Private Credit Report 2022 — The Floating Rate Trap

The initial reading of 2022 for private credit was straightforward and positive. Direct loans are predominantly floating rate — priced at a reference rate plus a spread — so when the reference rate rose sharply, so did the coupon. Portfolio yields increased substantially without any new lending.

This was presented, correctly, as evidence of the asset class's inflation and rate protection. Unlike fixed-rate bonds, which fell in price as rates rose, floating rate loans saw their income rise.

The incomplete part of that reading is the borrower's side of the same transaction.

The lender's higher coupon is the borrower's higher interest expense. A leveraged borrower paying a floating rate saw its interest cost rise by a very large proportion within roughly a year. That borrower's business did not change. Its revenue did not rise to match. Its debt did not shrink. But the cash required to service that debt increased substantially.

For a company underwritten at a leverage level that was comfortable at a low reference rate, the same leverage at a much higher one may not be serviceable at all. The loan did not become riskier because the business deteriorated. It became riskier because the cost of the loan rose.

That is the mechanism that defines 2022 for private credit, and it operates with a lag. Interest expense rises immediately, but distress takes time to manifest — cash reserves absorb the increase, sponsors inject equity, covenants are amended, and PIK arrangements defer the cash cost. The gap between when the cost arrives and when the stress becomes visible is where the asset class has been operating since.

Against this, 2022 also produced the asset class's clearest genuine win. When syndicated markets closed, private credit continued to lend, taking share in large transactions it had not previously competed for — and on terms considerably more favourable to lenders than the preceding vintage.

Why floating rate cuts both ways

The mechanism is simple and its consequences are consistently under-modelled.

A floating rate loan pays a reference rate plus a fixed spread, resetting periodically. When the reference rate rises, the coupon rises automatically.

For the lender, this is protection: income rises with rates, and the loan's value does not fall the way a fixed-rate bond's does, because the coupon adjusts.

For the borrower, it is the opposite: interest expense rises automatically, without any new borrowing or any change in the business.

The critical point is that these are the same cash flow. The lender's gain is the borrower's cost. A rate rise does not create value in the loan; it transfers cash from borrower to lender. If the borrower can afford it, the lender's position improves. If the borrower cannot, the lender's position deteriorates — because a borrower who cannot service the loan is a credit problem, and the higher coupon is worth nothing if it is not paid.

The relevant measure is interest coverage — earnings divided by interest expense — and it deteriorates arithmetically with the reference rate:

A borrower with $100m earnings and $500m of floating debt at a 5% all-in rate pays $25m of

interest. Coverage is 4.0×. Move the all-in rate to 10% and interest is $50m. Coverage is

2.0×. The business is identical. The leverage multiple is identical. The company is now

half as able to service its debt.

*Derived: illustrative arithmetic with round numbers, chosen to expose the mechanism.

Not a description of any actual borrower.*

This is why leverage multiples became a misleading measure in 2022. Debt-to-earnings did not change. A portfolio reported at the same leverage as the year before was substantially more stressed, and the leverage statistic did not show it.

The correct measure was coverage, and coverage deteriorated across the market without a single borrower taking on more debt.

Why the stress was slow to appear

Interest costs rose in 2022. Visible distress did not rise proportionally, for reasons that are worth enumerating because each defers recognition rather than preventing loss.

• Cash balances absorbed the increase first. Many borrowers had raised capital in the abundant 2021 market and held cash, which covered the higher expense for several quarters.
• Sponsors injected equity. A private equity owner facing a portfolio company with coverage problems can contribute capital. This is common and preserves the credit — at the cost of the sponsor's return, which is why it has limits.
• Amendments and waivers. Lenders and borrowers renegotiated rather than defaulting. An amended loan is not a defaulted loan in the statistics.
• PIK conversions. As the 2020 report describes, converting cash interest to PIK removes the immediate cash requirement entirely while increasing the eventual obligation.
• Weak documentation delayed the trigger. Following the covenant-lite shift described in the 2017 private equity report, many loans had no maintenance covenants. A deteriorating borrower does not breach a test that does not exist.

Each of these defers the moment of recognition. None of them changes the underlying economics. A borrower whose coverage has halved is in a worse position whether or not that position has been recognised in a statistic.

The measurement implication is the important one: default rates in 2022 and 2023 substantially understated the deterioration in credit quality, because the mechanisms above were absorbing it. Anyone assessing the asset class from default rates alone was reading a lagging and heavily managed indicator.

The better indicators, all publicly available from BDC filings:

• PIK income share — rising PIK means cash stress is being deferred.
• Non-accrual rates — a lender's own judgement that a loan will not be repaid, which precedes formal default.
• Amendment activity — disclosed in filings, and a direct measure of how many borrowers needed relief.
• Weighted average interest coverage of the portfolio — increasingly disclosed, and the most direct measure of the mechanism this report describes.

The genuine win: taking share when markets closed

2022 produced the strongest evidence yet for one of the asset class's claims, and it deserves equal weight with the criticism above.

