Buffett never published one formula — he revealed a process across 48 years of letters and meetings.
Buffett defined it in the 1994 Owner's Manual and repeated it dozens of times:
At the 1998 annual meeting he simplified it even further:
He also loved Aesop's framing: "A bird in the hand is worth two in the bush." Buffett said the real questions an investor must answer are:
Book value is an accounting artifact. Market price is a vote; intrinsic value is a weigh. He explicitly rejected EBITDA as a "flawed favorite of Wall Street" as recently as 2024.
Buffett introduced "owner earnings" in his 1986 letter as the only cash flow number that truly reflects what an owner actually gets to keep:
The critical distinction is maintenance CapEx vs. growth CapEx. Only the spending required to maintain the current business is deducted — money spent to grow is optional and belongs to future projections.
Once you have normalized owner earnings for the current year, you project them forward — typically 10 years — using a conservative growth rate based on historical performance, not management promises.
The 10-year stream of owner earnings plus a terminal value must be discounted to today's dollars. Buffett's discount rate choice is deliberately simple:
He doesn't use CAPM, beta, or weighted average cost of capital (WACC). His reasoning: if you've done the qualitative work correctly and only buy predictable businesses with moats, the business risk is already dealt with — you don't need to add a "risk premium" on top.
Even after calculating intrinsic value, Buffett never pays it. He inherited this principle from Ben Graham: since your estimate is always uncertain, you need a cushion. Buffett buys at a meaningful discount to his estimate of intrinsic value.
The margin of safety serves two purposes: it protects against errors in your DCF assumptions, and it provides extra upside when the market eventually recognizes true value. For businesses with very wide moats and highly predictable earnings (like See's Candy or Coca-Cola), he accepts a smaller margin. For murkier businesses, he requires a bigger cushion or simply passes.
Buffett adjusts his valuation informally by his confidence level in the cash flow projections. Think of it as a mental probability weight applied to the DCF output:
Buffett does the qualitative analysis first. If a business fails any of these tests, he doesn't bother with the DCF at all. The math is the last step, not the first.
Imagine evaluating a simple consumer brand (think See's Candy style):
| Year | Growth Factor (1.08)^N |
Owner Earnings (₹ Cr) |
Discount Factor 1 ÷ (1.045)^N |
Present Value (₹ Cr) |
|---|---|---|---|---|
| 1 | 1.0800 | 102.60 | 0.9569 | 98.18 |
| 2 | 1.1664 | 110.81 | 0.9157 | 101.47 |
| 3 | 1.2597 | 119.67 | 0.8763 | 104.87 |
| 4 | 1.3605 | 129.25 | 0.8386 | 108.38 |
| 5 | 1.4693 | 139.58 | 0.8025 | 112.02 |
| 6 | 1.5869 | 150.75 | 0.7679 | 115.76 |
| 7 | 1.7138 | 162.81 | 0.7348 | 119.64 |
| 8 | 1.8509 | 175.84 | 0.7032 | 123.65 |
| 9 | 1.9990 | 189.91 | 0.6729 | 127.79 |
| 10 | 2.1589 | 205.10 | 0.6439 | 132.07 |
| Sum of 10-Year Present Values | ₹1,143.83 Cr | |||