One Up on the Fed Model?

In their June 2003 paper entitled “A General Theory of Stock Market Valuation and Return”, Christophe Faugere and Julian Van Erlach contend that past stock returns are overstated and develop a market valuation formula that out-fits the Fed Model. Specifically, they show that:

• The long-term return from the stock market is determined by GDP growth and is much less than believed because dividends cannot be fully reinvested. Not enough new shares can be issued to absorb all dividends without suppressing share earnings/price (E/P). The dividend reinvestment rate is necessarily much lower than dividend yield.
• The total long-term return from the stock market (5.43% for 1926-2000) is very close to that of risk-free debt (5.28% for 10-year T-notes); there is really no risk premium.
• The stock market E/P ratio depends on inflation and real long-term per capita growth in GDP.
• Capital gains exhibit mean reversion at the market level. Monthly and quarterly volatility are due to changing expectations about earnings, inflation and taxes.
• Their valuation model produces a mean percentage tracking error of 21% versus the S&P 500 composite index during 1954-2002, compared with 32% for the Fed Model.

In summary, individual investors may be able to outperform through determined reinvestment of dividends and exploitation of capital gains mean diversion and reversion.