Market efficiency and index fund flow: An empirical study of the relationship between passive investment and broad-market efficiency

Sammanfattning: An observable rise in the popularity of index funds have caused the index funds to, in 2017, capture 20% of total fund assets globally. A cornerstone of such passive investment is a belief in an efficiently priced security market. This paper aims to relate index fund flows with market efficiency during the period 2000-2019. Using S&P500 returns we estimate a market efficiency measurement called the Hurst exponent, using two accredited methods: the rescaled range analysis (RS) and the detrended fluctuation analysis (DFA). We find similar estimations as previous studies, wherein the S&P500 index have exhibited a slight mean-reverting return process, close to theoretical market efficiency. We further relate this time-varying market efficiency measurement of S&P500 to its index fund flows. Using a correlation filtering method to find index funds in the US targeting the S&P500 index, and aggregating these mutual funds individual flow, we obtain aggregate index fund flow. Conducting a Granger causality test on both fractional flow and dollar flow, we find a causality that market efficiency Granger cause index fund flow. We further estimate that a lesser degree of market efficiency have a negative impact on flow: the more long-term memory the index experience, the smaller level of flow. These results hold stronger for dollar flow rather than fractional flow.