Finantrix
Subscribe
Deep DiveBanking & Fintech

Who Is Really Funding Private Credit?

Private credit was meant to take lending risk off the banks. Banks now lend $1.7 trillion to nonbank lenders and have promised $1.1 trillion more.

Finantrix Research 18 min readNovember 1, 2026

Key Takeaways

  • Insured banks lent $1.70 trillion to nonbank lenders and funds on June 30, 2026, 12.2% of all their loans, against 0.8% in March 2010.
  • $856 billion of it went to private credit funds, BDCs and private equity funds, and banks had promised those funds $489 billion more in undrawn lines.
  • At the same 517 banks, these loans grew 47% in eighteen months, against 10.6% for commercial and industrial loans.
  • Ten banks hold 73% of the lending; at Morgan Stanley Bank, State Street Bank and Goldman Sachs Bank USA it is more than half of all their loans.
  • BDC credit looks calm on average, but non-accruals at Golub Capital BDC rose from 0.6% to 2.9% of cost in nine months.
  • The 2025 failures of Tricolor and First Brands reached regional banks, a BDC, an investment bank's fund and a parts distributor.

On October 8, 2025, Jefferies Financial Group filed a short report with the SEC about a company most of its shareholders had never heard of. First Brands, an auto-parts maker, had filed for bankruptcy on September 29. Point Bonita Capital, a division of Leucadia Asset Management, part of Jefferies, managed money for outside investors and had bought the sums First Brands' customers owed it. Jefferies' own stake, it said six days later, was $43 million. The filing put a number on it: "the Point Bonita portfolio has approximately $715 million invested in receivables that are almost entirely due from Walmart, Autozone, NAPA, O'Reilly Auto Parts, and Advanced Auto Parts, with First Brands, as the servicer, responsible for directing the Obligors' payments to Point Bonita."

That sentence is the private credit market in miniature. A lender that is not a bank finances a company. The risk looks as if it sits with the fund and its investors. Then the borrower fails, and the question that matters is who stood behind the fund.

This deep dive asks that question across the whole market: who is really funding private credit? The popular answer goes like this. Since 2008, rules have pushed banks away from risky corporate lending. Pension funds, insurers and wealthy investors have filled the gap through private credit funds. Depositors are kept out of it. If a private loan goes bad, the loss belongs to people who chose to take it.

We tested that story against three kinds of record: the Call Reports every insured bank files with its regulators each quarter, the SEC filings of the 172 business development companies (BDCs) that lend to mid-sized American companies, and the filings in which banks and funds disclosed what the 2025 failures cost them. The answer is not that the story is wrong. It is that it stops one step too early. The risk left the banks' loan books as loans to companies. A large part of it came back as loans to the lenders.

The size of it

Start with the banks, because their numbers are the most complete. Every insured bank reports, on Schedule RC-C of the Call Report, its loans to "nondepository financial institutions": lenders, funds and finance companies that do not take deposits. On June 30, 2026, the total was $1.70 trillion. That is 12.2% of every loan made by every insured bank in the country.

Exhibit 1Banks and the lenders they lend to, June 30, 2026
$1.70Tloans from insured banks to nonbank lenders and funds
$1.09Tmore promised to them in credit lines not yet drawn
12.2%of all loans at insured banks

Source: FFIEC Call Reports, Schedule RC-C item 9.a and Schedule RC-L, all insured banks; Finantrix calculation. Data as of 2026-06-30.

Show the data
FigureWhat it counts
$1.70Tloans from insured banks to nonbank lenders and funds
$1.09Tmore promised to them in credit lines not yet drawn
12.2%of all loans at insured banks

The regulators themselves have described how fast this grew. When they proposed more detailed reporting in December 2023, they wrote that in March 2010 these loans "totaled approximately $56 billion and represented only 0.8 percent of gross loans," and that by June 2023 they had "increased significantly to almost $786 billion and represented 6.4 percent." In the three years since, the share has nearly doubled again.