When public credit markets deteriorated in 2022, syndicated loan issuance became difficult. A bank underwriting a large loan for later distribution faced the risk that the market would not absorb it — leaving the loan on the bank's balance sheet at a loss. Banks accordingly became unwilling to underwrite.

Private credit lenders faced no such risk. A direct lender holds the loan by design, so market conditions at the moment of syndication are irrelevant to them.

The result was that private credit financed transactions that would previously have gone to syndicated markets, including several very large ones that exceeded what the asset class had previously handled. This was a genuine structural advantage, demonstrated rather than claimed:

• Certainty of execution matters enormously to a borrower in an acquisition process, and a direct lender can provide it when banks cannot.
• Speed and confidentiality are real advantages in competitive processes.
• Scale had grown enough that the asset class could write cheques previously only available from syndicated markets.

The terms achieved reflected the negotiating position. Loans originated in 2022 and 2023 carry wider spreads, lower leverage, and stronger documentation than the 2021 vintage — because lenders had pricing power for the first time in years.

This is the most important qualification to the report's overall assessment. The asset class's problems are concentrated in loans originated in 2021 and earlier, at tight spreads and high leverage. The loans originated since are among the best-underwritten in the asset class's history. Vintage matters enormously here, and aggregate statistics blend the two.

Why hedging complicates the timeline

The coverage arithmetic above assumes the borrower's interest cost rose with the reference rate. Many borrowers held interest rate hedges, which changes the timing without changing the destination — and the timing is what made 2022's stress so hard to read.

What a hedge does. An interest rate cap or swap fixes or limits the borrower's effective rate for a defined period. A borrower who hedged in 2021, when hedges were cheap because rates were low and expected to stay low, was protected through 2022 and into 2023.

Why this matters for interpreting the data:

• Reported coverage held up better than the reference rate implied, because a portion of borrowers were paying a hedged rate rather than the market rate.
• The protection had an expiry. Hedges are purchased for a term — commonly two to three years. A hedge bought in 2021 expires in 2023 or 2024.
• Replacement is far more expensive. A hedge purchased when rates are already high costs substantially more than one purchased when they were low. The borrower either pays the higher premium, accepts the unhedged rate, or reduces the hedged proportion.
• So the coverage deterioration arrives on a schedule set by hedge expiry, not by the rate move itself.

This produces a specific pattern that was widely misread. Coverage statistics through 2022 and into 2023 looked more resilient than the reference rate move implied, which was read as evidence that borrowers were coping. They were coping because they had bought time, and the time was purchased for a fixed term.

The analytical consequence is that hedge expiry is a scheduled event. Like the depreciation schedule the 2026 global report describes for a different asset class, it is a cost that arrives on a calendar rather than contingently. Where disclosed — and BDC filings increasingly discuss portfolio hedging — it is one of the more predictable elements in an otherwise uncertain picture.

The rate rose in 2022. For hedged borrowers the cost arrived in 2024. A statistic that looked reassuring in between was measuring the hedge, not the credit.

What an allocator could act on

Use coverage, not leverage. Debt-to-earnings did not change in 2022 while the ability to service that debt halved for unhedged floating-rate borrowers. A portfolio reported at unchanged leverage was substantially more stressed and the leverage statistic did not show it. Weighted average interest coverage is increasingly disclosed in BDC filings and is the correct measure.

Track PIK income share quarterly. It leads default statistics, it is an accounting fact rather than a designation decision, and it directly measures stress being deferred rather than recognised. Distinguish contractual PIK, which was priced at origination, from amended PIK, which is a distress signal.

Read amendment activity as a stress indicator. Lenders amend rather than default when they expect recovery. Repeated amendment on the same credit is informative, and amendments are disclosed.

Analyse by origination vintage, not in aggregate. Loans from 2021 and earlier carry tight spreads and high leverage set in a low-rate environment; loans from 2022–2023 carry the best terms in the asset class's history. Aggregate portfolio statistics blend them and describe neither. Asking a manager for the split by origination year is reasonable, and reluctance to provide it is itself informative.

Give the execution-certainty advantage its due weight. When syndicated markets closed, private credit continued lending and took share in transactions it had not previously competed for. That is a genuine structural advantage, demonstrated rather than claimed, and it is the strongest evidence the asset class has produced for any of its distinguishing propositions.

Note that the deferral mechanisms are legitimate credit management. Cash buffers, sponsor support, amendments and PIK are what a lender should do for a borrower facing a temporary problem. The criticism is of statistics that do not capture them, not of the practices — and the correction is to read the disclosure rather than the headline default rate.

What 2022 established for private credit

• Floating rate protection was shown to be two-sided, transferring cash from borrower to lender rather than creating value.
• Interest coverage replaced leverage as the binding constraint, and leverage statistics became actively misleading.
• Stress recognition was shown to be deferrable through five distinct mechanisms, none of which changes the economics.
• Private credit's execution certainty was demonstrated genuinely for the first time, taking share from syndicated markets.
• The vintage split emerged — 2021-and-earlier loans carry the risk; 2022-and-later loans carry the best terms in the asset class's history.