by Dr Adelphe Ekponon, Research Associate, Cambridge Centre for Finance and Cambridge Endowment for Research in Finance
The very first asset pricing models (also called Capital Asset Pricing Models or CAPM) have postulated that the only risk that is needed to characterise a stock price is the contemporaneous correlation between the firm and the market portfolio returns. This implies that investors pay much more attention to information about the current economic conditions. Yet models that only incorporate this correlation risk tend to be unable to capture the dynamic of equity returns. The empirical asset pricing model proposed by Fama and French (1992) demonstrate that CAPM has no explanatory power to explain the cross-section of average stocks returns on portfolios sorted by size and book-to-market equity ratios.
An important trend of the literature has developed models to improve pricing performances of the CAPM via a consumption-based approach, CCAPM. The main innovation of CCAPM models lies in the introduction of macroeconomic conditions into asset pricing. According to these models, risk premia should be proportional to consumption beta (correlation between the firm’s profit and consumption). However, this line of CCAPM models are known to produce very little level of equity risk premium, less than one per cent for reasonable levels of risk aversion. These models are also rejected by several empirical tests.
Since then, two new features have been introduced in asset pricing. The first comes from the observation by Hamilton (1989) that shocks to the US economic growth are not i.i.d. as growth rates may also shift from periods of high to low levels. Secondly, a new class of utility functions introduced by Epstein and Zin (1989), allows to isolate, the aversion to future economic uncertainty from that of the current correlation risk.
Bansal and Yaron (2004) and recent papers have successfully developed consumption-based models in which the representative agent has Epstein-Zin type of preferences. These models pave the way to disentangle the impact of long-run vs. current correlation risks in stock prices. Additionally, they generate reasonable levels of equity risk premium and are able to explain some key asset pricing phenomena. Here, long-run risk (LRR) captures the unforecastable and persistent nature of future economic conditions and has two components, expected growth rate and volatility.
Constructed on this last trend of papers, Dorion, Ekponon and Jeanneret (2019) propose a consumption-based structural approach, with endogenous default and debt policies, that allows investigating both long-run and correlation risks individually and in tandem. This is the first study to isolate and quantify, conditional on the state of economy, the impact of LRR in equity prices.
They found an average risk premium of one per cent in expansion against six per cent in recession. The paper also predicts that long-run risk represents about three-quarters of this risk premium and that its impact is countercyclical, being more than 90 per cent in recession. To reduce the impact of LRR, managers lessen the optimal amount of debt to issue and lower the default barrier. Despite these adjustments, LRR still governs equity premium leading to the above predictions.
Using US stocks prices, consumption growth (correlation risk), and expected economic growth rate and volatility (long-run risk), over the period from 1952 to 2016, this study confirms that LRR is priced in US firms, particularly in bad times. These data show that the compensation for LRR represents around 70 per cent of excess returns in a zero-investment portfolio, consisting in shorting stocks which returns have a low correlation with expected growth rates (or high correlation with expected growth volatilities) and buying stocks with high correlation with expected growth rates (or low correlation with expected growth volatilities). These results imply that LRR is a priced risk factor for equity.
Hence, investors are compensated for trading/holding stocks based on their sensitivity to future economic conditions. This result provides a strong evidence that long-run economic outlook is an important driver of equity premium at the cross section.
References
Bansal, R. and Yaron, A. (2004) “Risks for the long run: a potential resolution of asset pricing puzzles.” Journal of Finance, 59(4): 1481-1509
Epstein, L.G. and Zin, S.E. (1989) “Substitution, risk aversion, and the temporal behavior of consumption and asset returns: a theoretical framework.” Econometrica, 57(4): 937-969
Fama, E.F. and French, K.R. (1992) “The cross-section of expected stock returns.” Journal of Finance, 47(2): 427-465
Hamilton, J. (1989) “A new approach to the economic analysis of nonstationary time series and the business cycle.” Econometrica, 57(2): 357-384