Though exchange-traded funds (ETFs) are similar to mutual funds, we identify several reasons how they are different based on their structure and trading characteristics. Therefore, we argue that the determinants of fund closure decisions for ETFs will not be the same as the mutual funds. We systematically explore those factors.
We use Cox Proportional Hazard model, which is considered a superior method, over the logistic regression models. All previous studies are based on logistic regressions.
We investigate the closure rate of ETFs over the 1995–2018 sample period. We find that the first three years are the most critical period for the survival of ETFs. Our full sample results show that early fund performance, the investment style of the fund, the expense ratio and fund family size are the most relevant factors influencing the likelihood of closure. When we consider equity-only funds, we find that key factors that influence fund closure are early fund performance, the expense ratio, failure to grow the fund's assets relatively quickly and the equity investment category of the fund.
Tracking error could be a significant factor. However, we have several missing values in the data. Therefore, we are forced to drop that variable. However, we use the SD of daily returns in lieu of that. Similarly, we were constrained by the availability of data for the equity style box scores.
Our study suggests that individual investors will be better off by investing in ETFs that are at least three-year to four-year old. If individuals want to invest in ETFs from the date of inception, the probability of survival is higher for an ETF within a larger fund family.
Hopefully, our research will attract the attention of CFPB and provide a warning to individual investors when they choose to invest in ETFs. More and more ETFs are getting included in retirement savings. So, abrupt ETF closures are likely to have large social implications for the future.
We are the first to use Cox Proportional Hazard model. We base our arguments from latest research on ETFs that the one earlier paper on ETF closure has missed. So, we examine the issue in a more systematic way.
