The historical equity premium
The US equity market has returned approximately 9–10% per year in nominal terms since 1900, while risk-free short-term Treasury bills have returned approximately 3–4%. This leaves an equity risk premium (ERP) of roughly 5–7% per year. The question: why does this premium exist? Standard expected utility theory suggests investors demand compensation for bearing risk — but the magnitude of the observed premium implies a coefficient of relative risk aversion of around 30–40, when empirical estimates of human risk aversion suggest values of 1–5. At risk aversion of 30–40, humans would refuse a coin flip where they win 50% of their wealth or lose 1% — manifestly implausible.
Proposed explanations
Survivorship bias: the US is one of history’s most successful economies. Averaging across all markets (including those that experienced war, hyperinflation, or political collapse) reduces the estimated ERP significantly. Rare disaster risk (Barro): if investors fear catastrophic economic collapses (25%+ consumption declines) that occur rarely but with devastating consequences, they demand a large premium even if such events don’t appear in the recent historical record. Behavioural explanations: loss aversion (Kahneman and Tversky) and myopic loss aversion (Benartzi and Thaler) suggest investors evaluate portfolios over short horizons and are overly sensitive to short-term losses, demanding excessive compensation for equity volatility.
Implications for expected future returns
Whether the historical ERP will persist in the future is deeply uncertain. If the high historical premium reflects survivorship bias, future returns will be lower. If it reflects rational compensation for disaster risk that didn’t materialise, it will continue. If it reflects behavioural biases that persist, it will continue as long as those biases do. Most practitioners use forward-looking ERP estimates of 4–5% (lower than the historical realised premium) for valuation purposes, acknowledging that higher equity valuations today imply lower future returns through mean reversion in valuation multiples.
“The equity premium puzzle is a reminder that the most important facts about financial markets are often the ones we cannot yet fully explain.”
What this means for you
The equity premium puzzle has a deeply practical implication: equities deliver substantially higher long-run returns than bonds, and the best explanation is that this premium compensates for genuine uncertainty — the kind that makes investors uncomfortable enough to demand 6%+ per year to hold stocks rather than bills. As a long-term investor who can genuinely hold through volatility (behavioural or rational), you capture a premium that short-horizon, myopically risk-averse investors leave on the table. The puzzle’s persistence suggests this premium is unlikely to be fully arbitraged away — maintaining the case for high equity allocation over long horizons.