Rebalancing costs are silent killer of ETF returns

Passive ETFs tracking US equity indices end up almost 10 bps a year worse-off than their active peers due to rebalancing inefficiencies

Jamie Gordon


The adage that all returns equal the market before costs has been a calling card for passive ETFs charging low fees, however, a recent paper from the University of Illinois has found hidden trading costs borne by rules-based strategies slowly erode long-term returns.

The research, titled Should Passive Investors Actively Manage Their Trades?and authored by PhD candidate Sida Li, stated regular rebalancing sees ETFs undergo a “significant amount of trading” in response to index constituent changes, IPOs, mergers and delistings.

In fact, Li found the median rate of portfolio turnover in US-listed ETFs tracking US equities was 16% in 2020, meaning the cost of portfolio reshuffles becomes a highly relevant factor when weighing up cost of ownership and performance drag.

Unfortunately for the 56% of US-listed ETFs passively tracking indices or employing “sunshine trading”, their methodologies mean they have sluggish rebalance processes, leaving them vulnerable to higher transaction costs and frontrunning by opportunists.

By tracking public indices that announce constituent changes five days prior to rebalancing, ETFs lose out on five days of what is normally upward price action, meaning the tendency to buy high is baked into their methodologies.

By trading at 4pm closing prices on rebalance days, Li said the ETF trades are large, abrupt, not driven by informational advantages and are fully predictable. Most importantly, stock prices rise by an average of 67 basis points (bps) over the five days prior to index rebalancing and fall 20 bps in the 20 days following. This results in an average performance impact of 14.6 bps per year.

“This high cost is especially surprising because ETF rebalance trades are generally rule-based and not information-driven,” Li said. “Given these poor execution strategies, these uninformed mechanical traders are paying higher execution costs than informed traders.”

In fact, Li argued the cost of this kind of mechanical rebalancing is “comparable to total management fees charged by ETF managers”.

What is damning is that active and less transparent ETFs in the US illustrate this kind of performance drag is not inevitable. Li terms these “opaque ETFs” and they either hide when or what they trade.

Those hiding when they trade only report their month-end holdings, meaning the pace at which they trade is unknown to investors. Crucially, these ETFs outperform their “sunshine trading” peers by 7.3 bps a year.

They do this by adjusting their baskets both in-between rebalance announcements and implementation dates, or by delaying rebalancing trades until after trading activity has normalised.

There are also opaque ETFs camouflaging what they trade by tracking their issuer’s own index – for instance, the Schwab 1000 ETF tracking the Schwab 100 index – which creates a considerable rebalance cost saving of 30 bps per trade versus passive ETFs. Considering an annual portfolio turnover rate of 16%, Li said the annual cost saving for these ETFs comes to 9.6 bps.

Implementing this cost-saving across all US-listed ETFs could save $1.7bn in rebalancing costs each year, Li said, while failing to save 9.6 bps in annual costs on a $2m retirement account could result in $29,000 losses at retirement.

“I provide evidence that even not-so-complex execution strategies. For example, simply camouflaging either the timing or the underlying stock of a trade can lead to considerable execution-cost savings for passive investors,” Li concluded.

However, while many would agree the gap between index announcements and rebalances is a free lunch for frontrunners, many ETF investors will not see the erosion of transparency as the correct countermeasure, especially considering this is one of the product class’s key selling points versus more opaque and often more costly fund structures.

While the tension between transparency and performance drag will continue, the costs borne by tracking public indices are becoming increasingly well-documented. In July, a Research Affiliates report, titled Revisiting Tesla’s addition to the S&P 500, found two-thirds of S&P 500 entrants had a higher price on the day the index committee announced the stock versus the day they were actually added. Meanwhile, 60% of deletions saw negative returns between an announcement and their removal.

Overall, due to public index announcements, new additions to the S&P 500 underperform deletions by an average of 20% after 12 months, Research Affiliates found, with the full effect of this being captured by “sunshine traders” who buy stock in bulk.

Related articles

Featured in this article


No ETFs to show.