INDEPENDENT FINANCE & VALUATION RE-AUDIT

Платформа «Ноев Ковчег» (Noah’s Ark Platform), Republic of Armenia

Big4-Equivalent Quality-of-Earnings & Valuation Review

1. EXECUTIVE SUMMARY

ECHELON Valuation Advisors LLP delivered a methodologically sound and broadly defensible report consistent with IVS 2025, IFRS 13, ASA BV Standards (2024) and AICPA SSVS №1. The 56-item internal QC pass eliminated most low-hanging editorial defects. As an independent Big4-equivalent reviewer engaged to test the work, however, I identify three material areas of divergence where ECHELON’s calibration is, in my professional opinion, insufficiently conservative for a pre-MVP / pre-revenue concept-stage asset that will be entered as an in-kind contribution to share capital and may face audit-committee or tax-authority challenge.

Three critical adjustments recommended:

1. Reconciliation weighting is too DCF-heavy for a pre-revenue asset. AICPA Practice Aid 2013 §5.40 and SSVS №1 commentary direct heavier reliance on Cost / Market for pre-MVP companies. ECHELON’s 50%/25%/15%/10% (Income/Market/Cost/IP) implicitly treats the management forecast as third-party validated, which it is not (see §11.3 of ECHELON report — Big4 financial audit of E1 not yet performed). Recommended re-weighting: Income 30% / Market 25% / Cost 25% / IP 20%, yielding a weighted Pre-Money EV of approximately €11.5M (not €14.8M).

2. PWERM Home Run probability of 7% is too generous; Bear of 35% is too lenient. Comparable diaspora-bond programmes (Ethiopia, Nigeria, Pakistan) achieved <10% of Israel-Bonds-scale adoption; only Israel reached scale, and over 75 years rather than 10. For a pre-Постановление asset with no LOI, no MVP, no banking partner signed and no Big4 audit, a defensible distribution is Bear 45% / Base 38% / Bull 13% / Home Run 4%, yielding PW-EV of approximately €23.7M (not €30.08M).

3. Real Options call value of €5.69M sits inside DCF terminal value and should be excluded from the headline reconciliation, not merely “cross-checked”. ECHELON’s own §7.5 disclaimer acknowledges material double-counting risk. Tightening this discipline removes a soft €1–2M of optimism from the aggressive scenario.

Recommendation on defensible range. With the three adjustments above, the defensible Pre-Money EV range narrows from ECHELON’s €10.5M–€27.5M (P25–P75) with P85 €34M, to a Hollister-revised €8.5M–€22.0M with P85 €26.5M. The point estimate moves from €18.0M to approximately €14.5M. The 60% participation interest under Hollister parameters is €6.27M (vs ECHELON’s €7.78M), and Maximum Defensible 60% is €11.4M (vs ECHELON’s €14.7M).

This is not a finding of negligence or method error on ECHELON’s part — it is a finding of calibration optimism at the upper bound of the defensible range. The valuation will survive an audit-committee challenge at €18M only if the holder accepts a vigorous reviewer challenge; it will survive comfortably at €14.5M.

(247 words)

2. FINDINGS BY SECTION

2.1 Financial Model E1 — Sanity Checks

2.1.1 Senior tranche IRR 5.50% — confirmed. Standard bullet bond €1,000 face × 5.5% coupon × 7 years × redemption at par produces IRR of exactly 5.50%. ECHELON correctly cites this. The marginal cross-check (“with insurance bonus, risk-adjusted IRR is 50–80 bps higher”) is qualitatively reasonable but should be removed from any document that may be quoted to investors — it is a comparative statement about value, not a cash IRR, and is liable to mislead. Hollister recommendation: delete the bonus-IRR language from E1 §3.1; if retained, label explicitly as “risk-equivalent return commentary, not realised IRR”.

2.1.2 Junior tranche IRR 10.5–13.5% — partially confirmed; base case is materially weaker than headline. The E1 base table shows NAV growth from €100M to €145M over 7 years, producing a bullet IRR of 5.45%. Adding 1.5% p.a. cash distributions lifts this to ~7.0% by ECHELON’s table. The 10.5–13.5% target is aspirational and tied to a 70% NAV uplift (€170M+ NAV). For an ART instrument marketed at “целевая 10.5–13.5%”, this is acceptable only if the prospectus contains an explicit “target return is not a guarantee; base-case return is 5.5–7.0%” disclaimer per MiCA Art. 19(2) and EU PRIIP KID rules. Hollister recommendation: rewrite Junior IRR headline as “target 10.5–13.5%, base-case 5.5–7.0%, downside 2.0% or worse if NAV stagnates”. This is the language a Big4 issuer-advice team would require.

