Where the Graham Number comes from
Graham proposed that a defensive investor should never pay more than 15 times earnings, and never more than 1.5 times book value. Multiply those two limits — 15 × 1.5 = 22.5 — and you get the constant in the formula: the square root of 22.5 × EPS × book value per share.
The square root is what makes it work. It takes the geometric mean of the two constraints, so a company must clear both bars to score well. A business with excellent earnings but almost no book value — most software companies — scores low, and so does an asset-heavy business earning nothing. That is deliberate: Graham wanted a margin of safety on both sides of the balance sheet.
Treat it as a filter, not a price target. It ignores growth entirely, which is why almost no high-quality growth stock passes it today. It works best on the boring, asset-backed, profitable middles of the market — which is exactly what Graham designed it for.
What the Graham number was built for
Benjamin Graham published this formula in the 1949 edition of The Intelligent Investor as a shortcut for defensive investors: multiply earnings per share by book value per share, take the square root, and scale by the product of his maximum acceptable P/E (15) and maximum price-to-book (1.5). It produces the highest price a conservative investor should pay for a company meeting both limits at once.
That context matters. Graham was writing for investors who wanted protection first and would happily forgo upside to get it. His screen assumed industrial companies with tangible assets on the balance sheet — factories, inventory, equipment — where book value meant something real.
Where it stops working
Asset-light businesses. Software, consulting and platform companies have little book value because their assets are people, code and brand, none of which sit on the balance sheet at anything close to their economic worth. Their Graham numbers come out near zero and stay there. This is a limitation of book value as a proxy, not evidence the stock is overpriced.
Banks and insurers. Here the problem runs the other way. These businesses hold financial assets marked close to market value, so book value is unusually meaningful — but they also run on leverage, which the formula ignores entirely. Use price-to-tangible-book for financials.
Companies with negative or depressed earnings. The square root of a calculation that produces a negative product is undefined, and a company in a temporary earnings trough produces a Graham number far below any sensible estimate of intrinsic value.
How to use the result
Treat the Graham number as a floor, not a target. A price below it means the stock passes one conservative test — it says nothing about growth, competitive position or whether earnings are sustainable. Graham himself paired the calculation with requirements for dividend payments, debt levels and earnings stability over ten years, all of which were filters against exactly the kind of value trap that a low multiple alone will happily walk you into.
Formula
Graham Number = √(22.5 × EPS × Book Value per Share)
FAQ
What is a good Graham Number result?
Graham wanted a 30% to 50% margin of safety — meaning the market price should be that far below the Graham Number before he would buy. A stock trading exactly at its Graham Number only meets the bare minimum, with no cushion for error.
Why does the formula use 22.5?
It is 15 × 1.5: Graham's twin ceilings of 15 times earnings and 1.5 times book value. The square root then blends the two into a single price limit rather than requiring both tests separately.
Does it work for tech and growth stocks?
Rarely. Companies with little tangible book value — software, services, anything built on intellectual property — fail the book value test even when their economics are excellent. The Graham Number was designed for industrial-era balance sheets and still suits banks, insurers, manufacturers and REITs.
Should I use tangible book value instead?
Strictly, yes — Graham was interested in assets that would survive a liquidation, and goodwill and intangibles would not. Most data sources report total book value, which is more conservative in the wrong direction. Using tangible book value gives a stricter, more faithful number.
Related calculators
- DCF Calculator — fair value from five years of cash flow plus terminal value
- Dividend Discount Model — Gordon Growth valuation for dividend payers
- Two-Stage DDM Calculator — high-growth then stable dividend valuation
- Reverse DCF Calculator — the growth rate today's price already assumes
More calculators
- WACC Calculator — weighted average cost of capital for discounting
- EPS Calculator — basic and diluted earnings per share
- P/B Ratio Calculator — price against book value, with ROE context
- Margin of Safety Calculator — discount between value and price
- Compound Interest Calculator — future value with periodic contributions
- DuPont ROE Calculator — break return on equity into its three drivers
- Position Size Calculator — how many shares to buy for a set risk
- CAGR Calculator — compound growth rate and forward projections
- Dividend Yield Calculator — yield, income per $10,000, capital needed
- Payout Ratio Calculator — dividend safety and sustainable growth
- P/E Ratio Calculator — multiple, earnings yield, PEG, target price
- Total Return Calculator — price change plus dividends, annualized
- EV/EBITDA Calculator — enterprise value against operating earnings
Guides
- How to Value a Stock — the full valuation sequence, with a worked example
- Free Cash Flow Explained — the input every DCF depends on
- Margin of Safety Explained — why the discount is not optional
- Dividend Investing Guide — yield, payout ratio and dividend safety
- How to Read a Cash Flow Statement — three sections, eight combinations
- Intrinsic Value vs Market Price — why the two diverge, and how to behave
- How to Read a Balance Sheet — the asset side, and the four ways it misleads
- What Is a Good P/E Ratio? — the arithmetic behind a justified multiple
- Dividend Yield vs Total Return — why a high yield is not a high return
- How to Compare Two Stocks — a five-axis framework, worked side by side
- WACC Explained — the discount rate, built from five sourced inputs
- The DCF Method Explained — the forecast half of a DCF, and checking the terminal value
- The Graham Number Explained — Graham's ceiling tested on two companies, and what it cannot see