The Dynamics of ETF Fees

Abstract

Despite widely publicized fee reductions, average expense ratios of ETFs have remained relatively steady. Thousands of new funds have not led to lower fees. Investors should examine all available opportunities before choosing specific funds.

Despite widely publicized fee reductions, average expense ratios of exchange-traded funds (ETFs) remained relatively steady between 2004 and 2018. Even though thousands of new funds entered the market during this period, the arrival of most ETF sponsors into a narrowly defined area has not generally led to lower fees for competing funds. Given the impact of fees on long-term investment returns, investors should carefully examine all available opportunities before choosing specific funds. Furthermore, as objectives for newer ETFs become increasingly specialized, investors must also consider whether the benefits of targeted strategies justify their higher prices.

Disclosure: The authors report no conflicts of interest.

Editor’s Notes

Submitted 5 September 2019

Accepted 1 November 2019 by Stephen J. Brown

Acknowledgment

We thank Stephen J. Brown, Luis Garcia-Feijóo, CFA, CIPM, and two anonymous reviewers for their helpful comments and suggestions.

Notes

¹ Our sample contained 2,120 ETFs across all years but only 2,028 in 2018.

² Asjylyn Loder, “Who Needs Free? Passive Fund Prices Flatline,” Wall Street Journal (19 February 2019). www.wsj.com/articles/who-needs-free-passive-fund-prices-flatline-11550592000.

³ According to Box, Davis, and Fuller (2019), fee changes—both increases and decreases—are much more common in the complex strategies. Specifically, ETFs whose portfolios consist of derivative securities that are rolled over every few months must adjust expense ratios to account for changes in transaction costs. For most derivative securities, transaction costs were relatively high during the global financial crisis period (2007–2009) before moderating in the latter part of our sample.

⁴ All 10 ETFs experienced a negative total return in 2018, and 6 had lower returns than the S&P 500 Index. One year of performance data is not sufficient to draw any conclusions, so time will tell whether these strategies end up generating better outcomes for investors.

⁵ Morningstar Direct provides a taxonomy that distinguishes funds by what they own as well as by their prospectus objectives and styles. The most granular classification scheme, known as Morningstar Institutional Categories, was intended to provide assistance to institutional investors as they identify real peers, create genuinely diversified portfolios, and achieve better insight into individual portfolio strategies. Thus, ETFs competing within the same Institutional Category can be thought of as close substitutes. Each Institutional Category also belongs to a less narrowly defined Morningstar Global Category, and each Morningstar Global Category belongs to a more expansive global broad category group.

⁶ For instance, some ETF sponsors may be known for their innovative product offerings whereas others might excel at customer service or risk management. For sufficiently brand-loyal investors, the prospective fee savings would have to be large to make switching funds or fund providers worthwhile.

⁷ All ETFs that were alone in their category-characteristic combination had a fractional ranking of 50%.

⁸ We also compared investment fees within a regression framework that included controls for ETF size, fund flows, returns, share creation, age, tracking error, and a variety of variables measuring the degree of competition within each classification. Results of our regression analysis are consistent with the graphical analysis presented in Figure 3. Specifically, results suggest that more than 70% of the variance in expense ratios can be explained by the six sponsor indicators, the performance and competition control variables, the 268 yearly Global Category fixed effects, and the 261 Institutional Category and characteristic combination fixed effects. Additionally, Wald tests confirmed that $Bran{d}_{it}^3$ and $Bran{d}_{it}^6$ ($Bran{d}_{it}^2$ and $Bran{d}_{it}^5$) set significantly lower (higher) net expense ratios than the other market leaders.

⁹ See, for example, John C. Bogle, “The Relentless Rules of Humble Arithmetic,” Financial Analysts Journal, vol. 61, no. 6 (2005). www.tandfonline.com/doi/abs/10.2469/faj.v61.n6.2769.