Risk Consistency and Professional Fund Managers’ Investment Behavior: Questionnaire-based Analysis on Japanese Market*

Masayuki Susai (Nagasaki University)
Hiroshi Moriyasu (Nagasaki University)

ALL PDF FILE

1.Introduction

Recently, a substantial number of papers have investigated the behavioral characteristics of investors based on research results from within the field of behavioral finance. In addition to theoretical analysis, empirical research has been aggressively carried out using both market data as well as information on individual investors. Within this vast research, a variety of behavioral characteristics have been found in investors that match characteristics implicated in behavioral finance studies.

Suto and Toshino (2005) summarized the results of a questionnaire distributed in 2003 to domestic institutional investors in Japan. This questionnaire serves as the basis for the questionnaire we utilize in our research. In the above-mentioned paper, Suto and Toshino found that fund managers tend to sell with a shorter horizon than is optimally desirable.

*We appreciate financial support from Grant-in-Aid for Scientific Research(B), 17330075 (Japan Society for the Promotion of Science). The earlier version of this paper was presented at the annual meeting of the Japan Society of Monetary Economics, 2007 and the comments from Y.Tsutsui and T.Misumi are gratefully acknowledged.

Furthermore, their research demonstrates that fund managers show herding behavior. Regarding these results, the authors point out that one explanation for such behavioral characteristics is the pressure placed on fund managers by their clients. Based on information from the same questionnaire, Toshino and Suto (2004) found that Japanese institutional investors occasionally predict optimistically (or bullishly) on market returns, and that this behavioral tendency is more apparent when their predictions are based on the domestic market.

Suto, Menkhoff and Beckmann (2005) analyzed the results of a questionnaire conducted in both the US and Germany that are identical to the survey we used. Their analyses indicated that fund managers in the US tend to be more myopic, show stronger herding behavior, and demonstrate higher risk aversion than their counterparts in Germany.

Hiruma and Ikeda (2006) have investigated the factors that affect time-discounting rates as well as the impact of time-discounting rates on individual behavior. They find that a time discounting rate increases when the amount of money that must be paid out is smaller (money amount effect). Furthermore, this rate is significantly higher if the inter-temporal choice is made at a nearer point in time as opposed to a later point in time (dual discount phenomenon).

Misumi, Shumway and Takahashi (2006) have empirically explored disposition effect using Japanese on-line investor data provided by a Japanese securities company. Their results suggest that disposition effect does exist, and that investor irrationality may be its cause.

The questionnaire we utilized for this paper contains questions pertaining to the subjective risk attitudes of individual fund managers as well as their corresponding objective risk attitudes generated by expected utility theory. A question we pose that directly asks fund managers whether they consider themselves to be risk lovers, risk neutral, or risk averters directly exposes each manager’s subjective risk attitude. As for the problem based on expected utility theory, we provide questions that pose hypothetical investment opportunities in which there is variance in expected return value. This allows us to measure the theoretical risk attitude of each respondent. We refer to this type of attitude as ‘objective risk attitude’.

Consistency of risk attitude can be determined in the following manner. First, we measure the subjective and objective risk attitudes of each respondent based on the method outlined in the previous paragraph. If these two types of risk attitude are the same, for example, when a subjective risk averter answers the questionnaire in a manner consistent with an objective risk averter, then he or she will be called a ‘risk consistent’ fund manager.

There is also the possibility that a fund manager is ‘risk inconsistent’. All the respondents to our questionnaire are professional fund managers. Therefore, we can predict that all the respondents have specialized knowledge concerning risk within the investment field. With this in mind, respondents should theoretically be ‘risk consistent’ fund managers. However, for example, if a respondent states that they are a subjective risk averter, yet demonstrates through their responses that they are an objective risk lover, they will be considered a ‘risk inconsistent’ fund manager. We found two types of risk inconsistency in our analysis. The fund manager mentioned above, whose subjective and objective risk attitudes don’t match each other represents the first type of ‘risk inconsistent’ respondent. As for the second type, some fund managers inconsistently answer multiple problems on the questionnaire based on expected utility theory. For example, a respondent will choose an answer that demonstrates risk aversion in first problem, but choose a risk neutral answer in the second problem. Therefore, there are two types of risk inconsistency that appeared in our research.

Later, we will explore the two different types of respondents, ‘risk consistent’ and ‘risk inconsistent’ fund managers. As mentioned in the previous paragraph, there are two varieties of ‘risk inconsistent’ managers. In our analysis, we take up only the second type of risk inconsistency, that is, we will focus our attention only on objective risk attitudes. The rationale for this is that we want to know how ‘risk consistent’ and ‘risk inconsistent’ managers choose from alternatives concerning risk. With this objective in mind, we think the second type of ‘risk inconsistent’ manager described above can be included within the first type of risk inconsistency.

Risk inconsistency in fund managers may cause some objective and subjective differences in their preferences regarding the most important factors affecting investment. Therefore, it might be expected for such fund managers to invest in irrational ways. At the very least, it would not be going too far to assume that ‘risk inconsistent’ fund managers might invest more irrationally than ‘risk consistent’ managers.

In this paper, we look at the characteristics of herding and disposition effect, which we consider to be irrational investment behavior. When facing an unanticipated situation, ‘risk inconsistent’ fund managers may not be self-conscious, thus they may pay too much attention to the reactions of their colleagues or other fund managers. If this is the case, other fund managers may affect ‘risk inconsistent’ fund manager’s investment actions in these situations. This may be the root of herding behavior. Disposition effect refers to asymmetric assessment for upward and downward price change. This is also an irrational investment act. Therefore, in this paper we will investigate the investment characteristics of Japanese professional fund managers by paying close attention to the two above-mentioned behavioral biases that have recently been the subject of a considerable amount of behavioral finance literature. The data we use is from a questionnaire-type survey conducted in 2005. The respondents were all fund section discusses the results.