Loss Aversion Preferences, Performance, and Career Success of Institutional Investors
University of Notre Dame – Mendoza College of Business
Michigan State University – Eli Broad Graduate School of Management; Centre for Economic Policy Research (CEPR); Gaidar Institute for Economic Policy; SITE
Abstract:
Using survey-based measures of loss aversion of mutual fund managers, we study the effects of institutional investor preferences on their investment decisions, performance, and career outcomes. Funds managed by managers with higher aversion to losses take on less downside risk and have lower risk-adjusted returns. More loss averse managers are more likely to have their contracts terminated. Our results indicate that fund management companies may improve quality of hiring decisions by screening prospective managers on the degree of their loss aversion to ensure better match between managerial characteristics and fund’s objectives.
Loss Aversion Preferences, Performance, and Career Success of Institutional Investors – Introduction
Growing experimental and empirical evidence suggests that economic agents are more sensitive to losses than gains. Investors with loss averse preferences care not only about risk and return, but also about the likelihood of losing money and are willing to pay a premium for securities with lower downside risk (Ang, Chen, and Xing, 2005). This should result in underperformance of their portfolios relative to traditional asset pricing benchmarks. Additionally, assets with a lower chance of suffering a drop in price would be primarily owned by more loss-averse investors.
Do institutional investors have loss-averse preferences? And if so, do these preferences for avoiding losses get reflected in their portfolio choice, performance, and career success? The literature traditionally maintained that loss aversion attributes primarily to individual investors; whereas professional asset managers are supposed to be immune to behavioral biases due to their greater sophistication, better resources, lower search and processing costs, regulation etc (Burns, 1985, Holt and Villamil, 1986). Indeed, Christensen-Szalanski and Beach (1984) and Bonner and Pennington (1991) state that experimental studies that employed professionals rather than students usually report behavior in line with expected utility paradigm.1 Haigh and List (2005) and Bias and Weber (2009), however, suggest that investment professionals may exhibit large aversion to losses (at times even greater than control groups of students). Evidence linking loss aversion of institutional investors to performance is also mixed. Locke and Mann (2005) document that professional traders hold on to losses significantly longer than gains, but find no evidence that it lowers their profits, a finding that they attribute to trading discipline. On the other hand, Garvey and Murphy (2004) conclude that investment professionals could increase their trading profits by mitigating loss averse behavior.
The key challenge to prior research investigating the effect of investment professionals’ attitudes toward risk on their decisions and performance is that on-the-job behavior of professionals is determined not only by their preferences, but also by incentives and constraints imposed by institution’s organization (e.g., Chevalier and Ellison, 1999a, Baks, 2003). This, on the one hand, raises concerns whether the results of experimental studies could be generalized to financial markets. Indeed, high aversion to losses exhibited by investment professionals in pressure-free experiments which involve hypothetical pay-offs (e.g., Haigh and List, 2005, Kaustia Alho, and Puttonen, 2008) may be moderated in high stakes real work environment where, for example, career concerns also play the role. On the other hand, it is unclear what conclusions could be drawn from studies which explore behavior of professional investors without explicitly controlling for both their preferences and incentives. For example, the lossaverse behavior following short-term underperformance may be caused either by investor’s preference to avoid losses or by institutionally imposed incentives to meet target performance (Chevalier and Ellison, 1997, Coval and Shumway, 2005).
Therefore, to establish a causal link between investment professionals’ loss aversion and their performance researchers ideally should be able both to directly relate loss-aversion of institutional investors to their actual decisions and to disentangle the effect of investors’ preferences from that of organizational incentives. We believe that our paper comes closest to achieving these objectives.
See full PDF here.
