Skip to main content

How Bond Yields Affect Equity Multiples

Updated Jun 24, 2026 at 2:35pm

Research Draft Medium 1,210 words

A stock's price-to-earnings (P/E) multiple is, at bottom, the present value of a stream of future earnings divided by current earnings — and bond yields are the benchmark "risk-free" rate against which that discounting is anchored. When the yield on government debt rises, the rate used to discount future corporate cash flows tends to rise with it, mechanically compressing what investors will pay per dollar of earnings; when yields fall, multiples have room to expand. The core tension is that this textbook discount-rate channel is real but conditional: bond yields move for different reasons (growth optimism, inflation fear, fiscal stress, policy tightening), and which reason dominates often matters more for equity multiples than the yield level itself.

How the mechanism works

The link runs through the dividend discount model. In its simplest Gordon Growth form, price equals next year's cash flow divided by (discount rate minus growth rate), where the discount rate = risk-free rate + equity risk premium (ERP). Holding cash flows and the ERP fixed, a higher risk-free rate (the bond yield) shrinks the denominator's distance and lowers the justified price — hence a lower multiple. Three sub-channels:

1. Discount-rate / present-value channel — the dominant one. Higher yields raise the rate applied to all future earnings, reducing their present value. This is the "long-duration" effect. 2. Cost-of-capital / earnings channel — higher yields raise corporate borrowing costs and can slow the economy, reducing the level of future earnings (the numerator), not just the multiple. 3. Competition-for-capital channel — when bonds offer higher yields, the relative appeal of stocks falls, pressuring valuations at the margin (the intuition behind the "Fed model," critiqued below).

Duration is why the effect is uneven. Growth stocks derive most of their value from cash flows far in the future; value stocks lean on near-term earnings. The present value of distant cash flows falls more sharply than near-term cash flows when rates rise, so high-multiple growth names are more rate-sensitive — they behave like long-duration bonds. This is the textbook explanation for why the 2022 rate-hike cycle hit the Nasdaq-100 and unprofitable tech far harder than value, with value outperforming growth as the 10-year Treasury climbed toward roughly 4% (T. Rowe Price; Direxion).

A commonly cited sensitivity from Goldman Sachs Research's macro valuation model: a change of roughly 50 basis points in real (inflation-adjusted) Treasury yields is associated with on the order of a 3% move in the S&P 500 forward P/E in the opposite direction (i.e., very roughly mid-single-digit percent per 100bp), all else equal — but treat this as one firm's point estimate that varies across their published notes, not a fixed coefficient. Real yields, not nominal, are generally treated as the cleaner driver because nominal earnings tend to rise with inflation.

How it's used in practice

Practitioners use the relationship in three ways:

  • Multiple framing / fair-value anchoring. Analysts back out an "implied" or "warranted" P/E by plugging a discount rate (built off the 10-year yield plus an assumed ERP) into a DDM, then compare it to the market multiple to judge over/undervaluation.
  • Sector and style rotation. Rising-yield regimes favor short-duration, cash-generative value, financials (which can benefit from higher net interest margins), and energy; falling-yield regimes favor long-duration growth and rate-sensitive sectors like utilities and REITs.
  • Earnings yield vs. bond yield comparison. Investors watch the equity earnings yield (E/P) against the 10-year yield — the "equity risk premium" spread — as a rough gauge of relative attractiveness.

Adoption, debate & evidence

The qualitative claim — that lower yields support higher multiples and vice versa — is mainstream and broadly supported by valuation theory and observed cross-sectional duration effects. What is genuinely contested is the simple, mechanical version of it: the Fed model, which holds that the equity earnings yield should equal the 10-year Treasury yield.

  • The correlation is real but highly regime-dependent. The contemporaneous correlation between the 10-year nominal yield and the aggregate equity earnings yield was strong in some windows (commonly cited figures include roughly 75% over 1995–2000 and around 0.81 over 1965–2001) but weak over the long sweep — only about 19% over 1881–2002 (and similarly low, ~0.18, over 1926–2001). The Fed model only "worked" reliably in two U.S. periods, roughly 1921–1928 and 1987–2000, and the relationship broke down after 2002 (Fed model literature; Wikipedia summary; Asness 2003; Estrada).
  • Asness's critique is decisive on theory. Cliff Asness ("Fight the Fed Model," 2003) showed the model compares a real number (earnings yield, which rises with inflation) to a nominal one (the bond yield), conflating the two via "money illusion." When inflation falls, discount rates fall — but so do future nominal cash flows, so the two effects partly cancel. Lower rates do not automatically justify higher multiples "all else equal," because all else is not equal.
  • Direction depends on why yields move. Goldman and others stress that equities can rise with yields when yields climb on stronger growth expectations, but struggle when yields rise on fiscal or inflation fears. The level of yields is less informative than the driver.

Honest summary: the discount-rate channel is sound and useful as a directional, conditional heuristic. The precise "earnings yield should equal bond yield" formulation is widely regarded by academics as theoretically flawed and a poor predictor of future returns.

Strengths & limitations

Works best when real yields move sharply and for clear reasons (e.g., a hawkish policy repricing), and for ranking relative rate sensitivity across styles (growth vs. value, long- vs. short-duration). The duration framework reliably explains cross-sectional dispersion even when it can't time the index.

Fails when used as a standalone valuation or timing signal. The ERP is not constant, growth expectations move simultaneously with yields, and the nominal-vs-real confusion (the Fed model's flaw) produces false signals — most notoriously implying stocks were "cheap" relative to bonds throughout high-inflation regimes.

The #1 misuse: treating the earnings-yield-minus-bond-yield spread as a precise fair-value gauge, ignoring that a high-inflation environment lowers both the justified multiple and future real cash flows. A second common error is assuming all stocks respond equally — they do not; duration dominates.

Sources

Dispute flagged: the Fed model's earnings-yield-vs-bond-yield equivalence is empirically and theoretically contested; treat the discount-rate channel as a conditional heuristic, not a precise rule. The Goldman real-yield-to-forward-P/E sensitivity (~3% per 50bp) is one firm's model estimate that varies across their notes, not a universal constant.