Do mutual funds time market liquidity? A study on US mutual fund performance

摘要

The purpose of my thesis is to study whether actively managed mutual funds successfully alter market exposure in anticipation of changes in aggregate liquidity. Specifically, I test whether funds increase (decrease) market exposure prior to increases (declines) in liquidity. Furthermore, I investigate whether certain fund characteristics are related to funds’ success in timing liquidity. As liquidity has been shown to be positively related to asset returns and negatively related to market volatility, mutual funds could benefit from the ability to predict and time changes in aggregate liquidity. Although market liquidity has been identified as an important risk factor affecting asset returns, it hasn’t received much attention in timing literature. My study adds to the so far scarce evidence on mutual funds’ liquidity timing ability. I use monthly return data on 21,500 actively managed US mutual funds, ranging from January 1980 to December 2010. Fund data come from the CRSP Survivor-Bias-Free US Mutual Fund Database. The sample contains open-ended equity, fixed income and balanced funds. The Sadka (2006) liquidity shock measure and innovations to average stock turnover and trading volume are used to proxy for market liquidity. My results provide additional evidence that mutual fund managers in fact adjust the level of market exposure prior to changes in aggregate liquidity. The results are surprisingly consistent across all three liquidity measures and hold in a variety of robustness checks. On average, mutual funds increase market exposure by 2.8% to 7.5% prior to a one-standard-deviation increase in market liquidity, depending on the liquidity measure used. Funds with riskier investment strategies outperform more conservative funds in timing ability. Evidence on the relationship between timing ability and fund characteristics is weaker, but it seems that successful timers charge higher fees, have higher fund flow volatility, possibly experience higher flows and and trade less. Fund age and size appear insignificant.