In this paper we provide new evidence that relates asset prices, consumption volatility and expected growth. In particular, we show that economic uncertainty (that is, consumption volatility) sharply predicts and is predicted by asset valuation ratios. Our evidence shows that a rise in economic uncertainty leads to a fall in asset prices, and that high valuation ratios predict low subsequent economic uncertainty. In addition, we show that there is a strong positive relation between aggregate earnings growth and asset prices. In all, our evidence suggests that fluctuations in economic uncertainty and expected growth are potentially important channels for interpreting asset markets and the variation in asset prices.
Why is this evidence important? First, this empirical evidence highlights an often discussed but not verified view that aggregate economic uncertainty (i.e., real aggregate consumption volatility) has sizable effects on asset valuations and that financial markets dislike economic uncertainty. Our empirical work for the U.S. and foreign economies suggests that the effects of fluctuating economic uncertainty on asset valuations are sizable. Second, the evidence regarding growth rates suggests that fluctuations in expected growth directly affect asset valuations and information regarding future expected growth is encoded in current asset valuations. Our overall evidence regarding economic uncertainty and expected growth suggests that a plausible interpretation of asset markets is based on these economic fundamentals. A rise in economic uncertainty increases expected returns and leads to a fall in asset valuations. A rise in expected growth, on the other hand leads to a rise in asset valuations. Both these effects can be interpreted from the perspective of general equilibrium models (see for example, Bansal and Yaron (2004)).
An alternative view of asset markets “shuts-off” the channels of expected growth rates and economic uncertainty, as growth rates in these models are assumed to be i.i.d. (e.g., Campbell and Cochrane (1999), Cechetti, Lam, and Mark (2000)). These models suggest that asset markets can be interpreted via the channels of fluctuating risk aversion and/or distorted beliefs. The empirical evidence provided in this paper does not exclude the possibility of time-varying risk-aversion; however, it does suggest that channels related to observable macroeconomic fluctuations (in expected growth and volatility) can by themselves go a long way to help interpret market movements.
Bansal and Yaron (2004) briefly discuss correlations between asset valuation and squared consumption residuals. In contrast, the main contribution of this paper is to provide a systematic and comprehensive empirical analysis of the economic uncertainty channel. By utilizing various volatility measures, alternative econometric techniques, multiple data sets, and time periods, we show that the uncertainty channel is robustly found in the data. An additional contribution is to exploit our evidence regarding volatility and predictability of cash flows to evaluate the plausibility of alternative preferences and models for asset markets.
In terms of the empirical evidence we find that consumption volatility predicts price-dividend and price-earnings ratios, with R2 in excess of 20%. This channel is robust to using alternative measures of volatility; we use GARCH and integrated volatility measures. Interestingly, it is difficult to find comparable evidence if one replaces consumption volatility with simple measures of market volatility. Future, realized consumption volatility is predicted by current valuation ratios, and at horizons of 4-8 quarters, the R2s from these regressions are about 6%. Current valuation ratios do not predict future realized market volatility. The slope coefficient in the regressions that link valuation ratios to consumption volatility is always negative and significant—as predicted by our economic model. Volatility measures and valuation ratios are quite persistent, this can make inference in finite samples difficult. To account for this we also provide, monte carlo based, finite sample empirical distribution for the various parameters and test statistics of interest. Finally, we also consider data from UK, Germany, and Japan and find very similar evidence for these economies as well. The overall message is that the data strongly supports the economic uncertainty channel.
Download
Interpretable Asset Markets?
