Berkshire Hathaway: 26 Years of Intrinsic Value vs. Market Price
Across 107 quarterly observations spanning more than 26 years, Berkshire Hathaway's market price has tracked estimated Intrinsic Value with remarkable consistency. Here is what 26 years of fundamental data reveals.
Most valuation debates are about the future: growth rates, margins, multiples. This article looks backwards instead, at 107 consecutive quarters of Berkshire Hathaway data from Q4/1999 to Q2/2026, and asks a simpler question: what happens when you strip out every growth assumption and compare the market price only with the company’s earning power?
107 Quarters, One Relationship
Over 26 years, Berkshire’s market price (P) and the Intrinsic Value estimated by the Sam Finn Valuation Model (IVP) have moved together with remarkable consistency. Roughly 96% of the variation in the share price moves together with the model’s estimate.
That fit held through the dot-com collapse, the financial crisis, the pandemic and the rate shock that followed. But does it also hold quarter to quarter, when the market swings wildly? That test is in the Historical Validation section of our Methodology.
No Bubble Territory
Across all 107 quarters, Berkshire’s price moved above and below the estimate, sometimes sharply, but it never detached into speculative bubble territory. It kept returning towards equilibrium.
So what does the same model show when market psychology detaches from the business entirely? Our Nike case study tells the opposite story: a multi-year bubble and its deflation.
Buffett’s Own Signal: Share Repurchases
Berkshire has stated that it repurchases its shares only when management believes the price is below a conservatively estimated intrinsic value. Berkshire never publishes that estimate, but its repurchases are reported to the SEC. That gives two independent views of the same question: the model’s Valuation, and management’s actual behaviour.
- 2020–2021: in Q2/2020 the model showed the lowest valuation in 26 years. Berkshire then repurchased about $51.8 billion of its own shares over two years ($24.7 billion in 2020 and $27.1 billion in 2021), its largest buyback period ever.
- 2025: with the price close to or above the model’s estimate, Berkshire repurchased nothing for the full year.
- 2026: as valuation fell back, repurchases resumed, reaching about $14.4 billion in January–September according to Berkshire’s Q3 10-Q.
Across the quarters with detailed repurchase data, the correlation between the model’s Valuation and repurchase amounts was strongly negative (r ≈ −0.70): the cheaper the stock relative to IVP, the more Berkshire bought.
This does not prove that Berkshire uses the same reasoning as the model. But two independent approaches, one built from reported financial data and one from the decisions of Berkshire’s management, point in the same direction.
What This Teaches Investors
The model does not forecast growth. It works from reported financial data and assumes no future growth, which removes the most uncertain input in most valuations.
Berkshire shows that, over long periods, earning power is what the market eventually pays for. Short-term prices can drift, but the gap to intrinsic value tends to close. The practical lesson is the one Berkshire’s own buybacks illustrate: pay attention to the price you pay relative to value, not to the story around the stock.
Historical statistical fit does not guarantee future performance, and nothing here is investment advice. Repurchase amounts are rounded figures based on Berkshire’s SEC filings (SEC EDGAR).