Two Liquidity Measures, Two Different Questions
Bank-loan and high-yield mutual funds cannot be assigned a single meaningful liquidity score from the amount of cash they hold. The focus here is how the Federal Reserve measures two different sides of the problem: immediately available liquid capacity and exposure to assets that are especially difficult to value. Readers who keep those measures separate can use the reported BL and HY trends as evidence without treating a proxy as a complete safety rating.
The distinction matters because the fund’s cash position answers a capacity question, while the rest of the portfolio raises a saleability question. A fund may be able to meet some withdrawals without selling assets, yet still face a difficult funding decision when that capacity is used. Conversely, a portfolio can contain assets that are not in the most extreme valuation category but are still hard to sell in size when market depth disappears.
The Federal Reserve’s August 2026 update is designed to examine those relationships over time. It expands earlier work that covered the ten largest bank-loan and high-yield mutual funds to all BL and HY mutual funds in its coverage. It also uses the more granular SEC Form N-PORT data that replaced Form N-Q in 2020 and compares fund flows around the April 2025 tariff announcement and the March 2020 onset of the pandemic. The data do not settle the safety question by themselves. They make the question more precise.
The Two Axes in the Federal Reserve Measure
The first measure is the liquidity ratio. In plain terms, it asks how much of a fund’s net assets are held in instruments treated as readily usable liquid resources. The Federal Reserve defines it as:
liquidity ratio = (cash and cash equivalents + U.S. Treasury bills + short-term investment vehicles) ÷ total net assets
Including Treasury bills and short-term investment vehicles is important. Calling the measure a “cash ratio” would be convenient, but it would be incomplete. The numerator includes assets that are intended to be close substitutes for cash in the funding decision. The denominator, total net assets, turns the amount into a comparable share across funds of different sizes.
That normalization also explains what the ratio does not say. A five-percent ratio does not mean that five percent of every possible redemption wave can be met without consequence. It means that, at the measurement date, the specified liquid instruments represented five percent of the fund’s net assets. The usefulness of that capacity depends on the size and speed of withdrawals, the fund’s other sources of cash, and the conditions under which the remaining portfolio would have to be sold.
The second measure asks a different question. The illiquidity ratio is the fraction of total net assets represented by Level 3 assets. Level 3 assets are valued using significant unobservable inputs, which require management’s best estimates and generally make the assets difficult to value. The measure therefore isolates a particularly extreme dimension of illiquidity and valuation uncertainty.
It is not a count of every asset that could be slow or expensive to sell. A security can have observable inputs and still be difficult to liquidate in a large position. A fund can have a small Level 3 share while holding substantial Level 2 positions or concentrated holdings in thinly traded securities. The Federal Reserve explicitly notes that its analysis does not observe market depth. That omission is not a minor technical footnote: market depth helps determine whether a sale can occur at a price close to the value used for the fund’s net asset calculation.
The two measures consequently have different meanings:
- The liquidity ratio is a signal about the resources that may be available before the fund must sell more difficult positions.
- The illiquidity ratio is a signal about the portion of the portfolio exposed to the greatest valuation difficulty, not a full inventory of slow-to-sell assets.
- Read together, the measures can show whether liquid capacity and extreme illiquidity are moving in the same direction or in opposite directions.
This is why the Federal Reserve treats the measures as monitoring tools rather than as a precise ranking system. They provide a consistent way to track movement across time, but consistency is not the same as completeness. A reader should ask what the numerator includes, what the denominator normalizes, and what the classification leaves outside the frame.
Classification practices create another limit. The Federal Reserve says it emphasizes time-series trends rather than point estimates partly because funds may classify similar assets differently. For example, some funds may include an illiquid asset such as commercial paper among cash and cash equivalents. That would bias the liquidity ratio upward on balance: the reported numerator would look more immediately usable than the underlying asset really is.
The Level 3 measure has a corresponding vulnerability. Assets can move between Level 2 and Level 3 classifications, and that “fluidity” can result in an understatement of the illiquidity ratio. A decline in the reported Level 3 share therefore does not necessarily prove that the portfolio became easier to sell. It may reflect a change in classification, valuation inputs, or the composition of the positions that remain outside Level 3.
These caveats do not make the measures useless. They determine how they should be used. A repeated trend in a consistent series can reveal that the balance between liquid capacity and hard-to-value exposure is changing at the sector level. It cannot, without additional evidence, tell us the exact liquidation cost of a particular position or the probability that a particular fund will face a run.
The data source also has a time boundary. Form N-PORT succeeded Form N-Q in 2020, and the Federal Reserve uses N-PORT’s more granular portfolio information to broaden the coverage and update the earlier monitoring exercise. That creates a stronger basis for observing recent patterns than a narrow sample of the largest funds would provide. It does not turn the result into a randomized test. The measures remain descriptive, and the classification choices remain part of the evidence’s uncertainty.
