covered Metal for the corresponding Monthly Payable Production minus the Base Spot Price for such Covered Metal, payable by the Seller to the Buyer within two (2) Business Days following the relevant Monthly Delivery Date.”

A2.10 We estimate the upside participation on an annual basis as follows:

Upside participation = 30% × Forecast annual payable production × (Forecast price – Base Spot Price)

A2.11 The Base Spot Price is, for each metal, the lower of (i) the average of the settlement prices for January 2017 and (ii) the five lowest settlement prices from 31 January 2017 to the date of payment of the first Tranche,³⁰⁵ being 9 August 2017.³⁰⁶ For the purposes of our calculation we use the same Base Spot Price as that used by Red Cloud.³⁰⁷

³⁰⁴ AC-04 Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017, Section 7(4) ↩

³⁰⁵ AC-04 Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017, page 3. Settlement calculated with reference to (i) LBMA Gold Price PM for gold, (ii) LBMA Silver Price for silver, and (iii) the Official LME Cash Settlement Price for copper, lead and zinc. ↩

³⁰⁶ Claimant's Memorial, paragraph 43 ↩

³⁰⁷ AC-15, sheet "fin_lupaka", rows 306, 316, 326, 336 and 346 ↩

[Page 83]

A2.12 Consistent with our reading of the PLI Loan terms, we assume that the upside participation would have been payable in each month from the start of production to the month in which the last gold repayment was made to PLI.³⁰⁸

A2.13 Based on the above approach, we estimate total upside participation to be USD 6.7m in the 355t/day Scenario.

B. PLI Loan in the 590t/day Scenario

Repayment of principal and interest

A2.14 Under the draft Third Amendment to the PLI Loan, Claimant would have received additional gross proceeds of USD 13.0m to fund the acquisition of the Mallay Plant. In addition, Claimant would have benefitted from a nine-month grace period between the closing date of the Mallay Plant acquisition and the first gold repayment.³⁰⁹ A repayment schedule appended to the draft Third Amendment to the PLI Loan assumed that the first gold repayment would have taken place in September 2019, based on a closing date of December 2018.³¹⁰

A2.15 Based on this repayment schedule, following receipt of USD 13m of additional financing, we model repayments of USD 32.3m, of which (i) USD 20m in principal (including Tranches 1-3); and (ii) USD 12.3m in interest at an effective interest rate of 16.1%, over the full term of the PLI Loan.

A2.16 Since Claimant’s purchase of the Mallay Plant was contingent upon Buenaventura obtaining approval of an easement agreement from the Mallay Community,³¹¹ which was not received until March 2019,³¹² Claimant considers that it would have signed and closed the purchase of the Mallay Plant by the end of March 2019 and that its repayment schedule would have been pushed back to January 2020 accordingly.³¹³ We therefore model the gold repayments starting in January 2020.

Upside participation

A2.17 We model the upside participation in the same way as in the 355t/day Scenario. We estimate total upside participation to be USD 9.9m in the 590t/day Scenario.

³⁰⁸ See AC-04 Second Amended and Restated Pre-Paid Gold Purchase Agreement dated 2 August 2017, Section 7(4): "The Seller shall pay [the upside participation] within two (2) Business Days following the relevant Monthly Delivery Date". The definition of Monthly Delivery Date in turn refers to the Scheduled Delivery Month, which is defined as "each of the 60 calendar months following the month in which the Gold Prepayment Amount is paid in three instalments in accordance with this Agreement on the First Effective Date, the Second Effective Date and the Third Effective Date, as applicable, including any months pursuant to Section 7(3) under this Agreement" (AC-04, p.19). ↩

³⁰⁹ AC-57 Draft Amendment and Waiver No. 3 to the Second Amended and Restated Pre-paid Forward Gold Purchase Agreement dated 5 October 2018, page 12 ↩

³¹⁰ AC-57 Draft Amendment and Waiver No. 3 to the Second Amended and Restated Pre-paid Forward Gold Purchase Agreement dated 5 October 2018, page 48 ↩

³¹¹ Claimant's Memorial, paragraph 93 ↩

³¹² C-233 Email from Buenaventura to Lupaka dated 11 March 2019 ↩

³¹³ Second Witness Statement of Gordon Lloyd Ellis dated 23 September 2022, paragraphs 38 & 39 ↩

[Page 84]

Appendix 3 Our updated damages model for the 590t/day Scenario

A3.1 In Section 6, we detail our updated assessment of the FMV of Claimant's investment under the 590t/day Scenario.

