Over the last 25 years, exchange-traded funds (ETFs) have earned a significant share of investment flows. The functionality of indexed ETFs as reliable investment vehicles, as risk-transfer mechanisms and even as price discovery beacons for their underlying asset classes has contributed to their healthy growth. Despite the long track record and performance of ETFs, some skeptics still believe that ETFs will fail in the face of extreme market volatility.

The performance of ETFs throughout the market turbulence spurred by the COVID-19 pandemic shows that ETFs passed yet another episode of extreme market volatility and served their function as investment vehicles, risk-transfer mechanisms and price-discovery beacons. This article explains the market events that transpired in March and April of 2020 and shows how most ETFs delivered liquidity and returns in line with stated investment objectives across a variety of underlying asset classes.

Equity markets plummeted due to the global pandemic and oil price war

Although typical pockets of market volatility occurred periodically in the years preceding 2020, no one predicted the perfect storm that hit global markets this year, which was largely confined to March. Before then, equity valuations in particular slowly climbed with little to no interruption. As an illustration of these calm markets, the Chicago Board Options Exchange Volatility Index (VIX)—a standard measure of U.S. stock market price volatility—averaged just 14.74 between January 2016 and December 2019, and it stayed in this range until the third week of February 2020, when concerns of a wider-reaching outbreak of COVID-19 grabbed headlines.

Then, as initial COVID-19 cases emerged in the U.S. and the spread in China intensified, the global oil market showed its first cracks on declining demand (see Exhibit 1). By the first week in March, Saudi Arabia and Russia were in a full-fledged oil price conflict that sent the price of crude oil into a tailspin (see Exhibit 1), and many other asset classes along with it. In a matter of three weeks, oil crashed and front-month futures went well into negative territory. Meanwhile, the Federal Reserve (Fed) lowered interest rates to zero in an emergency decision on Sunday, March 8, 2020. Most of the U.S. went into lockdown, the floor of the New York Stock Exchange closed temporarily and the VIX spiked to 82.69, the highest level since the 2008 Global Financial Crisis (GFC).

The cost of trading U.S. equities also shot up to the highest levels experienced since the GFC, as extreme price volatility and uncertainty rippled across Wall Street. The volume of stock trading hit all-time highs, but this happened in much smaller increments and at bid-ask spreads that were three to four times the “normal” levels of January 2020 (see Exhibit 2).

The Federal Reserve’s policies impacted U.S. fixed-income markets

Problems in the fixed-income markets began on the heels of the Fed’s decision to cut its overnight lending rate to zero and roll out a series of never-before-seen measures to inject liquidity into the financial system to prevent a potential collapse. Buyers ran away from longer-dated, older vintages of U.S. Treasury bonds and Treasury Inflation-Protected Securities (TIPS), as well as a majority of investment-grade corporate bonds. As a result, the difference between the price a buyer might pay and the amount an owner was willing to accept (the bid-ask spread) grew apart drastically for these categories of securities.

Treasury securities and corporate bonds trade over the counter in the secondary market, and most owners buy and hold these securities unless there is an arbitrage opportunity or an interest rate move to secure a higher yield. Furthermore, a vast majority of individual bonds change hands infrequently, even in stable market conditions. Add an illiquidity crisis, and only the most opportunistic of buyers will trade these bonds. As shown in Exhibit 3, this occurred in the second half of March and into the first part of April, but began to normalize in May. At the height of this illiquidity event, 30-year U.S.Treasury bonds traded with a bid-ask spread of 3.1%. Under “normal” circumstances, the bid-offer spread on these bonds is generally in the low single-digit basis point range.

Perception versus reality of ETF market price dislocations from net asset value

As all of this turmoil unfolded in the stock and bond markets, ETFs faced a related challenge: Would these instruments remain tethered to the value of their underlying investments, or would their market prices dislocate from this intrinsic value?

Transparency of global stock pricing helped ease this challenge among equity ETFs. Even when the stocks held in an ETF do not trade at the same time as the ETF shares (for example, a U.S.-listed ETF that holds Japanese stocks), the availability of cost-efficient correlated hedges generally enables these funds to trade continuously, without difficulty and with few exceptions.

