Using novel daily holding data for exchange-traded funds (ETFs), I identify three types of ETFs that adopt distinct approaches to rebalancing their portfolios, which generates meaningful return heterogeneity. First, 56% of ETFs track public indices that pre-announce their rebalances, and they trade entirely on reconstitution days at closing prices. Their large, uninformed trades pay 67 bps in execution costs, a figure that is three times higher than what is paid in similar-sized institutional trades. Second, 7% of ETFs spread out their trades across 10 days and save 34 bps per trade or 7.3 bps per year. Third, 37% of ETFs use self-designed indices to avoid pre-announcements of rebalancing stocks and save 30 bps per trade. The alternative rebalance schedule leads to a tracking error of 10.6 bps per year and an information ratio of 0.69. For a $2 million retirement account that accrues over 30 years, the transaction cost savings rise to $29 thousand at retirement.
Li, Sida, Should Passive Investors Actively Manage Their Trades? (November 18, 2021). Available at SSRN: https://ssrn.com/abstract=3967799 or http://dx.doi.org/10.2139/ssrn.3967799