Warren Buffett's Intrinsic Value

Yanik•04/19/2026•11 min read

Key Takeaways

Intrinsic value is the discounted value of all cash an owner can extract from a business over its remaining life
The inputs (owner earnings, growth rate, discount rate) matter far more than the formula itself
Margin of safety is non-negotiable. You buy at a discount to intrinsic value, not at intrinsic value
The tool below pulls historical filings automatically so you can test Buffett's framework on any public company

Formula

OE_t = NI_t + DA_t + SBC_t - Capex_t

V = \sum_{t=1}^{n} \frac{OE_0 \cdot (1+g)^t}{(1+r)^t} + \frac{1}{(1+r)^n} \cdot \frac{OE_0 \cdot (1+g)^n \cdot (1+g_\infty)}{r - g_\infty}

\text{Fair value / share} = \frac{V}{\text{Shares}} \qquad \text{Buy below} = \text{Fair value} \cdot (1 - \text{MoS})

OE: Owner earnings. Buffett's 1986 definition of distributable cash: net income plus non-cash charges minus capex. Closer to economic reality than GAAP earnings.
NI: Net income, bottom line of the income statement.
DA: Depreciation & amortization. Non-cash charge added back because no cash actually left the business.
SBC: Stock-based compensation. Non-cash from an accounting view, but a real cost to shareholders, added back here and then priced back in through share count.
Capex: Capital expenditures. Subtracted in full (FMP filings don't split maintenance vs. growth capex, so this is slightly conservative).
g: Growth rate applied to owner earnings for years 1 through n. Default is the historical CAGR of owner earnings.
g∞: Terminal growth rate beyond year n. Should not exceed long-run GDP growth (~2-3%).
r: Discount rate (required return). Proxies opportunity cost; Buffett uses the long-bond yield, adjusted upward.
n: Forecast horizon in years before the terminal value kicks in.
MoS: Margin of safety. The discount to fair value you demand before buying.

Two simplifications worth knowing

Buffett's full 1986 owner-earnings definition also subtracts changes in working capital required to maintain competitive position. This calculator drops that term. For mature, working-capital-light businesses (Coke, software) the impact is small. For working-capital-heavy growers it can swing owner earnings by 10 to 20 percent.

SBC is added back as a non-cash charge and then re-priced through diluted share count. That is defensible in theory, but if you do not project diluted shares forward rigorously this systematically inflates fair value for SBC-heavy companies. A safer alternative is to subtract SBC directly from owner earnings and treat it as the real economic cost it is.

What this calculator cannot do

The default growth rate is the historical CAGR of owner earnings. That is a backward-looking number. Run Apple in 2004 through this and you get a fair value of roughly 16 cents against a price of 67 cents (split-adjusted), which would have told you to walk away from a stock that returned more than 400x. The calculator is a discipline tool, not an oracle. When the future is structurally different from the past (a new product cycle, a category leader emerging from a turnaround, a regulatory shift), you have to override the historical growth rate yourself, and you need a written reason for doing so. Buffett avoided tech for thirty years for exactly this reason. He only bought Apple in 2016 once the cash flow profile looked like Coke.

The One Idea That Matters

Strip away the folksy quotes, the cherry Coke, and the annual letters. What is Warren Buffett actually doing when he buys a business?

He is comparing two numbers. Price is what the market is asking today. Intrinsic value is what the business is actually worth. If the gap is large enough in his favor, he buys. If not, he waits. That is the entire operating system.

Everything else (moats, management, circle of competence) exists to make the second number more reliable.

Intrinsic value can be defined simply: It is the discounted value of the cash that can be taken out of a business during its remaining life.

Warren Buffett, Berkshire Hathaway Owner's Manual

Notice what this definition does not say. It does not mention earnings, book value, P/E ratios, or DCF spreadsheets. It says cash that can be taken out. Not cash the business generates, but cash the owner can actually extract without damaging the earning power that produces future cash. That distinction is where most investors get lost.

Where the Idea Comes From

Buffett did not invent intrinsic value. He inherited it from Benjamin Graham, his professor at Columbia, and refined it with Charlie Munger over six decades.

Graham's version was quantitative and defensive. Buy net-nets, companies trading below liquidation value, and you cannot lose much. Munger pushed Buffett toward a different view: it is better to buy a wonderful business at a fair price than a fair business at a wonderful price. That shift, from cheap-and-mediocre to great-and-reasonable, is what made Berkshire Berkshire.

The common thread is this. Value is not what the market says it is. Value is what the business will produce for you, in cash, over its remaining life. The market price is a separate question entirely.

