Warren Buffett Retired Having Not Purchased His Favorite Stock -- a Company He Spent $78 Billion Buying Over 6 Years -- in 19 Months

Warren Buffett was a net seller of equities for 13 consecutive quarters (Oct 1, 2022–Dec 31, 2025), selling nearly $187 billion, and he had spent about $78 billion repurchasing Berkshire shares from July 2018–June 2024 but made no buybacks in the final 19 months before retiring on Dec. 31, 2025. Successor Greg Abel recommenced share repurchases on March 5 and Berkshire holds $373.3 billion in cash, cash equivalents and U.S. Treasuries; Class A shares traded at roughly a 44% premium to book on March 2 (a two-year low). Net: disciplined, valuation-driven capital allocation is likely to continue under Abel, which could modestly influence Berkshire’s share price but is unlikely to have broad market impact.

Analysis

Berkshire's shift from prolonged net selling to renewed repurchases is an idiosyncratic supply shock: buybacks mechanically shorten float, amplify per-share economics of its operating businesses, and make Berkshire itself a progressively more concentrated, closed-end vehicle for its legacy businesses. That concentration changes the investable universe for institutional allocators — managers who underweight private-like cash flows (insurance, rail) may find Berkshire an increasingly efficient proxy, reducing demand for some small-cap asset-intensive names and raising competition for high-grade deal flow.

The immediate market effect will be episodic rather than continuous. Repurchase announcements are binary catalysts that compress volatility and can trigger short-term squeezes, but multi-year returns still depend on the valuation backdrop when repurchases are executed and on underlying operating ROIC. A durable outperformance scenario requires repeated opportunistic repurchases at buyer-friendly prices or accretive M&A; absent that, repurchases simply reallocate returns to remaining shareholders without altering enterprise economics.

Key risks are governance drift (management stretches “intrinsic value” to justify expensive buybacks), a continued frothy market that keeps buybacks idle, or a macro shock that forces capital deployment into distressed acquisitions at unattractive multiples. Time horizons split: watch filings and repurchase notices for day-to-week alpha, but treat any BRK position as a multi-year capital-allocation bet tied to compounding and share-count dynamics.

The consensus focuses on headline repurchases; it underappreciates the interaction between decreasing float and passive index demand, which will mechanically raise tracking error for ETFs and create a feedback loop of higher liquidity concentration — a structural factor that should widen idiosyncratic return dispersion across insurers and capital-intensive conglomerates.

Key Decisions for Investors

• Long BRKB (or BRKA if size permits): initiate a 2–4% NAV position, tranche: 33% now, 67% on a 5–12% price dip. Hold 12–36 months. Target: 20–35% total return if repurchases accelerate and operating ROICs hold; cut to 50% position and stop at a 12% loss if management signals aggressive high-multiple M&A.
• Event-driven trade: small, tactical long (1% NAV) ahead of quarterly repurchase filings; hold 3–21 trading days. Use call spreads (12–24 month expiries) to cap downside if preferred — aim for 2–3x potential upside on a positive repurchase-size disclosure while limiting max loss to premium paid.
• Relative-value pair: long BRKB vs short NVDA (equal notional, hedge ~60–70% delta) for 3–12 months to isolate capital-allocation alpha vs secular momentum. Rationale: capture mean-reversion of value premium if buybacks persist; risk: extend stop if NVDA sustains earnings-led repricing, cap pair loss at 15% of NAV.
• Liquidity/flow hedge: reduce exposure to small insurers or regional banks with similar asset profiles by 30% and reallocate into BRKB when implied buyback yield > dividend-equivalent threshold. Timeframe: 6–24 months; reward is lower beta and higher idiosyncratic optionality, with a downside if market keeps compressing book multiples.