In large U.S.-style markets, stock return predictability comes mainly from future price gains, not future dividend growth. A high dividend yield signals higher expected returns because smoothed payouts barely move with yields.
Here dividend yield means annual cash dividends divided by stock price. That answer shifts in smaller markets, where yields often forecast dividend changes instead. Readers can use that split to judge timing claims, pick better payout measures, and keep return expectations grounded.
Table of Contents
- Why a high yield must predict something
- Why U.S. dividends stopped answering yields
- Where the signal flips abroad
- What investors should watch instead
Why a high yield must predict something
The logic starts with present value. A stock price equals expected future dividends, discounted for risk and time. When dividends look high against price, future returns must rise, future dividend growth must fall, or both adjust.
The Yale Cowles Foundation Discussion Paper 858 sets out the present-value identity. This tradeoff forces a choice on every yield signal. Investors cannot read a high yield as pure good news. They must ask whether it points to reward for holding risk or to coming payout weakness.
Why U.S. dividends stopped answering yields
After World War II, U.S. dividends barely responded to yields. John Cochrane found almost no link from yield to later dividend growth, and read that silence as evidence that yield moves reflect expected returns. The Review of Financial Studies via EconPapers publishes Cochrane's dividend-variability analysis.
Large firms smooth payouts to avoid cuts and sharp hikes. Chen, Da and Priestley argue that smoothing breaks the link between yields and dividend growth. That break explains why dividends looked forecastable before World War II but not after. Prices swing with news while cash dividends stay steady. Yield then moves because the denominator moves, not because managers signal payout change.
Where the signal flips abroad
In smaller markets, yields often predict dividend growth. Rangvid and co-authors find that split. Growth forecasts dominate where payouts move freely, while return forecasts dominate in large smoothed-dividend markets. Testing method also matters. Traditional regressions find return predictability in only the U.K.
and France across a broad country sample. Joint tests in the Cochrane style, reported in Revista Contabilidade & Finanças, find wider return predictability because they use the return-growth tradeoff directly. An investor cannot import a U.S. rule abroad. Local payout policy decides whether yield warns about cash flow or points to expected return.
What investors should watch instead
Cash dividends now miss much of shareholder payout. Boudoukh and colleagues show that buybacks replaced dividends, so total payout yield and net payout yield predict returns better than dividend yield alone. Even good in-sample signals often fail in live use. Goyal and Welch show in out-of-sample equity-premium tests that dividend-price, earnings-price and similar predictors generally do not beat the simple historical-average return forecast.
Readers should treat yield as context, not a timing trigger. A simple check helps: Do not judge equity gains from price alone. Advisor Perspectives reports the S&P 500 dividend comparison. That comparison puts the end-1929 to March 2012 annualized return at about 9.4 percent with reinvested dividends, against 5.2 percent on price alone.
- Track total payout yield when buybacks are large.
- Ask whether local firms smooth dividends before trusting a yield signal.
- Judge any timing rule against the historical-average return out of sample.
For the whole study explained, with every figure checked against the September 18, 2026 version of the paper, read The CAPE That Cried Wolf: What CAPE-H Says About Stock Valuations.