Skip to main content
Bond Funds and ETFs

Bond ETFs During Stress (March 2020)

Pomegra Learn

Bond ETFs During Stress (March 2020)

In March 2020, bond ETFs that investors thought were liquid became difficult to trade at fair prices. The authorized participant network seized up. Spreads between ETF prices and underlying bond values (NAV) widened to levels unseen since the 2008 financial crisis. The Fed had to step in to restore order. Understanding what happened—and why—is essential for any investor who holds bonds.

Key takeaways​

  • On March 18, 2020, corporate bond ETFs (LQD, HYG, ANGL) saw bid-ask spreads widen to 2–3%, versus typical 0.05%
  • The cause: massive inflows to Treasury ETFs and outflows from corporate ETFs, combined with frozen credit markets
  • Authorised participants couldn't finance arbitrage because their funding costs spiked and counterparty risk froze credit lines
  • The Fed's intervention (announcing bond-buying programs, supporting credit markets) restored functioning within days
  • The episode showed that ETF liquidity is real only when market structure holds; stress can break it

The week of March 9–13: early signs​

On March 9, 2020, the stock market fell 7.3%, the largest daily drop since 1987 (except for 2008). Investors panicked. Bond yields spiked (prices fell). The bond ETF market, which normally runs smoothly, began to show strain.

On March 12, BND (aggregate bonds) fell 2.2% in a single day—a decline matched by a widening of its ETF-NAV spread to 0.50%, versus the normal 0.05%. Investors who sold BND that day received a premium or large discount compared to the underlying bond values they were actually trading.

More alarming, some corporate bond ETFs (LQD, which holds investment-grade corporates) saw even wider spreads. Retail investors couldn't understand why: LQD's holdings were solid corporate bonds from companies like Microsoft and Apple. The bonds had intrinsic value. Yet the ETF was trading at a discount.

The reason: the market structure had seized up. Authorised participants couldn't step in to arbitrage the spread because:

  1. Funding stress: Credit markets froze. APs had to borrow money to finance their arbitrage positions, and interest rates spiked 500+ basis points overnight. The cost of financing was prohibitive
  2. Bond market dislocations: Bid-ask spreads on corporate bonds widened from 0.3–0.5% to 2–3%. APs trying to assemble the basket of bonds that backs LQD faced massive execution costs
  3. Counterparty concern: Major financial institutions were in distress. Investors worried about the solvency of large banks (including APs like Goldman Sachs and Citadel). Credit lines were tightening

The authorised participant network, which normally keeps ETF prices aligned with NAV, had broken.

March 18: the crisis peak​

On March 18, the situation reached its worst. News broke of a 23% corporate default probability (according to credit derivatives pricing). The Fed had not yet announced bold support.

On this day:

  • BND spread: 0.75% (retail investors trading at 3/4 of a percent discount to fair value)
  • LQD spread: 2.0%–2.5% (retail investors trading at 2–2.5% discount to fair value)
  • HYG (high-yield corporates) spread: 2.5%–3.0%
  • Other corporate ETFs: Similar stress

Investors who had believed they could exit at any time found they couldn't. Those who tried to sell at the market price got hit by massive discounts. Those who tried to buy with limit orders found shares scarce.

The psychological impact was severe: investors realized that the "liquid" ETF market was an illusion. The illusion worked because APs arbitraged constantly. Remove the APs, and the ETF was only as liquid as the underlying bonds, which were illiquid during stress.

The tragedy was that the bonds themselves were not in default. Investors didn't know if the companies were going bankrupt. They just knew that financing was expensive, credit lines were tightening, and the financial system was on edge.

The Fed's response and market recovery​

On March 18, after markets closed, the Fed announced emergency measures:

  1. Unlimited quantitative easing: The Fed would buy Treasury bonds in unlimited amounts, removing the biggest source of demand (Treasury ETF inflows)
  2. Secondary Market Corporate Credit Facility (SMCCF): The Fed would buy corporate bonds directly, supporting the corporate bond market

On March 19 (the next day), markets reversed:

  • Stock market rallied 5.2%
  • Bond ETF spreads collapsed
  • By March 20, corporate ETF spreads were back to 0.15–0.25% (not quite normal, but approaching it)
  • By late March, spreads were fully normal (0.05% or less)

What changed? Not the underlying bonds. Not the companies' credit quality (companies didn't suddenly become less risky). Instead, the Fed's backstop restored confidence in the financial system. Authorised participants believed credit lines would hold. Funding costs fell. APs returned to arbitraging.

