Who funds a private credit fund?
Banks do, in large part. A direct lending fund raises equity from its investors and then borrows against it, and the counterparty on the other side of that borrowing is usually a commercial bank. The regulator counts this as lending to non-depository financial institutions.
It is no longer a small line. Loans to non-depository financial institutions at domestically chartered commercial banks stood at 1,480.6 billion dollars in the week ended 22 July 2026, against 1,217.3 billion a year earlier, a rise of 21.6 percent (Federal Reserve H.8, not seasonally adjusted). The same series was 219.4 billion dollars when it began in January 2015.
So the pipe behind the non-bank lending market is now roughly half the size of the entire commercial and industrial loan book, and it has been the fastest-growing thing on bank balance sheets for several years. That makes the terms on which banks extend it a live question rather than a technical one.
It is also the part of the chain a borrower never sees. You negotiate with the fund. The fund negotiates with its banks. The second negotiation sets much of what the fund can offer you in the first.
What did the July survey actually say?
That standards on every queried category of loan to non-bank lenders are at the tighter ends of their historical ranges. The Federal Reserve asked banks to place their current standards against the full range those standards have spanned since 2011, and the answer came back tight across the board.
The survey is the July 2026 Senior Loan Officer Opinion Survey on Bank Lending Practices, published 3 August 2026. It covers the second quarter of 2026 and collects responses from 56 domestic banks and 18 United States branches and agencies of foreign banks, sent out on 17 June 2026 with responses due by 2 July 2026. The special questions on non-depository financial institutions are a second set, separate from the standards and demand questions that produce the survey's headline numbers.
The category that matters most to a middle-market company is business credit intermediaries, which is where the funds that lend to operating businesses sit. Half of the banks answering that question, 50.0 percent of 48 respondents, placed their current standards at or beyond somewhat tighter than the midpoint. Only 10.4 percent placed them anywhere on the easier side (Federal Reserve, July 2026 Senior Loan Officer Opinion Survey, table 1).
The timing question is the sharper one. Asked when their standards on loans to non-bank lenders had reached their tightest level since 2011, 53.3 percent of 45 respondents answered 2023 to the present, against 22.2 percent for 2020 to 2022 and 6.7 percent for 2016 to 2019. A majority of banks are saying the tightest they have ever been on this book is now.
| Type of non-bank lender | Easier than midpoint | Near midpoint | Tighter than midpoint | Banks answering |
|---|---|---|---|---|
| Consumer credit intermediaries | 6.5% | 39.1% | 54.3% | 46 |
| Business credit intermediaries | 10.4% | 39.6% | 50.0% | 48 |
| Private equity funds | 13.1% | 43.5% | 43.5% | 46 |
| Other non-bank lenders | 6.7% | 51.1% | 42.2% | 45 |
| Mortgage credit intermediaries | 9.1% | 54.5% | 36.4% | 44 |
“Borrowers read the credit market through the price they are quoted, which is the last thing in the chain to move. The funding line behind the quote moves first, and it is the better early indicator. When the banks behind the funds get careful, the funds get careful about two quarters later.”
Does this contradict the idea that banks are competing for my loan?
No, and holding both facts at once is the whole point. Banks are lending directly to middle-market companies on terms that have been getting easier, and tightening on the intermediaries that compete with them for the same borrowers. Those are different books with different regulatory treatment.
On the direct book the Federal Reserve's own summary language for the second quarter of 2026 is that banks left standards basically unchanged for commercial and industrial loans to firms of all sizes, and eased or left basically unchanged all queried terms. Commercial and industrial lending is the only category in the survey where standards sit easier than their midpoint since 2005 (Federal Reserve, July 2026 Senior Loan Officer Opinion Survey).
Put the two side by side and the shape is clear enough. The cheapest capital in the market is the one a borrower is most likely to skip, because the received view is that banks have retreated. On the evidence of this survey they have retreated from funding the competition, not from funding companies.
That is a favourable asymmetry for anyone prepared to run both processes. It is an expensive one to miss if a company goes straight to a fund because a banker told it three years ago that its business was too small.
Why does a fund's funding cost become the borrower's problem?
Because a levered lender passes its own terms through. A direct lending fund that borrows against its loan book at a higher advance rate can quote a borrower more aggressively than one that cannot. When the bank behind it reduces that advance rate or reprices the line, the fund's economics change before any borrower is told.
The effect shows up in three places. In price, because a fund with a costlier funding line needs a wider spread to hit the same return. In capacity, because a fund that cannot lever the position as far can write a smaller cheque against the same business. And in behaviour during a difficult quarter, because a lender managing pressure on its own facility is a less patient counterparty than one that is not.
None of that is visible in a term sheet. A term sheet shows the spread, the leverage and the covenant package, all of which are outputs. The funding structure behind them is an input, and it is knowable if a borrower asks.
This is also why the distinction between committed fund capital and redeemable vehicle capital keeps mattering. A fund drawing on a bank line that is being repriced, and facing redemption requests at the same time, has two calls on its liquidity before it gets to yours.
What should a borrower ask before signing?
Four questions, none of which are unusual and all of which a serious lender will answer. First, what proportion of this fund's lending capacity is levered, and who provides the leverage. Second, when does that facility mature or reprice, and against what advance rate.
Third, how much of the fund's existing book is already drawn, because a lender near its capacity is a lender that will struggle to fund an add-on acquisition eighteen months from now. Fourth, whether the capital behind the fund is committed for a term or can be redeemed by its investors, which determines whether the fund is managing its own outflows while managing your credit.
Then run a bank process in parallel, on the evidence above rather than on reputation. The comparison is not only about headline pricing. It is about which lender is structurally able to hold the position through a soft year, and a bank funding a loan from deposits and a fund funding one from a repriced facility are not in the same position on that question.
The survey does not say the non-bank market is in trouble. Volumes are still growing quickly, and the tightening is in the terms rather than in the taps being turned off. It says the terms behind that market are as tight as they have been in fifteen years, which is a reason to ask what your lender is paying for its own money before you agree what you will pay for yours.
As of August 2026
Sources: Federal Reserve, July 2026 Senior Loan Officer Opinion Survey on Bank Lending Practices, published 3 August 2026, for the survey period covering the second quarter of 2026, for the respondent count of 56 domestic banks and 18 United States branches and agencies of foreign banks, for the survey being sent on 17 June 2026 with responses due 2 July 2026, for the finding that standards on all queried types of loans to non-depository financial institutions are at the tighter ends of their ranges since 2011, for the statement that banks left commercial and industrial standards basically unchanged for firms of all sizes and eased or left basically unchanged all queried terms, and for commercial and industrial lending being the only category where standards sit easier than their midpoint since 2005; Federal Reserve, July 2026 Senior Loan Officer Opinion Survey, table 1, special questions on loans to non-depository financial institutions, for every figure in the table and for the category shares quoted in the text, being 50.0 percent of 48 respondents placing business credit intermediaries at or beyond somewhat tighter than the midpoint against 10.4 percent on the easier side, and for 53.3 percent of 45 respondents naming 2023 to the present as the period in which their standards on loans to non-bank lenders reached their tightest level since 2011, against 22.2 percent for 2020 to 2022 and 6.7 percent for 2016 to 2019; Federal Reserve H.8, Assets and Liabilities of Commercial Banks in the United States, series LNFDCBW027NBOG, not seasonally adjusted, for loans to non-depository financial institutions at domestically chartered commercial banks of 1,480.6 billion dollars in the week ended 22 July 2026, 1,217.3 billion dollars in the week ended 23 July 2025, the resulting rise of 21.6 percent, and 219.4 billion dollars at the start of the series in January 2015.