Mr. Market

Graham's metaphor, which Buffett still quotes in nearly every letter. Imagine a manic-depressive business partner who shows up daily offering to buy your share or sell you his, at whatever price his mood dictates. His price tells you nothing about the business. It tells you only about his mood. Your job is to exploit him, not listen to him.

The Definition Has No Formula

This is the part people hate. There is no Buffett Intrinsic Value Formula. Buffett has said so explicitly, every year, for forty years.

The calculation of intrinsic value, though, is not so simple. As our definition suggests, intrinsic value is an estimate rather than a precise figure, and it is additionally an estimate that must be changed if interest rates move or forecasts of future cash flows are revised.

Berkshire Hathaway Owner's Manual

Two investors looking at identical data will produce different intrinsic values. That is not a bug. It reflects genuine uncertainty about future cash flows. The honest answer is a range, not a point. Anyone quoting a single number to three decimal places is either selling something or fooling themselves.

The goal is not precision. The goal is to be approximately right rather than precisely wrong.

The Math Buffett Actually Uses

Strip the mystique away and intrinsic value is a discounted cash flow. The formula is 200 years old:

V = \sum_{t=1}^{n} \frac{CF_t}{(1+r)^t} + \frac{TV}{(1+r)^n}

Where CF_t is the cash flow in year t, r is the discount rate, and TV is the terminal value. Every valuation model in existence is a variation on this equation.

What makes Buffett's application different is not the math. It is the inputs, and the ruthlessness about admitting when he cannot estimate them.

Input 1. Owner earnings, not net income

Buffett coined the term owner earnings in his 1986 letter. It is closer to reality than either GAAP earnings or reported free cash flow:

PLAINTEXT

1Owner Earnings = Reported Earnings
2               + Depreciation, depletion, amortization
3               + Other non-cash charges
4               - Average annual maintenance capex
5               - Working capital required to maintain competitive position

The critical line is maintenance capex, what the business must spend just to stand still rather than to grow. Most companies do not disclose this separately. They lump maintenance and growth capex together, which flatters reported free cash flow. Estimating the split is judgment, not arithmetic.

Why this matters

A railroad that reports $5B in free cash flow but needs $4B annually to replace aging track is not a $5B cash machine. It is a $1B cash machine that will stop producing anything the moment the track degrades. Owner earnings force you to face that.

Input 2. A discount rate that reflects opportunity cost

Buffett has described his discount rate two ways across the letters. Sometimes as the long-term US Treasury yield, adjusted upward when rates are artificially low. Other times as a fixed hurdle of roughly 10 percent regardless of where Treasuries trade. The exact number matters less than the logic. If you can earn the long bond yield risk-free, why would you accept anything less from a business?

He does not add an equity risk premium the way textbooks do. His view is that risk is not volatility. Risk is permanent loss of capital. If you understand the business and buy with a margin of safety, you have handled risk at the source. This is the part of the framework people miss. Without an equity risk premium, the moat analysis is the risk adjustment. Skip the qualitative work and the whole approach collapses.

This is controversial. Most academic finance rejects it. Most practitioners quietly use it anyway, because beta does not actually predict permanent capital loss.

Input 3. A growth rate you can defend in a court of law

The tempting move is to plug in 15% growth for ten years because the company grew 15% for the last three. Buffett's discipline is the opposite. Assume regression to the mean unless you can articulate exactly why this business is structurally different.

Moat analysis is not a vibe. It is the justification for the growth rate in your DCF. If you cannot write three paragraphs explaining why this company will still be earning these returns in ten years, you do not have conviction. You have an extrapolation.

The Margin of Safety

Even if you do the math correctly, you will be wrong. Interest rates will move. Management will stumble. A competitor you did not see will appear. The business model will face a headwind you could not have anticipated.

The margin of safety is Graham's answer to this permanent uncertainty. You do not buy at intrinsic value. You buy at a large discount to intrinsic value, so that you can be wrong about the inputs and still not lose money.

The three most important words in investing: margin of safety.

Warren Buffett, paraphrasing Benjamin Graham

How much margin? Buffett's historical pattern suggests roughly 30 to 50 percent below his central-case estimate of intrinsic value. For a great business with predictable earnings, he will accept a smaller discount. For a cyclical business or one with macro sensitivity, he wants more.

The bridge analogy

Buffett often says: if you are building a bridge that trucks weighing 9,800 pounds will cross, you do not rate it for 10,000 pounds. You rate it for 15,000. Margin of safety is engineering discipline applied to capital.

A Worked Example. Coca-Cola, 1988

Berkshire began accumulating Coca-Cola in mid-1988 and continued through 1989, paying roughly $2.45 to $2.75 per split-adjusted share across the program. Critics called it expensive at about 15x earnings. Buffett called it cheap. What did he see?

