5 Formulas to Find the Fair Value of a Stock in 15 Minutes

Five quick formulas give a stock's fair value in two or three minutes each:

1. Graham number: √(22.5 × EPS × book value per share)
2. Graham's growth formula: EPS × (8.5 + 2 × growth) × 4.4 ÷ AAA bond yield
3. Peter Lynch's PEG rule: fair P/E = the growth rate
4. Earnings yield vs the 10-year Treasury: fair price = EPS ÷ Treasury yield
5. Free-cash-flow yield: fair price = FCF per share ÷ the yield you want

They rarely agree. We ran all five on 445 S&P 500 stocks: earnings yield calls 30% cheap, the Graham number 9%. Only 7 pass all five.

Watch it play outOne test calls 30 in 100 stocks cheap; all five agree on fewer than 2

Shrink the 445 S&P 500 stocks we tested to 100 dots. A formula lights a dot when the price sits below its fair value.

Earnings yield against the 10-year Treasury is the most generous: 30.3 in 100 look cheap.

Graham's growth formula: 22.2 in 100.

Free-cash-flow yield, asking for a 6% cash return: 21.3 in 100.

The Graham number: 9.4 in 100. Lynch's PEG rule lands almost in the same place, at 9.2.

Now ask for all five at once. Only 7 stocks pass, about 1.6 in 100.

More than half, 52.6 in 100, pass none of the five.

Replay

Fair value here means one formula's estimate of what a share is worth, built from a few numbers in the annual report. If the price is below it, the stock looks cheap by that formula. EPS, or earnings per share, is a year's profit divided by the number of shares. These are sanity checks, not full valuations. For discounted cash flow and the other slower methods, see how to calculate the intrinsic value of a stock.

We used each company's latest annual report and a recent price. "Can't value" means the formula gives no positive answer: a loss, negative book value, shrinking or missing earnings history, negative free cash flow, or (for the cash test only) a bank or insurer.

1. The Graham number: what is the formula?

Graham number = √(22.5 × EPS × book value per share). It's the most a cautious investor should pay under Benjamin Graham's rules. He told defensive investors to pay no more than 15 times earnings and 1.5 times book value, but allowed a trade-off between the two as long as the product stayed under 22.5 (The Intelligent Investor, chapter 14, via Wikipedia). Solve that rule for the price and you get the formula.

Book value is shareholders' equity: what the balance sheet says would be left for owners after paying every debt. Divide it by the share count to get book value per share.

Where to find the inputs in two minutes:

• EPS: the last lines of the income statement in the 10-K (the annual report), labelled "Diluted." Our stock pages, such as Verizon's, show it.
• Shareholders' equity: the balance sheet. Use the line for the company's own shareholders, not the one that includes outside owners of subsidiaries (noncontrolling interests), and subtract any preferred stock.
• Shares outstanding: the 10-K cover page.

Verizon passes: 10.9 times earnings times 1.78 times book is 19.4, under Graham's 22.5.

When to ignore it: after heavy buybacks. When a company spends more on its own shares than it has earned and kept, book value shrinks, sometimes below zero. McDonald's balance sheet carries $79.3 billion of repurchased shares, more than the $70.3 billion of profit it has kept over its history. Its shareholders' equity ended the latest year at −$1.79 billion, so it has no Graham number at all. It's one of 28 S&P 500 companies with negative equity, along with Lowe's, Starbucks, AbbVie, Booking, AutoZone and Philip Morris. Even where book value is positive it can be tiny: Apple's is $5.02 a share, which gives a Graham number of $29 against a recent price of $342.94. Brands, software and patents barely show up in book value, so the formula almost never likes them.

The Graham number goes blank for 56 S&P 500 companies

28have negative equityMcDonald's, Lowe's, Starbucks, AbbVie

28lost money in their latest yearIntel, Ford, Dow

23of its 42 cheap picks are financialsNo health care or consumer staples stock passes

2. Graham's growth formula: how much is growth worth?

Value = EPS × (8.5 + 2g) × 4.4 ÷ Y (Benjamin Graham formula, Wikipedia):

• 8.5 is the P/E (price-to-earnings ratio) Graham thought fair for a company with no growth.
• g is the yearly earnings growth you expect over the next seven to ten years, as a whole number: 10 for 10%.
• 4.4 was the average AAA corporate bond yield when he first published the formula.
• Y is that yield today. Higher bond yields mean lower stock values.

