Graham Number Calculator

The Graham Number is a formula that calculates the maximum price a defensive investor should pay for a stock, based on its earnings per share (EPS) and book value per share (BVPS). It was derived from principles outlined in Benjamin Graham's The Intelligent Investor.

The formula:

Graham Number = √(22.5 × EPS × BVPS)

Where the 22.5 comes from:

• Graham recommended that defensive investors should not pay more than 15 times earnings (P/E ratio of 15) for any stock. A P/E of 15 roughly represents a company earning a reasonable return on its market price.
• He also recommended that the price-to-book ratio should not exceed 1.5. This ensures you are not paying too much relative to the company's tangible net assets.
• Multiplying these two limits together gives 15 × 1.5 = 22.5. The square root converts this product back into a per-share price.

Interpretation: If a stock trades below its Graham Number, it meets both of Graham's criteria for defensive investing — it has a reasonable P/E ratio and is not overpriced relative to its book value. The bigger the gap between the Graham Number and the current price (with the Graham Number being higher), the larger the margin of safety.

Important constraint: The formula requires both EPS and BVPS to be positive. Companies with negative earnings or negative book value cannot be evaluated using the Graham Number — by design, Graham would have avoided these companies entirely for defensive portfolios.

The Graham Number is a useful screening tool, but it has important limitations that every investor should understand before relying on it for investment decisions.

Key limitations:

• No growth consideration — The formula uses trailing earnings and current book value. It does not account for future earnings growth, which is the primary driver of value for growth companies. A fast-growing company may appear “overvalued” by the Graham Number but actually be fairly priced when growth is factored in.
• Industry bias — The formula works best for asset-heavy, capital-intensive businesses (banks, utilities, manufacturers) where book value is meaningful. For asset-light companies (software, services, platforms), book value often understates true economic value because intangible assets are not fully captured on the balance sheet.
• Backward-looking — EPS reflects past performance, not future prospects. A company whose earnings are about to collapse might look cheap by the Graham Number, while a turnaround story might look expensive.
• Ignores cash flow quality — Earnings can be manipulated through accounting choices (depreciation methods, revenue recognition, one-time items). The Graham Number does not distinguish between high-quality recurring earnings and one-time gains.
• No debt adjustment — Two companies with identical EPS and BVPS but vastly different debt levels will produce the same Graham Number. In reality, the highly leveraged company is riskier and should be valued more conservatively.
• Excludes negative-EPS companies — Many high-growth companies (especially in tech and biotech) operate at a loss during their growth phase. The Graham Number cannot evaluate these companies at all.

Best practice: Use the Graham Number as one data point among many. It is most effective as a quick screening filter to identify stocks that might deserve deeper analysis — not as a standalone buy/sell signal.

The Graham Number was designed for a specific type of investing: defensive value investing in established, profitable companies. Understanding when it works best helps you apply it appropriately.

Stocks where the Graham Number is most useful:

• Financial institutions — Banks, insurance companies, and REITs have large balance sheets where book value is a meaningful measure of tangible assets. The Graham Number is particularly effective here.
• Industrial and manufacturing companies — Companies with significant physical assets (plants, equipment, inventory) tend to have book values that approximate their liquidation value, which is exactly what Graham intended.
• Utilities and mature businesses — Stable, slow-growing companies with predictable earnings are ideal candidates. The lack of a growth component in the formula matters less when growth is modest anyway.
• Cyclical stocks at the bottom of a cycle — When cyclical companies have depressed earnings, the Graham Number naturally produces a lower number, helping prevent value traps. However, you should use normalized earnings (average over a full cycle) for better results.

Stocks where the Graham Number is less reliable:

• High-growth tech companies — Companies like Amazon, Google, or Nvidia almost always trade well above their Graham Number because the market is pricing in future growth that the formula ignores.
• Asset-light businesses — Software companies, marketplaces, and service businesses often have minimal book value relative to their earnings power. The Graham Number will almost always flag them as overvalued.
• Pre-profit companies — Startups and early-stage companies with negative earnings cannot be evaluated by the formula at all.

Rule of thumb: If you are screening for classic value stocks — boring, profitable, asset-heavy businesses trading at reasonable multiples — the Graham Number is an excellent first filter. For growth stocks or asset-light businesses, pair it with a DCF model that accounts for future cash flow potential.

The 22.5 multiplier in the Graham Number formula is not an arbitrary number — it derives directly from two specific valuation limits that Benjamin Graham recommended for defensive investors.

The two Graham limits:

• Maximum P/E ratio of 15 — Graham argued that a defensive investor should not pay more than 15 times a company's trailing twelve-month earnings. A P/E of 15 implies an earnings yield of roughly 6.7%, which Graham considered a minimum acceptable return above prevailing bond yields at the time.
• Maximum P/B ratio of 1.5 — Graham recommended not paying more than 1.5 times book value per share. This limit ensures you are buying the company at a price close to its net tangible asset value, providing downside protection if the business deteriorates.

How they combine:

Graham allowed for some flexibility: a company could exceed one limit as long as the product of its P/E and P/B ratios did not exceed 22.5 (= 15 × 1.5). For example, a company with a P/E of 10 could have a P/B up to 2.25 (10 × 2.25 = 22.5), and a company with a P/B of 1.0 could have a P/E up to 22.5. The formula captures this tradeoff mathematically.

