Why passive ETFs are really active, and what that means for investors

Tracking errors result when exchange-traded funds (ETFs) pursue liquidity by deviating from their underlying indices, according to research co-authored by Wharton’s Yao Zeng

This article is republished with permission from Knowledge @ Wharton, the online business journal of the Wharton School at the University of Pennsylvania, which owns the copyright to this content.

Many exchange-traded funds (ETFs) that go by the moniker of “passive” are in fact active, and it is important that investors understand why and the implications, according to a new research paper titled Steering a Ship in Illiquid Waters: Active Management of Passive Funds, by experts at Wharton and elsewhere. The so-called passive ETFs get active to increase the liquidity of their shares, especially those that track relatively illiquid markets such as small-cap stocks and corporate bonds, the paper stated.

“Our punchline is that to be passive, an ETF has to be active,” said Wharton Finance Professor Yao Zeng, one of the paper’s authors. Prof. Zeng’s co-authors are Naz Koont, a finance doctoral student at Columbia University, Columbia University Finance Professor Yiming Ma, and University of Chicago Finance Professor Lubos Pastor.

Read more: Three megatrends reshaping the world of investment management

Authorised participants (APs), or financial institutions, manage liquidity in an ETF by “creating” or “redeeming” its shares, which they do by buying or selling the underlying securities and grouping them in baskets to suit the ETF’s structure. These baskets are carefully designed by the ETFs to ensure that ETF shares are liquid. The authors found evidence to support their research model that ETFs dynamically adjust their baskets and portfolios to create liquidity and maintain a balance between tracking the index and keeping the ETF shares liquid.

With those adjustments, ETFs “deviate substantially” from their underlying index, and “despite their passive image, ETFs are remarkably active in their portfolio management,” the researchers write. The implication of that is when ETF investors weigh their returns, they will face “tracking errors,” or differences between the returns of their ETF and those of the stated underlying indices.

The study focused on corporate bond ETFs, covering data from 2017 to 2020. The researchers selected bond ETFs because they feature a large difference between the liquidity of ETF shares and the liquidity of the underlying securities. “Fixed-income ETFs have been gaining popularity because through ETF shares, retail investors get the opportunity to invest in Treasuries and corporate bonds, which are typically less accessible to them than stocks,” said Prof. Zeng. That market segment has also grown rapidly, tripling over the past six years to reach US$342 billion by end-2021.

The study found that the passive ETFs in their sample achieved more liquidity by deviating from their stated structure with more cash and fewer bonds from their underlying index. “Both facts are costly to the ETF in terms of index tracking,” the paper stated. For example, the average bond ETF has a tracking error of 36 basis points annually relative to its underlying index, and the tracking errors are significantly larger for ETFs holding less liquid bonds. “This reflects the trade-offs that an ETF investor faces,” said Prof. Zeng. “If investors get liquid ETF shares out of a less liquid index, they should know that the ETF is likely tracking the index less closely. In that sense, fixed-income ETFs are a great innovation, but they are no free lunch.”

Takeaways for investors

The authors' aim to make that market practice transparent to ETF investors. “We think that it would be important for ETF investors to understand that their supposedly passive ETFs are actively designing creation/redemption baskets that are quite different from the underlying indices, which affect the ETF return they receive,” said Prof. Zeng. “They should be aware of that when they think about their returns.” The paper noted that since their arrival in 1993, ETFs have grown by end-2021 to command US$7.2 trillion in assets under management, and they account for about half the total assets of U.S. equity mutual funds. Among them, an estimate of between 95 and 98 per cent is labeled as “passive” in the sense that they are designed to track an index.

Subscribe to BusinessThink for the latest research, analysis and insights from UNSW Business School

In addition to the “liquidity transformation” in bond ETFs, the authors said their study emphasised their “active basket management, including the dynamic management of their cash balances and individual bonds.”

The study also found that a bond’s inclusion in an ETF basket has a significant effect on the bond’s liquidity. Bonds included in ETF baskets benefit from improved market liquidity in normal times. But they also face larger liquidity strains during bad times, such as during the COVID-19 crisis, when cash-strapped investors sold their ETF shares, causing bond redemptions to be “systematic and persistent.”

Passive ETFs that pursue liquidity could also cause distortions in bond prices. The portfolio adjustments they make to improve the liquidity of ETF shares may have left them with fewer bonds to redeem during a surge in fund outflows, such as during a pandemic scare. That imbalance in their portfolio could create “concentrated price pressure” for the bonds, Prof. Zeng explained.