"In bull markets - usually when things have been going well for a while - people tend to say 'Risk is my friend. The more risk I take, the greater my return will be. I'd like more risk, please.' The truth is, risk tolerance is antithetical to successful investing. When people aren't afraid of risk, they'll accept risk without being compensated for doing so... and risk compensation will disappear. But only when investors are sufficiently risk-averse will markets offer adequate risk premiums. When worry is in short supply, risky borrowers and questionable schemes will have easy access to capital, and the financial system will become precarious. Too much money will chase the risky and the new, driving up asset prices and driving down prospective returns and safety."
One of the interesting points that both Warren Buffett and Howard Marks have stressed over the years is that risk - viewed as the risk of losing significant amounts of money - moves in the same direction as valuations. So as valuations become rich, risk increases, and as valuations become depressed, risk declines. At the same time, rich valuations imply weak long-term prospective returns, while depressed valuations imply strong long-term prospective returns. As a result, both Marks and Buffett suggest that risk is lowest precisely when prospective returns are the highest, and risk is highest precisely when prospective returns are the worst.
Of course, most of finance theory is based on a positive relationship between risk and return. That is, higher risk is assumed to be required in order to achieve higher expected return. But the hidden assumption in finance theory is that the securities are "efficiently" priced. See, efficiency says that there is no way to adjust your portfolio in a way that produces greater expected return per unit of risk. It therefore follows that the only way to increase expected return, once your portfolio is efficient, is to take more risk. Still, even finance theory has no problem with the idea that you can reduce risk and increase expected return at the same time if you are starting with a poorly diversified portfolio or inefficiently priced securities. So neither Buffett nor Marks are proposing some "new" sort of finance theory. They're simply saying that they don't believe that stocks are always efficiently priced.
On the subject of expected returns, I noted last week that Jeremy Grantham views "fair value" to be about 920 on the S&P 500. That implies a price-to-revenue multiple of just under 1.0, which is about right historically. Since 1940, the average price-to-revenue multiple for the S&P 500 has been about 0.9. Given that revenues are still somewhat depressed here, it makes sense to bump up the current number ($962.71) somewhat, or to use a multiple closer to 0.95, which is consistent with what Grantham is getting, albeit with much different methods. As I noted above, on the basis of virtually every normalized fundamental we choose to examine (normalized net earnings, forward operating earnings, dividends, book values, Tobin's Q, or revenues), we estimate that prospective 10-year returns on the S&P 500 are only about 3.6%, which translates into a market that appears overvalued by roughly 45%.