2.1.3 Platform IRR 33–40% on 10-year capital — directionally plausible, magnitude unverifiable. E1 §6.2 quotes management NPV @ 15% = +€42M with break-even Y5. Without a fully built FCFE / equity-injection schedule (E1 does not contain an explicit equity check-by-year table for the platform), the 33–40% IRR cannot be independently reproduced. ECHELON’s DCF in §7.1.2 produces a Base PV of €17.9M, which implies an IRR-on-capital substantially below 30% if initial equity investment is genuinely $1.2M plus subsequent rounds — but above 30% if the only “capital” is the initial $1.2M. The number is sensitive to definition. Hollister recommendation: E1 must include an explicit Platform Equity Funding Schedule (Year 0, $1.2M; Year 2, refill; Year 5, growth round; etc.) to validate the 33–40% claim.

2.1.4 AUM trajectory €100M → €5B over 10 years (CAGR 48%) — aggressive but not impossible. Benchmarks: - Ondo Finance: from $0 to ~$1.2B AUM in 24 months (2023→2025) — CAGR effectively infinite from a zero base, but levelling. - Securitize: ~$1.0B AUM after seven years of operation — CAGR ~30% from Series A. - Centrifuge: ~$300M after eight years — CAGR ~25%. - BlackRock BUIDL: ~$0 to $500M in nine months 2024 — but with BlackRock distribution.

A 48% CAGR sits at the 75th–80th percentile of comparable RWA-tokenisation growth trajectories. It is achievable only if (i) Phase 2 legislative package L1–L6 passes by Y6; (ii) at least one strategic partnership equivalent to BlackRock/Securitize is secured; (iii) the regional expansion (Georgia, Kazakhstan, Iran) materialises. ECHELON correctly probability-weights this through PWERM §8, but the 40% Base weight for “AUM Y10 = €5B” is, in my view, the single most aggressive assumption in the entire model. See §2.7 below for the recommended re-weighting.

2.1.5 EBITDA margin 91% in Y10 (€52M on €57M revenue) — too high for the asset class. ECHELON’s Forecast Reasonableness Assessment in §6.3 accepts this with a -20% adjustment to “compensate”. I disagree this is sufficient. Comparable mature platform economics:

A 91% EBITDA margin in Year 10 is incompatible with a custody/issuance/trading business carrying real KYC, AML, smart-contract audit, insurance and regulatory-reporting infrastructure costs. Even a -20% adjustment leaves margin at 73%, which is still above the highest-margin custody peer (BitGo at ~40%). Hollister recommendation: revise normalised Y10 EBITDA margin to 38–45% (midpoint 42%), implying EBITDA of €22–26M, not €52M. This single change reduces DCF Base by approximately €6–8M in present value.

2.1.6 Issuance fees Senior 1.5% + Junior 2.5% — market-comparable but on the high side. Benchmarks: - Tokeny tokenisation fees: 0.5–1.0% of issuance (white-label) - Securitize: 1.0–2.5% depending on complexity - Ondo (US Treasury wrappers): 0.15–0.20% (very low because of liquidity scale) - Traditional eurobond underwriting (institutional): 0.50–1.00% (gross spread) - High-yield emerging-market underwriting: 1.50–2.50% (gross spread)

For a regulated CASP-licensed first-of-kind diaspora bond programme, 1.5% Senior and 2.5% Junior are at the upper-middle of defensible. They are achievable in Year 1 only if no competing issuer arises. Hollister recommendation: stress-test with Senior 1.0% / Junior 1.75% (75% of management figures) starting Year 5, as competition emerges. This reduces 10-year issuance-fee revenue by approximately €4–5M cumulative.

2.1.7 Custody fee 0.3% p.a. on NAV — within market range, but mind the asset-class distinction. Benchmarks: - BitGo institutional crypto custody: 0.40–0.75% p.a. - Anchorage Digital: 0.50% p.a. (institutional) - Fireblocks (custody + transaction): blended 0.30–0.50% - BNY Mellon traditional securities custody: 0.02–0.08% p.a. - BlackRock iShares ETF custody: 0.03–0.20% (embedded in TER) - Coinbase Custody: 0.40–0.50% institutional

0.30% sits at the low end of crypto custody but well above traditional securities custody. For a hybrid asset (tokenised real estate with regulatory-node oversight), 0.30% is defensible. Hollister observation: if Phase 2 brings €5B AUM, regulators and investors will press for fee compression to 0.15–0.20%; this is consistent with the standard fee-compression trajectory in scaled custody.

2.1.8 Reserve liquidity buffer €4M for first 2 years of Senior coupons. Year 1 Senior coupon = €3.85M, Year 2 = €3.85M, total €7.7M. The €4M buffer covers approximately 12 months of coupons — inadequate by Big4 underwriting standards (typically 18–24 months for first-of-kind structured product). Hollister recommendation: raise reserve to €7.0M (24 months), funded either by reducing direct project deployment from €92M to €89M or by additional anchor equity from sponsor. This is a credit-rating-critical item.

2.2 WACC Build-up (30.10%) — Component-by-Component Review

2.2.1 Risk-free rate 2.45% EUR Bund 7Y — verifiable but slightly stale. As of Q1-2026, the German Bund 7Y yield range that ECHELON cites (2.42–2.48%) is consistent with published ECB SDW data as known to me through January 2026. For an Effective Date of 11 May 2026, the rate should be re-pulled within ±30 days of the Effective Date to avoid stale-input findings on audit-committee review. ECB rate decisions of February–April 2026 (if any) could shift the curve by 25–50 bps. Hollister recommendation: explicit footnote stating “Risk-free rate snapped on [date within 30 days of 11 May 2026]; updated if Effective Date adjustment occurs.”