Owner earnings. Coke generated roughly $0.36 per share in pre-tax owner earnings in 1988, with maintenance capex well below reported depreciation. Actual distributable cash was higher than GAAP earnings suggested.

Durability. The brand, the distribution network, and the emerging-markets runway were all still decades ahead. The moat was not just wide. It was widening. A 10 to 15 percent long-term earnings growth rate was defensible, not aspirational.

Discount rate. The 30-year Treasury was around 9% in 1988. Buffett's required return was roughly equivalent, with the understanding that a great business growing at double-digit rates would comfortably exceed that hurdle.

Running those numbers through the DCF produced an intrinsic value well above the purchase price, even with conservative assumptions. By 1998, the position was worth over $13B on a $1.3B cost basis. The math was not clever. The discipline to trust the math and size the position accordingly was.

What the example does not say

Buffett got Coke structurally right and tactically early. He also held through a decade of flat returns in the 2000s when Coke's valuation normalized. Intrinsic value is a long-duration concept. If you cannot hold for a decade, the framework is not for you.

What Intrinsic Value Is Not

Equally important is what this framework explicitly rejects:

Not a relative valuation. P/E vs sector, EV/EBITDA vs peers. These tell you how the market prices this stock relative to other stocks. They say nothing about what the business is actually worth. Two expensive stocks do not make one cheap.
Not book value. Buffett writes it on the cover of every annual report, but he is explicit that book value and intrinsic value are different things. Book value is a snapshot of historical cost. Intrinsic value is a forecast of future cash.
Not a price target. A sell-side analyst's 12-month price target is a prediction about Mr. Market's mood. Intrinsic value is a statement about the business itself, independent of when (or whether) the market agrees.
Not a single number. Anyone who tells you the intrinsic value of Apple is $231.47 is selling confidence they do not have. A range, say $180 to $260, is the honest answer.

How to Apply This Without Pretending to Be Buffett

You are not going to buy 10% of Coca-Cola. You do not have Buffett's float, his network, or his sixty years of pattern recognition. What you can borrow is the discipline.

The minimum viable version

Stay inside your circle. If you cannot explain in plain English how the business makes money and what would kill it, your intrinsic value estimate is fiction. Skip it.
Estimate owner earnings, not headline EPS. Strip out non-cash charges. Subtract honest maintenance capex. Adjust for stock-based compensation, since it is a real cost even if cash did not leave the building this quarter.
Build a range, not a point. Bull case, base case, bear case. Each with its own growth assumption and margin assumption. Your intrinsic value is a range across these, not the average.
Demand a margin of safety. If your base case says the business is worth $100, do not pay $95. Wait until it trades at $65 to $70. This is where most investors fail. They understand the concept but refuse to wait.
Revisit when inputs change, not when prices change. A 20% drop in the stock is not a signal to lower your intrinsic value estimate. A 20% drop in owner earnings is.

Reverse-DCF as a sanity check

Rather than forecasting cash flows and solving for value, solve for the growth rate implied by today's price. If NVDA at today's price requires 25 percent annual owner-earnings growth for the next ten years, ask whether that is a reasonable base case or a heroic one. If KO at today's price implies 4 percent growth and the business has compounded at 6 percent for two decades, that is the gap you are looking for. Sometimes acceleration is real. Usually it is not.

Why It Still Works

Intrinsic value is not a trick. It is not proprietary. Graham published the framework in 1934. Buffett has explained it in every annual letter for sixty years. Every value-investing textbook teaches it.

So why does it still work? Because executing it is hard.

Most people want certainty. Intrinsic value gives you a range. Most people want to trade. Intrinsic value rewards holding. Most people want to feel smart when the price goes up. Intrinsic value asks you to ignore the price until it is absurd. Most people want a formula. Intrinsic value demands judgment.

The edge is not in the equation. The edge is in being willing to sit on cash for years waiting for the gap to open, and having done enough homework that when it opens, you know it is real.

You don't have to swing at everything. You can wait for your pitch. The problem when you're a money manager is that your fans keep yelling 'swing, you bum!'

Warren Buffett

Disclaimer

This post is for educational purposes only. Nothing here is investment advice. Always do your own research and consult a qualified financial advisor before making investment decisions.

Sources

Berkshire Hathaway Owner's Manual: Warren Buffett's definitional text on intrinsic value
Berkshire Hathaway 1986 Shareholder Letter: Introduction of the "owner earnings" concept
Berkshire Hathaway 1992 Shareholder Letter: Intrinsic value vs book value discussion
The Intelligent Investor by Benjamin Graham (1949): Origin of margin of safety and Mr. Market
Security Analysis by Graham and Dodd (1934): The foundational text on intrinsic value