Lesson 1 of 7
Graham and the defensive investor
Seven tests from 1973, all arithmetic on filed accounts. Which three a ten-year archive can check, and why debt to equity is not his leverage rule.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- Graham's defensive investor passes seven tests, every one arithmetic on figures a company files.
- His balance sheet floor is a current ratio of at least 2.0; the median of 3,673 US companies on 2026-09-07 was 1.86.
- Debt to equity is not his leverage rule. His compares long-term debt with working capital, a different question.
Graham's defensive investor is not one number. It is seven tests, set out in chapter 14 of the 1973 edition of The Intelligent Investor, and every one of them is arithmetic a reader can run on a company's own filings.
Where Graham's current-ratio floor sits in the US market today
Graham's floor, at least 2.0: between the median (1.86) and the 75th percentile (3.59).As a floor it admits between a quarter and a half of the cohort.
- 5th percentile: 0.45
- 10th percentile: 0.69
- 25th percentile: 1.1
- median: 1.86
- 75th percentile: 3.59
- 90th percentile: 8.06
- 95th percentile: 12.71
The seven tests, in order
| Test | What it asks |
|---|---|
| Adequate size | Excludes small companies, which have more ways to fail than a set of accounts shows |
| Strong financial condition | A current ratio of at least 2.0, and long-term debt below net current assets |
| Earnings stability | Positive earnings in each of the past ten years, not just on average |
| Dividend record | An uninterrupted dividend for at least the past twenty years |
| Earnings growth | Per-share earnings up at least a third over ten years, on three-year averages at each end |
| Moderate P/E | Price at most 15 times the average earnings of the past three years |
| Moderate P/B | At most 1.5 times book, or more if the P/E is low enough that the two multiplied stay at or under 22.5 |
Since ASC 842 (fiscal years from 2019), current lease liabilities lower the current ratio with no change in the business.
The order carries an argument. Survival comes first, record length second, and price last, which is the reverse of how most screens get built. Tests two and three ask whether the company survives a bad stretch. Four and five ask whether its record is long enough to be evidence rather than an anecdote. Six and seven ask whether the price already assumes the record continues.
Why earnings are averaged
The growth test and the P/E test both use three-year average earnings, not one year's. Graham's own reasoning, in chapter 12, is that a single year's per-share earnings carry whatever special charges, tax effects and accounting choices happened to fall in that year. Several years averaged together carry more information than one year reported alone. A cyclical business bought at the peak of its earnings looks cheapest exactly when it is least likely to stay that way, and averaging is Graham's defence against that trap.
Debt to equity is not his leverage rule
Graham's leverage test for an industrial company compares long-term debt with working capital, current assets minus current liabilities, which Graham also calls net current assets: could the borrowings be covered by what is left once current liabilities are paid. Debt to equity compares total debt with the book value of equity, a different question about how the capital structure is split. A company can pass one and fail the other. Where a published screen calls itself a Graham screen and reports debt to equity, it has substituted the second question for the first without saying so.
What a ten-year archive can actually check
Three of the seven tests are pure arithmetic on ten years of filed statements: financial condition (the current ratio and the debt comparison), the ten-year earnings record, and the growth test. Two more need a price added on top: the earnings and book multiples.
The dividend test needs twenty years of history, which the ten-year archive does not hold. It is answered from a different record: since 11 September 2026 the dividend streak is measured per share from declared dividends reaching back decades, and on 2026-09-13 182 US listings carry a run of twenty years or more. That run counts years the dividend was held or raised, which is stricter than Graham's uninterrupted payment, so it is a floor rather than his count.
The size test was set in 1973 dollars, and no modern equivalent is Graham's own.
Check yourself
Graham's balance-sheet floor for the defensive investor is a current ratio of at least what?
- 1.5
- 2.0
- 3.0
Show the answer
Answer: B.
Chapter 14 sets the floor at 2.0. The median of 3,673 US companies on 2026-09-07 was 1.86.Which two figures does Graham's own leverage test compare?
- Total debt and total equity
- Long-term debt and net current assets
- Total assets and total liabilities
Show the answer
Answer: B.
His test asks whether long-term debt is covered by net current assets. Debt to equity is a different, more common substitute.The twenty-year dividend test cannot be run on the ten-year filed archive. What is it answered from instead?
- An estimate built from the last ten years
- A separate per-share record of declared dividends reaching back decades
- Nothing; the test is still reported as unrunnable
Show the answer
Answer: B.
Since 11 September 2026 the streak is measured per share from the record of declared dividends, not from ten cash flow statements. 182 US listings carried a run of twenty years or more on 2026-09-13. That record is not a filing this product holds, so the source trail names the basis on every row.Sources
- Graham, The Intelligent Investor, revised edition 1973, chapter 14: the defensive investor's seven criteria.
- Graham, The Intelligent Investor, revised edition 1973, chapter 12: per-share earnings averaged over several years.
What this leaves out
- The current ratio distribution was built on 2026-09-07 from statements last verified then, and a later reader will see different breakpoints.
- The criteria are reported as Graham published them in 1973, for a different market. Passing them is not a statement that a company is worth owning.
- Only the current ratio and the ten-year earnings record are tested backwards across filed years here; the growth and price tests are not, and the dividend test is answered from the declaration record rather than from filed years.
- Nothing here is advice about any security.