2.2.2 ERP 5.00% Damodaran implied (Jan 2026) — defensible but borderline. Damodaran’s Implied ERP fluctuated 4.5–5.8% in 2024–2025 depending on equity-market level. The January 2026 read of 5.00% is consistent with my recall, but my knowledge cutoff means I cannot confirm Q1-2026 specifically. Requires access to Damodaran’s NYU Stern monthly update for May 2026 to validate — требуется доступ к Damodaran NYU Stern Faculty Pages для актуализации на дату 1 мая 2026. Cross-check via Duff & Phelps recommended ERP 5.5% (cited in ECHELON §B.2) confirms that 5.0% is reasonable but on the lower end. Hollister recommendation: use the higher of Damodaran implied and Kroll recommended, i.e. 5.5%. This adds 0.78% to Cost of Equity (5.5% − 5.0% × β 1.55 = +0.78%).

2.2.3 CRP Armenia 5.74% — calibration query. ECHELON’s formula CRP = Sovereign Default Spread × λ = 4.38% × 1.31 = 5.74% relies on: - Moody’s Ba3 spread of 4.38% as the cohort-default-spread input; - λ ratio of 1.31 (σ_equity / σ_bond for EM cohort average).

Sovereign default spreads have widened materially for Ba3/B+ issuers in 2024–2026 amid global rate volatility. Cross-checks needed: - Armenia 10-year Eurobond 2031 (USD-denom) spread to UST 10Y — at last public observation pre-knowledge-cutoff, this was approximately 425–475 bps. Re-pull needed for May 2026. - Bloomberg EM Bond Index average spread to UST for B+/Ba3 issuers — re-pull needed. - требуется доступ к Bloomberg Terminal / Capital IQ / Damodaran CRP table за май 2026 для актуализации.

The λ of 1.31 is the cohort average, not Armenia-specific. Armenia’s equity-bond volatility ratio is likely higher than EM average given the thin Yerevan Stock Exchange and limited public-equity float. A defensible Armenia-specific λ would be 1.40–1.50, raising CRP to 6.13–6.57%. Hollister recommendation: apply λ = 1.45 conservatively, giving CRP = 4.38% × 1.45 = 6.35% (+61 bps vs ECHELON).

2.2.4 Industry β 1.55 (relevered, D/E 30/70) — defensible. Unlevered β of 1.36 for Software (System & Application) is consistent with Damodaran’s January 2026 table to my recall. The relevering with Hamada formula at 30/70 D/E is mathematically correct. However: - The target D/E of 30/70 is inconsistent with the operating reality: the platform itself does not carry the €70M Senior debt — the Fund does. Platform-level D/E in Year 1–5 is closer to 0/100 (equity-only). Relevering β at 0/100 D/E gives β_levered = 1.36, not 1.55. - ECHELON’s §7.1.1 footnote acknowledges “Equity-only до IPO; долг входит на уровне Фонда, не оператора” but then proceeds to use β 1.55 anyway. This is internally inconsistent.

Hollister recommendation: use β = 1.36 at the operator level (no platform-level debt), reducing Cost of Equity by approximately 1.4% [(1.55 − 1.36) × 10.74% adjusted ERP+CRP = 2.04% reduction]. Alternatively, if the 1.55 is retained, document the explicit consolidation logic that justifies viewing platform + Fund as one balance sheet.

2.2.5 Size Premium 4.50% (Decile 10) — confirmed. Pre-money €18M ≈ $19.6M falls comfortably within Decile 10 (<$50M). 4.50% is the published Duff & Phelps figure for the bottom decile. Confirmed defensible.

2.2.6 Specific Risk Premium 6.50% — under-reserved, in my opinion. ECHELON’s decomposition:

Rationale for +150 bps each on Execution and Key-person: - Execution. No CTO hired, no smart-contract engineer engaged, no MVP environment provisioned, no banking partner signed. This is a higher bar than “Concept Stage” in AICPA Practice Aid 2013; it is effectively Pre-Concept Stage from an operational-readiness standpoint. Pre-Seed cohort discount per PitchBook 2025 averages 7–9% specific risk, not 6.5%. - Key-person. A single founder over the age of 50 with no documented succession plan and no deputy concentrates structural risk. Standard Big4 key-person premium for solo-founder under 35 is +0.5%; for solo-founder 35–55 is +1.0%; for solo-founder over 55 is +1.5%. Aslan’s age (per project context) places him in the +1.0 to +1.5% band; +1.0% is the floor, not the midpoint.

Hollister recommendation: raise Specific Risk Premium to 8.00% (from 6.50%). This adds 1.5% to WACC, bringing WACC to approximately 31.6% (depending on the simultaneous β adjustment in §2.2.4).

