حمایت حرفه ای از امور مالی رفتاری: آیا بر درک خود از بازارها و خودشان تاثیر می گذارد؟

This paper provides evidence on the hypothesis that many behavioral finance patterns are so deeply rooted in human behavior that they are difficult to overcome by learning. We test this on a target group which has undoubtedly very strong incentives to learn efficient behavior, i.e. fund managers. We split this group into endorsers and non-endorsers of behavioral finance. Endorsers do, indeed, view markets differently as they regard stronger influences from behavioral biases. However, when it comes to the perception of one's own behavior the endorsement of behavioral finance becomes almost meaningless, even though endorsers otherwise do adapt behavior towards their conviction.

The history of many anomalies in financial markets has shown that they disappear over time (Fama, 1998). This has raised the suspicion that markets may need time to recognize such anomalies – which largely motivate behavioral finance – but that they react consequently afterwards. According to this view one may assess behavioral finance being largely concerned with transitory phenomena. Others argue that many behavioral finance patterns are so deeply rooted in human behavior that they are difficult to overcome by learning (Alpert and Raiffa, 1982, Fischhoff, 1982b and Tversky and Kahneman, 1982, and more recently Hirshleifer, 2001). Obviously, these two views make contrary predictions on the persistence of behavioral phenomena although both camps agree that in the long run fundamentals drive prices. If rational learning is dominant then one could expect that insights into behavioral finance impact one's own behavior. If, however, learning mechanisms are weak in this respect the insights into behavioral finance will have a minor impact on one's own behavior. We provide a test of these competing views and find evidence in support of the latter hypothesis put forward by psychologists working in behavioral decision-making and more recently by behavioral economists.1 Our research aims for extending literature in psychology which has clearly revealed the “bias blind spot” (Pronin et al., 2002), i.e. the belief that one's own judgments are less susceptible to biases than the judgments of others.2 We extend this research by addressing an important objection that economists might voice against evidence based on student surveys: The probability that behavioral biases hold depends on how strong incentives are to learn and thus to overcome entrenched behavior. One can easily imagine circumstances where people adhere to behavioral patterns as forecast by behavioral finance just because their welfare is not at stake. It seems therefore advisable for an empirical examination of the hypothesis put forward by psychologists and behavioral economists to choose a target group that has undoubtedly strong incentives to learn. We have thus targeted professional fund managers. Their investment performance is on the one hand negatively affected by behavioral biases (see e.g. Shefrin and Statman, 1985, Coval and Shumway, 2005 and Biais and Weber, 2007) and on the other permanently monitored and linked to high performance-related bonuses. Among these fund managers we differentiate between “endorsers” of behavioral finance and others, who we call “non-endorsers”. Endorsers of behavioral finance are those fund managers who believe that the approach of behavioral finance truly reflects decision behavior in fund management (and who know the key messages of behavioral finance well). In contrast to them are non-endorsers, i.e. fund managers who are not that much convinced about the relevance of behavioral finance. We have surveyed more than 100 fund managers in Germany and classify them according to their self-assessment into these two groups. We then analyze whether fund managers’ endorsement of behavioral finance, i.e. being an “endorser”, impacts their perception of markets and themselves: do these groups respond in the same way to questionnaire items addressing, first, their perception of general fund managers’ behavior and, second, their perception of their own behavior with respect to issues being raised by behavioral finance research? We find a revealing split in responses: whereas endorsers recognize significantly stronger behavioral finance effects in other fund managers’ behavior than non-endorsers, the perception of their own behavior is largely unaffected by their insights. When endorsers are asked about their own behavior with respect to items being linked to behavioral finance, such as hindsight bias or disposition effect, they answer as non-endorsers do. However, there is one exception to this pattern: endorsers show less miscalibration with respect to forecasting the interval of a stock index. As less miscalibration here also means more correct answers, this suggests that endorsers’ conviction of behavioral finance may increase awareness for respective distortions and can improve decisions to some extent. In a final exercise, we analyze whether endorsers’ assessment of their own behavior may reflect the fact that behavioral finance does not influence their decisions in any respect. Therefore, we ask whether endorsers differ from non-endorsers regarding two further items of investment behavior, i.e. their preferred information sources and investment strategies. We find that both groups seem to differ clearly in their use of information sources although not to a statistically significant extent. The difference becomes significant, however, with respect to preferred investment strategies as endorsers rely more on momentum and contrarian strategies. We conclude that endorsers do not generally behave like non-endorsers, making the above found indifference between both groups with respect to their self-assessment more credible. Thus, we come back to the hypothesis – proposed by psychologists and behavioral economists – that many behavioral patterns are difficult to overcome by learning. Our findings provide support for his view as insight into behavioral finance and thus into the behavior of others does not easily change one's own behavior. We regard it as an important and innovative aspect of our research that this result has been found among fund managers because this target group has strong incentives to improve behavior. In this respect we also find a partial justification for Fama's (1998) optimism into learning processes in that the endorsement of behavioral finance reduces miscalibration. The remaining study is structured into four sections. Section 2 describes data of the questionnaire survey. The following sections present our analyses, starting with views on the market's behavior (Section 3), then views on one's own behavior (Section 4) and, finally, consequences on investment behavior (Section 5). Section 6 concludes.

Recent empirical research in the field of behavioral finance has shown that professional investors are (just as well as private ones) subject to irrational, psychologically motivated biases in their investment decisions (e.g. Glaser et al., 2005, Haigh and List, 2005 and Menkhoff et al., 2006). Even though professionalism may reduce biases to some degree it does not eliminate them entirely (Shapira and Venezia, 2001, Shapira and Venezia, 2006 and Feng and Seasholes, 2005). This paper offers another perspective on the role of behavioral biases in finance by examining the impact from endorsement of behavioral finance on the perception of markets’ and one's own behavior. We provide evidence on this issue by examining survey data of more than 100 German fund managers. Distinguishing the respondents in two groups, endorsers and non-endorsers of behavioral finance, we test for differences in their answers with respect to perceived market-wide biases, self-assessment, information processing and investment strategies. Interestingly, there is no difference in personal characteristics between the two groups. Our findings show that the endorsement of behavioral finance has a significant impact on professionals’ perception of markets, but that it hardly influences the view on one's own behavior. This finding holds although endorsers do behave differently from non-endorsers in the expected direction with respect to preferred information sources and investment strategies. So far our findings clearly support the hypothesis that many behavioral biases are difficult to overcome by learning, even though the fund managers analyzed here have very strong incentives to learn efficient behavior. However, results do not only feed pessimism about the capabilities of financial professionals. One indication of learning is the significantly lower degree of miscalibration shown by endorsers. Moreover, their stronger reliance on technical analysis and trading strategies may not conform to expectations derived from the efficient market hypothesis; however, this behavior seems consistent with their endorsement of behavioral finance. The results of our study demonstrate the persistence of behavioral patterns even with knowledge of their existence and thus their dominance over rational learning (in this sample). The failure to recognize own biases prevents from striving to correct them (Fischhoff, 1982b and Pronin, 2007). The challenge for professional fund managers, in particular for those with insight in behavioral finance, consists therefore in a more critical assessment of own investment behavior in order to discipline and rationalize it.