Investors asked private credit funds for $15.6 billion back last quarter. Managers returned under 40% of it.

Why would a borrower diligence a lender?

Because the diligence runs one way and the exposure runs both. The fund spends eight weeks on your quality of earnings, your customer contracts and your management team. You spend eight weeks negotiating a coupon. Then you are counterparties for five to seven years, through at least one period when something goes wrong, and by then the only thing that matters is what this particular lender does when a credit is under water.

That question has a range of answers, not one answer. A private credit facility is not a standardized product with a regulated counterparty behind it. It is a loan from a specific fund, with a specific vintage, a specific remaining life, specific liquidity terms owed to its own investors, and a specific track record in workouts that the borrower can partly establish and rarely tries to.

It matters more now than it did in the previous cycle because deterioration has stopped being exceptional. Lender takeovers of borrowers ran $24.2bn in 2025 and $15.2bn in the year to date, against $13.6bn across the three preceding years combined, with close to three quarters of the activity tracing to 2021 and 2022 vintage loans (Lincoln International, as of 31 March 2026). Whatever your lender does in a workout, it is doing it more often than it used to.

The good news is that most of what you need is establishable. Fund life, investment period, strategy, leverage and liquidity terms are all documented. Workout behaviour is knowable through references. Neither is confidential, and a lender that treats a borrower's questions about them as impertinent has told you something useful about the relationship on offer.

What does the fund's own funding position do to your facility?

It decides whether your lender can fund the next thing you need, which for a growing business is usually the acquisition line, the capital expenditure line or the accordion, not the initial term loan. A fund late in its investment period, or one managing redemptions, is a different counterparty from the one that underwrote you, even though the name on the loan agreement has not changed.

The liquidity data from this year is unambiguous. Investors sought $15.6bn of withdrawals from private credit vehicles in the second quarter of 2026 and managers returned $5.9bn, under 40% of what was requested (Wall Street Journal via SPP Capital Partners, July 2026). More than $14.5bn sits behind gates across roughly twenty funds, at a median redemption request rate of 8.7% (Financial Times via SPP, July 2026). One large perpetual vehicle received $4.7bn of redemption requests in the quarter and paid roughly $2.1bn of them under a 5% tender cap.

The knock-on for borrowers is already visible. Some business development companies are retaining capital to support existing stressed borrowers rather than funding new transactions (Private Equity Wire, 10 July 2026). Deployment fell accordingly: direct lending volume was $33.6bn in the second quarter, down 55% quarter over quarter across 154 deals, the weakest since the second quarter of 2023, even as fundraising rose to $16.25bn from $1.3bn in the first (Preqin and PitchBook LCD via Reuters, 10 July 2026).

Read those two facts together, because they are the whole point. Capital was raised and capital was not deployed. A lender can be well funded at the platform level and unable to support your incremental need, and the borrower who discovers that at the point of an add-on acquisition has discovered it too late to do anything but pay for it.

What does this lender do when a credit deteriorates?

The honest answer varies enormously by manager, and the industry-level statistics are close to useless for predicting an individual one. What the statistics do establish is that the question is live, that the tools being used have shifted, and that the same borrower situation now produces materially different outcomes at different funds.

Payment-in-kind is the first tool and the most informative. Payment-in-kind interest was present on 10.6% of loans and represented 8.9% of total interest income in the first quarter of 2026, the highest since the fourth quarter of 2020. What matters is the split inside it: so-called bad payment-in-kind, meaning interest deferred because the borrower cannot pay rather than because the structure planned for it, was 55.7% of loans carrying any payment-in-kind, or 5.9% of all loans, which the same analysis describes as a shadow default rate (Lincoln International, published 7 May 2026). Loan-to-value on those loans rose 33.5 points to 76.0% over the year to the first quarter of 2026.

The second tool is definitional. The same market currently prints default rates of roughly 1.5%, 2.51%, 3.1% and 6.0% depending on whether the measure is missed payments, documentation defaults, covenant defaults or a definition that counts maturity extensions, and the global regulator reaches the same conclusion independently, putting outright defaults near 1% rising to around 5% once selective defaults are counted (Financial Stability Board, Report on Vulnerabilities in Private Credit, 6 May 2026). When a lender quotes you its default rate, the first question is which of those it is quoting.

The third is enforcement, and here the size cut matters. Restructurings out of court remain concentrated at the larger end, while lower-middle-market lenders typically hold unassailable senior positions, clearer enforcement rights and more direct engagement with borrowers (First Eagle Investments, March 2026). A smaller borrower is less likely to be in a creditor group and more likely to be in a room with one lender, which raises the stakes on which lender it is.

Which questions actually get answered?

Five, reliably, and they are worth asking in writing before the term sheet is countersigned rather than during confirmatory diligence when the borrower has lost the ability to walk.

What is the vintage and remaining investment period of the fund lending to me, and what happens to my facility when it ends? A loan held by a fund entering its harvest period is a loan whose holder is now managing toward an exit rather than toward a relationship, and incremental needs get priced accordingly.

What is your hold size, and will you syndicate any of this? A single-lender facility is easier to amend than a club, and a borrower who learns after closing that two thirds of the paper moved to funds it has never met has lost the amendment mechanic it thought it had bought.

What are your own investors' redemption terms? Perpetual vehicles with quarterly tenders, closed-end funds with fixed lives and separately managed accounts behave differently under stress, and the answer is documented rather than confidential.

How many of your borrowers are on payment-in-kind, and how many have you taken control of? The market-level numbers above give you a benchmark to hold the answer against, which is the only reason to know them.

And last: give me two borrowers you have amended terms for, and one you have taken through a difficult period. A lender who cannot produce the third reference has either never had one or does not want you to talk to it, and both readings are informative.

What should a borrower do with the answers?

Price them, rather than score them. Lender behaviour in a downside is a term like any other, and a borrower who has established that one lender is materially more patient than another has established something worth paying a spread for, in the same way that covenant headroom is worth paying a spread for.

The comparison set makes that affordable, because the range is wide. A borrower below $10m of EBITDA faces bank senior at S+350 to 425 and non-bank senior or unitranche at S+550 to 750, while the same structure above $25m of EBITDA prices at S+275 to 350 and S+425 to 575 respectively (SPP Capital Partners, Market At A Glance, July 2026). Inside those bands there is room to choose a counterparty rather than a coupon.

There is one structural fact to weigh against all of it. Valuation dispersion is widest at the small end: 12% of borrowers in the $10m to $20m EBITDA bracket were carried below 90% of par, against 3% of borrowers above $100m of EBITDA (Alternative Credit Council and AIMA Quarterly Update, drawing on more than 70,000 loan valuations, as of March 2026). Smaller borrowers are marked more harshly and more variably, which means the identity of the institution doing the marking matters more, not less.

The practical test we apply is simple. If this facility needed to be amended eighteen months from now, on a set of facts we cannot currently predict, would we rather be sitting across from this lender or from the one 50 basis points behind it. That question has a real answer, and it is available before signing at no cost other than asking.