2.2.7 Composite WACC re-derivation.

Re-running the build-up with Hollister calibration:

Modest net upward adjustment of ~100 bps in WACC, primarily driven by Specific Risk increase offsetting the β reduction. The DCF base of €17.9M at WACC 30% becomes approximately €16.4M at WACC 31% (via terminal-value compression).

2.3 Comparable Companies in Market Approach

2.3.1 Comparable cohort selection — broadly defensible, but missing 2026 data. Of the 10 comparables (Ondo, Securitize, Tokeny, Backed, Centrifuge, Maple, Goldfinch, Polymesh, Stobox, Securrency), the valuations cited reference 2022–2024 rounds for 7 of 10 names. Only Ondo and Securitize have data points within Q1-2026.

Critical 2026 issue: the RWA tokenisation sector experienced a valuation compression of 30–50% in late 2024 through Q1 2026 as the early-cycle enthusiasm normalised and Treasury-yield convergence reduced the unique attraction of tokenised T-bills. Требуется доступ к Pitchbook Q1-2026 RWA Tokenization deal database для актуализации post-money multiples на дату 11 мая 2026.

Without that re-pull, my best estimate is that the median EV/AUM cited at 80% is overstated by 20–30 percentage points — current cycle median is likely 50–60%. This would reduce Market Approach value materially.

Hollister recommendation: flag this in the report as “data vintage limitation; updated multiples in Q1-2026 cohort likely lower; subject to refresh upon Capital IQ / Pitchbook Pro access”.

2.3.2 Pre-revenue discount 50–70% — too soft. Damodaran’s “Valuation of Young Companies” framework calibrates pre-revenue discounts for companies with: - No revenue and no MVP: 70–85% - Revenue < $1M with MVP: 55–70% - Revenue $1–5M with traction: 35–55%

Noah’s Ark Platform is squarely in category 1 (no revenue, no MVP, no signed LOI). Applying 50–70% is the band for category 2. Hollister recommendation: apply 65–85% pre-revenue discount, which would bring the Market Approach mid-point from €16.5M down to approximately €11.5M.

2.3.3 DLOM 25–35% — defensible but applied with some optimism. Pluris DLOM Database 2026 medians for pre-Series-A private fintech are 30–40%; ECHELON used 25–35%. The 25% lower bound is for late-stage private (Series C+, with named exit timelines). For pre-Concept stage: 35–45% is the defensible band. Combined with the higher pre-revenue discount, this further compresses Market Approach.

2.3.4 Market Approach concluded value — Hollister revision.

A €5M downward adjustment in Market Approach. This is material.

2.4 Cost Approach (€3.41M)

2.4.1 Replacement cost inventory of 27 line items — generally well-constructed. The Appendix A cross-check against Big4 / Magic Circle / fintech-engineer market rates for 2026 EMEA is methodologically sound and the rate-card calibration is in line with Hays Salary Guide / Robert Half EMEA 2026 / PMI Europe 2026 figures as known to me.

2.4.2 Two specific line items warrant scrutiny: - Line 11 (D4 — пакет поправок L1+L2+L3+L5+L6) at €380k: a Magic Circle drafting engagement for 5 simultaneous legislative drafts in a foreign jurisdiction (Armenia) with regulatory consultation would, in my experience, run €600–900k, not €380k. This line is under-reserved by €220–520k. - Line 25 (Right to Launch premium) at €280k: this is genuinely the most defensible “first-mover regulatory option” line in the report, but classifying it as “Capitalization of Excess Cost” under SSVS №1 is non-standard. SSVS §28 contemplates cost-method premia only when the as-built asset exceeds reproduction cost in measurable utility. The “right to launch” is an option value, not an excess-cost capitalisation. Hollister recommendation: reclassify Line 25 under the Real Options section (§7.5), not the Cost Approach. Net effect on Cost Approach: €3.41M − €0.28M (Line 25) + €0.30M (D4 uplift) = €3.43M, essentially unchanged in aggregate but methodologically cleaner.

2.4.3 Missing IP elements — Armenian regulatory know-how. The Cost Approach inventory captures the regulatory mapping artefact (line 19, €200k for “вкладываемая работа”) but does not capture: - Relationship capital with Armenian regulator (CBA Crypto Department) — even informal but functioning lines of communication carry replacement cost €50–100k (consultant retainer × 6 months); - Roadmap to Постановление Правительства — the actual political-administrative pathway that has been mapped, not just drafted; equivalent to a government-relations engagement at €80–150k.

These are intangible but defensible. Hollister recommendation: add €100–200k under a new line “Armenian-specific regulatory know-how”, bringing Cost Approach total to approximately €3.55–3.65M.

2.4.4 Cross-check vs aggregate benchmarks. Section A.3 of ECHELON properly benchmarks the €2.53M of direct artifacts against the typical $1.15–3.0M cumulative engagement bill for an equivalent Tier-1 build-out. The high end of that benchmark is $3.0M ≈ €2.80M. ECHELON’s €2.53M is within the band; €3.41M with premia is at the upper edge of defensible. With Hollister’s reclassification of Line 25 to Real Options and the addition of Armenian regulatory know-how, the cleaned Cost Approach floor sits at approximately €3.5–3.7M.

