How to calculate intrinsic value, step by step

A share price tells you what somebody was willing to pay this morning. It tells you nothing about what the business behind the share is worth. Working out that second number is the whole job, and the logic is simpler than the reputation of the subject suggests: figure out how much cash the company can hand its owners over the rest of its life, then decide what that future cash is worth to you today.

What you are actually estimating

The intrinsic value of a company is not a fact sitting somewhere waiting to be looked up. It is your estimate, built from your assumptions, and someone equally competent using different assumptions will get a different number. That is not a flaw in the method. It is the point. The exercise forces you to write down what you believe about a business, in numbers, where you can be checked later.

The standard machinery for doing this is discounting. A euro arriving in ten years is worth less to you than a euro today, because today's euro can be invested, and because the future one might not arrive at all. Discounting puts a price on both of those problems at once, using a single rate, and turns a stream of future payments into one number in today's money.

Start with cash, not accounting profit

Reported net income is an accounting opinion shaped by depreciation schedules, write-downs and one-off items. Cash is harder to dress up. So valuation normally runs on free cash flow, which you build from the cash flow statement: take cash generated by operations, then subtract the capital spending the business needs to keep running and growing. What is left is genuinely available to the people who financed the company.

Get those figures from the company's own filings rather than a summary site. Listed companies publish audited annual reports, and US issuers file the 10-K with the Securities and Exchange Commission, where anyone can read it free on EDGAR. Read three or four years at once. A single year's capital spending can be distorted by one factory, and you want the pattern rather than the snapshot.

Forecast only as far as you can defend

Most people building a discounted cash flow model project five to ten years of free cash flow explicitly. Not because the tenth year is knowable, but because you have to stop somewhere. Ask yourself what you actually know about this business: is demand growing, can it raise prices, is a patent expiring, is a competitor spending heavily to take its market? Turn those judgements into a growth rate you could defend out loud. If you cannot explain why the number is 6 percent rather than 12 percent, you have not made a forecast, you have made a wish.

Pick a discount rate, then admit how much it matters

The discount rate is the return you require for tying up money in this particular business rather than somewhere safer. In corporate practice it is usually the weighted average cost of capital, blending what the company pays lenders with what shareholders should demand for the risk they carry. Plenty of private investors skip that machinery and simply use the return they insist on earning, raised for a fragile balance sheet or an unpredictable industry.

Either way, understand that this one input dominates. It sits in the denominator of every year of the forecast, and its effect compounds the further out you go, so a small change in the rate produces a large change in the answer. If you want the mechanics of building the rate properly, the discount rate walkthrough takes it apart piece by piece.

Terminal value is where most of the answer hides

Your forecast stops at year ten, but the company does not. Everything after that gets compressed into one figure, the terminal value, usually by assuming the final year's cash flow grows forever at a modest rate and dividing by the discount rate minus that growth rate. The growth rate has to be low, no faster than the economy the company operates in, because a business growing faster than its economy forever eventually becomes the entire economy.

Here is an illustration with invented round numbers. A company throws off 100 million a year in free cash flow. You assume 5 percent growth for ten years, a 9 percent discount rate, and 2 percent growth thereafter. The ten forecast years discount back to roughly 820 million. The terminal value discounts back to a bit over 1,000 million. So more than half of the estimate comes from a period you did not forecast at all, driven by a growth rate you invented. That is the honest shape of a DCF, and it is why nobody should present the output to two decimal places.

From company value to a price per share

Adding the discounted forecast to the discounted terminal value gives you the value of the whole business, funded by both lenders and shareholders. Shareholders only own what is left after the debt. Subtract borrowings, add back cash and easily sold investments, and divide by the diluted share count, which includes shares that will exist once outstanding options and convertibles are exercised. Skipping that dilution step quietly flatters your answer. Compare the result with the current share price only at the very end, once your assumptions are already written down. Look first and you will find yourself nudging the growth rate until the model agrees with whatever you already wanted to believe.

Cross-check with a cruder method

One model is not a valuation, it is a single opinion with a spreadsheet attached. So test the answer a second way. Apply a sober earnings or cash flow multiple to a normal year and see whether you land in the same neighbourhood. Look at what similar businesses have actually been bought for. Benjamin Graham, writing in The Intelligent Investor, offered rules of thumb precisely because he distrusted elaborate forecasts, and the Graham number he is remembered for is a deliberately crude screen rather than a valuation. The purpose of the cross-check is not agreement. It is to catch the case where your model says a company is worth three times what the market and every comparable transaction say. When that happens, the model is usually the thing that is wrong, so go back and find the assumption doing the damage.

Report a range, then subtract a margin of safety

Run the model again with the discount rate one percentage point higher. In the illustration above, moving from 9 percent to 10 percent takes the estimate down by more than a tenth, without touching a single thing you believe about the business. Do the same with the growth assumptions, low and high, and you end up with a spread rather than a point. That spread is the real output.

Then refuse to pay anywhere near the top of it. A margin of safety is the gap you insist on between your estimate and the price you will actually pay, and it exists because your inputs are estimates and some of them will be wrong. A wide, stable, easily understood business earns a narrower margin. A cyclical one with heavy debt and a short track record needs a much wider one.

When this method does not work

Discounting assumes you can say something sensible about future cash. For a young company burning money with no visible path to generating it, a loss-making biotech waiting on one trial, or a bank whose accounts work on a completely different logic, the honest answer is that this tool does not apply. Forcing it produces a confident-looking number built entirely from guesses, which is worse than admitting you cannot value the thing. Banks are the clearest case: their borrowing is raw material rather than financing, so subtracting net debt from an enterprise value stops meaning anything and analysts value them off equity and book value instead. Knowing when to put the tool down is part of using it well.

Value one business you already understand, by hand, before you value ten. Write your assumptions down next to the estimate, because in a year you will want to know which one you got wrong, and a stored spreadsheet without reasoning attached teaches you nothing.

Written by

Federico Romaldi

Co-Founder, Worthmap

Published: August 14, 2026

Federico is a co-founder of Worthmap, a wealth-intelligence platform built for serious investors. With a background in software engineering and a long-standing passion for value investing, he created Worthmap to bridge the gap between net-worth tracking and investment analysis.