Benjamin Graham Formula — Intrinsic Value Explained
V = EPS × (8.5 + 2g): the quick intrinsic-value formula from The Intelligent Investor, with the revised version, a worked example, and Graham's own warnings.
The formula
In the 1962 edition of Security Analysis, Graham offered a shortcut for valuing growth stocks: V = EPS × (8.5 + 2g), where EPS is normalized trailing earnings per share, 8.5 is the fair P/E for a no-growth company, and g is the expected annual earnings growth rate over the next 7–10 years (entered as a whole number, e.g. 7 for 7%).
The revised formula adjusts for interest rates: V = [EPS × (8.5 + 2g) × 4.4] ÷ Y, where 4.4% was the AAA corporate bond yield when Graham wrote, and Y is the current AAA yield. Higher rates shrink the value; lower rates expand it.
Worked example
A company earning $5.00 per share, expected to grow 7% a year, with the AAA bond yield at 5.5%:
| Step | Calculation | Result |
|---|---|---|
| Growth multiplier | 8.5 + 2 × 7 | 22.5 |
| Base value | $5.00 × 22.5 | $112.50 |
| Rate adjustment | × 4.4 ÷ 5.5 | × 0.80 |
| Adjusted intrinsic value | $112.50 × 0.80 | $90.00 |
| With a 33% margin of safety | $90.00 × 0.67 | Buy zone below ≈$60 |
The formula is a screening heuristic. Graham added in a footnote that it was meant to produce figures close to more elaborate calculations — not to be a precise valuation model.
Graham's own warnings — and modern fixes
- The growth assumption dominates the result: 2g doubles the multiplier's sensitivity to your forecast.
- Many practitioners cap g (e.g. at 10–15%) and lower the 2× factor to 1× or 1.5× for conservatism.
- Use normalized EPS — a cyclical peak or trough year will distort everything downstream.
- It ignores balance-sheet risk: pair it with debt checks and Graham's defensive criteria.
- Treat the output as a rough anchor, then demand a margin of safety before any decision.
The $25 million lesson that changed Buffett
In 1972 the sellers of See's Candies asked $30 million; Buffett, trained on Graham's cigar butts, refused to go above $25 million — roughly three times book value, a price the old Buffett would never pay. The deal nearly died over $5 million. Charlie Munger pushed back: a beloved brand with loyal customers and genuine pricing power deserved a premium.
They paid the $25 million. See's went on to ship more than $1.65 billion of cumulative pre-tax earnings back to Omaha while needing almost no additional capital — raising prices every year without losing a single loyal customer. Buffett later called it 'the prototype of a dream business' and credits this one purchase with rewiring how he invests.
Quality compounds quietly for decades. Temperament and a clear circle of competence beat chasing the cheapest price.
The only deal Munger ever pushed
In the depths of the 2008 crisis, Munger did something he had never done before: he actively lobbied Buffett to invest in BYD, a Chinese battery and car maker led by engineer Wang Chuanfu — a man Munger described as "a combination of Thomas Edison and Jack Welch". The conviction came through Li Lu, whose circle of competence covered what Omaha's did not.
Berkshire paid $232 million for about 10% of the company — a tiny position by its standards, in an industry Buffett had publicly sworn off. At its 2022 peak the stake was worth roughly $9 billion, close to a forty-fold gain, before Berkshire began trimming.
A circle of competence can be extended honestly — by trusting someone who genuinely has one where you do not.
Related tools
Frequently asked questions
Where does the Graham Formula come from?
The 1962 edition of Security Analysis; it also appears in later editions of The Intelligent Investor. The revised interest-rate version was introduced in 1974.
What does the 8.5 mean?
Graham considered 8.5 a fair price-to-earnings ratio for a company with zero expected growth.
Is the Graham Formula reliable for tech stocks?
It was designed for established companies with steady earnings. For high-growth or unprofitable businesses, the growth term overwhelms the formula — a full DCF with conservative assumptions is more appropriate.
Valuefocus gives you Graham-style quick checks alongside full DCF models, in one workflow.
Download Valuefocus