Time Horizon: 12-18 months. BVN's three catalysts — Peru election (April 2026), San Gabriel ramp (2026-2028), Trapiche social license (H1 2026 decision) — all operate on different clocks. The election resolves in weeks. San Gabriel ramp is a 2-year story. Trapiche is 5-10 years. Coverage initiation sets the framework; actionability depends on which catalyst fires first.

Base Rate: Diversified EM miners with country-risk discounts and management transitions. Reference class: mid-cap miners trading at >30% P/E discount to sector due to jurisdiction risk. Historical base rate for closing the discount within 12 months: ≈25-30%. Most stay cheap because the discount is earned. BVN's discount decomposes into four specific, identifiable components (Peru country, governance, execution, cyclicality) — none of which have clear near-term resolution catalysts except Peru.

Base rate: EM miner at 40%+ P/E discount to sector → 25-30% close gap within 12mo
Prior odds: 0.36

Alpha vs Beta: This is the critical finding. BVN is 75% factor risk, 25% idiosyncratic.

1Y total return:              +154%
  Gold miners (GDX beta):      +62%  (0.57 × 108%)
  Peru country (EPU beta):     +54%  (0.59 × 92%)
  Market (SPY, negative):      −4%   (−0.22 × 18%)
  Cross-factor:                +12%
  Idiosyncratic alpha:         +30%  (13.5% ann., t = 0.56, NOT significant)

The idiosyncratic alpha is not statistically significant at any conventional level. Over a 2-year window, alpha reverses to −9.8%. If you own BVN, you own Peru and gold miners with a thin, unreliable layer of company-specific return on top. The stock is a leveraged factor expression, not an alpha opportunity — at least on trailing data.

B — Business Model

Peru's largest publicly-traded mining company. Seven operating mines producing gold, silver, copper, lead, and zinc, plus a 19.58% equity stake in Cerro Verde, one of the world's largest copper mines (FCX-operated, 415,000 tpd concentrator).

Revenue decomposition (FY2025 consolidated):

Including Cerro Verde's $4.2B revenue at BVN's 19.58% share, effective copper exposure rises to 45-50%. BVN is a copper company that reports like a precious metals miner.

The product form distinction matters. Dore bars realize ≈99% of spot metal price. Concentrates realize 54-93% after treatment charges, refining charges, and escalator deductions. San Gabriel producing dore — not concentrate — creates $42-78M/yr of incremental margin at steady state (120k oz/yr) versus equivalent concentrate production. This is not in headline production numbers.

Unit economics by mine (FY2024-2025):

Revenue concentration: Trafigura accounts for 40.6% of consolidated revenue (up from 28.2% in FY2023). Single-customer risk on the offtake side.

Cash flow structure: Revenue is entirely commodity-price-driven. FY2025 revenue growth of 50% was 80%+ metal prices, with gold volumes actually declining 19%. At 2023 metal prices with FY2025 volumes, revenue drops to ≈$1.17B (−32%). The CF_t stream is a leveraged call on metals with procyclical cost leakage (see Phi section).

Phi — Financial Trajectory

Revenue: g = dR/dt — Decelerating

Management guides FY2026 at $4,500 gold / $70 silver / $12,000/t copper. Spot is 7-9% above across all three metals. At spot, revenue hits the high end or above guidance. But consensus expects flat EPS ($3.07 forward vs $3.08 trailing), implying margin compression is priced in.

Margins: dm/dt — Compressing Despite Revenue Surge

Operating margin declined from 38.6% to 36.6% despite 50% revenue growth. Three structural drags:

1. Worker profit-sharing (8% of PBT) — Peruvian law, mandatory, procyclical. ≈$46M in FY2024.
2. Mining royalties (1-12% of operating profit) — progressive rate structure penalizes high metal prices.
3. Silver escalator deductions above $50/oz — CAS rose 616% QoQ in Q4 2025. Non-linear cap on silver revenue. At $75 silver, blended realization falls to ≈66% vs ≈79% at $30. Estimated drag: ≈$150M/yr versus linear extrapolation.

