The short answer: A discounted cash flow (DCF) estimates intrinsic value directly from a company's own projected cash flows, discounted back to today. A multiples approach estimates value by comparing a company's earnings, revenue, or cash flow to what similar companies trade for in the market. DCF is more rigorous but extremely sensitive to its assumptions; multiples are faster and grounded in real market prices but inherit whatever mispricing already exists in the comparison group. Careful analysis typically uses both and treats a large disagreement between them as a signal to dig deeper, not to pick whichever number is more convenient.
| DCF | Multiples | |
|---|---|---|
| What it estimates from | The company's own projected future cash flows | What comparable companies currently trade for |
| Main inputs | Growth rate, margins, discount rate, terminal value | A peer group and a multiple like P/E or EV/EBITDA |
| Biggest weakness | Small changes in assumptions swing the answer a lot | Assumes the peer group is itself fairly priced |
| Speed | Slow, requires building a full projection | Fast, a few inputs and a comparison |
| Best suited to | Businesses with predictable, forecastable cash flow | Businesses with clear, actively traded comparables |
A discounted cash flow model tries to answer the valuation question from first principles: project how much cash a business will generate in the future, then discount those future dollars back to what they're worth today. A multiples approach skips that entire exercise and instead asks a simpler question: what are similar businesses trading for right now, relative to their earnings or cash flow, and what does that imply this one should be worth.
Both methods are trying to get at the same underlying idea, what is this business actually worth, but they get there from opposite directions. DCF builds the number from the ground up, using nothing but the company's own expected future performance. Multiples borrow the number from the market's current opinion of comparable businesses, on the assumption that similar businesses should trade at similar valuations.
How a DCF actually works
You forecast a business's free cash flow for a number of years out, estimate a terminal value for everything beyond that window, and then discount the whole stream back to the present using a rate that reflects the riskiness of those cash flows. Every one of those inputs is an assumption, and DCF models are notoriously sensitive to them. Nudge the discount rate down by a percentage point or two, or push the growth assumption out a few more years, and the resulting value can move by a wide margin. The rigor of the method is real, but it can create a false sense of precision if you forget how much the output depends on inputs you're essentially guessing at.
How a multiples approach actually works
Pick a metric, earnings, revenue, EBITDA, free cash flow, find a group of genuinely comparable businesses, and see what multiple of that metric the market is currently paying for them. Apply a similar multiple to the company you're analyzing and you have a quick estimate of value. It's fast, and it's anchored to real transactions rather than a hypothetical future, which is its main appeal. Its weakness is baked into that same appeal: if the peer group is itself overvalued or undervalued, the multiple you're borrowing carries that mispricing straight into your answer.
- DCF answers 'what is this worth on its own merits,' independent of what the market currently thinks.
- Multiples answer 'what is the market currently willing to pay for something like this,' which isn't the same question.
- A DCF built on unrealistic growth assumptions can justify almost any price, which is why the assumptions deserve more scrutiny than the model's math.
- A multiple applied during a period when an entire sector is overpriced will simply reproduce that overpricing in your estimate.
“Intrinsic value is an estimate rather than a precise figure.”· Berkshire Hathaway Owner's Manual
Where each one tends to break
DCF breaks down fastest on businesses with unpredictable or lumpy cash flow, where a small change in near-term assumptions changes the whole trajectory, and on very early-stage companies where most of the value sits in a terminal value that's really just a guess wearing a formula. Multiples break down fastest during periods when an entire market or sector is being priced by sentiment rather than fundamentals, and on businesses without a genuinely comparable peer group, where the comparison itself is the weak link.
In practice, experienced analysts rarely pick one method and discard the other. A DCF forces you to be explicit about the assumptions driving a business's value, which is a useful discipline even when the output itself is imprecise. A multiples check keeps that DCF honest against what the market is actually paying for similar businesses today, and catches you if your model has quietly drifted into fantasy. Used together, each one catches the other's blind spot.