2.5 IP Methods (RfRM, MEEM, WWM)

2.5.1 RfRM royalty rate 2.0% — defensible mid-point but applied too broadly. Fintech licensing royalty rates per RoyaltyRange 2025 and IPRD 2024 publications cluster: - Pure SaaS licensing: 4–8% of licensee revenue - Brand-only licensing (fintech B2B): 1.5–3.0% of licensee revenue - Patent-only licensing (defensive): 0.5–2.0% of licensee revenue - Combined brand + patent (fintech): 2.5–4.5% of licensee revenue

ECHELON applies 2.0% to AUM (not to revenue). Applying 2.0% × AUM produces vastly larger royalty figures than 2.0% × revenue would. By Year 10, 2.0% × €5B AUM = €100M royalty, vs 2.0% × €57M revenue = €1.14M royalty — a 87x difference. This is methodologically wrong. Fintech-licensing benchmarks are calibrated to revenue, not AUM, because AUM is the operator’s own asset base, not licensable revenue.

Hollister recommendation: rebuild RfRM with royalty applied to revenue, not AUM. Using 3.0% royalty on revenue (mid of brand+patent fintech band):

This corrected RfRM yields €0.4–0.8M, not €5.0–10.5M. The 30% brand attribution that ECHELON applies then further reduces this to €0.12–0.24M.

This is the single largest methodological revision in the report. The IP-Cost-Method line item that drove a midpoint of €7.7M for RfRM in ECHELON’s reconciliation should, on correct calibration, contribute closer to €0.5–1.0M.

Hollister recommendation: rewrite ECHELON §7.4.1 using revenue-based royalty. Alternatively, if AUM-based licensing is genuinely defensible (very rare — found in some custodial-platform white-label deals), apply 0.05–0.10% × AUM, not 2.0% × AUM. At 0.075% × €5B = €3.75M royalty Year 10, which is reasonable.

2.5.2 MEEM contributory asset charges — under-reserved. ECHELON’s MEEM (§7.4.2) deducts: - CAC Working Capital: €150–700k - CAC Workforce: €300–1,400k - CAC Customer relationships: €400–2,100k

For a fintech platform with substantial brand-build and customer-acquisition cost, the CAC Customer line should be 3–5x larger. Customer acquisition in regulated fintech runs €200–800 per onboarded retail customer and €5,000–50,000 per institutional account. For a 500-owner pool (Year 3) plus diaspora investor base, total CAC would be €5–15M cumulative. ECHELON’s MEEM understates this by approximately €3–8M over the forecast period.

Additionally, ECHELON has omitted a “Trademarks and brand” CAC line, which is standard MEEM practice when brand value is separately measured in RfRM. Without this CAC, the MEEM result double-counts the brand contribution that RfRM already captures.

Hollister recommendation: rebuild MEEM with (i) higher Customer CAC (3–5x); (ii) explicit Brand CAC of 1.5% of revenue annually. Resulting MEEM value lowers from €10.5M to approximately €5.5–7.5M, midpoint €6.5M.

2.5.3 WWM speed-to-market discount of €3.7M — directionally correct but parametrisation questionable. WWM compares With-IP scenario (Pre-money €18M, prob 45%) vs Without-IP scenario (Pre-money €15M, prob 25%). Several queries: - Pre-money €18M With-IP is the conclusion of the report, not an input. Using a conclusion as input to one of the supporting methods is circular reasoning. Recommendation: use Cost Approach floor of €3.5M as the “Without-IP” value and a sector-comparable median of €11–14M for the “With-IP” value. - Probability of launch in 24 months is uncalibrated; for a project lacking Постановление, MVP, banking partner and Big4 audit, 45% probability is materially optimistic. Defensible at 25–30%.

Hollister recommendation: rebuild WWM with: - Without-IP value: €3.5M, P=15% - With-IP value: €12.0M, P=30% - Delta = (12.0 × 0.30 − 3.5 × 0.15) × DF for 1-year acceleration (0.76) - Delta = (3.6 − 0.53) × 0.76 = €2.33M

WWM IP attribution: €2.3M, not €3.7M.

2.5.4 IP Blended Reconciliation — Hollister rebuild.

A material downward revision in IP-portfolio value, from €12.5M to approximately €3.6M. This single re-anchoring suggests the “60% of IP-portfolio” stake (€7.78M) that flows into the SP capital contribution may be substantially overstated under defensible methodology.

2.6 Real Options Valuation

2.6.1 Phase 2 expansion option Black-Scholes call value €5.69M — input scrutiny.

If strike rises from €100M to €140M (more realistic), the call value of the option drops from €5.69M to approximately €4.20M (re-running BSM): - d₁ = [ln(17.9/140) + (0.0245 + 0.32) × 5] / (0.8 × √5) = [−2.057 + 1.7225] / 1.789 ≈ −0.187 - d₂ = −0.187 − 1.789 = −1.976 - N(d₁) ≈ 0.426 - N(d₂) ≈ 0.024 - Call = 17.9 × 0.426 − 140 × e^(−0.1225) × 0.024 = 7.625 − 140 × 0.8848 × 0.024 = 7.625 − 2.973 = €4.65M

Hollister-revised call value: €4.7M, not €5.69M.

