Southern Copper's FY2025 10-K is one of those filings where the financial statements scream success and the geological data screams trouble. Record revenue ($13.4B), record earnings ($4.3B), net cash cost of $0.58/lb — and production FELL. That contradiction is the story.

What the Filing Actually Says

Every major SCCO mine is declining in grade simultaneously. Not one mine having a bad year — all of them, and it's geological.

MineGrade (2023→2025)Change
Buenavista0.525% → 0.511%-2.7%
Toquepala0.531% → 0.490%-7.7%
Cuajone0.571% → 0.527%-7.7%
Buenavista leach0.248% → 0.218%-12.1% (1yr)

Management guides 2026 at 911,400 tonnes — down 4.7% from 2025. And 2027? "A little north of 900,000 tons." That's the CEO telling you this isn't a trough. It's the new trajectory.

The zinc concentrator at Buenavista makes it worse. When silver trades above $30/oz, economics favor routing ore through the zinc/silver circuit instead of the copper circuit. In 2025: Buenavista zinc surged 82%, Buenavista copper fell 4.3%. Same ore body, different door. This mechanism is permanent and price-driven — it persists as long as silver stays elevated.

Los Chancas (130kt/yr, a top-5 SCCO growth project) is blocked by illegal miners. $43.7M spent in 2025 on community programs, zero operational progress, no timeline offered. That's 130kt of copper that was supposed to come this decade. It's gone until 2031 at the earliest.

Tia Maria is the only real near-term addition: 120kt/yr, 24% complete, 2027 start date if everything goes right. Peru execution history says it won't go right — Tia Maria itself took 15 years of community fights to reach groundbreaking.

Cross-Ticker Corroboration: It's Not Just SCCO

We pulled Q4 transcripts and 10-Ks from every major Chile/Peru copper producer. Grade decline is industry-wide.

ProducerOperationTrendRecovery?
FCXCerro Verde0.34→0.30% (3yr)None stated
RIO/BHPEscondida (≈5% global supply)"Expected grade decline 2026"Unclear
AngloCollahuasi (≈1.7% global)"Much lower grade 2026"2027 (Rosario pit)
AntofagastaLos Pelambres2yr lower, recovering 2026Yes
HudbayConstanciaPampacancha depletingPartial (throughput)

Six producers, all declining, in the same 24-month window. The world's core copper-producing region is moving through lower-grade ore simultaneously. The recoveries are partial and years away — Antofagasta recovers in 2026, Anglo in 2027, FCX never stated one.

Antofagasta's CEO said the quiet part: new greenfield grades are "materially lower than current installed capacity." The mines replacing today's production will structurally cost more to operate.

The Options Market Already Figured This Out

Here's where it gets interesting. The options market is doing the factor decomposition in real time:

  • COPX (copper miners ETF): P/C ratio 0.45 — calls 2.2x puts. Bullish.
  • SCCO specifically: P/C ratio 2.92 — puts 2.9x calls. Bearish.

Translation: long copper price, short SCCO relative to it. The market agrees the supply thesis is real. It just says SCCO at 40x P/E with the worst grade profile in the industry is the wrong way to play it.

SCCO June max pain sits at $180 — 13% below current. Street mean target is $150 — 27% below. Two directors sold $3M in the last 60 days. Zero insider buying across the entire copper mining sector. Twitter confirmed it: "Copper has absolutely no insider purchases at all."

Producers are selling into $5.90 copper. They see the depletion curves in their own mine plans.

Factor Decomposition

Nine factors drive SCCO returns. Edge on almost none:

FactorWeightOur Edge
Copper price60-65%None — consensus
Grade decline10-15%Weak — sell-side sees it
Byproduct economics10-15%Moderate — cross-factor coupling
Project pipeline5-10%None — public
China demand10-15%Weak — was edge in 2024
Jurisdiction/governance/FX/tariffs15-20%None

80%+ of SCCO returns are in factors where we have zero informational advantage. $170B market cap, 16 analysts, +155% in 12 months. Not our edge zone.

The One Genuine Insight: Bidirectional Silver-Copper Coupling

Standard models treat silver supply as a one-way function of copper production. The SCCO 10-K reveals the reverse: copper supply is ALSO a function of silver/zinc prices.

When silver rises above $30, SCCO's Buenavista diverts ore to the zinc/silver circuit. Copper output falls. Copper tightens. Copper price rises. More copper mining globally. More silver byproduct. Silver stabilizes. Diversion reverses.

This is a negative feedback loop that dampens copper supply response to high prices. Nobody models it because copper analysts and silver analysts sit in different teams at every bank. The coupling is bidirectional and self-reinforcing — silver supply is even more tightly constrained by copper economics than linear models predict, AND copper supply is constrained by silver economics.

This mechanism is confirmed SCCO-specific (no other producer has an integrated zinc/copper concentrator on the same ore body routing ore based on relative prices). It affects ≈4% of global copper supply. Material, not dominant. But it's a structural feature, not a one-time event.

Predictions

CallProbDeadline
SCCO misses 911kt production guidance65%Feb 2027
SCCO underperforms COPX (12mo total return)60%Mar 2027
Tia Maria slips past 2027 production start50%Jan 2028
LME copper holds above $4.50/lb through 202670%Jan 2027

Verdict

SCCO: PASS. Copper-supply factor: CONFIRMED AND STRENGTHENED.

The filing's value is in the factor knowledge, not a stock thesis. The copper supply constraint is real, industry-wide, and worsening. But it's consensus — that's what +155% in 12 months means. The options market, the sell-side, and the insiders all agree: copper good, SCCO overpriced.

The residual insight flows to silver (SLV watchlist — byproduct coupling confirms structural deficit from the producer side) and to the copper-supply factor base (31 evidence items, 7+ tickers, institutional quality — ready to deploy if a vehicle with idio edge ever materializes).

The filing told us something the market partially knows (grade decline), something the market has priced (record earnings at record copper), and something neither copper nor silver analysts model correctly (the bidirectional coupling). Only the third one is information.