Skip to content
ruly

Benjamin Graham criteria: original thresholds, limits and updates

Benjamin Graham's criteria for the defensive investor, with their original thresholds, what has changed since and how to adapt them, applied to a fictitious company.

The Benjamin Graham criteria are the filters Graham sets out for the defensive investor in his book “The Intelligent Investor”: size, financial condition, earnings stability, dividend record, earnings growth, a moderate P/E and a moderate price to book value. They remain a sound base, but several thresholds need updating because of intangible assets and share buybacks.

What are Graham’s criteria for the defensive investor?

Graham describes two profiles. The enterprising investor spends time looking for special situations. The defensive investor wants above all to avoid serious mistakes and keep the effort low. For this second profile, Graham sets seven selection criteria, each with a precise threshold.

The thresholds below are those of the last edition Graham revised himself, in the early 1970s, for US industrial companies.

CriterionOriginal threshold
SizeAt least $100 million in annual sales
Current ratioCurrent assets at least 2 times current liabilities
Long-term debtNo more than net current assets
Earnings stabilityPositive earnings in each of the last 10 years
Dividend recordPaid without interruption for at least 20 years
Earnings growthEarnings per share up at least one third over 10 years
Moderate P/E15 at most, on average earnings of the last 3 years
Price to book1.5 at most
P/E × price to book22.5 at most

The current ratio and long-term debt together make up the financial condition criterion. Net current assets, or working capital, are current assets minus current liabilities.

For growth, Graham compares the average of the first three years of the period with that of the last three, which smooths out an exceptional year. For public utilities (electricity, gas, water), he sets other thresholds: at least $50 million in total assets, and debt no more than twice the stock equity.

Why multiply the P/E by the price to book?

Graham accepts a price to book above 1.5 when the P/E is low enough to make up for it. He suggests multiplying the two ratios: the product should not exceed 22.5, that is 15 × 1.5.

A company at a P/E of 9 can therefore be priced at up to 2.5 times its book value (9 × 2.5 = 22.5). The P/E × price to book entry explains the calculation and its pitfalls.

What has changed since Graham?

The principle still stands: pay a moderate price for a solid, steady company. What has changed is the listed companies themselves and their accounting.

Intangible assets are missing from the balance sheet

Software, brands, patents, customer lists: a large part of today’s corporate value is intangible. Much of the spending that creates it (research, marketing, training) is expensed and never shows up as an asset.

These companies therefore have low book value and a high price to book, even when their price is moderate relative to their earnings.

Buybacks replace part of the dividend

Many companies now return part of their cash by repurchasing their own shares. This affects three of Graham’s criteria:

  1. the dividend record misses the cash returned this way,
  2. buybacks reduce equity, which inflates the price to book,
  3. earnings per share rise when the share count falls, even with flat total earnings.

Some sectors fall outside the framework

A bank or an insurer does not split current assets and liabilities the way Graham meant: the current ratio makes no sense there. Companies paid in advance by their customers (subscriptions, large retailers) often run below a current ratio of 2 without any particular fragility.

The size threshold is in 1970s dollars

With inflation, $100 million of that time is worth far more today. Taken as is, the threshold lets through small companies Graham would have excluded.

How do you adapt Graham’s criteria today?

Criterion by criterion, here is one possible adaptation. The thresholds quoted for the Value preset strategy are those of the app, to keep or change in your own copy.

  1. Size. Replace the amount with a relative rule, such as being among the largest market capitalizations in your market, or convert the threshold into today’s money.
  2. Current ratio. Keep 2 for industrial companies. Elsewhere, 1.5 is a common benchmark: it is the green threshold of the Value strategy, with orange from 1. For a bank or an insurer, tick “Not applicable”.
  3. Long-term debt. Debt measured against the ability to repay it says more: net debt / EBITDA, green up to 2 and orange up to 3 in the Value strategy.
  4. Earnings stability. Keep it as is. Ten years without a loss remains a demanding filter that is easy to check in annual reports.
  5. Dividends. Count buybacks too: a shorter record can be enough if the share count falls year after year.
  6. Earnings growth. Keep the one-third-over-ten-years rule (a little under 3% a year), but check the share count: if it has dropped sharply, part of the increase comes from buybacks.
  7. P/E. The three-year average smooths out exceptional years. The Value strategy uses the P/E of the last fiscal year or the trailing twelve months, with green up to 12 and orange up to 18.
  8. Price to book and the product. Keep them for companies with tangible assets (industry, banks, insurers). The Value strategy keeps 1.5 and 22.5 for green, with orange up to 2.5 and 30. For a company rich in intangibles, tick “Not applicable” and look at enterprise value / EBIT or FCF yield instead.

