Fed Model proponents argue that there is an equilibrium relationship between the earnings yield of a stock index and the 10-year government bond yield. When the earnings yield is below (above) the 10-year government bond yield, the stock market is overvalued (undervalued). In his January 2006 paper entitled “The Fed Model: The Bad, the Worse, and the Ugly”, Javier Estrada recaps the (lack of) theoretical basis for the Fed Model and tests its empirical support in the markets of 20 countries. Using both actual (trailing) and projected (forward) earnings for total market indices over various periods ending in June 2005, he concludes that:
- The Fed Model is implausible because stocks and bonds are not comparable assets. Companies do not pay out most earnings as dividends; earnings tend to grow; and, investors should require a greater return from stocks than bonds. Further, stock earnings yields and bond yields should react differently to inflation; the former is real, the latter nominal.
- Proponents of the Fed Model use carefully chosen and limited evidence. A broader view reveals its weakness. (See charts below.)
- Although there is a correlation between earnings yields (both actual and projected) and bond yields in many countries, there is not an equality. The departures from equality are larger than those one should expect from an accurate model.
- In most countries, the projected price/earnings ratio (P/E) outperforms the Fed Model in predicting stock market returns. Exceptions are Austria and the U.S. Actual P/E outperforms the Fed Model in all countries.
The following charts, extracted from the paper, show the trailing earnings yield of the S&P 500 (E/P) and the yield on 10-year Treasury notes (Y) over the relatively recent period of January 1968 through June 2005 (left-hand chart) and the longer period of January 1871 through June 2005 (right-hand chart). The apparently close relationship visible since 1968 breaks down in the more distant past. The author does not discuss whether changes in the investing climate (financial/technological/regulatory/cultural) might have recently introduced an E/P-Y relationship.
In summary, practitioners who use the Fed Model are simpletons. The model is theoretically implausible and empirically challenged.
