Two angles on the same picture
Both DCF and multiples try to answer "what's this stock worth?" They look very different — DCF projects cash flows, multiples compare to peers. But under the hood, they're connected. A multiple implies a DCF, and a DCF implies a multiple. Same picture, different angles.
The connection is most visible in P/E. Recall Gordon Growth (with earnings as the cash flow):
Divide both sides by E:
So P/E is just 1 / (r − g). A P/E of 20 implies r − g = 1/20 = 0.05 = 5%. Whatever combination of r and g gives r − g = 5% is consistent with that P/E. So a P/E embeds an implicit assumption about discount rate minus growth rate.
Why higher P/E = higher implied growth (or lower required return)
From the formula above:
- P/E = 10 implies
r − g= 10% - P/E = 20 implies
r − g= 5% - P/E = 50 implies
r − g= 2% - P/E = 100 implies
r − g= 1%
If r is fixed (interest rates don't change), then higher P/E directly implies higher g. If g is fixed, higher P/E implies lower r (lower required return — usually because the company is perceived as safer). In practice both move together for the same company.
Tech stocks trading at P/E = 30-50 implicitly require very high g (or very low r, or both) to justify those prices. Utility stocks at P/E = 12-15 imply modest g. The market is making implicit assumptions about each via the P/E it assigns.
What to do when DCF and multiples disagree
You compute a DCF. You separately compute the implied price using peer multiples. The numbers don't match. Now what?
The disagreement is informative. It means either:
- Your DCF inputs (FCF projections, r, g) are off.
- The peer multiple is wrong (wrong peers, or peers themselves mispriced).
- The market is pricing the company as if your DCF is wrong.
None of these is automatically "right." But the gap forces you to identify which input is suspect and stress-test it. This is why professional analysts compute both — not to get a single number, but to surface where assumptions are most fragile.
The takeaway
DCF and multiples are two angles on the same picture. P/E = 1 / (r − g) for the simple Gordon case — so multiples implicitly embed assumptions about r and g. Higher P/E = higher implied growth (or lower required return). When DCF and multiples disagree, the disagreement points to which assumption you should stress-test. Use both for sanity, not just one.