Exhibit 2Loans to nonbank lenders as a share of all bank loansYear-ends 2010 to 2023, then quarters. The reporting rules changed in December 2024, so the two lines are not joined.
Before the 2024 revisionAfter the 2024 revision0.0%5.0%10.0%15.0%2010201420182022Sep 2024Jun 2025Jun 2026Before the 2024 revision, 2010: 0.9%Before the 2024 revision, 2012: 1.5%Before the 2024 revision, 2014: 2.7%Before the 2024 revision, 2016: 3.4%Before the 2024 revision, 2018: 4.1%Before the 2024 revision, 2020: 5.3%Before the 2024 revision, 2022: 6.1%Before the 2024 revision, 2023: 6.5%Before the 2024 revision, Sep 2024: 7.1%After the 2024 revision, Dec 2024: 9.1%After the 2024 revision, Jun 2025: 10.6%After the 2024 revision, Dec 2025: 11.6%After the 2024 revision, Jun 2026: 12.2%After the 2024 revision

Source: FFIEC Call Reports, all insured banks, item 9.a over total loans and leases; Finantrix calculation. Data as of 2026-06-30. Between September and December 2024 the total rose from $900 billion to $1,161 billion, the quarter banks first reported under the revised instructions and the same quarter their commercial and industrial loans fell by $131 billion. That jump is reclassification, and no growth figure in this report spans it.

Show the data
Before the 2024 revisionAfter the 2024 revision
20100.9%–
20121.5%–
20142.7%–
20163.4%–
20184.1%–
20205.3%–
20226.1%–
20236.5%–
Sep 20247.1%–
Dec 2024–9.1%
Jun 2025–10.6%
Dec 2025–11.6%
Jun 2026–12.2%

Now the other side. The part of private credit that files with the SEC is the business development company, a fund that lends to mid-sized businesses and must publish its portfolio. On June 30, 2026, the 162 BDCs that report their portfolio in structured data held $535 billion of investments at fair value. Measured on the same 129 BDCs, the portfolio grew 81% in the two and a half years from December 2023. And on average, a BDC now carries about a dollar of liabilities, mostly borrowing, for every dollar of its own net assets.

Exhibit 3The part of private credit that files with the SEC
$535Bportfolio of 162 business development companies, at fair value
+81%growth in the same 129 BDCs' portfolios since December 2023
$1.02of liabilities for every $1 of net assets, all BDCs combined

Source: SEC XBRL company facts for 172 BDCs (file numbers 814-), 10-Q and 10-K filings; Finantrix calculation. Data as of 2026-06-30. BDCs began filing structured financial data around 2022, so growth is measured only on the same BDCs at both dates. Most of the market, private funds that do not file with the SEC, is not in these numbers.

Show the data
FigureWhat it counts
$535Bportfolio of 162 business development companies, at fair value
+81%growth in the same 129 BDCs' portfolios since December 2023
$1.02of liabilities for every $1 of net assets, all BDCs combined

Those two numbers meet in the middle. Banks report how much they lend to "business credit intermediaries," the category the regulators created for BDCs, direct lenders and private credit funds. On June 30, 2026, it was $446 billion. That is more than the combined liabilities of the 170 BDCs that report them, $283 billion. The bank money reaches well past the listed funds, into private ones that publish nothing.

Myth one: the banks stepped back

Here is the received view at its strongest. Banks have been squeezed out of leveraged and mid-market lending by capital rules, and private credit took the business. In the narrow sense this is true: those loans now sit in funds, not banks.

But the banks did not step back from the risk. They moved to the next rung up. In December 2024 the regulators began requiring banks with $10 billion or more of assets to split their lending to nonbank financial institutions by kind of borrower. The split shows where the money went.

