Market Expectations in the Cross‐Section of Present Values

Abstract

Returns and cash flow growth for the aggregate U.S. stock market are highly and robustly predictable. Using a single factor extracted from the cross-section of book-tomarket ratios, we find an out-of-sample return forecasting R 2 of 13% at the annual frequency (0.9% monthly). We document similar out-of-sample predictability for returns on value, size, momentum, and industry portfolios. We present a model linking aggregate market expectations to disaggregated valuation ratios in a latent factor system. Spreads in value portfolios' exposures to economic shocks are key to identifying predictability and are consistent with duration-based theories of the value premium.THE MOST COMMON APPROACH to measuring aggregate return and cash flow expectations is predictive regression. As suggested by the present value relationship between prices, discount rates, and future cash flows, research shows that the aggregate price-dividend ratio is among the most informative predictive variables. Typical in-sample estimates find that about 10% of annual return variation can be accounted for by forecasts based on the aggregate book-tomarket ratio, but find little or no out-of-sample predictive power.1 In this paper we show that reliance on aggregate quantities drastically understates the degree of value ratios' predictive content for both returns and cash flow growth, and hence understates the volatility of investor expectations. Our estimates suggest that as much as 13% of the out-of-sample variation in annual market returns (as much as 12% for dividend growth), and somewhat more of the insample variation, can be explained by the cross-section of past disaggregated value ratios.To harness disaggregated information we represent the cross-section of assetspecific book-to-market ratios as a dynamic latent factor model. We relate disaggregated value ratios to aggregate expected market returns and cash flow growth. Our model is based on the idea that the same dynamic state variables driving aggregate expectations also govern the dynamics of the entire panel * Kelly is with Booth School of Business, University of Chicago, and Pruitt is with the Board of Governors of the Federal Reserve System. The view expressed here are those of the authors and do not necessarily reflect the views of the Federal Reserve System or its staff.1 See Cochrane (2005) and Koijen and Van Nieuwerburgh (2011) for surveys of return and cash flow predictability evidence using the aggregate price-dividend ratio. Similar results obtain from forecasts based on the aggregate book-to-market ratio.

Does a lot help a lot? Forecasting stock returns with pooling strategies in a data‐rich environment

“…6 Due to the fact that principal component estimates might change their signs and we additionally take overlapping returns into account, the sample period over which parameters are estimated must be reduced by the degree of the forecasting horizon (i.e., one observation for monthly returns, three observations for quarterly returns, etc.). We additionally follow Kelly and Pruitt (2013) and reduce the sample size according to the forecasting horizon in the first and third pass of the 3PRF filter to obtain out-of-sample forecasts from target-relevant factor models. Note.…”

The Cost of Capital of the Financial Sector

“…We normalize size by the total size of the financial system to obtain a stationary forecaster. Finally, we add the dividend yield as a forecasting factor, as that is the baseline variable that Kelly and Pruitt (2013) use. PLS is designed to extract the linear combination of this large number of right hand side variables as the single factor that best forecasts the respective variable.…”

“…In implementing the partial least squares (PLS) filter of Kelly and Pruitt (2013), we allow the expected return to depend on the following five characteristics of all financial stocks: leverage growth, ROE, size normalized by the size of the financial system, book-tomarket, and dividend yield. We use leverage growth and ROE as we had earlier documented that these were significant forecasting variables using the panel regressions (see Table 1).…”

“…We use the methodology from Kelly and Pruitt (2013) to forecast the value weighted return on the financial sector over the risk free rate. We use this forecast as an estimate of the conditional risk premium on the financial sector at any point in time.…”

“…Here, we estimate conditional expected returns on the entire financial sector using the Kelly and Pruitt (2013) partial least squares method. We study how the expected returns in the financial sector vary over time, providing us with a conditional cost of capital for the financial sector.…”