Bond ETF providers may have played dirty to prevent a corona-inspired exodus last March. This is what the Bank of International Settlements seems to suggest in a paper earlier this month when it tried to explain why ETFs were trading at such huge discounts to their underlying assets at the height of the covid panic last year. But understanding exactly what they’re getting at here requires that we have a grasp on how the ETF market functions.

Why does this matter? The popularity of bond ETFs has surged over the past decade, with funds now managing more than USD 1.2 tn assets, compared to less than USD 10 bn in 2009.

The basics: ETFs are funds that are designed to track the performance of another asset or market, be that an equity index, a commodity or a currency. This means that in theory the value of an ETF will ebb and flow with the value of the underlying asset, or net asset value (NAV).

As we know from last year though, this doesn’t always happen. Sometimes the market will price an ETF below the value of the underlying assets (called a discount to NAV) and other times it might be more expensive (otherwise known as a premium to NAV).

“Authorized participants” are the most important players in this space: Before we can get a handle on how ETF prices are adjusted, we need to understand a bit about the principal ETF market makers. Large banks otherwise known as so-called “authorized participants” (APs) are responsible for creating and redeeming shares of an ETF, and act as intermediaries between investors and the ETF provider.

Creating and redeeming ETF shares: This group of financial institutions purchase the assets the ETF provider wants to have in its portfolio and hands them over to the provider, receiving ETF shares in return which they then sell on to investors. The redemption process works in the opposite way: investors sell ETF shares to the APs which then either hold onto them or hand them back to the ETF provider, receiving either the underlying asset or cash in return.

It is through this mechanism that ETF prices are aligned with the underlying assets: APs take advantage of cheap ETFs by purchasing shares in the secondary markets and redeeming them with the ETF provider, netting a profit when they sell the more expensive underlying asset. This brings the price of the ETF back into alignment by trimming the supply of ETF shares and raising the supply of underlying assets. The same principle applies when the fund is trading at a premium: the APs are incentivized to decrease the supply of the underlying asset and create more ETF shares, thus lowering the price of the ETF.

So what happened last March? At the height of the sell-off in the financial markets in March 2020, huge gaps opened between the share price of bond ETFs and their NAV, indicating that the arbitrage mechanism in the market had broken down. The two biggest bond ETFs — Vanguard’s USD 55 bn total bond market ETF and BlackRock’s USD 71 bn iShares core US aggregate bond ETF — traded at a 6.2% discount and a 4.43% discount respectively on 12 March.

And why did the arbitrage fail? In an effort to protect themselves from the market sell-off, the BIS believes that ETF providers handed over low-quality bonds to the APs. This does two things: not only does it dissuade the APs from redeeming more of their shares, but it also shores up confidence among investors by removing risky assets from the portfolio.

A shock absorber that comes at a price: Author of the report Karamfil Todorov describes this tactic as a “shock absorber” deployed by ETF providers during sell offs but warns that it also puts providers’ reputations at risk. Investors, seeing the provider short-changing the APs with junk bonds, may convince them to abandon the fund, he writes.

WANT MORE ON THE BASICS OF ARBITRAGE? Check this brief explainer brought to you by Khan Academy (watch, runtime: 2:50).