[This article appears in our April 2020 edition of ETF Report.]
U.S. exchange-traded funds debuted in 1993 as passive equity-index-based products tracking broad markets, but in the past several years, ETFs have evolved to reach into harder-to-access, illiquid sectors.
Additionally, investors’ interest in ETFs picked up after the 2008 global financial crisis, reflected in assets under management that total $4 trillion.
Except for a few smaller corrections in intervening years, ETFs haven’t been tested in a serious financial crash. That’s particularly true for ETFs in smaller and less liquid markets.
A recent paper by two finance professors looks at whether ETFs could cause systemic risk in financial markets, how robust markets are to the risks, and if certain markets are more prone than others to such risks.
The idea that ETFs are just big, market-following index funds is 10 years out of date, explains Maureen O’Hara, the Purcell finance professor in the Samuel Curtis Johnson Graduate School of Management at Cornell University.
She co-wrote the paper “ETFs and Systemic Risks” with Ayan Bhattacharya, assistant professor in the Bert W. Wasserman Department of Economics and Finance at the Zicklin School of Business at Baruch College. It was published in January by the CFA Institute Research Foundation.
“ETFs are a varied zoo of creatures, and some of them are very big, and some of them are quite dangerous, and I think that’s what the paper is trying to talk about,” O’Hara said in an interview with ETF Report.
O’Hara points out that she’s “a huge fan” of ETFs, and in the paper, she and Bhattacharya write that there are some instances where ETFs appear to amplify market movements in times of stress and uncertainty. They offer a few examples.
One is the 2013 “taper tantrum” when then-Federal Reserve Chairman Ben Bernanke suggested that the Fed should taper its asset purchases under the quantitative easing program. They suggest that ETF bond outflows at the time may have affected bond yield spreads.
The authors cite academic research from 2017 analyzing the event, which they say “found that ‘a 1 standard deviation increase in ETF Tantrum outflows [led] to a 12.4 basis point greater increase in the yield spread of corporate bonds in September 2013.’”
They also reviewed the impact leveraged ETFs had in 2018 when the Cboe Volatility Index (VIX) rose by 20.01 points, reversing the trend of falling volatility. The authors cited the rush to exit two popular exchange-traded products, the VelocityShares Daily Inverse VIX Short-Term ETN (XIV) and the ProShares Short VIX Short-Term Futures ETF (SVXY).
“The data from this episode need more careful analysis, but anecdotal evidence seems to suggest a feedback loop at play due to the rebalancing needs of the ETFs, which amplified the buying pressure on the VIX,” they wrote.
The authors also infer that ETFs could be causing more volatility during end-of-day trading, noting how the percentage of trades that occur in the final hour—or even final minutes—of the day has increased. That’s due in part to index fund managers needing to keep their positions in line with the index.
The problem with heightened volatility at the end of the day is that “if there are problems at the close, there’s really no time to fix them. You can’t attract the liquidity to offset any problems before the market closes,” O’Hara said, adding that this is “particularly scary” with leveraged ETFs.
New Products’ Scant Study
Because ETFs have grown so quickly, there’s been little study of some of the newer products and how they would act during a time of crisis. O’Hara remarks that ETFs are largely exempted from the SEC’s liquidity requirements. That may be fine for a large cap index-following ETF, but perhaps not for smaller funds.
The paper raises a concern about using ETFs as cash substitutes by money market funds and other investment products, which could heighten the chance of problems in ETFs spreading to other markets.
“Treasuries are liquid; you can always sell them. Is that always going to be true for bond ETFs?” she asked.
That answer is unknown, because it’s unclear how many ETF cash holdings aren’t in actual cash or Treasuries, and because these ETFs haven’t been stress-tested.
“I think we need to know the dimensions of the problem. It may not be a problem,” she said. “Regulators should have a better idea of that.”
Step-Away Risk Problem
Another concern is “step away” risk by authorized participants, particularly in illiquid markets where APs are often also the dealers in the underlying markets, the authors say. O’Hara and Bhattacharya explain that it can sometimes be difficult for APs in ETFs based on hard-to-trade illiquid assets to present the entire basket of underlying assets to the issuer during creation/redemption. When that occurs, only part of the basket might change hands when new ETF shares are issued.
Their concern is that if APs hold a dual role in illiquid markets, it can have a significant effect on arbitrage activity, citing a previous study from other researchers, who found that a “1-standard deviation increase in bond market illiquidity generates a 10-40% decline in AP arbitrage sensitivity.”
Since APs aren’t required to undertake creation/redemption activity, O’Hara and Bhattacharya suggest these deviations may occur in particular illiquid markets. Step-away risk can have a domino effect, because it affects money managers holding ETFs in other types of funds, especially those who use ETFs for cash management purposes.
“This practice of using ETFs as cash equivalents is only appropriate, however, if the ETFs can always be turned into cash immediately and relatively without cost. Disruptions in the bond market, leading to disruptions in the creation and redemption process for fixed income ETFs, would undermine this ability,” the authors said.
O’Hara reiterates that ETFs in general are a good thing, but that as the vehicle has evolved, the industry and regulators need to look at some possible weakness, and if necessary, fix these before they inadvertently cause harm.
“The paper is trying to raise some issues and say, ‘we’re not sure yet whether these are risks, but at the moment, we don’t even have the data to know that,’” she said.