A3.2 Our underlying model is provided in electronic format.

[Page 85]

Appendix 4 Our updated damages model for the 355t/day Scenario

A4.1 In Section 6, we detail our updated assessment of the FMV of Claimant's investment under the 355/day Scenario.

A4.2 Our underlying model is provided in electronic format.

[Page 86]

Appendix 5 Our updated analysis of the other indicators of value

A5.1 In Section 7, we set out and respond to AlixPartners' comments on the other indicators of value of the Invicta Project presented in the First Accuracy Report as benchmarks for our damages assessments at the Valuation Date.

A5.2 Our updated calculations relating to these other these indicators of value are provided in electronic format.

[Page 87]

Appendix 6 Comparison of selected interest rates to inflation

A6.1 In the figure below, we compare the US Consumer Price Index (“CPI”) against indexed returns applying pre-award interest rates of (i) LIBOR +2% (as we were instructed in the First Accuracy Report); and (ii) UST +2%, proposed by AlixPartners (all figures indexed to 100 at the Valuation Date).

A6.2 The figure demonstrates how that applying pre-award interest at either of the two rates above between the Valuation Date and 31 August 2022 (being a proxy for the date of this report) results in returns significantly below inflation. In other words, pre-award interest at 31 August 2022 using either of these rates would be negative, in real terms.

Table A6.1. Comparison of US CPI to LIBOR +2% & UST +2%

120.0

115.0

110.0

105.0

100.0

95.0

Aug-19
Sep-19
Oct-19
Nov-19
Dec-19
Jan-20
Feb-20
Mar-20
Apr-20
May-20
Jun-20
Jul-20
Aug-20
Sep-20
Oct-20
Nov-20
Dec-20
Jan-21
Feb-21
Mar-21
Apr-21
May-21
Jun-21
Jul-21
Aug-21
Sep-21
Oct-21
Nov-21
Dec-21
Jan-22
Feb-22
Mar-22
Apr-22
May-22
Jun-22
Jul-22
Aug-22

CPI                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                           &-

Source: Appendix 6, tab "CPI vs rates +2%”

A6.3 For the purposes of this report, we are instructed to apply alternative pre-award interest rates of (i) LIBOR +4%; and (ii) UST +5% (see our updated damages assessment in Section 6).

[Page 88]

A6.4 Applying pre-award interest at either of these rates between the Valuation Date and 31 August 2022 results in overall returns broadly consistent with inflation over the calculation period. In other words, Claimant would more or less “break even” in real terms at 31 August 2022, were either of these rates to be applied.

Table A6.2. Comparison of US CPI to LIBOR +4% & UST +5%

120.0

115.0

110.0

105.0

100.0

95.0

Aug-19
Sep-19
Oct-19
Nov-19
Dec-19
Jan-20
Feb-20
Mar-20
Apr-20
May-20
Jun-20
Jul-20
Aug-20
Sep-20
Oct-20
Nov-20
Dec-20
Jan-21
Feb-21
Mar-21
Apr-21
May-21
Jun-21
Jul-21
Aug-21
Sep-21
Oct-21
Nov-21
Dec-21
Jan-22
Feb-22
Mar-22
Apr-22
May-22
Jun-22
Jul-22
Aug-22

CPI                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               &-

Source: Appendix 6, tab "CPI vs revised rates"

A6.5 Our calculations are set out in electronic format in Appendix 6a.