Price Is What You Pay, Value Is What You Get
A stock quote tells you what the last buyer paid, not what the business is worth. Graham spent four decades at Columbia and on Wall Street arguing that the two figures drift apart in both directions: optimism pushes prices above anything the underlying cash flows justify, and despair pushes them far below. His answer was to estimate value from fundamentals first, then compare — never to start from the quote.
Intrinsic value, in the strict sense, is the present value of every dollar the business will ever hand its owners: future earnings, dividends, and the terminal stake. Full models discount explicit forecasts year by year and inherit every error in them. The Graham formula trades that precision for robustness, collapsing the whole exercise into three inputs you can verify from public filings in minutes.
Keep the framing straight when you move between tools. This formula prices equity per share — what a share of the earnings stream is worth to you. Whole-company math lives elsewhere: an enterprise value calculator prices the entire firm, debt included, which is the right frame for acquisitions and net-debt-heavy names.
The Graham Formula, Term by Term
The revised formula is V = EPS × (8.5 + 2g) × 4.4 / Y. The 8.5 is the price-to-earnings multiple Graham judged fair for a business with no growth at all — a claim on stable earnings, worth roughly what a mature bond-like equity pays. Each point of expected annual growth, g, adds two points to that multiple, so 5% growth buys an 18.5x multiple and 7% growth a 22.5x multiple.
The EPS input deserves more care than any other. Use trailing diluted earnings per share, stripped of one-off gains, asset sales, and tax quirks — a bad year's figure or an inflated one distorts everything downstream. If earnings swing hard across the cycle, a three-year average restores a defensible baseline. An EPS calculator rebuilds the figure from net income and share count when you only have the raw financials.
The formula has a sibling worth knowing. The classic Graham number calculator computes the balance-sheet ceiling — the square root of 22.5 times EPS times book value per share — derived from Graham's 15x earnings and 1.5x book limits. The two tools answer different questions: the Number asks what a defensive buyer may pay for the assets, this formula asks what the growth trajectory is worth at today's interest rates.
Why the Formula Divides by the Bond Yield
Graham published the original version, V = EPS × (8.5 + 2g), in the 1962 edition of Security Analysis. After the interest-rate turmoil of the early 1970s he appended the correction that matters most today: multiply by 4.4 and divide by Y, where 4.4% was the average AAA corporate bond yield in 1962 and Y is the current yield. The factor ties what investors pay for a dollar of stock earnings to what a riskless-grade bond pays.
The leverage is large. Take $6.00 of EPS at 6% expected growth — a 20.5x multiple, or $123.00 of value at the baseline 4.4% yield. Push the AAA yield to 7% and the value falls to $77.31, a 37% markdown delivered by the bond market alone. In 1981, with AAA yields near 15%, the same earnings were worth about $36 — which is close to what the stock averages of that era actually did.
Pull the input from a real series rather than guessing. Moody's Seasoned AAA corporate bond yield is published free on FRED and has run near 5% in recent years, so 5.0 to 5.5 is a defensible entry today. When you want to compare a specific bond's promised return instead of the index average, a bond YTM calculator prices the yield to maturity on the exact instrument competing for your capital.
Growth Rate: The Input That Moves Everything
Doubling the growth input nearly doubles the answer. Moving g from 5% to 10% lifts the multiple from 18.5x to 28.5x — a 54% jump in intrinsic value from a five-point change in an opinion. No other input in the formula carries that leverage, which is why Graham clamped the value at 15% and why honest growth work matters more than decimal-point precision elsewhere.
Ground the forecast in history. A ten-year EPS CAGR, trimmed toward the industry mean, beats last year's growth rate or a bull-case analyst number; buybacks and one-time margin spikes flatter both. The EPS growth calculator computes the compound rate from two endpoint earnings figures, which keeps the input anchored to what the business actually delivered rather than what a pitch deck projects.
Wall Street's consensus five-year growth estimates run optimistic by a wide margin on average, so haircut them before entry. Graham's own cutoff was blunt: he considered double-digit decade-long projections speculative and built the 15% cap as an honesty mechanism. A business whose bull case requires 20% growth for a decade scores the same as one at 15% here — deliberately, because the difference between those two futures is not forecastable.