Don't skip Y. Moody's Aaa corporate bond yield was 6.08% recently, so 4.4 ÷ 6.08 cuts every Graham value by 28% compared with his baseline. At today's yield, a company with no growth is worth about 6.2 times earnings.

For g, we used past growth as a stand-in: the yearly growth rate of diluted EPS over the past four years. The free CAGR calculator turns two EPS figures into that rate.

Worked example, Microsoft. Diluted EPS was $17.95 in its latest fiscal year, up from $9.65 four years earlier: 16.78% a year.

• 8.5 + 2 × 16.78 = 42.07
• $17.95 × 42.07 = $755
• $755 × 4.4 ÷ 6.08 = $546

At a recent price of $533.52, Microsoft sits 2% below its Graham value. That "cheap" verdict rests almost entirely on growth continuing. Change one input and see.

Microsoft's Graham value swings from $240 to $1,050 on two inputs

Pick the growth rate and the AAA bond yield

EPS growth a year

5%10%16.8%25%

AAA bond yield

4.4%5.0%6.08%

Graham value per share$332$292$240$512$450$370$755$665$546$1,050$924$760

−37.8%−45.3%−55.0%−4.0%−15.7%−30.6%+41.5%+24.6%+2.3%+96.8%+73.2%+42.5%below Recent pricebelow Recent pricebelow Recent pricebelow Recent pricebelow Recent pricebelow Recent priceabove Recent priceabove Recent priceabove Recent priceabove Recent priceabove Recent priceabove Recent price

Recent price $533.52

$240$1,050

Past four yearsOur capGraham's baselineRecently

Where it misleads: fast growth. The 2g term takes over. At 25% growth the formula pays 42 times earnings even at today's bond yields, and past growth from a low starting point can be far higher than that (more on that in the PEG section). We capped growth at 25%. Graham himself called his projections illustrative and warned readers not to treat them as reliable (Wikipedia). It also can't run without a growth rate, which rules out 54 S&P 500 companies: 50 that lost money in the latest year or four years earlier, and 4 recent listings without the history.

In our run it was the most generous formula to big tech. It's the only one of the five that calls Microsoft, Alphabet and Amazon cheap.

3. Peter Lynch's PEG rule: is the P/E below the growth rate?

Fair P/E = growth rate, so fair price = EPS × growth rate. Lynch wrote that "the P/E ratio of any company that's fairly priced will equal its growth rate" (PEG ratio, Wikipedia). The PEG ratio is the P/E divided by that growth rate. Below 1 looks cheap; above 1 looks expensive.

For dividend payers, Lynch added the dividend yield to growth and divided by the P/E. On that version, under 1 is poor, 1.5 is okay and 2 or better is what you want (One Up on Wall Street, quoted by Stock Talk).

Worked example, Alphabet. Diluted EPS was $10.81 in the latest year and $5.61 four years earlier: 17.8% growth a year.

• P/E at a recent price of $344.18: 31.8
• PEG: 31.8 ÷ 17.8 = 1.79
• Lynch's fair price: $10.81 × 17.8 = $193
• With dividends: the $0.83 dividend is a 0.24% yield, so (17.8 + 0.24) ÷ 31.8 = 0.57, "poor" on Lynch's scale

Alphabet trades 79% above Lynch's fair price

Lynch fair price$193

Recent price$344+79%

$150$200$250$300$350$400

The trap: shrinking or erratic earnings. If EPS fell, the rule gives a fair P/E of zero or less, which means no answer. That's the case for 137 S&P 500 companies, which earn less per share than four years ago: UnitedHealth's EPS went from $18.08 to $13.23, and Pfizer's from $3.85 to $1.36. Add the loss-makers and four recent listings with no history, and the PEG rule can't value 191 of the 445 stocks, 43% of the index. The opposite trap is a tiny starting point. Delta earned $0.44 a share four years ago and $7.66 last year, which works out to 104% growth a year. That pace says more about the starting year than about the business. Adding dividend yields, as Lynch suggested, only lifts the count of cheap stocks from 41 to 44.

