The interest is being paid in paper, not money

That private credit is simply banking that moved, with better underwriting, tighter covenants and lenders who know their borrowers. Yields look high, defaults look low, and the asset class has grown through a rate cycle without visible damage.

The yields are not in dispute. What the low default rate is actually measuring is.

Two things that compound each other. The first is payment in kind: interest a borrower does not pay in cash, added instead to the principal. It is recognised as income the moment it accrues. A borrower too stretched to pay therefore produces a rising reported yield rather than a missed payment, and the loan keeps performing in every sense that gets reported.

The second is how the loans are valued. There is no screen for most of this paper. Marks come from models maintained by the manager whose fee depends on them, reviewed by valuation agents that manager engages. That is not an accusation of bad faith. It is the observation that price discovery only happens on exit, and exits have been scarce.

Together these produce a book that looks like it is performing precisely because nothing has been tested. Rising payment in kind share, lengthening hold periods and a slowing pace of realisation are one signal read together, not three read separately.

Not as a default wave. As a realisation problem. Funds reach the end of their lives and have to sell, and the gap between the carrying mark and the clearing price stops being an opinion and becomes a number.

The second order effect is larger than the first. Many of these vehicles are sold to investors expecting periodic liquidity. A redemption window opening against assets with no market is how a valuation question turns into a forced seller.

Start with the business development companies, because they are the only part of this market that files. Ares Capital, FS KKR, Blue Owl Capital Corporation, Blackstone Secured Lending, Golub and Prospect all publish an investment income note, a non accrual list and a net asset value per share every quarter. Read across them and the dispersion is the finding: the same asset class, the same vintage, and payment in kind running at very different shares of total investment income from one book to the next.

Three numbers, taken together rather than separately. Payment in kind as a share of total investment income. Non accruals at fair value rather than at cost, because cost flatters the ratio once a loan has already been marked down. And net asset value per share across several quarters, which is the only running record of whether the marks have ever moved.

Then look at where the paper is funded. The perpetual non traded vehicles are the ones to watch, because they sell monthly subscriptions and quarterly liquidity against assets that trade neither monthly nor quarterly. The repurchase caps are published in the offering documents and they are small, usually a few percent of net asset value a quarter. A cap is not a problem until it is hit, and the quarter it is hit is public information.

And look at the insurance channel. Several of the largest managers now own or control an annuity writer, which turns a fee business into a balance sheet that holds the paper to maturity. That is a genuine structural advantage against a redemption cycle, and it is also the place where a mark that is wrong stays on somebody's books the longest.

The cleanest expression is not a directional bet on the asset class. It is a bet on which part of the structure gets paid.

Own the fee taker, not the balance sheet. The managers collect on gross assets and on the income recognised, including the income that arrived as paper rather than cash. The vehicles carry the credit. Those two things have been correlated on the way up and have no reason to stay correlated through a realisation cycle. Long the manager against the lending vehicle is the pair, and it does not require the loans to go bad, only for the marks to be tested.

Within the lenders, the dispersion is the trade. Senior secured books with low payment in kind and a long record of realisations at or near carry are a different instrument from books where the income is increasingly non cash and the holds keep lengthening. The market has mostly priced them as one thing, through the lens of dividend yield, which is exactly the metric payment in kind inflates.

Discount to net asset value is the market making its own mark. When a lending vehicle trades at a persistent discount while reporting a stable net asset value, one of those two numbers is wrong, and the one set by a screen is usually the one worth believing. A wide discount on a book that genuinely is senior and cash paying is the long. A tight price on a book carrying rising payment in kind is the other side.

The timing marker is not a default print. It is the first quarter a large perpetual vehicle meets its repurchase cap, or the first large secondary sale of a portfolio at a disclosed price. Either one converts an opinion into a comparable, and comparables are what force everybody else to remark.