Gold cost inflation is structural: Orcopampa CAS +16%/yr, Tambomayo +26%/yr (FY2022-2024 trend). Depleting mines get more expensive to operate.

Combined, approximately 12-15% of incremental profits from higher metal prices leak to mandated procyclical costs before reaching shareholders.

FCF Conversion

CapEx peaks 2025-2026 (San Gabriel completion at ≈$160M, plus Uchucchacua expansion). If Trapiche doesn't commence, sustaining CapEx drops to $200-220M and FCF yield improves to 13%+ at current metal prices.

CapEx guidance for FY2026 ($385-415M) surprised above the Investor Day projection disclosed just months earlier. Management blamed "rainy season" timing. This is the second guidance credibility issue after the San Gabriel production cut.

Balance Sheet

Clean. $530M cash, $180M net debt, 0.22x leverage. $148.5M in 5.50% notes mature July 2026 — will refinance from cash. $650M in 6.80% notes due 2032. Interest coverage 13-15x.

Contingencies:

• SUNAT tax receivable: $454M paid under protest. BVN filed an amparo (constitutional appeal) in April 2024 challenging the assessment. Expected write-off: ≈$168M after tax (1.8% of market cap), probability-weighted across full recovery (20%), partial (35%), full loss (35%), and already-priced (10%) scenarios.
• Mine closure provisions: $317M (+22% YoY). Additions jumped 6x from $11M (FY2023) to $66M (FY2024), reflecting Orcopampa nearing closure, San Gabriel startup provisions, and Uchucchacua commitments. Letters of credit $186M — quasi-pledged cash.

Capital Allocation

• Dividend: $0.99/share (40% payout, 2.72% yield). Highly cyclical — at 2023 metal prices, this collapses to ≈$0.03/share.
• Buyback: None. 21.2M treasury shares (8%) held static since FY2022.
• Asset sales: Delayed. Management is "reanalyzing" as gold prices rise. No timeline.
• Exploration: Budget increasing to $90-100M (from $73M). Heavily brownfield-focused.

K — Competitive Position

Moat Classification

BVN is the only major Peruvian-controlled miner. All other large miners (SCCO, FCX/Cerro Verde, Antamina/BHP-Glencore, Las Bambas/MMG) are foreign-controlled. This is an underappreciated competitive asset for government relations, permitting, and community engagement.

Reserve Replacement: STRONG

San Gabriel reserves (1.83 Moz Au at 3.71 g/t) were booked at a $1,900/oz gold reserve price vs $4,900 spot — substantial reserve upside if re-estimated at current prices.

Uchucchacua's geology is mesothermal (not epithermal). Mesothermal deposits form at greater depth and higher temperatures, conferring greater depth extent. The 62% silver reserve increase over two years confirms the deposit extends downward. This mine has substantial remaining life.

Market Share: dS/dt — Currently LOSING

Gold production declined from 139.6k oz (FY2024) to 121.3k oz (FY2025), a 13% drop. BVN is losing Peru mining share while SCCO and Cerro Verde expand. Share reverses only when San Gabriel reaches steady state (≈2028).

G — Governance

Verdict: BELOW AVERAGE. The 3-4 P/E turn discount is earned.

ROIC vs WACC:

• Operating ROIC: 8.7% (direct operations only)
• Including Cerro Verde equity income: 12.4%
• Estimated WACC: 10-11% (BB credit, Peru sovereign risk premium)
• Marginally value-creating only at elevated metal prices. Value-destructive in FY2022.

Governance weaknesses are Peruvian-structural, not company-specific: no class action or derivative suit rights, weaker disclosure standards than NYSE (Item 16G, 20-F), limited minority shareholder protections. These don't improve without regulatory change.

Beta — Factor Profile

The Regression

Seven model specifications tested (251 daily observations, 1Y trailing). The preferred specification:

r_BVN = 0.57 x r_GDX + 0.59 x r_EPU - 0.22 x r_SPY + alpha

R-squared = 0.75 | %Idio = 25% | alpha_orth = 13.5% (t = 0.56, NOT significant)
Orthogonal Sharpe = 0.57

Variance Decomposition

At 25% idiosyncratic variance, BVN is 3x below the 75% target. This is a factor trade.