What the BL and HY Trends Say—and Do Not Say
The sector comparison gives the two ratios practical meaning. Bank-loan and high-yield mutual funds are both corporate-debt funds, but they are not interchangeable portfolios. Bank-loan funds primarily hold bank loans; high-yield funds primarily hold non-investment-grade corporate bonds. Both asset classes are relatively illiquid, yet their aggregate size and measured trends differ.
From the fourth quarter of 2019 to the fourth quarter of 2025, bank-loan mutual-fund net assets declined by $19 billion, or 21 percent, to $71 billion. High-yield mutual-fund assets increased 13 percent to $263 billion. The scale figures are not a resilience ranking. They show that a sector’s cash percentage is embedded in a changing pool of assets. A ratio observed in a shrinking sector and a ratio observed in a growing sector may have different implications for market absorption, portfolio turnover, and the range of positions a manager must fund.
The bank-loan pattern is the clearest example of why the two axes must be read together. The median BL liquidity ratio increased after 2020 and reached almost 10 percent of net assets in 2021. It then declined and remained around 4.5 percent from 2024 onward. The range between the fifth and ninety-fifth percentiles narrowed after 2021, suggesting less dispersion in the reported liquidity ratio across the observed funds than during the earlier rise.
If that were the only series available, a reader might describe the recent position as stable. But the illiquidity ratio adds a different signal. The median BL illiquidity ratio rose to above 1 percent in 2020, declined afterward, and since 2025 has approached the levels seen during the pandemic-era increase. The Federal Reserve’s interpretation is that liquidity-transformation risk in the median bank-loan fund has increased on balance in recent years: the liquid ratio has been relatively stable while the measured share of Level 3 assets has ticked up from post-pandemic lows.
That conclusion is carefully limited. It does not mean that the median BL fund has crossed a specified safety line. It does not mean that its cash holdings provide no protection. It means that the protection represented by the liquid ratio has not risen in step with the extreme illiquidity signal. If the fund must meet redemptions while the residual portfolio contains more difficult-to-value assets, the same liquid capacity may have a more demanding job.
The Level 3 subsample also matters for the denominator of the conclusion. The Federal Reserve restricts the illiquidity analysis to funds that reported Level 3 assets in at least one reporting quarter. As of the fourth quarter of 2025, those funds represented 84 percent of total bank-loan mutual-fund assets. That is broad coverage, but it is not the same as saying the Level 3 measure describes every dollar in the sector. The coverage qualification should travel with the trend claim.
High-yield mutual funds show a different pattern. The median HY liquidity ratio stayed about 4 percent from 2020 onward. The median illiquidity ratio was relatively flat, and the percentile range narrowed notably after 2020. On that evidence, the Federal Reserve says liquidity-transformation activity appears to have declined on balance in recent years for HY funds.
Again, “appears” is doing useful work. A flat Level 3 ratio and a stable liquid ratio are consistent with a lessening of the measured transformation pressure, but they do not demonstrate that every high-yield fund can withstand a run. The Level 3 subsample covered 97 percent of HY-fund assets as of the fourth quarter of 2025, which gives the aggregate observation substantial reach. It still leaves the market-depth and concentration questions unresolved.
The contrast between the sectors is more informative than either sector’s number alone. For BL funds, stable liquid capacity accompanies a renewed rise in the extreme illiquidity proxy. For HY funds, both measures are comparatively stable, with the illiquidity distribution narrowing. The difference argues against a portable rule such as “four percent cash is safe” or “ten percent cash is safe.” A percentage becomes meaningful only in relation to what lies outside it and how that portfolio behaves under stress.
The comparison also shows why medians should not be mistaken for individual-fund descriptions. A median identifies the middle of the observed distribution. It does not tell us whether a particular fund is above or below it, whether its holdings are concentrated, or whether its investor flows resemble the sector average. A narrowing percentile range indicates less dispersion in the reported measure, not the disappearance of tail risk.
Nor does the Federal Reserve’s use of Level 3 assets imply that other liquidity categories are safe by default. Level 3 is an extreme valuation category. Market depth, trading concentration, position size, and the ability to sell without significantly changing market value can create pressure before an asset reaches that category. The reported ratios are therefore best understood as two windows onto a larger problem. One window looks at resources available near the start of a redemption response; the other looks at the portfolio most exposed to unobservable valuation inputs.
The practical reading rule is simple but demanding. Start with the liquidity ratio, but immediately ask what it excludes. Read the Level 3 ratio, but ask what may sit in Level 2 or in concentrated positions. Compare trends, not just levels, because the authors warn that classification differences can distort point estimates. Keep BL and HY separate because the sectors have different portfolios and different aggregate paths. Finally, treat the result as evidence about liquidity transformation, not as a prediction of investor behavior.
The next test is behavioral rather than purely balance-sheet based. If a higher ex-ante liquidity ratio represents protection, funds with more of it might be expected to experience smaller outflows when stress arrives. If it represents precautionary management or exposure to investors who are more likely to redeem, the relationship could look different. The Federal Reserve’s April 2025 and March 2020 comparisons provide the evidence needed to examine that tension.