How to Value a Stock: A Step-by-Step Guide

Three Methods, One Goal

Professional investors do not rely on a single valuation method. They triangulate -- using multiple approaches to estimate intrinsic value and looking for convergence. If three different methods all suggest a stock is worth roughly $100 and it trades at $65, your confidence in the margin of safety is far higher than if only one method supports that conclusion.

Here are the three most important valuation approaches, from simplest to most rigorous.

Method 1: The Graham Number

Benjamin Graham designed this as a quick defensive screen. It calculates the maximum price a conservative investor should pay based on two fundamental metrics: earnings per share and book value per share.

The 22.5 comes from Graham's maximum acceptable P/E ratio of 15 multiplied by his maximum acceptable price-to-book ratio of 1.5. Any stock trading below its Graham Number passes the most basic value screen.

Limitation: The Graham Number only works for profitable companies with positive book value. It tends to undervalue asset-light businesses like software companies and overvalue asset-heavy businesses with mediocre earnings.

Method 2: Graham's Growth Formula

For companies with earnings growth, Graham proposed a formula in the 1962 edition of Security Analysis:

Where V is intrinsic value, EPS is trailing twelve-month earnings, and g is the expected annual growth rate for the next 7-10 years. The 8.5 represents the P/E Graham considered fair for a zero-growth company.

Be warned: Graham himself cautioned against treating this formula as gospel. Small changes in the growth assumption produce large changes in the output. If you estimate 10% growth instead of 8%, the intrinsic value jumps by roughly 14%. Always use conservative growth estimates -- cap at 15% even for fast growers, because reversion to the mean is one of the most powerful forces in business.

Method 3: Discounted Cash Flow (DCF)

The DCF model is the most rigorous approach. It projects future cash flows and discounts them back to present value using a discount rate that represents your minimum acceptable return.

Step 1: Start with real earnings

Use trailing twelve-month EPS or, better yet, Buffett's "owner earnings" (net income + depreciation - required capital expenditures). This is your anchor, based on actual numbers rather than analyst forecasts.

Step 2: Estimate growth conservatively

Look at the company's EPS growth over the past 5-10 years, but assume future growth will be lower. A company that grew at 20% per year is unlikely to sustain that rate. Cap your estimate at 15% for even the fastest growers.

Step 3: Project and discount

Project earnings forward 10 years, then discount each year's earnings back to present value:

Then add a terminal value for all earnings beyond year 10. A standard approach uses the Gordon Growth Model with a perpetual growth rate of 2-4% (roughly GDP growth).

Step 4: Sum and compare

Add all ten discounted annual values plus the terminal value. This is your estimated intrinsic value. Compare it to the current market price.

The Sensitivity Problem (Garbage In, Garbage Out)

Here is the most important thing to understand about any valuation model: small changes in inputs produce large changes in output. Using the same company above, watch what happens when you adjust just two variables:

• Growth 10%, Discount 12%: Intrinsic value roughly $118
• Growth 12%, Discount 10%: Intrinsic value roughly $185
• Growth 8%, Discount 14%: Intrinsic value roughly $72

Same company. Same current earnings. Three wildly different "answers." This is why Buffett has said he would rather be approximately right than precisely wrong. The DCF model is a thinking framework, not a calculator that spits out truth. Use it to establish a range of reasonable values, not a single target price.

Putting It All Together

Run all three methods. If the Graham Number says $52, the growth formula says $57, and the DCF says $60, you have a convergent estimate around $55-60. If the stock trades at $38, you have a compelling case. If only one method supports a discount while the others suggest fair value or overvaluation, your conviction should be much lower.

Then apply the margin of safety. Even your converged estimate is still an estimate. Graham and Dodd recommended buying at 30-50% below intrinsic value. In practice, most disciplined value investors look for at least 25%.