To see what happens when growth arrives but the P/E shrinks anyway, try the P/E multiple compression calculator.

4. Earnings yield vs the 10-year Treasury: does the stock out-earn a bond?

Fair price = EPS ÷ 10-year Treasury yield. Earnings yield is the P/E flipped over: EPS ÷ price, the profit you're buying per dollar. A 10-year Treasury recently paid 5.13%. A stock that earns less than that on its price needs growth to make up the gap. At 5.13%, the break-even P/E is 19.5. This comparison is often called the Fed model.

Worked example, Coca-Cola. EPS was $3.04 last year. At a recent price of $93.42, that's an earnings yield of 3.25%, below the Treasury's 5.13%. Fair price at the Treasury yield: $3.04 ÷ 0.0513 = $59.26. Coca-Cola trades 58% above it, so the price assumes its profits keep growing.

That's typical. The median S&P 500 stock in our run earns 3.7% on its price, so most of the index fails this test.

Of these eight, only Verizon and JPMorgan out-earn a 10-year Treasury

Verizon9.20%

JPMorgan Chase6.00%

10-year Treasury5.13%

ExxonMobil3.76%

Microsoft3.36%

Coca-Cola3.25%

Nvidia2.26%

Apple2.18%

Earnings yield is diluted EPS from the latest annual report divided by a recent price.

Where it fails: a boom year. A bond's coupon is fixed; a company's earnings are not. Cyclical stocks, like steelmakers, can look cheapest when their profits are near the top. Nucor shows how far one year can move the answer.

Boom-year earnings made Nucor look three times as valuable

Earnings from four years ago$23.16 a share

• Earnings yield at today's price: 8.9%
• Fair price at 5.13%: $451

Looks cheap

vs

Latest year's earnings$7.52 a share

• Earnings yield at today's price: 2.9%
• Fair price at 5.13%: $147

Looks expensive

The test also can't value the 28 companies that lost money, and it favours banks: 67% of financials pass, against 15% of health care.

5. Free-cash-flow yield: what price gives you the cash return you want?

Fair price = free cash flow per share ÷ the yield you require. Free cash flow (FCF) is the cash a business brings in from operations minus what it spends on buildings and equipment (capital spending, or capex). It's the cash left to pay dividends, buy back shares or pay down debt. Both numbers sit on the cash flow statement: "Net cash provided by operating activities" and a line such as "Purchases of property and equipment" under investing.

The yield you require is your call. The earnings-yield test only asks whether a stock beats a bond; this one asks what you'd want, so we set the bar a little higher. We used 6% for the S&P 500 run, about one point above the Treasury, because a company's cash is less certain than a bond's coupon. A 6% hurdle means paying at most 16.7 times FCF.

Worked example, Apple.

Apple keeps $98.8 billion of its $111.5 billion operating cash

Operating cash flow$111.5B

Capital spending−$12.7B= $98.8B

Free cash flow$98.8B

$6.73 of FCF a share against a recent price of $342.94 is a 1.96% yield. To earn 6%, you'd pay $6.73 ÷ 0.06 = $112. Apple costs three times that.

When to ignore it: during a building spree. Amazon's operating cash flow was $139.5 billion last year, but it spent $131.8 billion on capex. That leaves $7.7 billion, or $0.72 a share, and a "fair" price of $12 against a recent $262.45. The formula reads heavy investment as a lack of cash, whether or not the spending pays off. Utilities hit the same wall: 22 of 29 spend more than they bring in, and none passes. We skipped banks and insurers (40 companies), because their operating cash flow mixes in movements of customers' money: deposits, loans and insurance premiums.

How do you find the fair value of a stock in 15 minutes?

Run the two tests tied to today's interest rates first, then the Graham number, then the two that need a growth rate, the shakiest input. Fifteen minutes is realistic if a stock page lists the numbers for you; pulling them from two 10-Ks the first time takes longer.