What the Regression Corrected

The preliminary G-zero screen estimated Peru at 3-8% of variance. The regression says 40%. This is the single largest correction from the initiation process. Silver (SLV) and copper (COPX) are NOT independent factors after controlling for GDX and EPU — the commodity exposures run through the country/miner ETFs. You don't hedge BVN with SLV or COPX. You hedge with GDX + EPU.

Factor Stability

Rolling 90-day alpha ranges from −29% to +67% — extremely unstable. On a 2-year window, alpha reverses to −9.8%. The positive alpha observed over the trailing year may not persist.

BVN has negative SPY beta (−0.22, t = −2.0). It is a diversifier versus US equities. This is unusual and potentially valuable for portfolio construction purposes regardless of alpha.

Options-Implied Positioning

The options market is pricing a bimodal distribution consistent with the doorway state: ≈27% chance below $28, ≈56% chance $28-40, ≈17% chance above $40 over 75 days (risk-premium adjusted). Heavy deep ITM puts ($50-60) at June/September with 3,200+ OI suggest institutional holders own the stock but have bought crash protection through the election.

Call resistance at $40-42 (June, 1,011 combined OI) caps near-term upside. No meaningful put support below $34. The mechanical structure is asymmetric: hard ceiling, soft floor.

Delta — Expectations Gap

Price-Implied Parameters

Gap Table — Ranked by |Delta|

Key Finding

The largest gap is a coin flip (Peru election), not an analytical edge. Peru = 40% of BVN's return variance. The equity price embeds ≈55-60% benign probability, which is close to an uninformed estimate of 55%.

Stripping out the election, the remaining gaps are small and offsetting:

• Positives (Cerro Verde hidden value, reserves, dore margin shift) are latent or distant.
• Negatives (silver escalator, cost inflation, SUNAT) are real but each individually small.
• The net is approximately zero. The price is roughly correct for the information set.

Consensus vs Management vs Research

The sharpest disconnect: consensus likely models silver revenue linearly while the escalator creates a concave payoff. But this is offset by consensus using stale gold price decks below spot. The two errors approximately cancel, producing a flat EPS estimate for the wrong reasons on both sides.

Key Risks

1. Peru Political Risk (40% of variance)

Six presidents in ten years. April 2026 elections are imminent. An anti-mining government could impose higher royalties, windfall taxes, or revoke/delay permits (Trapiche most exposed). The market prices some discount but may underweight the tail — a hostile outcome could de-rate BVN to 8-9x on current earnings.

2. San Gabriel Execution (largest idiosyncratic risk)

First dore bar produced December 2025. But FY2026 guidance cut 31-39% (from 70-90k to 48-55k oz) in four months after a ventilation accident in late December. Management disclosed zero details about the accident — type, cause, casualties, or regulatory investigation. The 3/6 level restriction means mining at ≈2.3-2.7 g/t versus 3.71 g/t reserve average. Every 0.5 g/t deviation costs ≈$50M in annual gold revenue.

The mine uses underhand drift-and-fill in a massive carbonate replacement body (3,000m x 250m x 170m). This method is inherently ventilation-intensive — dead-end headings trap blasting gases. "Triple the air pressure" means more fans, ductwork, electrical capacity — capital and operating cost that wasn't in the original budget.

3. Orcopampa Closure (46% of gold production)

The 20-F explicitly states Orcopampa has approximately one year of mine life remaining. It produced 55,632 oz gold in FY2025 — 46% of total. San Gabriel at 48-55k oz barely replaces it. Until San Gabriel reaches 120k oz/yr steady state (≈2028), BVN faces a gold production trough.

4. Silver Escalator (concave payoff above $50/oz)

Concentrate offtake contracts include proportional deductions above $50/oz silver. The exact percentage is not disclosed — the single largest information gap. At $75 silver, blended realization falls to ≈66% vs ≈79% at $30. Q4 2025 silver CAS rose 616% QoQ. The market almost certainly models silver revenue linearly; reality is concave.

