Market Models

The CXO Advisory Group LLC offers this summary of projections from the Reversion-to-Value (RTV) Model and the Real Earnings Yield (REY) Model of the U.S. stock market as alternative ways of thinking about current equity valuation.

The RTV Model hypothesizes that the forward aggregate earnings yield for equities relates positively to future stock market returns. From the model description: the RTV Model appears to offer perhaps a little information about S&P 500 Index returns over the next few quarters, but it is not very reliable. Longer-term projections are more reliable than shorter-term.

The REY Model hypothesizes that the gap between the forward aggregate earnings yield for equities and the forward total inflation rate relates positively to future stock market returns. From the model description: the REY Model appears to offer some information about S&P 500 Index returns over the next few quarters. Longer-term projections are more reliable than shorter-term.

Both models depend on a purely technical Earnings Forecast based on hypotheses that the aggregate corporate operating earnings growth rate exhibits: (1) short-term momentum/acceleration; and, (2) long-term reversion related to degree of deviation from trend.

The REY Model depends also on a purely technical Inflation Forecast based on hypotheses of both persistence and momentum in the inflation rate.

Per analyses in “Comparing Lagged and Forward RTV and REY Models”, the summary includes projections from an alternative REY-A Model using the gap between the forward aggregate earnings yield for equities and the lagged total inflation rate.

The following table summarizes returns for the three, six and 12 months from the end of June 2010, as projected by the RTV, REY and REY-A Models.

Note the following sources of uncertainty and potential inaccuracy in the models, as stated in the detailed model descriptions:

• Investors may employ methods of estimating forward earnings and forward inflation rate that differ materially from the Earnings Forecast and Inflation Forecast used in the models.
• The U.S. stock market is not a closed system. Other, substantially uncorrelated asset classes may compete.
• Decision factors other than the ones used in the models may be important to investors.
• There may be considerable randomness in market behavior.
• Consideration of multiple models increases the probability of discovering “lucky” results.
• The samples used in some aspects of development of the models are fairly small for the tests employed.
• The relationships between S&P 500 Index future returns and the key metrics of the models found for the development sample periods may not persist.
• Non-normality of the stock market returns disrupts interpretation of the “normal” statistics used in the models.

Changes:

11/30/09: Extended the regressions and rolled the projections by one month.

12/3/09: Added REY-A Model projections as noted above.

12/31/09: Extended the regressions and rolled the projections by one month.

1/29/10: Made a moderate revision to the Earnings Forecast model (mostly to simplify calculations for momentum and reversion), incorporated (a little early) the actual Standard & Poor’s earnings results for the fourth quarter of 2009, and extended regressions and rolled the projections by one month. The Earnings Forecast (rather than the Inflation Forecast) is driving projections, with both earnings momentum strongly positive and earnings reversion positive.

2/26/10: Extended the regressions and rolled the projections by one month.

3/31/10: Extended the regressions and rolled the projections by one month. The combination of a forecasted earnings rebound and low lagged and forecasted inflation indicates strong market appreciation.

5/1/10: Incorporated the actual Standard & Poor’s earnings results for the first quarter of 2010, and extended the regressions and rolled the projections by one month. The Earnings Forecast (rather than the Inflation Forecast) is driving projections, with earnings momentum strongly positive and earnings reversion positive but approaching neutrality.

5/28/10: Extended the regressions and rolled the projections by one month.

6/30/10: Extended the regressions and rolled the projections by one month.

See also “How Well Do the REY/RTV Models Catch Turning Points?” and “Valuation Metric Map and Critique” for other perspectives on usefulness of the models.

Please see the disclaimer.