Exhibit 4Bank loans to nonbank financial institutions, by kind of borrower ($ billions)
Private credit and BDCsPrivate equity fundsMortgage lendersConsumer lendersOther

Source: FFIEC Call Reports, Schedule RC-C Memorandum items (loans to business credit intermediaries, private equity funds, mortgage credit intermediaries, consumer credit intermediaries, other). Data as of 2026-06-30. Banks with $10 billion or more of assets report the split, so these totals are slightly below the all-bank figure. In the first half of 2025 some banks re-sorted loans out of Other into the named categories, so growth in a single category overstates new lending; the total is not affected.

Show the data
Private credit and BDCsPrivate equity fundsMortgage lendersConsumer lendersOther
Dec 2024$212.8B$201.9B$222.1B$107.5B$395.2B
Jun 2025$341.9B$312.5B$322.6B$135B$245.5B
Dec 2025$402B$377B$357.2B$148.9B$253.9B
Jun 2026$446.4B$409.9B$378.9B$163.8B$273B

Loans to private credit funds and BDCs, and loans to private equity funds, together came to $856 billion by June 2026. Because some banks re-sorted their loans during 2025, the cleanest test of growth is the total, on the same banks. The 517 banks that reported these loans at both dates lent $1,152 billion to nonbank financial institutions in December 2024 and $1,693 billion in June 2026, a rise of 47% in eighteen months. Over the same period their commercial and industrial loans grew 10.6%, and all bank loans 9.5%.

Exhibit 5Eighteen months of growth at the same banks, December 2024 to June 2026

Source: FFIEC Call Reports, banks reporting at both dates; Finantrix calculation. Data as of 2026-06-30.

Show the data
Eighteen months of growth at the same banks, December 2024 to June 2026
Loans to nonbank financial institutions47.0%
Commercial and industrial loans10.6%
All loans and leases9.5%

That is the first myth, and the Call Reports bust it. The fastest-growing loan a bank makes is no longer a loan to a business. It is a loan to someone else who lends to a business.

Who lends to the lenders

The lending is concentrated. The ten largest lenders hold 73% of all bank loans to nonbank financial institutions, and 75% of loans to private credit and private equity funds. Wells Fargo is the largest lender to those funds, followed by JPMorgan Chase and Bank of America.

Exhibit 6The lenders' lenders: the twelve largest bank lenders to private credit and private equity funds ($ billions)
Loans to private credit and BDCsLoans to private equity fundsUndrawn lines to both

Source: FFIEC Call Reports, Schedule RC-C Memoranda and Schedule RC-L, by bank charter. Data as of 2026-06-30. Ranked by loans to private credit and private equity funds combined. Bank charters, not holding companies.

Show the data
Loans to private credit and BDCsLoans to private equity fundsUndrawn lines to both
Wells Fargo Bank$72B$56.7B$63.8B
JPMorgan Chase Bank$56.5B$68B$60.7B
Bank of America$31.5B$37.3B$37.6B
Goldman Sachs Bank USA$27.4B$34.6B$30.1B
PNC Bank$34.8B$27.1B$34.5B
Citibank$51.4B$9B$24B
First-Citizens Bank & Trust$2.3B$35B$15B
State Street Bank and Trust$22.3B$10.7B$12.1B
Morgan Stanley Bank$20.9B$12B$21.1B
U.S. Bank$14.2B$14.4B$18.1B
Truist Bank$22.7B$5.4B$23.4B
Citizens Bank$6.1B$9.8B$13.4B

For some banks this is no longer a sideline. At Morgan Stanley Bank, loans to nonbank financial institutions are 72% of all its loans; at State Street Bank, 66%; at Goldman Sachs Bank USA, 53%. At EverBank, a $47 billion bank, they are 43%. Among the largest commercial banks the share is lower but not small: 23% at Wells Fargo, 18% at Citibank, 17% at Bank of America and 16% at JPMorgan Chase.

Exhibit 7Loans to nonbank financial institutions as a share of each bank's own loans

Source: FFIEC Call Reports, item 9.a over total loans and leases, by bank charter. Data as of 2026-06-30. The fifteen largest bank lenders to private credit and private equity funds. The share covers all lending to nonbank financial institutions, including mortgage and consumer lenders.

