What the 2025 and 2020 Stress Episodes Reveal

The Federal Reserve’s comparisons of bank-loan and high-yield mutual funds during April 2025 and March 2020 test a specific proposition: do funds with more liquid assets before a shock experience smaller redemptions when investors become fearful? The answer is not constant. This chapter reconstructs what the tables actually show, keeping the ex-ante grouping, time window, sector, and statistical qualification attached to every result. That makes it possible to distinguish an association between a buffer and later flows from proof that cash protected a fund—or caused investors to leave it.

The comparison is useful because the two episodes put the same balance-sheet idea into different market settings. In each case, the Federal Reserve separates funds by whether their pre-shock liquidity ratio was above or below the median. The ratio is measured before the event, so it is treated as a starting condition rather than a response to the withdrawals being studied. Net flows are then reported for a defined period. Negative net flows mean that redemptions exceeded new money during that interval.

That design does not make the groups identical in every other respect. Funds above the median may differ in their assets, investors, management choices, and expectations about risk. The comparison therefore tells us how the groups behaved on average in a particular episode. It does not tell us what would have happened to the same fund with a different cash position.

April 2025: A Higher Buffer Did Not Mean Smaller Outflows

The April episode followed the April 2 “Liberation Day” announcement and the tariff-induced market volatility that followed. The relevant liquidity measure was taken in the first quarter of 2025, before the event window. This timing matters. If liquidity had been measured after the announcement, a higher ratio might have reflected a fund’s emergency response. Instead, the comparison asks whether funds that started with more of the specified liquid assets subsequently experienced different flows.

For bank-loan mutual funds, the above-median group had an average liquidity ratio of 8.8 percent of net assets, compared with 2.4 percent for the below-median group. There were 19 funds in the first group and 20 in the second. During April 2–9, average net flows were -4.2 percent for the above-median group and -2.7 percent for the below-median group. In other words, the group that started with more measured liquidity had the larger average outflow by 1.5 percentage points. The difference between the group averages was not statistically significant.

That last sentence is not a technical afterthought. The point estimate runs against a simple cash-protection story, but the reported BL difference is not strong enough to rule out ordinary sampling variation within the source’s comparison. It is accurate to say that above-median BL funds experienced larger average outflows in that week. It is not accurate to say that higher liquidity reliably predicted larger outflows for bank-loan funds in the episode.

The high-yield comparison points in the same direction with a somewhat stronger statistical qualification. The above-median HY group had an average first-quarter liquidity ratio of 8.5 percent, while the below-median group averaged 2.1 percent. Each group contained 63 funds. During April 2–9, their average net flows were -2.1 percent and -1.1 percent, respectively. The difference was -1.0 percentage point, and the table marks it as statistically significant at the 10 percent level. The sample represented 86 percent of total high-yield mutual-fund net assets in April 2025; the bank-loan sample represented 63 percent of that sector’s assets.