ETFs and their role of liquidity in bond markets are still a gray area, with supporters of the wrapper claiming they can act as a price discovery tool during times of market stress, while critics argue that they provide an “illusion of liquidity” by providing easy access to illiquid parts of the market.
In the period up to March, ETF liquidity was still of great concern to some investors. An Invesco survey found that 80% of banks in developed markets viewed “liquidity concerns” as a major barrier to increasing their exposure to ETFs.
However, when the survey was released later in the year, Invesco was quick to point out that while there were some concerns before the coronavirus, “ETFs turned out to have worked like planned”.
While ETFs remained relatively bullish in equity markets during the March turmoil, the same couldn’t be said of fixed income ETFs which started trading at record discounts when underlying market liquidity subsided. dried up.
According to an analysis conducted by Citi, around 80% of premium bond ETFs have started trading at unprecedented discounts to their net asset value (NAV), including the largest ETFs in the bond market such as Vanguard Total Bond Market ETF (BNP) which traded at 6.2% reduction from net asset value.
Discounts like this had never been seen before in the ETF structure and critics were quick to highlight the flaws in the ETF wrapper during times of market stress.
Highlighting this, an academic study found that increasing ownership of corporate bonds in the ETF space has an impact on bond market liquidity.
In particular, Efe Cotelioglu, a doctoral student at the Swiss Finance Institute, said in the study that corporate bonds that are heavily held by ETFs often share similar liquidity characteristics, which could pose problems for investors when they seek to exit positions during times of extreme volatility.
In contrast, he found that mutual funds did not have the same impact as “managers have the discretion to respond to investor flows by protecting liquidity and trading securities selectively.”
However, this is where the role of the secondary market comes in for ETFs, as the Bank of England (BoE) pointed out in May which stated that the ability to trade in the secondary market, unlike a mutual fund, reduces the risk of fire. sale.
One need only look at the Neil Woodford saga in 2019 or the UK real estate fund gating to realize the liquidity mismatch between the structure of mutual funds and illiquid assets. Any large amount of redemptions will force the manager to suspend trading in a fund, leaving investors trapped.
Highlighting mutual fund flaws, however, does not leave ETFs in the clear despite the advantages of the secondary market.
As the BoE said in the same report: “If ETF liquidity deteriorated in a crisis, it would pose risks to all market participants who depended on them for liquidity and price discovery.”
The idea of the price discovery role of ETFs needs to be deepened. When ETFs started trading at record discounts, ETF promoters, such as issuers and pundits, argued that this reflected the role ETFs play in finding prices in the market.
As BlackRock said in a report, “Rather than exposing a loophole in the structure of ETFs, these discounts have highlighted how fixed income ETF prices can provide a window into the underlying conditions. market, transmitting real-time information and providing price information to market participants. “
Analysis of the five largest European fixed income ETFs during the coronavirus pandemic
In other words, when liquidity dried up, the real-time pricing function of fixed income ETFs reflected the true fundamental value of the market versus the “stale” NAV which is only calculated. the end of the trading day by the pricing agents.
However, new research has come to light that shows the situation to be more nuanced than what BlackRock and other ETF issuers have argued.
In a study titled When the Selling Goes Viral: The Debt Market Disruptions in the COVID-19 Crisis and the Fed’s Response, Valentin Haddad and Tyler Muir of UCLA and Alan Moreira of the University of Rochester found that ETFs were not a price discovery tool but in fact a measure of the liquidity premium that investors were willing to pay. pay at that time.
Where the price discovery argument collapsed, they said, was in the behavior of the iShares 20+ Year Treasury Bond (TLT) ETF which started trading at a premium to to NAV during heightened volatility. Was this a sign that the fundamental value of Treasuries was going to rise? No, it showed that the most liquid instrument was the ETF versus the basket of underlying bonds.
“In this case, the NAV has undergone larger movements than the price of the ETF,” the authors said. “This suggests that the gap that widened between ETF prices and bond prices in their basket was not related to a slow update in NAV, but rather to the more liquid asset,” the ETF in this case, trading at lower prices than the less liquid asset. , the basket of individual bonds. “
Figure 1: ETF and NAV prices. The figure shows the cumulative log returns of various ETFs as well as their implied cumulative log returns of the contemporary NAV for an investment grade corporate bond (LQD) ETF, a high yield corporate bond (HYG) ETF, a Treasury ETF (TLT), and an S&P 500 ETF (IVV).
Source: Haddad, Muir, Moreira (2020)
Following the March turmoil, ETF issuers highlighted how Authorized Participants (APs) remained in the market despite the volatility. According to a study by Invesco, the same 16 access points were active in its ETF range in February and March.
Despite this, the role of PAs in the creation-redemption process remains an area of interest for regulators. In particular, APs are not obligated to create and redeem ETFs, therefore, in theory, during times of market stress, ETFs could be forced to trade only in the secondary market, resulting in premiums or significant discounts on the net asset value as an investment. trust.
For the European Systemic Risk Board (ESRB), the large discounts represented friction in the ETF’s intermediation chain, because “the APs would normally align prices via an arbitrage mechanism”.
“Low levels of liquidity in the underlying assets may have hampered this arbitrage mechanism, which normally closes these spreads, as ETF prices internalize the low level of liquidity of the underlying bonds,” said the ESRB.
Haddad, Muir and Moreira added: “The [March] The episode could be the result of a combination of strong selling pressure exerted by investors on covered debt ETFs and limitations on the ability of brokers to engage in arbitrage.
“For the latter, PAs might have been reluctant to accept these transactions because of the balance sheet space they occupy, or the adverse selection and risk of volatility that ensue.”
Their findings follow research conducted by Yao Zeng of the University of Pennsylvania, who argued that there was a “conflict of interest” for the major banks responsible for ETF and ETF arbitrage. management of their own inventory.
“When banks’ balance sheets are tight, they can opt out of ETF arbitrage,” Zeng continued. “But worse, because ETFs are easier to trade than bonds, APs can even create ETFs for the purpose of managing their illiquid inventory of corporate bonds rather than to fill the price gaps between the markets of the ETFs and bonds, further distorting ETF arbitrage and leading to even larger price differentials. “
There is still work to be done in this area to ensure that the buyout process runs as smoothly as possible during periods of dwindling liquidity.
While liquidity did not entirely disappear during the coronavirus sale, it has certainly provided a fascinating case study of how fixed income ETFs behave during times of market stress.
As an industry source put it, “No one is saying that the ETF wrapper is perfect, but it’s more perfect than other structures.”