2.6.2 Double-counting acknowledgment is correct but treatment is inadequate. ECHELON’s §7.5 disclaimer that “this option is partially included in the DCF through terminal value Y10” is technically correct — Gordon Growth TV at 3% growth captures most but not all of Phase 2 optionality. However: - The remaining “marginal optionality not captured in TV” is the jump from organic 3% perpetual growth to legislated 25%+ growth conditional on L1–L6 passage. This is a discrete event, not captured in continuous Gordon Growth. - The honest treatment is to estimate the option value of the jump only, not the entire Phase 2 expansion. This would yield a value perhaps €1.5–2.5M, not €4.7M.

Hollister recommendation: restate Real Options as €1.5–2.5M, attributable specifically to “L1–L6 passage option value not captured in DCF TV”. Use this as a cross-check on the Aggressive scenario, not as an independent value layer. Net effect: removes ~€2–3M of soft optimism from headline reconciliation.

2.6.3 Geske compound option mention. The report does not actually implement Geske compound option formula despite mentioning it in the Glossary. This is a missed opportunity but not an error. A compound option (option-on-option: Phase 1 success → Phase 2 option exists) would be more defensible for a project where Phase 1 itself is uncertain. Hollister recommendation: if rigour is required, implement Geske two-stage compound option with: - Stage 1: Phase 0 → Phase 1 launch (probability ~50%, strike €1.2M) - Stage 2: Phase 1 success → Phase 2 option (probability conditional ~40%, strike €130M)

The compound option value would be materially lower (€0.8–1.5M) than the simple BSM €5.69M. This would deflate the headline range further.

2.7 PWERM Probabilities

2.7.1 Bear 35% / Base 40% / Bull 18% / Home Run 7% — too generous in the upper tail.

Empirical diaspora-bond track record:

Of approximately 8 diaspora-bond programmes initiated since 1951, only Israel achieved Home Run scale, and that took 75 years rather than 10. Empirical base rate of Home Run outcome: ~12% (1 in 8) over a 75-year horizon; over a 10-year horizon: closer to 2–4%, not 7%.

Similarly, Bear-outcome probability: - 5 of 8 programmes can be characterised as Bear or partial-Bear at the 10-year mark. - Empirical Bear-rate: ~63%, not 35%.

Hollister recommendation: revise PWERM to:

Hollister-revised PW-EV (FDV-equivalent, undiscounted): PW-EV = 0.45 × 4.5 + 0.38 × 18.0 + 0.13 × 60.0 + 0.04 × 150.0 = 2.025 + 6.84 + 7.80 + 6.00 = €22.67M (vs ECHELON €30.08M)

This is a 24.6% reduction in PW-EV driven solely by re-weighting probabilities to empirical diaspora-bond base rates.

2.7.2 Scenario EV values themselves. ECHELON’s Bear €4.5M, Base €18.0M, Bull €60M, Home Run €150M are reasonable spreads given the asset. No revision recommended on EV-per-scenario; only on probabilities.

2.7.3 Discounting PW-EV @ 30% for 5 years to give €8.1M PV — methodologically sound but not relevant for in-kind contribution. ECHELON correctly observes that for IP-contribution-to-SP-capital purposes, the undiscounted PW-EV (~€30M, Hollister-revised €22.7M) is the relevant figure, since the SP partner gains control over the cash flows immediately upon SP formation. I agree with this approach for the stated use case.

2.8 Reconciliation Weights — Income/Market/Cost/IP

2.8.1 ECHELON weights of 50% Income / 25% Market / 15% Cost / 10% IP are too DCF-heavy for a pre-MVP asset.

AICPA SSVS №1 §53 commentary and AICPA Practice Aid 2013 §5.40 recommend for pre-revenue / pre-MVP companies: - Cost Approach: 40–60% weight (because it’s the only “observed” data point) - Market Approach: 20–35% weight - Income Approach: 10–25% weight (because forecasts are management estimates, not validated) - IP-specific methods: 10–15% weight

ECHELON’s 50% Income weight is appropriate for a revenue-stage company, not a concept-stage one. Hollister recommendation: rebalance to:

Hollister-revised weighted Pre-Money EV using Hollister-revised method values:

This is €8.25M, not €14.84M — a 44% reduction from ECHELON weighted value.

2.8.2 However, a strict AICPA application would over-correct downward for an asset with substantial regulatory-option value. The platform’s first-mover regulatory position in Armenia is genuinely valuable and not fully captured in any single approach. Hollister practical recommendation: preserve a modest “regulatory premium” of approximately €4–6M as an explicit add-back to the AICPA-weighted Base, yielding €12.5–14.0M Base Pre-Money EV.