Fifteen minutes, sturdiest test first

1. 0 to 3 minCollect the inputs
2. 3 to 5Earnings yield
3. 5 to 7Free-cash-flow yield
4. 7 to 9Graham number
5. 9 to 12PEG rule
6. 12 to 14Graham growth formula
7. 14 to 15Tally and check the traps

1. Minutes 0 to 3, collect. From the latest 10-K: diluted EPS, shareholders' equity, shares outstanding, operating cash flow and capex. From the 10-K four years back (or a stock page with history): diluted EPS then. From the web: a recent price, the 10-year Treasury yield and Moody's AAA yield.
2. Minutes 3 to 5, earnings yield. EPS ÷ price. Above the Treasury yield? Note pass or fail.
3. Minutes 5 to 7, FCF yield. (Operating cash flow − capex) ÷ shares ÷ price. Above your hurdle?
4. Minutes 7 to 9, Graham number. Equity ÷ shares, then √(22.5 × EPS × that). Skip it if equity is negative.
5. Minutes 9 to 12, PEG. Growth in percent = ((EPS now ÷ EPS four years ago) to the power 0.25 − 1) × 100, so 0.168 becomes 16.8. If the share count jumped several times over between the two reports, there was a stock split: divide the old EPS by the split ratio first. Cap growth at 25 and skip the rule if it's zero or less. Then P/E ÷ growth.
6. Minutes 12 to 14, Graham's growth formula. EPS × (8.5 + 2 × growth) × 4.4 ÷ AAA yield, with the same capped growth (use 0 if it's negative).
7. Minute 15, tally and check the traps. Count the passes. Then ask four questions: Is this a peak year for profits? Was the starting EPS tiny? Is capex unusually high or low? Has buyback spending wiped out book value? A yes on any of them means that formula's answer doesn't count.

How often do the five formulas agree?

Almost never. Across 445 stocks, 211 pass at least one test, but only 7 pass all five, and more than half pass none.

More than half the S&P 500 fails all five tests; 7 stocks pass every one

No formula234

One101

Two51

Three34

Four18

All five7

Number of the five formulas that call each of 445 S&P 500 stocks cheap at a recent price.

These are the seven:

Return is the share price change plus dividends over roughly the past 12 months. Over the same stretch, SPY, an S&P 500 index fund, returned 13.6%.

Six of the seven fell while the index rose. That's the catch: a stock that passes every backward-looking test is usually one investors have been selling. Check the latest year before trusting a P/E of 3 or 4. Charter's is low because its price fell, not because profit jumped: EPS was $36.21, up from $34.97 the year before. Comcast's is different. Its pre-tax profit includes $10.1 billion of other non-operating income, against $0.5 billion the year before, and that line alone is 39% of the total.

The household names on most watchlists mostly fail. Bold marks a fair value above the recent price.

Alphabet clears Graham's growth formula by about a dollar, which is noise, not a margin of safety. When the formulas split like this, the better question is what growth the price already assumes. The free reverse DCF calculator answers it in a minute.

Which sectors does each formula favour?

The Graham number is mostly a bank screen, earnings yield favours financials, communication services and utilities, and free-cash-flow yield favours consumer discretionary companies and shuns utilities.

Share of each sector's stocks the formula calls cheap. Financials score low on FCF yield because we didn't run that test on banks and insurers.

The skews follow from the inputs. Banks carry most of their value on the balance sheet, so book-value formulas suit them. Drugmakers and software companies keep theirs in patents, code and brands that book value ignores, so the Graham number passes 2% of tech stocks and none in health care. Utilities borrow to build, so their free cash flow is negative. Running one formula across a whole portfolio turns it into a sector bet.

Our read

These formulas are filters that tell you where to look, not price targets. Passing all five means "look here first," not "buy."

If we kept two, they'd be earnings yield against the Treasury and free-cash-flow yield, run together. Earnings yield can value 94% of the index and free-cash-flow yield 81%, even after skipping banks and insurers, against 57% for the PEG rule. Both use today's interest rate as the bar to clear. And they check each other: accounting can flatter earnings, a capex burst can hide cash, and a stock that clears both is less likely to be flattered by either. Only 53 S&P 500 stocks do. The Graham number is the one we'd drop first: it punishes buybacks, can't see what brands and software are worth, and more than half its picks are financials.

