This paper measures the extent to which persistence in expected returns moves asset prices. Over the period 1871–1997 persistence is found to vary greatly. For example, 1 unit of news about expected returns can have an impact between 2% and 30% on asset prices depending on the time period. Variations in persistence appear to be correlated with variations in the riskiness of the economy. That is, at times of high risk, news effects asset prices more than at times of low risk. These results have implications for the estimation of asset pricing models, tests of market efficiency and tests of present value models in general. Interestingly, we find that with the exception of the period surrounding the great depression and the second world war, volatility in stock prices is fairly constant and the time-variation in expected returns is driven by variations in the price of risk.

[The output gap, a production-based macroeconomic variable, is a strong predictor of U. S. stock returns. It is a prime business cycle indicator that does not include the level of market prices, thus removing any suspicion that returns are forecastable due to a "fad" in prices being washed away. The output gap forecasts returns both in- sample and out-of-sample, and it is robust to a host of checks. We show that the output gap also has predictive power for excess stock returns in other G7 countries and U. S. excess bond returns.]

We study the predictability of stock returns using a pure macroeconomic measure of the world business cycle, namely the world's capital to output ratio. This variable tracks variation in expected stock returns in a group of the major industrial economies in the presence of world financial market–based predictor variables. The world's capital to output ratio exhibits strong out-of-sample predictive power in almost all countries studied. This is in contrast to financial market–based variables that almost never have out-of-sample forecasting power. Using the stock return predictability that we uncover, we find that international versions of conditional asset pricing models perform well. The world capital to output ratio also predicts bond returns, interest rate changes, and credit spreads. The results highlight the importance of world business conditions for financial markets.

Using a new variable based on a model of dividend smoothing, we find that dividend growth is highly predictable and that cash flow news contributes importantly to return variability. Cash flow betas derived from this predictability are central to explaining the size effect in the cross section of returns. However, they do not explain the value effect; this is explained by noise betas. We also find that the relative importance of cash flow news in explaining recent stock price run-ups and subsequent declines increases when cash flow news is estimated directly.

We analyze the dividend behavior of the aggregate stock market. We propose a model that assumes managers minimize the costs of adjustment associated with being away from their target dividend payout. The target is expressed as a function of lagged stock prices and permanent earnings, generalizing previous models of dividend behavior. We present a new method for measuring unobserved permanent earnings based on the Kalman filter. Our specification of dividend behavior is strongly supported by the data both relative to alternative models and over time. We find significant evidence of dividend smoothing and dividends conveying information regarding unexpected positive changes in current permanent earnings. We also find that both the speed of adjustment of dividends to target dividends and tests of signaling are sensitive to the specification of the model.

Characteristics play a similar role in describing returns in private firms as in public firms. This evidence suggests a causal effect of optimal investment underlying the role of characteristics, as private firms do not have stock prices to over- or under-react on. Common factor models largely describe the cross section of investment returns of both types of firms, suggesting that the common factors are likely aggregate risk factors. Finally, the cost of capital and firm valuations are similar across private and public firms.

We ask to what extent the negative relation between investment and average stock returns is driven by risk. We show that: (i) the average return spread between low and high asset growth and investment portfolios is largely accounted for by their spread in systematic risk, as measured by the loadings on the Chen, Roll, and Ross (1986) factors; (ii) as predicted by q-theory and real options models, systematic risk falls during large investment periods; (iii) the returns of factors formed on the investment-to-assets, asset growth, and investment growth all forecast aggregate economic activities. Our evidence suggests that risk plays an important role in explaining the investment-return relation.

The output gap, a production-based macroeconomic variable, is a strong predictor of U.S. stock returns. It is a prime business cycle indicator that does not include the level of market prices, thus removing any suspicion that returns are forecastable due to a “fad” in prices being washed away. The output gap forecasts returns both in-sample and out-of-sample, and it is robust to a host of checks. We show that the output gap also has predictive power for excess stock returns in other G7 countries and U.S. excess bond returns.

Time variation in the discount rate affects investment and employment decisions in a manner consistent with Q -theory predictions. This evidence is uncovered when using cyclical consumption as a proxy for the discount rate. The results, which are consistent across both U.S. and international data, suggest that firms respond rationally to variations in the cost of capital and that the discount rate has a substantial impact on macroeconomic dynamics and hence business cycle fluctuations.