Show the data
Loans to nonbank financial institutions as a share of each bank's own loans
Morgan Stanley Bank72%
State Street Bank and Trust66%
Goldman Sachs Bank USA53%
EverBank43%
First-Citizens Bank & Trust27%
Wells Fargo Bank23%
PNC Bank20%
KeyBank19%
Citibank18%
Bank of America17%
JPMorgan Chase Bank16%
Citizens Bank15%
U.S. Bank13%
Truist Bank13%
Capital One9%

Where the risk went: the unused trillion

A loan shows up on a balance sheet. A promise to lend does not. Schedule RC-L of the Call Report asks banks for their unused commitments, and since December 2024 it has asked for them by kind of nonbank borrower. On June 30, 2026, banks had promised nonbank financial institutions another $1.09 trillion that had not yet been drawn. Of that, $489 billion was promised to private credit and private equity funds.

Exhibit 8Drawn and undrawn: bank lending to nonbank financial institutions by kind of borrower ($ billions)
Drawn loansUndrawn commitments

Source: FFIEC Call Reports, Schedule RC-C Memoranda and Schedule RC-L unused commitments. Data as of 2026-06-30.

Show the data
Drawn loansUndrawn commitments
Private credit and BDCs$446.4B$214.5B
Private equity funds$409.9B$274.7B
Mortgage lenders$378.9B$103.8B
Consumer lenders$163.8B$88.9B
Other$273B$390.2B

Credit lines matter because of when they are used. A fund draws its bank line when it needs cash: to make loans quickly, to meet investors' redemption requests, or to carry a portfolio through a bad stretch while waiting for borrowers to pay. In other words, the undrawn trillion becomes drawn exactly when conditions are worst. A bank that looks at its drawn balance sees what it has lent. It has to look at the undrawn line to see what it may be asked to lend next.

The picture of how a dollar reaches a private borrower now has several steps, and a bank sits near the top of it.

Exhibit 9How bank money reaches a private credit borrower
  1. $1.70TInsured banks lend to nonbank financial institutions
  2. $856Bof it to private credit funds, BDCs and private equity funds
  3. $1.02of liabilities per $1 of net assets at the average BDC
  4. $535Blent onward by BDCs alone to mid-sized companies

Source: FFIEC Call Reports, June 30, 2026; SEC XBRL company facts, 172 BDCs. Data as of 2026-06-30. Bank lending to private credit and private equity funds also reaches private funds that do not file with the SEC, so the steps are not a closed loop.

Show the data
StepFigureWhat it is
1$1.70TInsured banks lend to nonbank financial institutions
2$856Bof it to private credit funds, BDCs and private equity funds
3$1.02of liabilities per $1 of net assets at the average BDC
4$535Blent onward by BDCs alone to mid-sized companies

Myth two: the loans are fine

If banks are lending to the lenders, the next question is the quality of what the lenders hold. The industry's answer is that defaults are low and the loans are senior and well secured. The BDCs' own filings let us check part of that, and on the whole the answer is closer to right than the critics allow.

Start with marks. The 158 BDCs that report both numbers held their portfolios at 98.1% of what they paid for them on June 30, 2026, a markdown of about $10 billion on $544 billion. That is not the picture of a market in trouble.

Next, payment in kind, the practice of letting a borrower pay interest by adding it to the loan rather than paying cash. It is the measure critics watch most closely, because a borrower who cannot pay cash may simply be deferring a default. Across the BDCs that report it, PIK made up 6.2% of investment income in the first half of 2026, down from 7.1% in 2024.