Periods of elevated price volatility have historically caused equity ETF bid-ask spreads to widen, but by a factor typically lower than experienced in the underlying securities market. The percentage of total equity market turnover attributable to ETFs also historically increases during these periods of stress. This increase in trading market share is precisely because of the efficiency of the risk-transfer mechanism inherent in the ETF structure. We saw this again during the COVID- 19-induced period of market volatility (see Exhibit 4).

Fixed-income ETFs, on the other hand, faced fundamentally different challenges compared with equity ETFs that led to the appearance of malfunction despite the fact that these products did exactly what they were built to do, which is to deliver the returns in line with their underlying indices. Beyond that function, during the March 2020 illiquidity crisis, bond ETFs availed real ( i.e., executable) prices in the underlying fixed-income markets and the higher trading volumes seen suggest they were used for the repositioning of portfolios.

Despite meaningful differences between the net asset value (NAV) and market price, investors continued to trade fixed income ETFs in large quantities, and at considerably tighter bid-ask spreads than the underlying baskets of bonds comprising the ETFs. During the height of the illiquidity crisis, “discounts” of between 4.4% and 6.3% to NAV were recorded in the aggregate bond ETF space (see Exhibit 5). These discounts took over a week to dissipate, in part because of the NAV catching up to the market price as opposed to arbitrage trading by the market-making community. Exhibit 5 shows an example of the bond ETF market’s efficiency.

Two very compelling metrics—volume traded and relatively tight bid-ask spreads—suggest that fixed income ETFs delivered valuable liquidity to investors. However, by another metric—discount to NAV—these same ETFs appear to be flawed. A closer examination into the drivers of the market price divergence from NAV may help illustrate why this perceived flaw is actually not a flaw.

One measure that many investors and academics use to gauge whether a fixed-income ETF is operating efficiently is to compare end-of-day NAV and closing price. On a particular day, if the market price closes below the NAV, the ETF is trading at a discount. If the market price closes above NAV, then the ETF is trading at a premium. The variability between the two should be relatively low over time, but in practice, it is not, especially among fixed-income ETFs. The three primary factors that confuse the relevance of this comparison in bond ETFs are:

1. A timing mismatch for valuation. Bond ETFs trade on exchange until the 4 p.m. ET U.S. equity market close, but in many ETFs, the individual bond prices for determining NAV are calculated at 3 p.m. ET.
2. Inconsistent pricing conventions for the calculation of bond ETF NAVs. ETF NAVs may use mid-market prices, bid-side prices or even offer-side prices for the underlying bonds held in the portfolio. Meanwhile, the closing price for an ETF is also inconsistent across the three listing exchanges. It is typically the last traded price (which could be either a buy or a sell), but it may be a derived price based on bids or offers in the minutes preceding the close of the continuous trading session.
3. NAVs for bond ETFs are often calculated on evaluated bond prices provided by third-party pricing vendors. Fewer than a quarter of the bonds in the Bloomberg Barclays U.S. Aggregate Bond Index trade on a typical day; many may not trade for weeks. Pricing vendors use proxy prices for these bonds to determine a daily price for use in a NAV calculation. When markets are gapping down in a low liquidity environment, such as in March 2020, these evaluated prices used in NAV calculations often do not reflect the current market conditions’ impact on the value of bonds.

Conclusion

Global markets have endured bouts of extreme market volatility as a result of the dot-com crash, the global financial crisis and, most recently, the COVID-19 pandemic sell-off. While the markets have stabilized since the initial downturn in March, no one knows what lies ahead. But this latest bout has again illustrated that ETFs have continued to serve investors well by delivering liquidity and market performance during these periods of uncertainty. While investors are still navigating the uncertain outcomes of COVID-19-inspired changes in our economy, which range from very low inflation and ultra-low interest rates to murky corporate earnings guidance, ETFs have once again proven themselves as resilient and reliable. So much so, that the Fed made an unprecedented policy move to purchase corporate bond ETFs in an effort to support the corporate lending markets and provide liquidity for the investment-grade bond market. This policy, along with the history of ETFs’ efficiency in various market cycles, helps reaffirm their value in a diversified investment portfolio.