The Equity Risk Premium Puzzle
Why the size of the historical equity premium is larger than standard economic models predict it should be — and what that might mean.
Siegel engages directly with what economists call the "equity premium puzzle" — the observation, made prominently by economists Rajnish Mehra and Edward Prescott, that the historical size of the equity premium (the extra return stocks have delivered over bonds) is considerably larger than standard economic models of rational risk-aversion predict it should be, given how much genuine additional risk stocks appear to carry relative to bonds over reasonable investing horizons.
The book surveys several proposed explanations without fully endorsing a single one: that investors are more loss-averse than standard models assume, making them demand extra compensation for stocks' volatility beyond what pure risk math would justify; that rare, severe economic disasters (which are absent or underweighted in the observed historical sample) should rationally push the required premium higher even if they did not actually occur in the specific sample studied; and that transaction costs, taxes, and other frictions in earlier eras made stocks harder to hold than the raw return data alone suggests, requiring a higher premium to compensate.
The puzzle isn't purely an academic curiosity — it bears directly on whether an investor should expect the historical equity premium to persist going forward. If the premium is explained mostly by investors being more loss-averse than rational models predict (a persistent feature of human psychology), it would reasonably be expected to continue; if it is explained mostly by frictions and disaster risk specific to the historical sample studied, a smaller premium going forward would be more plausible. Siegel does not claim to fully resolve which explanation dominates, but treats the puzzle as a genuine reason for some humility about how confidently the exact historical premium should be extrapolated into the future, even while still expecting stocks to outperform bonds over the long run in some form.
- The historical equity premium is larger than standard rational-risk-aversion models predict it should be — the "equity premium puzzle."
- Proposed explanations include greater investor loss-aversion than standard models assume, disaster risk absent from the observed historical sample, and transaction costs and frictions specific to earlier eras.
- The puzzle is a reason for some humility about extrapolating the exact historical premium forward, even while still expecting a meaningful long-run premium to persist.