Example: Company A run through Graham’s criteria

Company A is fictitious. It makes industrial pumps and reports in euros. Its figures from the latest annual report:

  • revenue: €2.4 billion,
  • current assets: €900 million, current liabilities: €500 million,
  • long-term debt: €300 million,
  • positive earnings in each of the last 10 years,
  • a dividend paid without interruption for 14 years,
  • average earnings per share: €2.10 over the first three years of the period, €3.00 over the last three,
  • share price: €42, book value per share: €26.25.
CriterionCompany AResult
Size€2.4 billionMet
Current ratio900 / 500 = 1.8Not met (2)
Long-term debt300 against 400 of working capitalMet
Earnings stability10 years out of 10Met
Dividends14 yearsNot met (20)
Earnings growth3.00 / 2.10, or +43%Met
3-year P/E42 / 3.00 = 14Met
Price to book42 / 26.25 = 1.6Not met (1.5)
P/E × price to book14 × 1.6 = 22.4Met

The price to book exceeds 1.5, but the product stays below 22.5: by Graham’s own rule, the low P/E makes up for it. Two gaps remain. The dividend is only 14 years old, but Company A has cut its share count by 2% a year for five years: with the adaptation suggested above, the criterion is met. The current ratio of 1.8 is the real point to dig into: check whether current assets are mostly inventory, which is slower to turn into cash.

With the Value preset strategy, the P/E is based on the last fiscal year. With earnings per share of €3.30, it comes to 12.7: “To monitor”. The price to book of 1.6 is also “To monitor”, while the product 12.7 × 1.6 = 20.3 is “Criterion met”, like the current ratio of 1.8.

How do you apply Graham’s criteria in ruly?

The Value preset strategy includes four of Graham’s criteria in an adapted form: the P/E, the price to book, their product, which the app calculates from the two values you enter, and the current ratio. It adds net debt / EBITDA, enterprise value / EBIT, FCF yield, the margin of safety, a financial strength score and a text criterion on the value trap risk.

To get all seven criteria back, duplicate it. In your copy, you change the thresholds (2 for the current ratio, for example) and add “Yes or no” criteria for the ten profitable years and the twenty years of dividends. Each criterion weighs from 1 to 5 in the score. The strategies are described in detail on the features page.

Frequently asked questions

Are Graham’s criteria still usable?

Yes, as a starting filter. Earnings stability, ten-year growth and the three-year P/E apply as they are. Size, dividends and price to book need the adaptations described above, especially outside industry.

Where does the 22.5 figure come from?

From the product of Graham’s two ceilings: a P/E of 15 multiplied by a price to book of 1.5. It lets a slightly high ratio be offset by a lower one.

How many companies pass all of Graham’s criteria?

Few, mainly because of the price to book and the twenty years of dividends. That is intended: Graham was after companies without surprises, at a moderate price. If your list stays empty, relax the criteria tied to today’s accounting first, not earnings stability.

What is the difference between the defensive and the enterprising investor?

The defensive investor seeks safety with as little effort as possible and applies strict, simple criteria. The enterprising investor is willing to spend time studying more special situations. The criteria in this article are those of the first.

Choose this strategy

Start from the preset strategy “Value”: its criteria and thresholds are ready, all that is left is to enter your figures.

Free for up to 3 sheets, with no time limit. An email address is all you need, and your sheets stay private.

Read next

Sources

  • Benjamin Graham, "The Intelligent Investor" (criteria for the defensive investor)