What is intrinsic value (fair value)?
Intrinsic value is an estimate of what a business is worth based on the cash it can produce over its life, independent of today’s share price. Berkshire Hathaway defines it as the discounted value of the cash that can be taken out of a business during its remaining life. On this page, intrinsic value and fair value mean the same thing: your estimate of what one share is worth.
Price is what the market charges today; value is what you think you get. The drawing at the top shows how they interact: price swings above, through and below an estimated band. Above the band the stock looks expensive, inside it roughly fair, below it cheap, if your estimate is right. That “if” is why the margin of safety exists, and why Benjamin Graham and Warren Buffett built their approach on the gap; see the Warren Buffett strategy.
- Conservative fair value
- The low end of your fair-value range, built from cautious assumptions. It is the number many value investors compare with the price.
- Normalized earnings
- Earnings for a typical year, with one-off gains, unusual losses and cycle peaks smoothed out. A valuation built on a peak year overstates value.
Method 1: earnings per share × a fair P/E ratio
The quickest method multiplies earnings per share by a P/E ratio you consider fair for that company: fair value = EPS × fair P/E. Both inputs need care. The EPS should come from a normal year and follow GAAP rather than an adjusted figure, and the P/E should come from evidence, not hope.
Two sources of a fair P/E are common, as the drawing below shows. The first is the stock’s own history: the median P/E it traded at over the last five to ten years. The second is its sector: the median P/E of its direct competitors. Use the lower of the two so the estimate stays conservative. The P/E ratio guide explains why a P/E is only comparable within a sector and with the same type of EPS.
Hypothetical: XYZ earned $5.00 per share in a normal year. Its own 10-year median P/E is 18, and its sector median is 15. $5.00 × 18 = $90; $5.00 × 15 = $75. The conservative estimate is $75.
Methods 2 and 3: discounted cash flow and book value
A discounted cash flow (DCF) model forecasts the company’s free cash flow for several years, discounts each year back to today at the return you require, adds a terminal value for the years after that, and divides by the number of shares after adjusting for debt and cash. It is the method closest to the definition of intrinsic value, and the most sensitive to its assumptions: in the hypothetical example in the discounted cash flow guide, moving the discount rate by one percentage point changes the result by 15% to 19%.
Book value per share is shareholders’ equity divided by the number of shares. It suits banks, insurers and asset-heavy businesses, whose balance sheets hold mostly financial or tangible assets, and it can act as a rough floor elsewhere. For software or brand-driven companies it says little, because their most valuable assets barely appear on the balance sheet; see valuation ratios for the P/B ratio.
Why intrinsic value is a range, not a number
Every input is an estimate: next year’s earnings, the multiple the market will pay, growth, the discount rate. Small differences add up. For XYZ, EPS of $4.50 to $5.50 and a P/E of 15 to 18 give values from $67.50 to $99, a spread of almost 50%. Anyone who quotes a fair value to the cent is hiding that spread, not removing it.
So build a range, check it with more than one method, and treat the overlap as your fair-value range. In the drawing below, the earnings method gives $75 to $90, a cautious DCF gives $70 to $100, and book value is $35 a share, a floor far below the others because XYZ owns few physical assets. The methods overlap between $75 and $90.
Worked example: is XYZ cheap, fair or expensive?
Take the overlap, $75 to $90, and use its low end, $75, as the conservative fair value. Then compare the price with it, hypothetical like everything here:
- At $100, XYZ trades above the whole range: expensive, unless your assumptions are too low.
- At $85, it sits inside the range: roughly fair, with no cushion if you are wrong.
- At $60, it trades 20% below the conservative value: (75 − 60) ÷ 75 = 20%. That gap is the margin of safety.
Common mistakes when estimating fair value
A cheap result is where the work starts, not where it ends. Before trusting a low price, check why it is low: falling earnings, rising debt or a shrinking market can make a stock look cheap on last year’s numbers while its value is falling, the pattern described in value trap. Re-run the estimate after every earnings report.
Keep the estimate conservative, because the next steps depend on it: the margin of safety is measured from it, and value is one of the three gates in the UpTickFlow Strategy. Most inflated valuations come from a handful of habits:
- Peak earnings: multiplying a record year, or a cyclical company at the top of its cycle, by a normal P/E.
- Adjusted EPS: using a figure that leaves out real, recurring costs such as stock-based compensation.
- Borrowed multiples: applying a fast-growing peer’s P/E to a slower company, or a P/E from another sector.
- Ignoring the balance sheet: the P/E method says nothing about debt, and two companies with the same EPS and very different debt are not worth the same.
- Anchoring: starting from the current price or an analyst’s price target and working backward.