“和而不群”还是“卓尔不群”?——基于基金网络信息使用的视角

如何识别出优秀的基金管理者,理解其信息决策机制,对于优化资源配置、提升市场效率具有重要意义。本文探究了基金经理对自身所处基金网络中的共享信息的使用程度与其管理能力之间的关系。具体而言,本文通过基金的重仓持股构建了基金网络,采用基金自身交易与其所处网络中其他基金平均交易的偏离程度作为该基金对基金网络中信息使用的衡量。研究发现:(1)对基金网络中信息使用程度较低的基金的业绩要显著好于对基金网络中信息使用程度较高的基金。(2)更高的超额收益主要来源于基金经理优异的选股能力,虽由此承担了更多的异质性风险,却并未增大总体风险水平。(3)基金经理更换数据表明基金对网络内信息的使用程度更多地与基金经理特征相关而非与基金特征相关。(4)网络内信息使用程度直接反映了基金私有信息含量,因此更可能与基金经理能力相关。

Fund Managers' Skill and Use of Shared Information from the Social Network

Fund research has usually focused on whether fund managers have the skill to obtain excess returns. In recent years, the application of social network theory in capital market research has provided us with a new perspective for examining fund managers' skill. Fund managers may frequently benefit from their social network by sharing information with each other. A consequence of this practice is the convergence of fund transactions within the given network if managers use shared information to guide their investment. Nevertheless, some funds may deviate from the common trades in their network. We conjecture that managers of these funds rely less on shared information and more on valuable private information that they possess.
Specifically, when a fund manager receives shared information in his network, he will weigh the value of both his private information and shared information and then choose the information that generates the most profitable trading opportunity. That said, if the fund manager chooses to use shared information, it implies that his own private information generates less profit than the shared information. It may also reflect the fund manager's poor ability to obtain private information. Conversely, if the fund manager chooses not to use shared information, this indicates that his own private information generates more profits than the shared information. This may also reflect the fund manager's ability to obtain private information. Overall, we conjecture that the less information a fund manager uses from his network, the stronger his ability to manage the fund.
In this paper, we first create a fund-level network, denoted as the fund's information network, using the semi-annual large equity holdings of China's open-end mutual funds during the period from 2005 to 2017. We then calculate the deviation of the fund's transactions from the common trades of other funds in the network. The deviation thus measures the extent to which the fund manager uses shared information in the network. We find that funds with higher DFN(Deviation from Network) generate significantly higher risk-adjusted returns and also attract larger fund flows. Moreover, the better performance of funds with high-DFN is not driven by managers' shifting into more risky portfolios. On the contrary, it stems from fund managers' superior ability to select stocks with higher idiosyncratic risk.
We also find that the extent to which a fund uses shared information is a natural response to its own comparative advantage. Specifically, large funds and old funds have a wider network and their managers communicate more frequently with each other. Hence they use more information shared in the network. Managers of small funds and new funds communicate less frequently. However, these fund managers have new knowledge structures and more open minds. Thus, they are more likely to adopt an active management strategy. We also find that both fund characteristics and fund manager characteristics have a significant impact on the extent of using shared information. Further, by examining changes in the use of shared information before and after fund managers' turnover, we show that the use of shared information is more related to fund manager characteristics than to fund characteristics.
We further investigate heterogeneity in the relationship between the degree of using shared information and fund future performance under different economic conditions. We find that the relationship is stronger during the period when equity returns are more dispersed. This is also the time period when fund managers' ability is more easily recognized and rewarded. Lastly, we find that the degree of using shared information is positively related to a fund's private information. This finding provides further evidence that the degree to which fund managers use shared information is closely related to their ability.
This study expands identification of fund managers' ability, enhancing our understanding of the role played by fund managers in asset management. The results also suggest that improving fund managers' active management skills is essential for the healthy development of China's capital market.