Historical context: Graham developed these limits in the mid-20th century when average stock market P/E ratios hovered around 10-15. Today, the S&P 500 typically trades at 20-25x earnings. This means the Graham Number will flag many stocks as overvalued by modern standards. Some practitioners adjust the multiplier upward (e.g., to 30 or 37.5) to reflect the current interest rate environment and higher market multiples, though purists argue that Graham's conservatism is precisely the point.

Bottom line: The 22.5 multiplier enforces Graham's core philosophy: only buy stocks where you are paying a reasonable price for both the company's earnings power and its underlying asset base. When both conditions are met, your margin of safety is maximized.

The Graham Number and a discounted cash flow (DCF) model are both tools for estimating intrinsic value, but they approach the problem from fundamentally different angles and are suited to different use cases.

Graham Number:

• Inputs: Only two — EPS and book value per share
• Time horizon: Backward-looking (uses trailing earnings and current book value)
• Growth: Does not factor in future growth at all
• Complexity: One formula, instant result
• Best for: Quick screening of value stocks, defensive portfolio construction
• Weakness: Misses growth value entirely, ignores cash flow quality and capital structure

DCF Model:

• Inputs: Revenue projections, margins, capital expenditures, working capital, discount rate, terminal value assumptions
• Time horizon: Forward-looking (projects 5-10 years of future cash flows)
• Growth: Explicitly models future revenue and earnings growth year by year
• Complexity: Requires detailed assumptions and financial modeling expertise
• Best for: Deep fundamental analysis, understanding how assumptions drive value
• Weakness: Output is only as good as the inputs — garbage in, garbage out

How to use them together: Start with the Graham Number as a quick filter. If a stock passes (trades below its Graham Number), it deserves deeper analysis. Build a DCF model to understand whether the stock is cheap because the market is wrong (opportunity) or because the business is deteriorating (value trap). The Graham Number tells you a stock might be worth investigating. The DCF model tells you what it is actually worth.

Margin of safety is the central concept in Benjamin Graham's investment philosophy. It refers to the difference between a stock's intrinsic value (what it is worth) and its market price (what you pay). The bigger the gap, the larger your cushion against errors in analysis, unexpected business setbacks, or market downturns.

How margin of safety works with the Graham Number:

• Margin of Safety formula — (Graham Number - Current Price) / Graham Number × 100%. A positive margin of safety means the stock trades below the Graham Number. A negative margin means it trades above.
• Graham's recommendation — Graham generally looked for a margin of safety of at least 30-50%. This means he wanted to pay no more than 50-70 cents for each dollar of intrinsic value.
• Example — If a stock has a Graham Number of $50 and trades at $35, the margin of safety is 30%. You are paying $35 for a stock that, by Graham's criteria, is worth up to $50.

Why margin of safety matters:

• Protects against errors — No valuation model is perfectly accurate. If you buy with a large margin of safety, you can be somewhat wrong about the company's value and still not lose money.
• Accounts for the unexpected — Recessions, competitive disruptions, management mistakes — a margin of safety provides a buffer against negative surprises that could hurt the business.
• Improves risk-adjusted returns — By insisting on a discount to intrinsic value, you tilt the odds in your favor. Even if some picks do not work out, the ones that do should generate enough returns to compensate.

Practical tip: Use this calculator to find stocks with a positive margin of safety, then investigate further with a full DCF model to confirm the opportunity is real. A stock trading 40% below its Graham Number is a starting point for research, not an automatic buy signal.

The short answer is yes, but with caveats. The Graham Number was developed in an era when stocks generally traded at lower multiples, interest rates were higher, and the economy was more manufacturing-oriented. Modern markets have evolved, and the formula must be applied with context.

Why it is still relevant:

• Value investing outperforms long-term — Academic research (Fama-French, Lakonishok et al.) consistently shows that stocks with low P/E and low P/B ratios outperform over long periods. The Graham Number is a simple way to screen for these characteristics.
• Behavioral biases persist — Investors still overpay for exciting growth stories and neglect boring, profitable businesses. The Graham Number helps you avoid this trap.
• Downside protection — In bear markets, stocks trading below their Graham Number tend to fall less than expensive growth stocks. The built-in margin of safety acts as a cushion.

Modern adjustments to consider:

• Higher average multiples — The S&P 500 now commonly trades at 20-25x earnings, compared to 10-15x in Graham's era. Some practitioners adjust the multiplier upward (e.g., using 30 instead of 22.5) to reflect the lower interest rate environment.
• Intangible-heavy economy — Modern companies derive much of their value from intangible assets (brands, IP, network effects) that do not fully appear on the balance sheet. Book value understates the true asset base for many companies today.
• Sector selectivity — Apply the Graham Number primarily to sectors where book value is meaningful (financials, industrials, utilities). For tech and service-based companies, complement it with cash flow-based models like a DCF.

Bottom line: The Graham Number is not outdated — it is a tool with a specific purpose. It excels at screening for classic value opportunities and keeping you disciplined about price. Just do not expect it to capture the full picture for every type of company in the modern market.