Why corporate bonds suffered more than Treasuries​

During the March 2020 stress, Treasury ETFs (SHV, IEF, TLT) saw mild stress (spreads widened to 0.10–0.20%, vs. normal 0.03–0.05%). But corporate ETFs (LQD, HYG, ANGL, JNK) saw severe stress (spreads up to 2.5%).

The reason: Treasuries are the ultimate safe asset. When investors panicked, they bought Treasuries regardless of price. Demand was so strong that Treasury ETF spreads were pulled tight by sheer volume.

Corporate bonds are riskier. Demand for them dried up entirely as investors rushed to safety. With no buyers, corporate bond dealers couldn't move inventory. The bid-ask spread (what dealers charge) widened. Authorised participants trying to assemble a basket of corporate bonds faced massive transaction costs.

Additionally, corporate bond investors faced default risk. Investors didn't know if companies would survive the pandemic shock. The probability of recession-driven defaults was elevated. This risk premium widened spreads.

By contrast, Treasuries have no default risk. Investors knew the U.S. government would not default. So Treasury spreads, while widening, stayed much tighter.

Lessons for investors​

The March 2020 episode taught several lessons:

  1. ETF liquidity is contingent on market structure: The normal tight spread (0.05%) exists only when APs are actively arbitraging. During stress, this structure breaks. Spreads can widen 10–50x in severe cases
  2. Authorised participants have limited balance sheets: APs can absorb some arbitrage positions, but not unlimited amounts. During extreme stress, they exhaust their capital or credit lines
  3. Funding stress is the main culprit: APs can finance positions in normal times (cheap credit). During stress, credit is expensive or unavailable. The cost of carrying the position exceeds the arbitrage profit
  4. Corporate bonds are more fragile than Treasuries: Corporate bonds depend more on market structure and APs. Treasuries depend less on liquidity (always have buyers willing to pay near-fair prices)
  5. The Fed can fix the problem: When the Fed steps in and buys bonds directly (or backstops credit), it removes the shortage of buyers and restores normalcy

How to prepare for future stress​

For investors, the March 2020 episode suggests some practical guidelines:

  1. Don't assume ETF spreads stay tight: Keep some cash on hand (or other liquid assets) so you're not forced to sell during stress
  2. Avoid leveraged or complex bond ETFs during normal times: Inverse bond ETFs, leveraged Treasury ETFs, and commodity funds are more fragile than simple broad-based ETFs
  3. Aggregate bonds (BND, AGG) are safer than narrow corporate bond ETFs (LQD, HYG): Aggregates have more liquidity and are less dependent on corporate credit health
  4. If you need funds in the next 1–2 years, use short-duration Treasury or CD ladder, not bond ETFs: Avoid the stress risk
  5. For long-term portfolios (10+ years), the March 2020 stress was a temporary dislocation: Investors who held and didn't panic-sell recovered fully within months

Paradoxically, the investors who suffered most were those who panic-sold bond ETFs at 2–3% discounts. Those who held or bought the dip captured the recovery when spreads normalized.

The policy implications​

March 2020 exposed a structural fragility in the ETF ecosystem: the dependence on authorised participants' balance sheets and access to credit. The event sparked discussions among regulators and industry about:

  1. Should APs have more capital? Should they be required to maintain larger balance sheets to absorb stress?
  2. Should the Fed backstop bond markets automatically? Should it have pre-announced a willingness to buy bonds?
  3. Should ETF redemptions be restricted during stress? Should the fund gates redemptions (refuse to redeem) during extreme spreads?

As of 2024, no major regulatory changes have been implemented, but the risk is recognized. The Treasury market experienced a new stress episode in September 2019 (repo market dysfunction) and August 2024 (volatility spike), both of which highlighted similar fragilities.

The bottom line: ETFs are a great tool for retail investors in normal times, but they are not risk-free. Stress periods expose market structure risk. Investors should understand this and plan accordingly.

Next​

The March 2020 episode revealed the limitations of using standard bond ETFs as your only vehicle. The next article explores an alternative: defined-maturity ETFs (also called BulletShares), which offer a different structure for investors who want more predictability and less dependence on APs.