加载中...
Research on the Proper Size of Investment Funds: Evidence from Chinese Mutual Funds
过去十年间,中国公募基金市场发展迅猛,但与发达市场相比,中国公募基金仍相对较小,发展空间巨大。因此,基金规模是否可以无限扩张、还是应当受到限制,值得研究。已有研究存在若干局限:第一,文献中计算的最优基金规模大多是投资者视角的最优规模,而非基金经理视角的最优规模;第二,仅有少数文献研究基金经理的最优规模,但未考虑基金规模的持续扩张可能因规模报酬递减而导致投资者大规模赎回,从而使基金经理的最优规模在实践中无法实现;第三,本文表明委托代理冲突、规模报酬递减和投资者大规模赎回在基金市场中客观存在,并对基金规模决定发挥作用,但已有文献尚未将这些约束纳入最优基金规模的确定之中。针对上述问题和研究局限,本文发现由于基金市场存在规模报酬递减,基金规模不能无限扩张。基于识别出的三重约束——委托代理冲突、规模报酬递减和投资者赎回,本文提出了基金适当规模区间的概念,并改进经典的Berk and Green模型,构建了基金适当规模区间及其评价指标的理论模型。确定基金适当规模区间具有现实意义,因为它同时考虑了投资者和基金经理的利益以及市场对基金规模的约束。本文还提出了一种判断基金规模是否与基金经理能力相匹配的方法。本文构建的适当规模评价指标能够有效识别和判断规模过大、规模适当与规模过小的基金。本文以2011—2019年中国股票型和混合型基金为样本,实证检验了基金规模的适当性。主要结论如下:第一,2011—2015年中国公募基金的平均规模总体偏大,2016—2018年规模适当,2019年规模偏小;基金业绩与基金规模之间存在显著负相关关系,表明基金市场存在规模报酬递减;长期来看,平均净超额收益趋于零,且净超额收益非负基金发生大规模赎回的概率远低于净超额收益为负的基金。第二,随着市场冲击成本逐年下降,适当规模区间的上下界均下降,而适当规模区间的宽度则增大;原因在于适当规模区间的上下界受基金经理能力和冲击成本影响,而适当规模区间的宽度仅受冲击成本影响。第三,规模适当基金的业绩远好于规模过大和规模过小的基金,规模过大基金的净超额收益为负。第四,规模偏小基金占比最大且逐年上升;由于新成立基金数量快速增长而规模适当基金数量增长相对缓慢,规模过大基金占比位居第二且逐年下降,而规模适当基金占比最小。总之,规模适当的基金具有更好的业绩和资金流。因此,基金公司为提高规模适当基金的市场份额,应不断挖掘基金经理的潜力并持续提升其能力,同时应及时给予基金经理与能力和业绩相匹配的薪酬激励。监管机构应持续完善基金行业评价体系,不仅关注基金业绩,还应考虑投资者利益,评估基金规模与基金经理能力的适当性。
Abstract
Over the last decade, the Chinese mutual fund market has grown enormously. However, compared with the developed mutual fund market, Chinese mutual funds are still relatively small and have much more room for development. Thus, it is worth asking whether a fund's size can expand indefinitely or should be restricted. Previous studies have several limitations. First, the optimal mutual fund size calculated in the literature is mostly the optimal size for investors rather than fund managers. Second, only a few papers study the optimal size for fund managers, but they fail to consider that the continuous expansion of fund size may lead to large-scale redemption by investors due to diminishing marginal returns to scale, which makes the optimal size for fund managers impossible to achieve in practice. Third, this paper shows that principal-agent conflict, diminishing marginal returns, and large-scale redemption exist objectively in the fund market and play a role in determining fund size, yet their impacts have not been included in the determination of optimal fund size in the literature. In view of the above problems and limitations, this paper finds that fund size cannot be expanded indefinitely due to diminishing marginal returns to scale. Based on the three identified constraints—principal-agent conflict, diminishing marginal returns to scale, and redemption of investors—this study proposes the concept of a proper fund size interval. It also improves the classical Berk and Green model and proposes a theoretical model of the proper fund size interval and its evaluation index. Determining the proper fund size interval is practically relevant because it takes into account the interests of both investors and fund managers and the market constraints on fund size. This paper also develops a method for determining whether the fund size matches its managers' ability. The evaluation index of proper fund size constructed in this paper can effectively identify and judge oversized funds, funds with the proper size, and undersized funds. Using Chinese stock and hybrid funds between 2011 and 2019 as a sample, this paper empirically examines the appropriateness of fund size. The main conclusions are as follows. First, the average size of Chinese mutual funds is generally too large from 2011 to 2015, is appropriate from 2016 to 2018, and is too small in 2019. There is a significant negative relationship between fund performance and fund size, indicating diminishing marginal returns to scale. In the long run, the average net excess return tends to approach zero, and the probability of large-scale redemption of funds with a nonnegative net excess return is far less than that of funds with a negative net excess return. Second, as the market impact cost decreases annually, the upper and lower bounds of the proper size interval decrease, whereas the width of the proper size interval increases. The reason is that the upper and lower bounds of the proper size interval are influenced by managers' ability and impact cost, while the width is only influenced by the impact cost. Third, the performance of funds with the proper size is much better than that of oversized and undersized funds, and the net excess return of oversized funds is negative. Fourth, the proportion of undersized funds is the largest and increases each year; due to the rapid growth in the number of start-up funds and the relatively slow growth in the number of funds with the proper size, the proportion of oversized funds is the second largest and is decreasing annually, while the proportion of funds with the proper size is the smallest. In summary, funds with the proper size have better fund performance and fund flow. Therefore, for fund companies to increase the market share of funds with the proper size, they should constantly harness the latent potential of fund managers and keep improving their ability, while fund managers should be given timely payoff incentives to match their ability and fund performance. Regulators should continue to rapidly improve the evaluation system of the fund industry, and rather than paying attention only to fund performance, they should also consider investors' investment interests and evaluate the appropriateness of fund size and fund managers' ability.
