Bond ETFs were stress tested in March’s market meltdown. Whereas discounts to net asset value (NAV) hadn’t gone above 0.1% by January, in some instances they shot beyond 6% in the COVID-19 induced turmoil in early spring.

The reason for this blowout—greater than that which occurred with their equity peers—is that while ETFs are very liquid, many of their underlying bond holdings are not. In such cases, daily prices at best are proxies (marked to market); and in the worst case, the price for the last trade.

There is therefore more of a chance that bond ETF prices will diverge from those of their less liquid underlying securities. This in turn begs the question: while this relationship means that ETFs can aid in bond price discovery, could it be a problem at times of extreme market stress? And, lastly, how have investors responded in the wake of these events?

Bunfight at the ETF Coral

All of which is benefitting at least one group—finance PhD students, who are getting a whole load of material. There’s been a veritable bunfight going on in the academic and central bank literature. It’s an ill wind, after all.

The Bank of International Settlements noted that these discounts reflected a number of factors: because of the relative illiquidity of corporate bond markets, NAVs incorporate information more slowly than prices, making deviations more likely during such times; dealers provided less support to corporate bond liquidity, potentially limiting the arbitraging of NAV discounts; and that there had been a possible rebalancing from short-duration investment ETFs to money market funds during March, contributing to the stress. All of which could be viewed as red lights for investors.

Another paper warns that “higher ETF ownership of investment-grade corporate bonds can reduce the ability of investors to diversify liquidity risk [which] may result in facing higher transaction costs and significant impact on bond returns, and even, not being able to trade during stress times”.

Again, hardly a glowing recommendation.

Bond ETFs were of course able to trade in March’s stress times. Indeed, 62% of professional investors in one survey conducted after March found trading ETFs during the violent market conditions easier than trading the underlying bonds. Indeed, the point has been made that, because they trade on a secondary market, in principle they should be more liquid than a bond mutual fund, which has no such advantage.

Quick recovery

Refinitiv Lipper figures show that, while March saw outflows of £679m in UK bond ETFs, the market quickly bounced back (Chart 1). In April, inflows totaled £479m, and in June bond ETFs took £1.4bn, its strongest month of the year by far. From April to October, bond ETFs took in £2.7bn. It’s worth noting that the value of physical bonds being traded over this period is likely much smaller, so this—not least the ability to trade throughout March—testifies to the robustness of the vehicle.

Chart 1: UK bond ETF flows, 2020, £bn

Source: Refinitiv Lipper

Total UK bond ETF assets are £180bn, out of total £5.12trn mutual fund and ETF assets. That barely scratches the surface. Even as a proportion of passive fixed income funds—£664bn—they’re still very much the minor partner.

While mutual funds also dominate passive bond flows (£1.48bn v £499m for ETFs), ETFs are taking greater market share. Between March and October, UK bond ETF assets went from £145bn to £180bn—and increase of 24%. That £35bn more than accounts for the increase in total passive bond assets over the period of £30bn.

Bond ETFs came through March’s crucible intact. For now, at least, it looks like investors have more than given them the benefit of the doubt.