Go deeper with member research

Comcast's $10.1 billion of other income is the problem in one line: every formula here divides by a single year, so a gain or a peak makes a stock look cheap and a bad year makes it look expensive. The member chapter Normalize before you compare shows how to build a representative year: strip out one-off gains and charges, carry the change through to EPS, spot a cyclical peak before you trust a low P/E, and keep a written bridge from reported to normalized earnings. It's one of 40 chapters in the fundamental research course. When you're ready to turn a fair value into a buy price, Margin of safety and expected return is the next step.

Keep reading

• 5 ways to calculate the intrinsic value of a stock: DCF, reverse DCF, multiples, dividend discount and book value, worked on real filings.
• Free stock valuation calculators, tested: which tools give a sensible fair value for the same stock.
• 6 numbers to check before buying a stock, with what's normal across the S&P 500.
• What the P/E ratio tells you, and what it doesn't.

Questions readers ask

What does the fair value of a stock mean?

An estimate of what one share is worth, based on the company's earnings, assets or cash, set against the market price people pay today. Each formula gives its own estimate.

Is there a fair value of a stock calculator?

All five formulas fit on a phone calculator. For a fuller model, our free DCF calculator values a stock from its future cash flows, and our calculator roundup compares other free tools.

Which stocks trade below their Graham number?

In our run, 42 of 445 S&P 500 stocks, 23 of them financials. They include Bank of America, Wells Fargo, Citigroup, Verizon and AT&T.

What's the difference between fair value and market value?

Market value is the price times the number of shares, also called market capitalization. Fair value is an estimate of what those shares should be worth. The gap between the two is what every formula on this page tries to measure.

How we built this

• Universe: 445 of the 460 S&P 500 companies in our SEC fundamentals dataset (built from the SEC's XBRL frames API, synced 2026-10-19). We dropped 15 whose latest annual EPS isn't filed in SEC's standard XBRL tags or was a year out of date: Visa, Berkshire Hathaway, Hershey, Monster Beverage, Freeport-McMoRan, KKR, Exelon, Ares Management, Kimco Realty, Regency Centers, Constellation Brands, VeriSign, Halliburton, Lennar and Baker Hughes.
• Company data: each company's latest annual report, as of 2026-10-22. That's mostly fiscal years ending in 2025; 26 companies whose fiscal year ended in mid-2026, including Microsoft, Costco and Procter & Gamble, use that newer report from the SEC XBRL frames API. Diluted EPS, shares outstanding and operating cash flow come from the dataset above. Shareholders' equity (the parent company's share, less preferred stock), capital spending and dividends per share come from SEC XBRL frames and companyfacts. For eight companies that tag capex with their own labels (Cigna, ConocoPhillips, DTE Energy, Federal Realty, NextEra Energy, Phillips 66, United Rentals and Valero) we read it from the 10-K's XBRL file.
• Growth: compound annual growth in diluted EPS from four fiscal years before the latest one (mostly 2021 to 2025, which skips the 2020 pandemic trough). We adjusted for stock splits by chaining restated EPS across successive 10-Ks (21 companies, including Nvidia, Broadcom and Netflix). No growth rate when either year was a loss (50 companies) or history was missing (4 recent spin-offs and listings). Growth is capped at 25% (45 companies hit the cap). Graham's growth formula floors growth at zero; the PEG rule needs growth above zero.
• Rates: 10-year Treasury 5.13% (US Treasury daily par yield curve, 2026-10-21). Moody's Aaa corporate yield 6.08% (FRED series DAAA, 2026-10-20). Free-cash-flow hurdle: 6%.
• Prices: the latest close in our Yahoo Finance price file, 2026-10-19. Yahoo's API was rate-limiting requests during the run, so we used the stored closes. Past-year returns use the same file's dividend-adjusted monthly closes, from the end of October 2025.
• Exclusions from the cash test: 24 banks (companies reporting deposits) and 16 insurers (financials reporting earned premiums).
• Cheap means the recent price is below the formula's fair price. Sectors are the GICS sectors on our stock pages.