DCF vs Multiples: When Each Is Right

Last updated October 29, 2026·5 min read

Two families of valuation divide the field. Discounted cash flow builds a value from the company's own projected cash. Multiples borrow a price from the market, taking what investors pay for similar earnings or cash flow and applying it to this company. Neither is the correct method. Each is correct somewhere, and this guide runs both on the same companies so the differences are visible, then gives the conditions that should steer the choice.

What Each Method Actually Assumes

A DCF assumes you can say something defensible about a company's future cash: how much it produces now, how fast that grows, and what return you require. Its answer is intrinsic. It does not care what the market thinks of the sector this quarter.

A multiple assumes the market's pricing of comparable companies is roughly sane, and that this company's earnings or cash flow deserve a similar price. Its answer is relative. It inherits the market's mood, including the manic parts.

That single difference drives every practical consequence below. When a whole sector is mispriced, multiples import the mispricing with a clean conscience. When a company's future is genuinely unknowable, a DCF manufactures precision the situation does not contain.

One Company, Both Methods

Apple, as Stock Valuation Studio printed it on October 27, 2026, at a price of $342.58:

• The DCF, projecting free cash flow growing at 9.3 percent fading toward 3 percent, discounted at 9 percent, printed $236.87.
• The multiples model printed $168.31, the average of two legs: fiscal 2025 earnings of $7.46 a share priced at a fair price-to-earnings ratio of 20, which is $149.20, and free cash flow of $9.37 a share priced at a fair multiple of 20, which is $187.40.
• A third relative, the forward estimates model, priced next year's consensus earnings of $9.55 a share at the same fair multiple and printed $191.09.

The market price stood above all three. The methods disagree with each other by more than 40 percent, from $168.31 to $236.87, and that spread is itself the finding: Apple in late 2026 was a company whose value hinged on growth that multiples, anchored to today's earnings, could not see, and that a DCF could only reach by projecting it. When the methods split like this, the argument is not about arithmetic. It is about which assumption you are willing to own.

A Second Company, Where They Converge

Coca-Cola in the same week traded at $92.82. Its earnings of $3.04 a share put the price at 30.5 times earnings, which reads expensive for a company growing slowly. But its dividend models, built in the dividend guide in this series, printed values from $89.75 to $96.56 depending on the form, bracketing the price. For a mature payer with a long record, the income-based methods and the price were telling roughly the same story, while the raw earnings multiple, with no sense of the payout or the balance sheet behind it, was the outlier. Convergence among unlike methods is the strongest signal either family can send.

A Live Case of the Cycle Problem

Microsoft shows the split inside a single company. Net income rose from $88,136 million in fiscal 2024 to $133,749 million in fiscal 2026, while free cash flow fell from $74,071 million to $66,987 million over the same two years, because capital spending climbed to $115,948 million for data centers. Priced off earnings, the stock in late October 2026 stood at 29.5 times earnings. Priced off free cash flow per share of $9.02, it stood at 58.8 times. Both multiples are honest arithmetic on filed figures. They disagree because one denominator is spending through a building cycle and the other is not, and a DCF built on the current cash flow inherits the same trough. Which denominator deserves the weight is the entire argument about the stock, and no method settles it. The methods frame it.

When the DCF Is Right

• The company produces real, recurring free cash flow today, so the base is a measurement rather than a hope.
• Its economics set it apart from peers, so the market's average price for the sector is the wrong anchor: unusual margins, an unusual moat, an unusual capital structure.
• The question is long-horizon: what is this business worth if held for a decade, through whatever mood the market is in this year.
• The whole sector looks mispriced. A relative method cannot detect a bubble it is standing inside.

When Multiples Are Right

• The company is young, acquisitive, or mid-transition, and its cash flow path is a story rather than a trend. A DCF here is a novel with equations.
• The decision is comparative: ranking five similar banks, grocers, or landlords against each other, where one consistent yardstick beats five bespoke models.
• Cash flow is distorted by an investment cycle. A company building capacity prints thin free cash flow now for real reasons; its earnings multiple may carry the steadier signal, read with care.
• You need a sanity check on a DCF. When a careful projection lands at three times what any peer commands, the projection is the first suspect.

The Working Habit

Professionals do not choose a family once. They run the DCF for the intrinsic claim, run the multiples for the market's claim, and interrogate the gap. A gap that traces to a stated, arguable assumption, such as growth the peers have not achieved, is information. A gap that traces to an input nobody can explain is an error looking for a victim.

Sector routing still governs both. For banks, insurers, and real estate investment trusts, the cash flow version of each method bends or breaks, and the sector guide in this series covers the replacements. The models hub gathers all 31 models the Studio runs, cash flow and multiples side by side, and shows which one carries the headline for a given company and why.

Sources

• Stock Valuation Studio prints for Apple, October 27, 2026: DCF $236.87, multiples $168.31 (legs $149.20 and $187.40), forward estimates $191.09, price $342.58. The Coca-Cola Company figures from fiscal 2025 filings: earnings per share $3.04, price $92.82 on October 29, 2026. Filings are public at SEC EDGAR