Benjamin Graham spent his career on one question: what is this business actually worth, and am I paying less than that? Enter three numbers below to run his formula and see how much margin of safety you'd be buying. Tap the ? under any field to learn what it means and exactly where to find it.
In plain English: the company's annual profit divided by the number of shares. If a business earned $100m and has 50m shares, EPS is $2 — your slice of one year's profit for each share you own.
Where to find it: the bottom of the income statement in any annual report (10-K) or quarterly filing (10-Q), free at the SEC's EDGAR database. Every stock quote page also lists it, usually labelled "EPS (TTM)" — TTM meaning the last twelve months.
Watch out for: one-off items. A big legal settlement or asset sale can inflate or crush a single year's EPS without telling you anything about the underlying business. If the latest figure looks strange against the last five years, use a normalised average instead.
In plain English: how fast you think profits will grow each year over roughly the next seven to ten years. Not next quarter — the long haul.
Where to start: look at how fast earnings actually grew over the past five and ten years. Past growth is not a promise, but it's a far better anchor than a forecast. If we fetched data, the historical figure appears just above.
Be conservative on purpose. Analyst consensus estimates skew optimistic, and this formula is extremely sensitive to this input — the sensitivity table below shows exactly how much. For most established businesses, something in the 2–6% range is realistic. Above 15% you are making a forecast that very few companies sustain.
Graham's warning: obvious growth prospects don't produce investor profits, because the price already contains them. You get paid for being right about what the price hasn't priced.
In plain English: what one share costs right now.
Where to find it: any quote page, or filled automatically above.
The trap: seeing the price before you've formed a view on value anchors your thinking to it. You'll find yourself reverse-engineering a growth rate that justifies whatever the market happens to be asking. Ideally, settle on your growth assumption first, then reveal the price.
Margin of safety is the gap between what a business is worth and what you pay. If a stock is worth $100 and you buy at $70, you have a 30% margin. That buffer is what protects you when — not if — your estimate turns out to be wrong.
Graham's threshold was roughly 25%. Below that, a modest error in your growth assumption or one bad quarter wipes out your cushion entirely. This is the single idea he called "the central concept of investment."
Everything above rests on your growth guess. Here's what happens to the valuation if you're wrong in either direction — this table is the most useful thing on this page.
| Growth assumption | Intrinsic value | Margin of safety | Passes 25%? |
|---|
If the verdict flips from pass to fail across this range, your thesis depends on being right about growth — which almost nobody is, consistently. Widen your required margin accordingly.
A cheap price on a deteriorating business is a trap, not a bargain. Graham paired every valuation with a financial-strength check. Fetch a ticker above and the measurable ones tick themselves.
In plain English: everything the company will turn into cash within a year, divided by everything it owes within a year. Above 1.0 means it can cover its short-term bills; Graham wanted comfortably above that.
Where to find it: the balance sheet, in the 10-K or 10-Q. Divide "total current assets" by "total current liabilities". Most stock screeners list it directly as "current ratio".
Rule of thumb: above 1.5 is sound, above 2.0 is conservative, below 1.0 means it depends on refinancing or asset sales to keep the lights on — precisely when lenders are least helpful.
In plain English: compares money borrowed against money put in by shareholders. A ratio of 1.0 means for every dollar of owners' capital there's a dollar of debt.
Where to find it: the balance sheet — total debt divided by total shareholders' equity. Listed on most screeners as "debt/equity" or "D/E".
Why it matters: debt is what turns a disappointing year into a permanent loss. A business with no debt can survive a bad decade; a leveraged one can be wiped out by two bad quarters. Note that banks and utilities normally run much higher ratios for structural reasons — compare within an industry.
In plain English: annual profit as a percentage of shareholders' money in the business. ROE of 15% means every $100 of owners' capital generates $15 of profit a year.
Where to find it: net income divided by shareholders' equity, both on the financial statements. Every screener reports it as "ROE".
Rule of thumb: consistently above 15% suggests a genuine competitive advantage. Below 10% suggests a business that struggles to earn its cost of capital. Check whether high ROE comes from real profitability or simply from heavy borrowing shrinking the equity base.
In plain English: the company made a profit in every single one of the last ten years — not "on average", every year.
Where to find it: the "selected financial data" section of a 10-K, or the annual income statement history on any financial data site. You want ten separate figures, all positive.
Why Graham insisted: averages hide the years that would have frightened you. A business that lost money in a recession will lose money in the next one, and you need to know that before you own it rather than during.
In plain English: the company has paid shareholders a cash dividend every year for at least a decade, without cutting or suspending it.
Where to find it: the investor relations section of the company's own website usually publishes full dividend history. Financial data sites list it under "dividend history".
Why it's a quality signal: reported profit can be shaped by accounting choices. A dividend cheque cannot — the cash either exists or it doesn't. A long unbroken record is evidence of both real cash generation and a management that treats it as an obligation.
Fair exception: some genuinely excellent businesses reinvest everything and pay nothing. That's defensible — but then you need another way to verify the cash is real, such as consistent free cash flow.
In plain English: you could explain to someone else, without jargon, who pays this company, what for, and why they'd keep doing so.
How to test yourself: try writing it in three sentences. If you can't, you don't understand it well enough to judge whether a low price is an opportunity or a warning.
Why it can't be automated: no data feed can answer this. The boundary of what you understand matters far more than its size — Graham's point was never that you must be an expert in everything, only that you must know where your knowledge stops.
Graham published this formula in 1962 and was openly ambivalent about it — he offered it as a rough sanity check against the growth multiples of his day, not as a valuation method. It is genuinely crude: it ignores debt, cash, capital intensity, cyclicality and share count entirely. Two companies with identical EPS and growth get identical values even if one is drowning in debt.
Use it as a first-pass filter that tells you where to spend your research time — never as the final word. The sensitivity table above exists to keep you honest about how much the answer moves.
Educational tool only. Nothing here is investment advice, and no output constitutes a recommendation to buy or sell any security. Automatically fetched figures come from third-party data providers and may be delayed, restated or wrong — always verify against primary filings. Do your own research and consider speaking with a licensed financial adviser before making investment decisions.
The same logic as this page, as a TradingView script: LL Graham Screen plots the Graham criteria as a pass/fail table on any symbol.
