Berkshire's 2025 10-K is the last annual report with Warren Buffett as CEO. He signed it as Chairman. Greg Abel signed it as CEO. The filing captures one era ending and another beginning — and the two men are already telling you different things.

Two Animals

Buffett's 2025: Zero buybacks. Net equity seller ($13.8B). Built cash to $369B — 34% of market cap sitting in T-bills. Seven consecutive quarters without repurchasing a single share. The clearest "market is expensive" signal the most respected allocator alive has ever sent.

Abel's first 63 days: Resumed buybacks on March 4 (8-K filed March 5). Committed $15M/year of personal stock purchases for twenty years. Called out BNSF's operating ratio gap to Union Pacific — 5.7 points, "$230M per point" — as unacceptable. Closed OxyChem for $9.7B.

The cash hoard wasn't Berkshire policy. It was Buffett at 95 being Buffett at 95. Abel's first act was to start spending it.

Abel's Playbook: Industrial Operator, Not Financial Investor

OxyChem tells you everything. Top-3 North American PVC and chlor-alkali producer. 21 plants. ≈4,000 employees. $9.7B all-cash, no material goodwill. At the time of the deal, every public chemical peer was unprofitable:

Abel bought a cyclical trough in a commodity chemical business from a motivated seller (OXY deleveraging to hit <$15B debt target). No premium. No synergy promises. No press conference. Just a check for a cash-generating industrial asset trading below replacement cost while the entire sector can't earn its cost of capital.

This is not Buffett-style financial engineering (Oxy preferred at 8%, warrants at $59.59). This is an industrial operator buying a factory. Different animal.

The Crack: PacifiCorp Is Tracking PG&E

The filing contains language I've never seen from a Berkshire subsidiary:

"PacifiCorp may be unable to obtain the necessary funding to meet its liquidity needs."

That's going-concern-adjacent from a subsidiary of the most solvent company on Earth.

The numbers are bad and getting worse. The 10-K disclosed $2.85B accrued, $1.2B unpaid, ≈1,500 plaintiffs in trial queue under CMO No. 11. Then February happened:

• Feb 22: $575M federal settlement (six Oregon/California wildfires)
• Feb 26: $305M jury verdict (16 plaintiffs, 15th trial to conclude)
• Total February: ≈$900M in new liability in one month

Trial verdicts have exceeded $1B cumulative. 167 more trials scheduled through 2027. Another 1,400 plaintiffs queued through 2028. Bond math: $606M posted for 109 adjudicated plaintiffs (≈$5.6M average). Extrapolate to 1,500 and you're looking at $8.4B in potential bond exposure.

PacifiCorp is already selling assets — Washington state operations to Portland General for $1.9B. Insurance recoveries exhausted ($530M, all received). Oregon has no wildfire fund (California has SB 254). The Oregon Supreme Court petition is the key binary — filed November 2025, decision expected 2026.

For context: PG&E's $30B wildfire liability forced bankruptcy. PacifiCorp can't go bankrupt — it's a Berkshire subsidiary. Which means either Berkshire parent absorbs it, rate increases get pushed through (politically toxic), or asset sales continue until there's nothing left to sell.

Berkshire explicitly does NOT guarantee BHE or PacifiCorp debt ($59.3B outstanding). The ring-fence is legal, not reputational.

The Moat Is Real

Insurance float: $176B at negative cost. Up from $138B five years ago. That's $38B in free permanent leverage growth while being PAID to hold it. Pre-tax underwriting profit: $9.5B. Statutory surplus: $333B. No P&C competitor on Earth has comparable scale.

But GEICO is losing the growth war. Premium growth 5.3% versus Progressive's 12%, despite a 34% expense surge ($1.9B in advertising). Progressive's bundled auto+home strategy is winning — 4.2 million policies added, 15% policy count growth. GEICO is spending more to grow less. The moat is float scale, not competitive positioning.

The Macro Signal

Forget BRK-A for a second. The $369B cash pile is the most expensive macro opinion in corporate history. One man put $369 billion behind "equities are overvalued."

He's not alone. Druckenmiller (March 1, 2026): rotated to 25-30% gold allocation, fully exited AI, called "high probability" of flat US equities for ten years. S&P 500 buyback breadth at a 10-year low — only 34 companies announced buybacks in Q3, participation dropped from 76.8% to 66.6%. Shiller CAPE ≈39. Five prior instances above 30 all preceded >20% declines.

Two of the most credible capital allocators alive independently saying the same thing. The aggregate buyback number ($1T trailing) masks concentration — mega-caps buying back while smaller companies quietly pull back. Record headline, deteriorating breadth.

Abel's buyback resumption muddies the signal slightly. But he started nibbling at 1.50x book — not backing up the truck. The cash pile is ammunition, not policy.

What's Priced, What's Not

BRK-A at $748K: 1.50x book, ≈24x operating earnings, consensus NEUTRAL. Three analyst ratings total (UBS Buy $866K, KBW Underperform $695K). Beta 0.69, 80% idio variance, RSI 48 — middle of everything.

Priced in: Smooth succession (yes), fortress balance sheet (yes), steady operating earnings (yes).

Underpriced: PacifiCorp trajectory. The 10-K accrual of $2.85B already looks stale given February's $900M. The bond extrapolation to $8.4B isn't in anyone's model. And Abel's operational discipline — BNSF has a $1.3B opportunity in closing the OR gap to UNP, and Abel is the first CEO to say so publicly.

Overpriced: Nothing obvious. BRK-A is efficiently priced for what the market knows. The question is whether the market is correctly weighting PacifiCorp tail risk against Abel's operational upside.

Bottom Line

This is a filing about a regime change, not a business change. Operating earnings are flat. The balance sheet is a fortress. The moat is intact. What changed is WHO makes the decisions and HOW they think about capital.

Abel is more aggressive, more operational, and more willing to deploy than 95-year-old Buffett. OxyChem proves he'll buy cyclical troughs. The buyback resumption proves he disagrees with Buffett's extreme caution. The BNSF call-out proves he'll challenge subsidiaries.

PacifiCorp is the fly in the ointment — and it's growing. $900M in one month, with 1,400+ plaintiffs still waiting. At $369B cash, Berkshire survives anything. But "survives" and "unimpaired" aren't the same thing.

The real alpha from this filing isn't in BRK-A. It's the macro read: when the most patient capital allocator in history and the best macro trader alive both run to cash and gold simultaneously, the signal-to-noise ratio is unusually high.