Andrew Y. Chen

External Habit in a Production Economy

A unified framework for understanding asset prices and aggregate fluctuations is critical for understanding both issues. I show that a real business cycle model with external habit preferences and capital adjustment costs provides one such framework. The estimated model matches the first two moments of the equity premium and risk-free rate, return and dividend predictability regressions, and the second moments of output, consumption, and investment. The model also endogenizes a key mechanism of consumption-based asset pricing models. In order to address the Shiller volatility puzzle, external habit, long-run risk, and disaster models require the assumption that the volatility of marginal utility is countercyclical. In the model, this countercyclical volatility arises endogenously. Production makes precautionary savings effects show up in consumption. These effects lead to countercyclical consumption volatility and countercyclical volatility of marginal utility. External habit amplifies this channel and makes it quantitatively significant.

Accounting for the Anomaly Zoo: A Trading Cost Perspective

We zero in on the expected returns of long-short portfolios based on 120 stock market anomalies by accounting for (1) effective bid-ask spreads, (2) post-publication effects, and (3) the modern era of trading technology that began in the early 2000s. Net of these effects, the average anomalys expected return is a measly 8 bps per month. The strongest anomalies return only 10-20 bps after accounting for data-mining with either out-of-sample tests or empirical Bayesian methods. Expected returns are negligible despite cost optimizations that produce impressive net returns in-sample and the omission of additional trading costs like price impact.We study the post-publication trading costs of 120 stock market anomalies. Trading costs use effective bid-ask spreads from high-frequency ISSM and TAQ data when available and average four low-frequency proxies otherwise. The average equal-weighted long-short portfolio nets -3 bps per month post-publication after costs. Optimized cost mitigation using valueweighting and buy/hold spreads dramatically improves net returns insample but nets only 4 to 12 bps post-publication on average. The strongest cost-optimized anomalies in-sample net just 10-20 bps post-publication. These results show that the average investor should expect tiny profits (alternatively, a tiny risk premium) from investing in any individual anomaly. ∗First posted to SSRN: November 2017. This paper originated from a conversation with Svetlana Bryzgalova. We thank Marie Briere (HFPE discussant), Victor DeMiguel, Yesol Huh, Nina Karnaukh, Alberto Martin-Utrera (FDU discussant), Andy Neuhierl, Nitish Sinha, Tugkan Tuzun, Michael Weber, and seminar participants at the Federal Reserve Board, Penn State University, University of Georgia, the 11th Annual Hedge Fund and Private Equity Research Conference, 2019 Finance Down Under Meetings, and 2019 Eastern Finance Association Meetings for helpful comments. The views expressed herein are those of the authors and do not necessarily reflect the position of the Board of Governors of the Federal Reserve or the Federal Reserve System.

A Likelihood-Based Comparison of Macro Asset Pricing Models

We estimate asset pricing models with multiple risks: long-run growth, long-run volatility, habit, and a residual. The Bayesian estimation accounts for the entire likelihood of consumption, dividends, and the price-dividend ratio. We find that the residual represents at least 80% of the variance of the price-dividend ratio. Moreover, the residual tracks most recognizable features of stock market history such as the 1990s boom and bust. Long run risks and habit contribute primarily in crises. The dominance of the residual comes from the low correlation between asset prices and consumption growth moments. We discuss theories which are consistent with our results.

External Habit in a Production Economy: A Model of Asset Prices and Consumption Volatility Risk

A standard real business-cycle model with external habit and capital adjustment costs matches a long list of asset price and business-cycle moments: equity, firm value, and risk-free rate volatility; the equity premium; excess return predictability; consumption growth predictability; basic moments of consumption, output, and investment; among others. The model also generates endogenous consumption volatility risk. Precautionary savings motives make consumption sensitive to shocks in bad times, leading to countercyclical volatility, even with homoscedastic technology shocks. Habit acts as countercyclical leverage, which amplifies this channel. Habit also implies high risk aversion, which amplifies the stock price response.Received April 21, 2016; editorial decision February 3, 2017 by Editor Stijn Van Nieuwerburgh.

Semi-Parametric Restrictions on Production-Based Asset Pricing Models

Matching asset price volatility in production economies is difficult. This paper shows that this difficulty can be summarized by three nested restrictions. First, matching asset price volatility requires volatile investment returns. Second, volatile investment returns require either large capital adjustment costs or volatile investment specific technology shocks. Third, large costs or volatile shocks require a low elasticity of intertemporal substitution. I quantify these restrictions and show that they apply to a broad class of models which spans many assumptions about preferences, beliefs, market completeness or the stochastic structure of shocks.

Habit, Production, and the Cross-Section of Stock Returns

Solutions to the equity premium puzzle should inform us about the cross-section of stock returns. An external habit model with heterogeneous firms reproduces numerous stylized facts about both the equity premium and the value premium. The equity premium is large, time-varying, and linked with consumption volatility. The cross-section of expected returns is log-linear in B/M, and the slope matches the data. The explanation for the value premium lies in the interaction between the cross-section of cash flows and the time-varying risk premium. Value firms are temporarily low productivity firms, which will eventually experience high cash flows. The present value of these temporally distant cash flows is sensitive to risk premium movements. The value premium is the reward for bearing this sensitivity. Empirical evidence verifies that value firms have higher cash-flow growth. The data also show that value stock returns are more sensitive to risk premium movements, as measured by consumption volatility shocks.

Precautionary Volatility and Asset Prices

Many theories of asset prices assume time-varying uncertainty in order to generate time-varying risk premia. This paper generates time-varying uncertainty endogenously, through precautionary saving dynamics. Precautionary motives prescribe that, in bad times, next periods consumption should be very sensitive to news. This time-varying sensitivity results in time-varying consumption volatility. Production makes this channel visible, and external habit preferences amplify it. An estimated model featuring this channel quantitatively accounts for excess return and dividend predictability regressions. It also matches the first two moments of excess equity returns, the risk-free rate, and the second moments of consumption, output, and investment.

Selection Bias and the Cross-Section of Expected Returns

We propose an estimate of expected returns that accounts for selective anomaly submission and publication, and apply our adjustment to a broad cross-section of anomaly hedge portfolio returns. Selection bias accounts for a modest 10 to 15% of the typical in-sample return. This small selection bias is due to fact that the dispersion of in-sample returns is nearly twice as large as the typical standard error, indicating a significant amount of variation in true returns. Since the out-of-sample decline in returns is much larger than the selection effect, these results imply that investors learn about mispricing from academic research. Estimations on simulated data show that our bias adjustment is robust.