Wising up to the equity premium

Bryan Caplan, one of my favourite bloggers, is so convinced of the equity premium that he invests accordingly.
I'm not sure this is wise. The problem is that if the equity premium is a genuine anomaly, it should disappear as investors wise up to it; there might be $10 bills on the sidewalk occasionally, but they don't stay there for long.
And investors have had two decades to learn about the equity premium. It's 21 years ago this month that Mehra and Prescott published their original paper (pdf) on it. In those 21 years, US equities have returned 13% a year, against 8.2% for bonds. As a result, according to Robert Shiller's figures, the US market is now highly priced relative to history; a price-earnings ratio of 18 compared to a long-term average of 13.8 using 12 months' earnings, or a PE of 26.6 against a long-run average of 14.9 using 10-year earnings.
These high prices might be a sign that investors have partially wised up to the equity premium, and have driven prices up to levels from which subsequent returns will be lower.
Put it this way. Let's assume share prices rise in line with money GDP growth over the long-run. Call this 6% a year - say, 3.5% real and 2.5% inflation. Add in a dividend yield of 1.7% a year (I'll ignore share buy-backs), and equities offer  a total return of 7.7%. That's just over 3 percentage points better than bonds. That's better than Mehra and Prescott say the premium should be, but less than the 4.9pp we've had historically.
If we use Robert Merton's equation for allocating wealth between a safe and risky asset (in this postscript paper), and assume equity volatility of 20% a year, this premium suggests a risk-neutral investor should have three-quarters of their wealth in equities. Averagely risk-averse investors  (like me) would have less than half this.
I reckon a better way to generate a case for big investment in equities would be not to assume a higher equity premium than this, but rather to integrate human capital into the asset allocation decision. If your returns on human capital are uncorrelated with those on equities, you can afford to have a big chunk of your financial wealth in shares, because labour income partially insures against stock market losses.
But that's a different story...

Comments

[Chris M.]

Judging from the shocking numbers that Jack Bogle presents about how much investors actually keep after expenses and costs, maybe there isn't much equity premium left? That's assuming the marginal investors is a regular person getting ripped off by Wall Street. It wasn't until fairly recently that you could even own the index, which is what we're talking about when we compare stocks to bonds.

[DF]

equity premium (a six & a seven letter word)
due to
pyramid scheme (a seven and a six letter word)
coincidence?
i don't think so.