Graham Defensive-inspired

What this checklist is trying to capture

Benjamin Graham, the teacher of Warren Buffett, wrote The Intelligent Investor for ordinary savers who wanted good results without spending their lives studying the market. For this "defensive" investor he set out seven requirements: the company should be large, financially strong, consistently profitable, a reliable dividend payer, growing over the long run, and available at a moderate price. Each one is meant to remove a way of losing money, not to find the next big winner.

Graham wrote these rules for a very different market, and he knew that strict versions rule out most large companies when prices are high. A low score here usually means "priced above what Graham would pay", not "a bad business".

The tests, and exactly how we compute them

"Year t" is the latest fiscal year in the company's SEC filings.

Size and financial strength

1. Annual revenue above $2 billion: revenue in year t. Graham set a minimum company size in the money of his day; $2 billion is our modern equivalent, chosen by us, not by Graham.
2. Current ratio of at least 2: current assets divided by current liabilities at the end of year t, meaning short-term assets cover short-term bills twice over.
3. Long-term debt no more than net current assets: long-term debt (excluding the current portion) divided by current assets minus current liabilities is 1.00x or less. A company with no long-term debt passes.

Record

4. Profitable in each of the last 10 years: diluted earnings per share above zero in every one of the last ten fiscal years.
5. Dividends paid every year for 15 years: cash dividends paid in each of the last 15 fiscal years without a break. Graham asked for 20 years, but the machine-readable SEC data we use only starts around 2009, so we test the longest run the data can show.
6. Earnings up at least a third over 10 years: average EPS over the latest three years is at least one-third higher than average EPS over the three years ending ten years earlier. Using three-year averages smooths out a single unusually good or bad year, as Graham intended.

Price

7. P/E of 15 or less on 3-year average earnings: today's share price divided by average diluted EPS over the last three fiscal years.
8. P/E times price-to-book of 22.5 or less: the P/E above multiplied by market value divided by shareholders' equity. Graham allowed a higher P/E if the price-to-book was low, and the other way round, as long as the product stayed under 22.5.

Reading the score

The first six tests describe the company; the last two describe today's price. Many high-quality companies pass most of the first six and fail both price tests when markets are expensive. That pattern is exactly what Graham's defensive investor was told to wait out.

Limitations

• Graham excluded utilities from the current-ratio test and judged them on different debt rules; we apply one rule to every company, so utilities usually fail tests 2 and 3.
• Banks and insurers don't report current assets and liabilities in the usual way, so those tests show "n/a" for them.
• Reported earnings aren't adjusted for one-off gains or charges, which can move a three-year average.

See how we score for data sources and update timing.