From View From The Top, Issue #32
The Repricing Nobody Wants to Mark
Private Credit’s Quiet Problem — And Why Residential Real Estate Credit Is Different. Mostly.
Jamie Dimon reached for an old analogy recently. Cockroaches. When you see one, there are more.
He was talking about private credit — after Tricolor and First Brands collapsed with marks that hadn’t moved until the moment they did. Jeffrey Gundlach went further, calling private credit the leading candidate to start the next financial crisis. The SEC has been hosting roundtables on valuation governance. Redemption requests from private credit vehicles ran roughly five times their normal rate in the first quarter.
The asset class that spent a decade as institutional capital’s favourite child is having its first genuinely uncomfortable year. And the discomfort centres on one question.
Are the marks real?
I run a private credit platform in London. So let me try to answer that question honestly — including the parts of the answer my own sector would prefer to skip over or sugar coat.
WHAT’S ACTUALLY BEING ALLEGED
Strip away the headlines for a moment. Private loans rarely trade. There is no market price. The value of the loan — the mark — is produced by a model, informed by assumptions the manager controls or influences: discount rates, borrower performance, recovery expectations.
In good times, nobody pays much attention to this. The loan pays its coupon, the mark stays near par, everyone moves on.
The problem emerges under stress. A borrower deteriorates. In a liquid public market, the price falls immediately and visibly. In private credit, the mark is an opinion — and opinions under pressure tend to stay optimistic no matter the asset class. If I look at historic stats the headline default rate in private credit has held below 2% for years, but analysts who count selective defaults and liability management exercises put the true figure closer to 5% today. The gap between those numbers is the gap between what’s happening and what’s being marked.
The critics’ charge, in one sentence: private credit’s smoothness is not stability. It’s latency. The losses exist — they just haven’t been recognised- let’s use acknowledged -yet.
Some of that criticism is fair. Which is exactly why it’s landing.
DOES THIS APPLY TO RESIDENTIAL REAL ESTATE CREDIT?
Thing is I’d say, mostly, no. And the reasons are structural, not cultural.
They don’t depend on residential lenders being better people. They depend on the asset class making sustained self-deception harder.
1. The collateral is physical.
A development loan is secured against land and buildings. Not a projection of future earnings — a thing that exists, at an address, that a valuer can visit. The valuation is produced by an independent RICS valuer under Red Book standards, with professional liability attached, not by the lender’s internal model. And as we covered in Issue 29, the current audit environment — sharpened by the KPMG litigation — is pushing those valuations toward conservatism, not away from it.
2. The duration is short.
A development loan runs 18 to 24 months. It is self-liquidating: units get built, units get sold, the loan repays or it doesn’t. The truth arrives on a schedule. There is no five-year fog for a stale mark to hide in. Every loan in the book resolves itself, in cash, within a timeframe an investor can actually observe.
3. The exits are public.
House prices are published. Land Registry data is public. Sales comparables are checkable by anyone. The gap between a lender’s carrying value and reality can be independently tested in a way that enterprise value never can. Transparency isn’t a policy choice in residential credit. It’s ambient.
Three structural features. Three reasons the mark-to-model critique loses most of its force when the collateral is a building rather than a belief.
WHERE THIS ISN’T DIFFERENT
But here is the part my sector talks about less.
Residential credit has its own version of the disease. It doesn’t live in the marks as much but It lives in the maturities.
And it’s a common practice – it’s called extend-and-pretend.
A development loan reaches term. The scheme is late, or the units aren’t selling at the underwritten prices, or the developer’s equity is exhausted. The honest response is to recognise the position: default the loan, revalue the security at today’s number, take the loss if there is one.
The tempting response is quieter. Extend the term. Capitalise the interest. Keep the original valuation on file — it’s only eighteen months old. The loan stays “performing.” The book stays clean. The problem compounds at 10% a year while everyone waits for the market to rescue the position.
Anyone operating in UK development finance right now knows lenders doing exactly this. Extended loans carried against valuations that predate the correction. LTVs that have quietly drifted through covenant levels — if anyone re-ran the numbers, which nobody is in a hurry to do. Interest rolling up on schemes that will not, at today’s prices, repay it.
The short duration of residential credit only tells the truth if the lender lets the loan mature. Extension is how you manufacture long duration out of short loans — and reconstruct, by choice, exactly the fog that corporate direct lending has by design.
So the honest answer to the question is this:
The asset class is structurally more honest. But only if the platform chooses to be.
WHAT DISCIPLINED MARKING ACTUALLY LOOKS LIKE
The difference between an honest book and a hopeful one shows up in specific, checkable practices and being clear. If you’re an investor evaluating a residential credit manager — ours included — these are the questions worth asking.
How old are the valuations?
Every loan in the book should be carried against a valuation that post-dates the market correction. A 2022 valuation on a 2026 loan book is not a data point. It’s a decision.
What happens at maturity?
Ask for the extension rate. A book where a third of loans have been extended past original term is telling you something, whatever the marks say. Ask how extensions are approved, whether they trigger revaluation, and whether extended loans are flagged separately in reporting.
Is interest being paid or accrued?
Cash interest received is a fact. Rolled-up interest is a forecast. A book that relies on the second has changed its risk profile without changing a single mark.
Who sees the loan tape?
Platforms confident in their book let back leverage providers, auditors, and investors look through to loan-level detail. Opacity is a choice, and it’s rarely chosen by people with nothing to smooth.
Were losses ever taken?
This is the strangest signal, but the truest: a lender that has operated through 2022 to 2026 with zero recognised losses is either the best underwriter in Britain or hasn’t looked. There were losses in this market. A lot of them. Everyone knows it. The question is only whose books they’re on.
None of this is exotic. It’s just discipline, applied where discipline is optional.
HARD LANDING
The private credit valuation debate will run through 2026, and some of the cockroaches Dimon is warning about will duly appear. Most will be found in corporate direct lending, where the structure makes hiding easy and the vintage years of cheap money made hiding necessary.
Residential real estate credit will mostly be fine — and will be tempted to say so loudly.
But we should resist the victory lap.
Because the structural honesty of this asset class — the physical collateral, the short duration, the public exits — is an inheritance, and inheritances get spent. Every extension granted to avoid recognising a loss spends a little of it. Every stale valuation left unchallenged spends a little more. Until it’s a conversation no one wants to have.
The platforms that come through this cycle with their credibility compounding will be the ones that marked the repricing nobody wants to mark. That defaulted loans when default was the truth. That took the small loss early instead of the large one late.
Discipline isn’t inherited from the collateral. It’s practiced. Painful as it is sometimes. Or it isn’t.
