The Predictive Ability of the Bond-Stock Earnings Yield Differential Model

Does the Bond-Stock Earning Yield Differential Model Predict Equity Market Corrections Better than High P/E Models?

“…We start with a standard 95% one-tail confidence interval based on a Normal distribution. This definition is consistent with earlier works including Ziemba and Schwartz (1991); Schwartz and Ziemba (2000); Berge and Ziemba (2003);Berge et al (2008) and Lleo and Ziemba (2012). A standard confidence level does not provide a robust threshold because the tails of the distribution of the measures may not be approximately Gaussian.…”

“…The main difference between this paper and Ziemba and Schwartz (1991), Lleo and Ziemba (2014) or Lleo and Ziemba (2012) is our emphasis on testing statistically the predictions of the BSEYD under several sets of specifications and against another three measures over a longer period of time. This paper also complements Berge et al (2008) as their methodology focused on comparing the Sharpe ratio of a BSEYD-based investment strategy with that of a "buy and hold" strategy. A limitation of this paper is that it only considers the US market.…”

“…First, do all four measures perform similarly on international financial markets? Studies by Ziemba and Schwartz (1991), Berge et al (2008) and Lleo and Ziemba (2012) show that the BSEYD has had a successful track record predicting crashes in international markets. A large scale international study focused on some of the models discussed in this paper could also provide panel data useful to enhance understanding of the behavior of the models over time and to solve the puzzle of the large shift in significance between the subperiod 1964-1981 and the subperiod 1982-2012.…”

“…For example, Berge et al (2008) found that the strategy "move to cash when the model indicates a crash and invest in an equity index the rest of the time" provides about twice the final wealth with less variance and a higher Sharpe ratio than a buy-and-hold strategy over the period 1975-2005 and 1980-2005 in five countries considered in the study. On the other hand, long-term prediction models, such as the Campbell-Shiller model, were not intended as a crash prediction model but as a prediction model for long-term returns.…”

“…For long-term investors, the ability to forecast market corrections and crashes improves risk-adjusted performance (Berge, Consigli, and Ziemba, 2008). The ability to predict crashes should also matter to economists because the S&P500 is a leading indicator of economic growth (Klein and Moore, 1982;Hertzberg and Beckman, 1989;Ozyildirim, McGuckin, and Zarnowitz, 2003).…”