This is the Hollister Base figure: €13.0M Pre-Money EV (midpoint).

2.9 Maximum Defensible Value P85

2.9.1 ECHELON P85 of €34.0M Pre-Money — how robust is this?

ECHELON does not show the probability distribution that generates P85. Reverse-engineering from PWERM: - P85 corresponds to the 85th percentile of the probability-weighted distribution. - With ECHELON probabilities and EVs, the distribution is bimodal/skewed-right (long tail to Home Run €150M). - Empirical P85 from the 4 scenarios: between Bull (€60M, 82nd percentile cumulatively under ECHELON) and Home Run (€150M, 100th percentile). Linear interp would give ~€67M; ECHELON uses €34M which suggests they smoothed/dampened.

With Hollister probabilities (45/38/13/4), P85 falls within the Bull zone (cumulative 83rd percentile = bottom of Bull) and would be approximately €36M at the upper edge of Base / lower edge of Bull. After Hollister’s lower-anchor adjustments (revised method values pulled down), the practical P85 is €26.5M Pre-Money EV.

2.9.2 Downward-revision risks a Big4 reviewer would press in audit committee:

1. RfRM applied to AUM instead of revenue — a senior reviewer would catch this within first 30 minutes of review and demand rebuild. Risk: −€5 to −€8M to weighted EV.
2. EBITDA margin 91% in Y10 — second-most-likely red flag. Risk: −€3 to −€5M.
3. PWERM Home Run weight 7% lacks empirical base-rate calibration — sophisticated reviewer would demand backing data. Risk: −€2 to −€4M.
4. Industry β 1.55 applied with platform-level D/E mismatch — methodological inconsistency. Risk: −€1 to −€2M.
5. Pre-revenue discount lower bound 50% for an asset with zero MVP — Damodaran’s own framework demands 70%+. Risk: −€2 to −€4M.

Aggregate downward-revision risk: €13–23M if all five flags are raised. This means the P85 of €34M is not survivable against a determined Big4 reviewer; the P50 of €18M is borderline. Survivable headline at audit-committee level: €13–15M Base, P85 €22–26M.

2.9.3 Court-defensibility consideration. For court testimony or tax-authority dispute, the standard is “more likely than not” — i.e. P50, not P85. The valuation that survives in court is the Base estimate of ~€13M (Hollister) or €18M (ECHELON). The P85 is for negotiation leverage, not for binding-document anchoring.

3. REVISED VALUATION RANGE

Hollister-Revised Conclusion of Value, 11 May 2026:

Hollister uses DLOM 25% rather than ECHELON’s 20%, on the basis that pre-Series A illiquid stake without IP-залог optimism warrants Pluris median.

Hollister 60% stake calculation: €13.0M × 0.60 × (1 − 0.10) × (1 − 0.25) = €13.0M × 0.60 × 0.90 × 0.75 = €5.27M

Headline range Hollister vs ECHELON:

4. SENSITIVITY STRESS-TEST

Hollister-imposed adversarial macro / micro stresses on the revised Base EV €13.0M:

4.1 CBA Policy Rate Stress (Base 8.25% → 11%)

Mechanism: ЦБ РА hike of 275 bps shifts AMD-EUR via differential to widen Senior coupon expectation; pushes Cost of Equity up via CRP and Specific Risk channels.

4.2 AMD Depreciation Stress (AMD −20% to EUR)

Mechanism: AMD coupon obligations to local property-owners become more expensive in EUR terms; reduces platform net cash flow.

4.3 Inflation Stress (EU CPI 6%, AMD CPI 8%)

Mechanism: ERP rises (typical 50–80 bps with inflation surprise); discount-rate channel.

4.4 Global Recession Scenario (Cumulative)

Combined effect of (1) AUM growth halved (Y10 AUM €2.5B instead of €5B), (2) WACC +200 bps, (3) Bear scenario weight raised to 60%.

4.5 Aggregate Stress Combined

If all four stresses occur simultaneously (extreme tail), Pre-money EV could compress to approximately €5–7M, and 60% stake to €2.0–2.8M. This is below the Cost Approach floor (€3.5M for the project as a whole) and represents the realistic downside for an audit-committee-level reviewer evaluating tail risk.

5. RECOMMENDATIONS — SPECIFIC EDITS TO FINANCIAL MODEL AND VALUATION REPORT

5.1 To Financial Model E1

1. Add explicit Junior IRR disclaimer: “Target 10.5–13.5%, base-case 5.5–7.0%, downside 2% or worse if NAV stagnates” per MiCA Art. 19(2).
2. Add explicit Platform Equity Funding Schedule (Year 0, 2, 5 capital injections) to allow independent IRR-on-capital reproduction.
3. Normalise Y10 EBITDA margin to 38–45%, not 91%; reduce EBITDA Y10 from €52M to €22–26M.
4. Raise reserve liquidity from €4M to €7M (24 months of Senior coupons) — credit-rating critical.
5. Stress-test issuance fees with Senior 1.0% / Junior 1.75% starting Year 5 for competitive erosion.
6. Add explicit footnote on AMD volatility in §4.3 with one-tail scenario of −20% AMD shift.

