How to Use the Graham Number to Value Stocks

Benjamin Graham spent decades trying to answer a deceptively simple question: how do you know if a stock is cheap?

His answer wasn't complicated — and that's exactly the point. In a world full of intricate discounted cash flow models and multi-variable regression analyses, Graham believed in formulas an intelligent non-expert could actually apply. He wanted tools that gave individual investors a fighting chance.

The Graham Number is one of those tools. It's a single formula that estimates the maximum price a value investor should consider paying for a stock. It's been around since the mid-20th century. It's still in use today. And if you understand both what it does and what it doesn't do, it becomes a genuinely useful part of your valuation toolkit.

---

The Formula

The Graham Number is defined as:

Graham Number = √(22.5 × EPS × BVPS)

Where:

• EPS = Earnings per share (typically trailing twelve months)
• BVPS = Book value per share
• 22.5 = A constant derived from Graham's maximum acceptable ratios (15× earnings × 1.5× book value = 22.5)

That's it. Square root of 22.5 times earnings per share times book value per share.

If a stock is trading below its Graham Number, it passes Graham's basic valuation screen. If it's trading significantly above it, Graham would say the price has moved into territory that requires justification beyond what he was comfortable assuming.

---

Where Does 22.5 Come From?

Graham wasn't arbitrary about the constant. In The Intelligent Investor, he laid out two rules for what he called the "defensive investor":

1. The P/E ratio should not exceed 15
2. The price-to-book ratio should not exceed 1.5

Multiply those two together: 15 × 1.5 = 22.5. The square root formula then works out so that a stock trading at its Graham Number is simultaneously at a P/E of exactly 15 and a P/B of exactly 1.5 — right at the edge of what Graham considered acceptable.

Anything below that threshold was, in his view, reasonably priced for a conservative investor.

---

A Concrete Example

Suppose you're looking at a company with:

• EPS: $4.00
• BVPS: $30.00

Graham Number = √(22.5 × 4.00 × 30.00) = √(2,700) = $51.96

If the stock is trading at $38, it's meaningfully below its Graham Number — a potential value signal. If it's trading at $80, it's significantly above the Graham Number, and Graham would say the price is relying on growth expectations he wasn't willing to assume.

You can run this calculation for any stock using the Graham Number Calculator at valueofstock.com — no spreadsheet required.

---

What the Graham Number Actually Tells You

Think of the Graham Number as a floor, not a ceiling.

Graham designed it to answer a specific question: at what price is a stock conservative enough that I don't need heroic assumptions about the future?

It's not telling you what a company is worth. It's not projecting growth. It's not building a discounted cash flow model. It's simply drawing a line and saying: if you pay below this price, you have a margin of safety based on what the company earns today and the assets it holds today. Growth is bonus, not requirement.

That's a fundamentally different approach than most modern valuation. Today's analysts often build elaborate models where most of the value lives in years 6 through 20 of a forecast. Graham thought that was speculation dressed up as analysis. His formula intentionally grounds you in the present.

---

The Real Limitations (And Why They Matter)

The Graham Number is a blunt instrument. Used carelessly, it can mislead. Here's where it breaks down:

It doesn't work for companies with negative earnings or negative book value. If EPS or BVPS is negative, the formula produces an imaginary number — literally, you'd be taking the square root of a negative. This immediately excludes most early-stage companies, many tech businesses, and firms currently running at a loss.

Book value is increasingly meaningless for many businesses. Graham developed his formula in an era when companies' most valuable assets were physical — factories, equipment, inventory. For a retailer or manufacturer, book value roughly captured what the business was worth if you liquidated it.

Today, many of the most valuable companies in the world are built on intellectual property, brand equity, software, and human capital. None of these show up on the balance sheet in any meaningful way. A company with proprietary software and 10,000 engineers might have a book value that reflects little about its actual earning power. The Graham Number would dramatically understate fair value for such businesses.

It treats all earnings as equal. Graham's formula doesn't distinguish between a company earning $4/share this year because of a temporary tailwind and a company that has earned $4/share every year for the last decade. High-quality, predictable earnings deserve a premium that the formula doesn't grant.

Growth isn't factored in. If a business is compounding earnings at 20% per year, paying 15× this year's earnings is an excellent deal. The Graham Number doesn't capture this — it effectively assumes zero growth, which for many good businesses is a deeply conservative assumption.

It was built for a different interest rate environment. Graham developed his rules when returns on safe assets were higher. A P/E of 15 implies an earnings yield of about 6.7%. With 10-year Treasuries at various rates throughout history, that implied a different real premium than it might today.

---

How to Use It Correctly: As a Floor, Not a Target

Here's how sophisticated value investors actually use the Graham Number in 2026:

Use it as a first filter. Stocks trading well below their Graham Number deserve a second look. You haven't found a buy — you've found a reason to investigate further.

Combine it with business quality assessment. A stock can trade below its Graham Number because it's genuinely cheap, or because the business is deteriorating. The number can't tell you which. Your job is to figure that out.

Don't use it for asset-light businesses. Software, media, financial services — apply it only where book value actually means something. For asset-heavy industrials, real estate, banking (with adjustments), and some consumer staples, it holds up better.

Compare within industries. The Graham Number is most useful as a relative tool. If five companies in the same sector have Graham Numbers of $40, $45, $35, $50, and $38, and the stock prices are all over the map, that tells you something about relative valuation that a single data point can't.

Use it alongside other metrics. Pair it with free cash flow yield, EV/EBITDA, and return on invested capital. When multiple metrics are pointing toward a stock being cheap, that convergence is more meaningful than any single number.

---

The Deeper Lesson from Graham

What makes the Graham Number valuable isn't the arithmetic. It's the philosophy baked into it.

Graham was obsessed with margin of safety — the idea that you should never pay so much for a stock that you're depending on everything going right. Every investment should have enough cushion that even if you're wrong about some of your assumptions, you still come out ahead.

The Graham Number operationalizes that philosophy. It says: here's a price that requires no optimism about the future. If the business just maintains its current earnings and doesn't collapse, you're probably okay.

That conservatism is unfashionable in bull markets. It means you'll pass on a lot of companies that go on to perform well. But over a full market cycle — through corrections, recessions, and the inevitable disappointments — that discipline tends to protect capital in ways that more optimistic approaches don't.

---

The Bottom Line

The Graham Number is one of the oldest stock valuation formulas still in regular use, and it's earned that longevity. But treat it as what Graham intended: a floor below which stocks are almost certainly cheap, not a target above which they're certainly overvalued.

For asset-heavy, profitable businesses with stable earnings, it's a genuinely useful first filter. For software companies, high-growth businesses, or firms with negative book values, look elsewhere.

Use the Graham Number Calculator at valueofstock.com to quickly run the calculation on any stock you're analyzing — and pair it with the other fundamental metrics available there to build a more complete picture.

Graham's insight was that you don't need a precise valuation to invest intelligently. You need to know whether something is cheap enough that you have a cushion. Start there, and you'll avoid most of the worst mistakes.