November 15, 2014
Daniel Chen, ETH Zurich, and
Tobias Moskowitz and Kelly Shue, University of Chicago and NBER
Can misperceptions of what constitutes a fair process lead to unfair decisions? Previous research on the law of small numbers and the gambler's fallacy suggests that many people view sequential streaks of 0's or 1's as unlikely to occur even though such streaks often occur by chance. Chen, Moskowitz, and Shue hypothesize that the gambler's fallacy leads agents to engage in negatively autocorrelated decision-making. The researchers document negatively autocorrelated decisions in three high-stakes contexts: refugee asylum courts, loan application review, and baseball umpire calls. This negative autocorrelation is stronger among more moderate and less experienced decision-makers, following longer streaks of decisions in one direction, and when agents face weaker incentives for accuracy. The authors show that the negative autocorrelation in decision-making is unlikely to be driven by potential alternative explanations such as sequential contrast effects, quotas, or preferences to treat two teams fairly.
Tom Chang and David Solomon, University of Southern California;
Samuel Hartzmark, University of Chicago; and
Eugene Soltes, Harvard University
Chang, Hartzmark, Solomon, and Soltes present evidence that markets fail to properly price information in seasonal earnings patterns. Firms whose earnings are historically larger in one quarter of the year ("high seasonality quarters") have higher returns when those earnings are usually announced. Analyst forecast errors are more positive in high seasonality quarters, consistent with the returns being driven by mistaken earnings estimates. The researchers show that investors appear to overweight recent lower earnings following a high seasonality quarter, leading to pessimistic forecasts in the subsequent high seasonality quarter. The returns are not explained by announcement risk, firm-specific information, increased volume, or idiosyncratic volatility.
Jon Kleinberg and Chentao Tan, Cornell University;
Sendhil Mullainathan, Harvard University and NBER; and
Thomas Zimmermann, Harvard University
Machine learning methods have proven useful in a variety of applications from artificial intelligence to genetics. Though these techniques focus on black box prediction, Kleinberg, Mullainathan, Tan, and Zimmerman propose a method for using them to test theories. In most theory tests, which the authors call deductive, researchers will control for the known theories. In contrast, the authors in this paper develop a simple model that illustrates how machine learning can be used to conduct what they call an inductive test. These tests allow for one to control not just for known theories but some unknown theories, as long as they are covered in some way by the data. They then apply this method to a classic problem in behavioral finance: that realization utility and nominal loss aversion leads to the disposition effect. The authors test this using Odean's (1998) original data. A deductive test replicates his original result. An inductive test, however, rejects the disposition effect, and suggests that the propensity of stocks in the gain to be sold more often is merely a proxy for some other theory, not realization utility and nominal loss aversion.
Francesco D'Acunto, University of California at Berkeley;
Marcel Prokopczuk, Zeppelin University; and
Michael Weber, University of Chicago
D'Acunto, Prokopczuk, and Weber look at the geography of historical Jewish persecution to proxy for localized distrust in finance. Households in German counties where Jewish persecution was one standard deviation higher are 7.5% to 12% less likely to invest in stocks. The results hold when comparing only geographically close counties, and counties that hosted documented Jewish communities in the distant past. Current antisemitism, discriminatory beliefs, generalized trust, or supply-side forces do not explain the effect, which instead is consistent with a norm of distrust in finance, transmitted across generations. The forced migrations of Jewish communities across the German lands in the Middle Ages help assess if the effect of Jewish persecution on stockholdings is causal.
Yihui Pan, University of Utah;
Stephan Siegel, University of Washington; and
Tracy Yue Wang, University of Minnesota
Does culture shape risk preferences? While economic models of the origins of preferences point to an important role of culture, supporting empirical evidence is largely missing for risk and time preferences. In this study, Pan, Siegel, and Wang exploit variation in cultural heritage across CEOs of public U.S. companies and demonstrate an important effect of CEOs' culturally transmitted risk preferences on corporate physical investment. CEOs' uncertainty avoidance negatively affects corporate investment, and the effect is larger for acquisitions than for capital expenditures (Capx). The researchers' finding is robust to controlling for economic and institutional differences as well as genetic differences across countries of origin, and it does not depend on first-generation immigrant CEOs. CEOs' risk preferences seem to have a causal influence on riskier and more discretionary corporate decisions such as acquisitions. But the association between CEO risk preferences and more routine investment decisions such as Capx is largely explained by firm-CEO matching. The authors' results provide novel evidence of important social transmission of risk preferences, their effect on corporate investment policies, and the interplay of the culturally transmitted preferences of CEOs, corporate boards, and other top executives.
Chunxin Jia and Yaping Wang, Peking University, and
Wei Xiong, Princeton University and NBER
This paper uses the segmented dual-class shares of Chinese firms--A shares traded inside mainland China by local investors and H shares traded in Hong Kong by foreign investors--to compare reactions of local and foreign investors to the same public news. Jia, Wang, and Xiong find that local investors react more strongly to earnings forecast revisions by local analysts, while foreign investors react more strongly to forecast revisions of foreign analysts. This finding cannot be explained by local investors being more informed about local firms, earnings forecasts of local analysts being more precise, or local investors having better access to forecasts of local analysts. Instead, it supports the notion that local investors have more trust for local analysts while foreign investors have more trust for foreign analysts, and highlights social trust as an important force driving people with different social backgrounds to react differently to the same information.