CFOs Project the Equity Risk Premium

How do the corporate experts most responsible for assessing the cost of equity currently feel about future stock returns? In their August 2010 paper entitled “The Equity Risk Premium in 2010″, John Graham and Campbell Harvey update their report on the views of U.S. Chief Financial Officers (CFOs) on the prospective U.S. equity risk premium relative to the 10-year U.S. Treasury note (T-note) yield, assuming a 10-year investment horizon. Based on 41 quarterly surveys on this topic over the period June 2000 through June 2010 (an average 333 responses per survey), they find that:

• CFOs on average currently expect an annualized raw return of 6.31% for the S&P 500 index over the coming decade (the lowest in survey history. This expectation implies an equity risk premium of 3.00%, below the overall series average of 3.40%.
• The standard deviation of individual responses in the most recent survey is 3.07%, a relatively high level of disagreement for the series.
• Associated personal interviews indicate that CFOs follow the stock market closely. They interpret the equity risk premium as the long-term return above the T-note yield of buying and holding equities, so their average estimate indicates a geometric rather than arithmetic mean return.
• Informal statistical analyses suggest that the survey-based prospective equity risk premium relates:
  • Weakly negatively to prior-year stock returns.
  • Non-linearly to the aggregate stock price-earning ratio.
  • Weakly positively to the real interest rate.
  • Strongly positively (correlation 0.60) to the implied volatility of the S&P 500 index option (VIX). (See the chart below.)
  • Strongly positively (correlation 0.54) to the credit spread (Moody’s Baa rated bond yield less the T-note yield).

The following chart, taken from the paper, shows the contemporaneous relationship between the survey-based prospective equity risk premium and the VIX across all 41 quarterly surveys. The correlation between the the two series is a notable 0.60. Both series have declined sharply from a peak in early 2009.

How should investors interpret the survey data?

The next chart, constructed using quarterly data in the paper, summarizes Pearson correlations between S&P 500 Index returns over the next 63 and 126 trading days and each of four metrics: (1) the average CFO forecast for total stock market return; (2) T-note yield; (3) the average CFO forecast for the equity risk premium (ERP); and (4) VIX. The sample is winnowed for the 126-day horizon so that return intervals do not overlap. Results suggest that:

• CFO forecasts for the equity market total return may be a contrary indicator of returns at horizons of three and six months.
• This contrary information derives from variation in the risk-free rate (T-note yield) and not from expectations for the equity risk premium. Specifically, a relatively high (low) T-note yield suggests relatively weak (strong) stock market returns at the selected horizons over the sample period.

The sample is small for this kind of analysis (40 instances for the 63-day horizon and 20 for the 126-day horizon), so a few additional instances could change results substantially.

Note that the CFOs are making predictions at a ten-year horizon, so these results are not a test of those forecasts.

In summary, a current survey of U.S. CFOs indicates a prospective U.S. equity risk premium that has declined from a recent peak to a relatively low level, with a near record level of uncertainty. Historical data suggests that this outlook is uninformative for stock market return over the next quarter or two (but the T-note yield may be informative).

You May Also Enjoy...

• The 2008 Equity Risk Premium from Academia
• The 2010 Equity Risk Premium from Practitioners
• The 2010 Equity Risk Premium from Academia
• The Prospective “Academic” Equity Risk Premium
• Honing in on the Prospective U.S. Equity Risk Premium