تاثیر تعامل علمی بر انتخاب سهام و عملکرد: شواهد از ژاپن

Using mutual fund holdings data and fund manager demographic data, this paper examines whether academic interactions between fund managers and board members affect fund manager investment decisions and fund performance. I show that mutual fund managers are more likely to hold academically related stocks. Performance tests provide empirical evidence that academic interactions are beneficial to earn more profits. In addition, I show that mutual fund managers seem to lose profitable opportunities due to academic interactions because of their investment styles. Overall, this paper shows that mutual fund managers seem to take advantage of academic interactions to earn greater profits.

This paper examines whether an academic link between mutual fund managers and board members affects managers’ investment styles and performance. Based on an analysis of mutual funds’ equity holdings data, the results show that mutual fund managers favor academically related stocks. This paper also shows that the academic link between mutual funds and board members is important with regard to yielding higher performance. In the field of finance, how information has been incorporated into securities prices has been rarely discussed in the past few decades despite the common sense that information moves prices. To understand the mechanism through which information affects asset prices, academics have recently tried to relate the effect of information on prices to social interactions. In keeping with previous studies, this paper examines whether one type of social interaction, namely, the academic interactions of fund managers, impacts fund managers’ preferences for stocks and the subsequent performance of stocks. The contribution of this paper is its implications for stock market efficiency. Under the traditional efficient market hypothesis (hereafter EMH), no investor can beat the market if the stock market is efficient in terms of information. The finding of profitability due to academic interactions indicates a violation of the traditional EMH. Instead, as documented in Grossman and Stiglitz (1980), it implies that some investors can outperform the market because of an informational advantage. This study considers one kind of social interaction, namely, academic interaction. I expect that fund managers who engage in academic interactions are more likely to hold academically related stocks that perform well due to informational advantages.1 To test whether academic interactions have important effects on stock selection and performance, I use holdings data on 79 active mutual fund managers. As these data are produced semi-annually or annually, I can observe trading decisions by observing changes between the two reporting dates. Combining these data with demographic data regarding fund manager’ academic backgrounds obtained from the Morningstar website, I examine whether academic backgrounds influence stock preferences and fund performance. I construct variables to measure the strength of the academic interactions of managers. If a fund manager graduated from the same university as one of the board members in a listed firm, the firm is regarded as “academically related” to the fund manager. By measuring academic interactions between managers and board members, I examine whether academic interactions impact the stock preferences of fund managers. In the analyses of stock picking by mutual fund managers, I find that they tend to hold academically related stocks even after controlling for the effects of investment style, manager-specific characteristics, and time-series fixed effects. However, in the analysis of portfolio weight decisions, fund managers do not always tilt their investment toward academically related stocks, which implies that mutual fund managers cannot take advantage of the full information obtained from academic interactions. To test whether holding academically related stocks is beneficial, I compare the performance of academically related stocks to that of academically unrelated stocks. I create calendar-time portfolios that mimic the aggregate portfolio allocation of mutual funds in related and non-related stocks. The results show that academic interactions yield economically and statistically significant profits for the mutual funds. Academically related stocks outperform academically unrelated stocks by 0.018 percent per day, that is, 4.5 percent per annum. The results hold in some robustness checks. In addition, I examine whether mutual fund managers can take full advantage of information induced from academic interactions by comparing the performance of academically related stocks held by them to that of these stocks not held by them. The results show that academically related stocks held by mutual fund managers do not always outperform academically related stocks not held by them, which implies that mutual fund managers seem to be unable to take advantage of full information included in academic interactions. The comparison of academically related stocks held by mutual fund managers to stocks not held by them also shows that stocks in the latter group are more likely to be concentrated among smaller and value stocks, which are less likely to be held by mutual funds. Because of the investment styles that fund managers employ, these managers might be unable to take advantage of full information from academic interactions. Overall, the empirical evidence shown in this paper is consistent with the prediction that one type of social interaction, namely, academic interaction, drives informational advantages of mutual funds and yields higher performance. The remainder of the paper is organized as follows. In Section 2, I describe the background and related literature based on theoretical and empirical evidence. In Section 3, I describe the data. Section 4 provides empirical results on holding decisions. The empirical results on performance tests are reported in Section 5. In the last section, concluding remarks are documented.

This paper examines whether academic links between mutual fund managers and board members affect fund managers’ investment styles and performance. In the analyses of stock-picking decisions, I show that mutual fund managers are more likely to hold academically related stocks. However, they do not always seem to overweigh academically related stocks. These two combined results imply that mutual fund managers cannot take advantage of full information obtained from academic interactions. Performance tests provide empirical evidence that academically related stocks outperform academically unrelated stocks by about 4.5 percent per annum, which is measured by a 5-factor model. This empirical evidence holds when I consider some other ways in which fund managers might acquire informational advantages. In addition, I show that mutual fund managers lose profitable opportunities due to academic interactions by comparing the performance of academically related stocks within mutual funds’ positions to that of academically related stocks that fund managers do not hold. Overall, these results imply that mutual fund managers seem to take advantage of academic interactions to earn greater profits. The results in this paper are consistent with previous studies, such as Cohen et al. (2008) and Massa and Simonov (2005). This paper suggests one way to explain how informed investors acquire informational advantages and yield higher performance.