Margin of Safety: Turning Value Into a Buy Price
An intrinsic value estimate is a claim about an uncertain future, built from inputs that can each be wrong by a third. Graham's answer was never to pay full price: demand a discount wide enough that ordinary forecast errors still leave you whole. He suggested 25% for stable businesses and closer to 33% for cyclicals, and Buffett still calls the margin of safety the three most important words in investing.
The arithmetic is one multiplication. The default run values the stock at $81.40; a 30% margin sets the buy price at $56.98, meaning the market must hand you the shares roughly a third below fair value before you act. That threshold is the entire output worth acting on — a large gap between value and quote is a prompt for research, never an automatic signal that the market is wrong and you are right.
Choose one lever for conservatism and justify it. Cutting EPS to a mid-cycle average, capping growth at 10%, and demanding a 40% margin all at once can push every buy price below any quote the market ever prints — a discipline so strict it becomes inaction. A different route for cash-flow-based estimates is to raise the return you demand through a discount rate calculator, which lowers value at the source instead of stacking haircuts on top.
How the Formula Compares With DCF and the Dividend Discount Model
A DCF valuation calculator builds intrinsic value the long way: explicit free cash flow forecasts for a decade, a terminal value, and a discount rate to bring it all to present value. It captures the real economics of the business and inherits every error in its assumptions. The Graham formula needs three numbers and one minute; the DCF needs a model and an afternoon. They disagree most on businesses whose cash flows diverge from reported earnings.
The dividend discount model calculator prices the cash shareholders actually receive — next year's dividend divided by the spread between required return and perpetual growth. It suits utilities, telecoms, and consumer staples with decades of payout history, and it breaks on companies that retain most of their earnings. The Graham formula needs no dividend at all, which makes it usable for growers that reinvest everything.
The tools work best as a sequence rather than a competition. Screen a whole watchlist with the Graham formula in minutes, then push survivors through a full DCF and a dividend model before committing capital. Where the three disagree, the disagreement is diagnostic: a high DCF value against a low Graham value usually means the story lives in forecast cash flows rather than current earnings, and that is exactly where valuation mistakes hide.
Worked Examples Across Three Companies
Slow industrial: $3.00 of EPS, 2% expected growth, AAA yields back at the 4.4% baseline. The multiple is 8.5 + 2×2 = 12.5x and the yield factor is 1.0, so intrinsic value is $37.50 per share. At a 30% margin the buy price is $26.25. The lesson: with no growth and no rate penalty, value collapses to a bond-like 12.5 times earnings — cheap by modern market standards, normal by Graham's.
Quality compounder: $8.00 of EPS, 7% growth, 5.5% AAA yield. The multiple is 22.5x, the yield factor 4.4/5.5 = 0.8, and intrinsic value lands at $144.00 — a 30% margin sets the entry at $100.80. Notice the 22.5x multiple is identical to the constant inside the Graham Number; the formulas rhyme even when their inputs differ, which is deliberate on Graham's part.
Cyclical at mid-cycle: $10.00 of normalized EPS, 3% growth, 6% yield gives $106.33 of value and a $63.80 buy price at a 40% margin — the wider cushion fits earnings that can halve in a recession. For whole-company questions — a private acquisition, a partnership stake, a sale of the business — a business valuation calculator extends the same discipline with multiples and cash-flow methods built for deals rather than listed shares.
Where the Formula Breaks Down
Unprofitable companies fail outright — negative EPS multiplied by any multiple produces a meaningless number, and the honest path for loss-makers runs through a full cash-flow model with recovery assumptions. Asset-light technology names with 20%+ growth also sit outside the tool's territory: the 15% cap exists precisely because Graham distrusted decade-long projections above that pace, and a capped multiple will undervalue the rare company that actually delivers them.
Rate shocks distort every output at once. When AAA yields lurch — as they did in 2022, when the series moved from under 4% toward 5.5% in a year — the formula marks down the entire universe of earnings-heavy stocks simultaneously, sound businesses and weak ones alike. That is the mechanism working as designed, but it means comparing values computed at different bond yields is meaningless. For pricing how violently an individual name moves in such repricings, a stock beta calculator measures the volatility that the value formula ignores.
The formula is blind to balance-sheet risk: two companies with identical EPS and growth get identical values even if one carries net cash and the other debt at six times EBITDA. It also prices nothing about earnings quality — receivables build-up, aggressive capitalization, or pension holes all pass through untouched. Run it as a screen with judgment attached, and let the deeper filings have the final word before any position gets opened.