5. Metal Price Dependence

FY2025 revenue was 80%+ price-driven. At $3,000 gold (a 40% decline from spot), Julcani becomes unprofitable, Tambomayo becomes marginal, and FCF collapses. The portfolio is only robust at $2,500+ gold. The dividend ($0.99/share at 40% payout) requires sustained elevated metal prices — it was $0.03/share in FY2023.

6. Governance and Succession

No insider buying at 11x P/E. CEO and CFO own zero shares. Chairman Benavides is 70 years old with 46 years on the board and no disclosed succession plan. Compensation Committee is not fully independent (Chairman sits on it). Peruvian corporate governance standards offer limited minority protections. Van Eck, the largest institutional holder, is reducing its position.

What to Watch

Natural decision point: April 29, 2026. Q1 earnings and Peru election resolution collapse the two largest uncertainties simultaneously. The information value of the next 26 days exceeds the opportunity cost of early positioning.

Steelman Bear Case

The strongest argument against BVN is not Peru or governance — the market already prices those. The steelman is this: BVN is a procyclical cost trap masquerading as a cheap miner.

The math: operating margin declined from 38.6% to 36.6% despite 50% revenue growth. Worker profit-sharing (8% of PBT, Peruvian law), progressive mining royalties (1-12%), silver escalator deductions, sol appreciation, and structural gold cost inflation (+16-26%/yr at legacy mines) combine to create a system where 12-15% of every incremental dollar of metal price appreciation leaks to government and labor before shareholders see it.

This means BVN's earnings leverage to metal prices is substantially lower than headline P/E suggests. An investor buying BVN at 11x P/E expecting 18-24x sector-peer upside is implicitly betting that either (a) metal prices stay elevated forever, or (b) the discount closes for non-fundamental reasons. But the discount is structural — it reflects real, quantifiable, legally mandated margin leakage that doesn't exist for NEM, AEM, or other major gold miners operating in more favorable jurisdictions.

Additionally, Orcopampa (46% of gold production) closes within a year, San Gabriel is ramping at 27-38% below reserve grade, and the CEO owns zero shares. The market isn't wrong to discount this stock — it may be too generous.

This argument is partially conceded by the data. The margin compression is real and structural. The procyclical cost leakage is legally mandated and cannot be managed away. The gold production trough (2026-2027) is a timing gap with real P&L impact.

Where the bear case weakens: it assumes metal prices mean-revert. If gold holds above $4,000 and copper above $10,000/t, BVN generates $400-600M/yr in FCF even after procyclical leakage. The balance sheet is clean. Cerro Verde throws off $180-200M/yr in equity income. The procyclical costs are a tax on the upside, not a threat to solvency. The bear needs metal prices to break for this to become a capital-destructive story.

Kill Criteria

For coverage monitoring (not a position — no trade thesis established):

Escalate to BEARISH if:
- Peru election produces anti-mining government with royalty increase >5pp
- San Gabriel Q1 2026 production < 8,000 oz (execution failure)
- Gold sustains below $3,500 for >60 days (breaks FCF model)
- SUNAT rules against BVN on full $454M (3.5% mkt cap hit)

Escalate to BULLISH if:
- Peru election benign AND San Gabriel Q1 > 15,000 oz
- Insider buying by Benavides family or CEO (governance shift)
- Trapiche column test results disclosed with viable recovery rates
- Gold re-rates above $5,500 with BVN still at 11x (pure factor lag)

Coverage maintained NEUTRAL until:
- Peru election resolves (April 2026)
- San Gabriel Q1 data available (April 29, 2026)

LR Signal

LR = 0.95 — Mildly bearish. Analysis reveals structural negatives (procyclical cost trap, silver escalator concavity, gold production trough, governance deficiencies) that slightly outweigh structural positives (Cerro Verde hidden value, reserve quality, dore margin shift). The largest variable (Peru election) is a coin flip where we have no edge. The net effect is a minor negative revision to the prior — the stock is approximately fairly priced but the undisclosed risks (escalator magnitude, San Gabriel grade dilution) tilt slightly bearish.

This is not a high-information memo. The analysis confirms market pricing rather than diverging from it. The discount is earned across every dimension. The optionalities are real but latent.

Evidence