Loss Aversion and the Equity Premium Puzzle

The equity premium puzzle is an enduring anomaly in financial theory: stocks have historically returned roughly 5–7 per cent per year more than Treasury bonds, far more than classical asset-pricing models can justify given their volatility relative to bonds. Loss aversion and myopic loss evaluation explain this gap. Investors so deeply dislike short-term volatility and loss that they demand an enormous premium to own equities, even over very long horizons. This psychological requirement drives a persistent wedge between observed and theoretical returns.

The anomaly: observed versus predicted

Classical asset-pricing theory (principally the Capital Asset Pricing Model, or CAPM) connects risk to return through a single parameter: beta, the covariance of an asset’s returns with the market. Higher beta means higher risk and therefore higher expected return. But the theory says the relationship should be roughly proportional: twice the volatility warrants twice the risk premium.

Empirically, this does not hold. A broad stock index might have a beta of 1.0 and return, over the long run, 9–10 per cent annually. A Treasury bond might return 3–4 per cent annually with far less volatility (lower beta). The gap between the two—the equity risk premium—is 5–6 per cent per year.

On the surface, this seems reasonable: stocks are riskier, so they should return more. But when you do the mathematics, the premium is too large. The volatility of stocks relative to bonds does not fully explain the return gap. This is the puzzle: either the market is inefficient, or classical theory is incomplete.

For decades, economists searched for rational explanations. Perhaps investors face restrictions (no short-selling, or limited access to bonds). Perhaps the sample is biased (we observe the most successful market, not all attempts). Perhaps corporate income tax or labour income risk explains the gap. None of these explanations fully resolved the anomaly.

The myopic loss aversion solution

Behavioural finance, specifically work by Shlomo Benartzi and Richard Thaler, offers a compelling alternative: investors are myopic. They evaluate their portfolios too frequently and apply loss aversion at short intervals.

Here is the key intuition:

Suppose you face a long-term investment horizon (30 years) but you check your portfolio monthly. Stocks are volatile; in any given month, they fall about 40 per cent of the time. Bonds are stable; they rarely fall month-to-month. Loss aversion means losses loom roughly twice as large as gains. So a 2 per cent stock market decline in month 3 causes more pain than the pleasure of a 2 per cent gain in month 7.

If you evaluate frequently (monthly), you will experience dozens of losses over your 30-year horizon. Each one triggers loss aversion pain. The rational long-horizon investor should ignore monthly noise and focus on 30-year returns. But the behavioural investor, checking monthly, accumulates a huge subjective disutility from short-term volatility.

To compensate for this accumulated pain, investors demand a much higher long-term equity premium. They are willing to own stocks only if the expected long-term return is so much higher that it justifies the emotional cost of monthly volatility. This creates an endogenous premium: the market price of stocks is depressed (and bond prices are inflated) because investors irrationally dislike short-term volatility.

The mathematics of myopic loss aversion

Benartzi and Thaler formalised this. Suppose an investor evaluates returns monthly and suffers loss aversion. A 10 per cent annual expected return translates roughly to a 0.8 per cent monthly return (compounding). In about 40 per cent of months, returns are negative. These losses, multiplied by the 2× loss-aversion weight, create significant disutility.

Now, the investor requires sufficient additional expected return to offset this monthly loss pain. If the equity premium were only 1 per cent (what CAPM would predict), the monthly disutility would be uncompensated, and the investor would choose bonds instead. The investor thus demands much higher expected returns—the 5–7 per cent premium we observe—to justify holding equities while enduring monthly losses.

Notably, if the investor switched to an annual evaluation (checking the portfolio once per year), the frequency of losses drops to roughly 30 per cent of years, and the required premium declines accordingly. If the investor committed to a 20-year evaluation cycle (no checking), the frequency of losses would be much lower, and the premium could shrink substantially. Over centuries, stocks have realised higher returns than bonds; a truly patient investor would expect this and demand little premium at all.

Empirical support and limitations

The myopic loss aversion model explains several real-world observations:

• Young investors underweighting equities. Retirement surveys show that younger investors (who should have long horizons and should hold mostly equities) often hold large allocations to bonds or cash, even though the myopic model predicts they would if they evaluate frequently.

• Declining participation in equity markets among retail investors. As retail investors gain the ability to check portfolios daily (via smartphones and trading apps), equity-market participation and allocation to stocks have declined in some segments, consistent with the myopic model.

• The size of the premium. The 5–7 per cent premium observed historically is roughly consistent with the magnitude of premium required to compensate monthly loss aversion, under reasonable parameter assumptions.

However, the model has limitations. Not all observed equity premium can be explained by myopia alone; some likely reflects genuine risk factors not captured by simple volatility (tail risk, or non-normal distributions). Additionally, many investors (institutions, long-term funds) do not check portfolios monthly and should rationally accept lower premiums, yet markets serve all investors simultaneously.

Consequences for portfolio construction

The myopic loss aversion explanation has practical implications:

Long-term investors should exploit the premium. If much of the equity premium exists because of investor myopia and short-term loss aversion, a disciplined long-term investor who does not check their portfolio monthly can exploit this. By committing to a longer evaluation horizon (annually or once every 5 years), the investor can rationally accept a lower expected return from equities and still be better off than the myopic investor. Alternatively, the long-term investor can accept the full premium as compensation for patience.

Asset allocation depends on evaluation frequency. An investor’s optimal asset allocation should depend partly on how often they check their portfolio. A daily trader should hold more bonds; a 20-year buy-and-hold investor should hold more equities, and can rationally demand a lower premium for doing so.

Bonds are overvalued in equilibrium. If the equity premium is artificially high due to myopia, bond prices are correspondingly high (bond yields are low). Long-term investors who dislike frequent evaluation may find bonds to be poor long-term investments, because the myopic demand for bonds has inflated their prices beyond fundamental value.

Volatility reduction tools matter. Products that smooth short-term volatility (dividend ETFs, low-volatility strategies) might appeal to myopic investors and command a premium in the market, even if they reduce expected long-term returns. These tools reduce the frequency of losses and thus the loss-aversion pain.

The unresolved tensions

The myopic loss aversion explanation is compelling but not universal. Several tensions remain:

• Institutional sophistication: Large institutions and pension funds are not myopic; they check portfolios regularly but hold equity allocations consistent with long-term horizons. The myopic model does not fully explain their behaviour.

• Asset-class specificity: The myopic model predicts different premiums for different asset classes based on their volatility and loss frequency. Empirically, the premium varies, but not always in the predicted direction.

• Time variation: If myopia alone drove the premium, it should be more stable over time. Yet the observed equity premium has varied significantly across decades, suggesting other factors (changing risk aversion, changing expectations, institutional structure) play a role.

The most likely resolution is that myopic loss aversion explains much of the premium, but not all. Other factors—genuine long-tail risk, labour income correlation, changing demographics—also contribute. The puzzle remains only partially solved.

See also