中国公募基金规模适当性随时间变化,2011—2015年偏大,2016—2018年适当,2019年偏小。
以2011—2019年中国股票型和混合型基金为样本,基于改进的Berk和Green模型计算适当规模区间[q*I, q*M]及适当性指标I,发现基金平均规模在2011—2015年总体偏大,2016—2018年规模适当,2019年则偏小,表明中国公募基金规模与基金经理能力的匹配度呈现阶段性变化。
基金业绩与规模显著负相关,基金市场存在规模报酬递减。
回归结果显示,基金净超额收益与基金规模显著负相关,说明基金规模扩张会导致边际收益递减。样本期内净超额收益为负的基金占比更高,2011—2019年净超额收益为负的基金平均占比约8.52%,而净超额收益非负的基金占比约1.65%,且大规模赎回概率在负收益基金中显著更高。
市场冲击成本下降使适当规模区间上下界下移、区间宽度增大。
基于Amihud非流动性指标估计冲击成本b,2011—2019年b从0.0026降至0.0020,导致适当规模区间下界q*I从上界q*M均同步下降,而区间宽度q*M-q*I由7.21升至9.38,说明市场流动性改善扩大了规模适当的容错空间。
规模适当基金的业绩显著优于规模过大和规模过小的基金。
按I值将基金分为规模过小(I<0)、规模适当(0≤I≤1)和规模过大(I>1)三类,规模适当基金的平均净超额收益为0.0105,明显高于规模过小基金的0.0061和规模过大基金的-0.0393,且规模过大基金净超额收益为负,验证了适当规模区间的合理性。
规模偏小基金占比最大且逐年上升,规模适当基金占比最小。
样本期内规模过小基金平均占比51.83%,规模过大基金占比37.14%,规模适当基金仅占11.03%。由于新成立基金数量快速增长,规模适当基金占比最小且增长缓慢,规模过大基金占比逐年下降,规模偏小基金占比则逐年上升,2019年达到74.14%。
核心解释变量
基金规模(ln q_{t-1}),以基金净资产的对数衡量;同时引入基金规模平方项以检验非线性关系;另以适当性指标I_t作为规模适当程度的代理变量。
被解释变量
基金业绩(r_t),采用经CAPM和Fama-French模型调整后的净超额收益衡量,即基金收益率扣除无风险利率和市场基准收益后的残差。
样本与数据
2011年1月至2019年12月中国股票型和混合型开放式基金,剔除ETF、LOF、QDII及分级基金,样本量随年份从111只增至897只,平均基金数量约196只。
识别方法 / 模型设定
通过改进Berk and Green(2004)模型,在规模报酬递减、委托代理冲突和投资者赎回三重约束下推导适当规模区间[q*I, q*M]及适当性指标I_t=(q_t-q*I)/(q*M-q*I),将基金规模划分为过小、适当和过大三类,并利用面板回归和分组比较检验规模与业绩的关系。
内生性及稳健性检验
采用Amihud非流动性指标估计冲击成本参数b,并利用Goyenko et al.(2009)等替代流动性指标进行稳健性检验;通过分年度子样本回归、替换基金业绩度量方式(CAPM与Fama-French模型调整)以及改变样本筛选标准等方式验证结论的稳健性。
更多相关数据正在补充