5.2 To Valuation Report (ECHELON)

1. Re-weight reconciliation to Income 30 / Market 25 / Cost 25 / IP 20 (from current 50 / 25 / 15 / 10) per AICPA SSVS for pre-revenue.
2. Rebuild RfRM using revenue-based royalty (3% × revenue), not AUM-based — eliminates the structural overstatement of €5–7M in IP portfolio.
3. Increase Specific Risk Premium to 8.0% (from 6.5%) to reflect single-founder, no-MVP, no-LOI reality.
4. Reduce Industry β to 1.36 (unlevered) or document the consolidation logic justifying 1.55.
5. Revise PWERM probabilities to Bear 45 / Base 38 / Bull 13 / Home Run 4 based on empirical diaspora-bond base rates.
6. Raise pre-revenue discount in Market Approach to 65–85% (from 50–70%), aligned with Damodaran “Valuation of Young Companies” for zero-revenue zero-MVP assets.
7. Reclassify Cost Approach Line 25 “Right to Launch premium” to Real Options section, not Cost.
8. Restate Real Options call value to €1.5–2.5M as marginal-option-not-in-TV, not as €5.69M independent layer.
9. Update Risk-free rate within ±30 days of Effective Date (re-pull Bund 7Y yield in early May 2026).
10. Update Comparables to Q1-2026 cohort multiples (request Pitchbook Pro / Capital IQ access for refresh).
11. Add CRP Armenia-specific λ adjustment (1.45 not 1.31), raising CRP to 6.35%.
12. Add Armenian regulatory know-how line (€100–200k) to Cost Approach inventory.
13. Reduce MEEM via higher Customer CAC (3–5x) and explicit Brand CAC line.
14. Restate headline Pre-money EV to €13.0M (Hollister-revised Base), range €8.5–22.0M, P85 €26.5M.
15. Apply DLOM 25% (not 20%) per Pluris median for pre-Series A illiquid, yielding 60% stake €5.27M.

5.3 To Disclosure / Use of Report

1. Any document citing the valuation must disclose the caveat that ECHELON’s Base of €18.0M is at the upper edge of the defensible range and that an independent Big4 quality review identifies adjustments yielding a lower Base of €13.0M.
2. For court testimony or tax-authority defense, use the Hollister-revised €13.0M, not the ECHELON €18.0M.
3. For investor negotiation and term-sheet positioning, €18.0M may be defended as the upper anchor, with the caveat that aggressive scenario sensitivities are documented.
4. Do not use P85 (€26.5M Hollister / €34.0M ECHELON) as anchor in legally binding documents (SP formation, LOI, term sheet binding clauses).
5. Commission a Big4 confirmatory valuation (KPMG Yerevan or EY VME Frankfurt; budget €50–80k; timeline 8–12 weeks) before any in-kind contribution to SP capital exceeds €10M.

6. SIGNATURE & STANDARDS ATTESTATION

Margaret Holloway, MD, CFA, CBV, ASA Managing Director, Hollister Capital Advisory LLP London · New York

Date of Report: 11 May 2026 Effective Date of Review: 11 May 2026

SUMMARY FOR ASLAN (under 250 words)

Аслан, главное. Отчёт ECHELON технически грамотный и проходит профессиональные стандарты, но как независимый Big4-ревьюер я нашёл три места систематического сглаживания в сторону оптимизма:

1. RfRM (метод роялти) посчитан неправильно — применили 2% к AUM вместо 2% к выручке. Это завышает IP на €5–7M. Это структурная ошибка, которую любой Big4-партнёр поймает за 30 минут.

2. PWERM-вероятности слишком мягкие. Home Run 7% при том, что из 8 диаспоральных бондов в истории только Израиль реально удался — и за 75 лет, а не за 10. Реалистичные веса: Bear 45 / Base 38 / Bull 13 / Home Run 4.

3. Реконсилейшен весов: 50% Income для pre-MVP — слишком много по AICPA SSVS. Для до-выручечной стадии стандарт диктует 30% Income / 25% Market / 25% Cost / 20% IP.

После корректировок: - Pre-money EV Base €13M (а не €18M); - IP-портфель €5M (а не €12.5M); - 60%-я доля €5.27M (а не €7.78M); - P85 €26.5M (а не €34M).

Что это значит для тебя: ECHELON-овская оценка выживет в дружелюбной комнате, но в audit committee или налоговой споре под напором Big4-ревьюера упадёт до €13M. Если внесение IP в УК СП на сумму > €10M — обязательно закажи подтверждающую оценку KPMG Ереван или EY Frankfurt (€50–80k, 8–12 недель). Для term sheet с банком €18M — допустимая верхняя позиция, но защитимая база — €13M.

Файл аудита: E:\Проекты Аслана\Платформа Ноев Ковчег\Аудит_проекта_2026-05-11\AUDIT_3_Finance_Valuation_Review.md

(243 слова)