Exhibit 10Interest paid in kind, as a share of BDCs' investment income
0.0%2.5%5.0%7.5%10.0%20212022202320242025H1 2026PIK share of investment income, 2021: 8.2%PIK share of investment income, 2022: 7.5%PIK share of investment income, 2023: 6.9%PIK share of investment income, 2024: 7.1%PIK share of investment income, 2025: 6.6%PIK share of investment income, H1 2026: 6.2%

Source: SEC XBRL company facts: InterestIncomeOperatingPaidInKind and GrossInvestmentIncomeOperating, BDCs reporting both for the period; Finantrix calculation. Data as of 2026-06-30. BDCs reporting both items: 28 in 2021, rising to 86 in the first half of 2026. Some large non-traded BDCs do not report PIK income as a separate item and are not included.

Show the data
PIK share of investment income
20218.2%
20227.5%
20236.9%
20247.1%
20256.6%
H1 20266.2%

The trouble is not in the average. It is in the spread. Non-accrual loans, those on which a lender has stopped recording interest because it no longer expects to be paid, are low at most of the largest BDCs and rising quickly at a few.

BDC Non-accrual, at cost Non-accrual, at fair value Earlier
Golub Capital BDC 2.9% 1.9% 0.6% / 0.3% (Sep 2025)
Blue Owl Capital Corp 2.8% 0.8% 2.3% / 1.1% (Dec 2025)
Ares Capital 2.4% 1.4% 1.8% / 1.2% (Dec 2025)
FS KKR Capital — 3.8% —
Apollo Debt Solutions BDC 0.8% 0.4% 0.8% / 0.3% (Dec 2025)
HPS Corporate Lending Fund — 0.7% — / 0.74% (Dec 2025)
Ares Strategic Income Fund 0.3% 0.2% —
Goldman Sachs Private Credit Corp 0.1% 0.1% 0.2% / 0.1% (Dec 2025)
Golub Capital Private Credit Fund 0.1% 0.1% —
Ares Core Infrastructure Fund none none none

As of June 30, 2026, from each fund's 10-Q. Golub's funds report on a September fiscal year. Funds whose 10-Q does not state the share as a percentage are left out rather than estimated. Sources are listed at the end.

Exhibit 11Non-accrual loans at cost, selected large BDCs

Source: Each BDC's 10-Q for the quarter ended June 30, 2026 (sentences quoted in the Finantrix research file). Data as of 2026-06-30. Golub Capital BDC went from 0.6% to 2.9% at cost in nine months. FS KKR reports 3.8% of its portfolio at fair value on non-accrual but no cost figure, so it is not in this chart.

Show the data
Non-accrual loans at cost, selected large BDCs
Golub Capital BDC2.9%
Blue Owl Capital Corp2.8%
Ares Capital2.4%
Apollo Debt Solutions BDC0.8%
Ares Strategic Income Fund0.3%
Goldman Sachs Private Credit0.1%
Golub Capital Private Credit0.1%

Two things are worth seeing in that table. Golub Capital BDC's non-accrual share at cost went from 0.6% to 2.9% in nine months, five times as much. And the gap between cost and fair value is wide: at Blue Owl Capital Corp, loans on non-accrual are 2.8% of the portfolio at cost but only 0.8% at fair value. Blue Owl has already marked those loans down to well under half of what it lent. That is prudent accounting. It is also a reminder that "losses are low" and "the loans are fine" are different statements.

So on credit quality the critics are mostly wrong about the market and partly right about parts of it. The averages are calm. The problems are concentrated in particular funds and particular borrowers, and they can grow quickly. That is exactly the pattern that makes a lender's lender pay attention to which funds it backs.

When something breaks

The second half of 2025 showed what concentrated problems look like from the outside. Two companies failed within days of each other. Neither was a household name. Both had borrowed from lenders that were not banks, and the losses still reached banks.

Exhibit 12Two failures in 2025, as the filings disclosed them
  1. Sep 10, 2025Tricolor Holdings, a subprime auto lender that packaged its loans into asset-backed securities, files for Chapter 7 liquidation. Origin Bancorp discloses about $30.1 million of loan commitments to it the same day.
  2. Sep 11, 2025Renasant says Tricolor owes its bank about $22.5 million. Triumph Financial says its bank is agent on a $60.5 million facility to Tricolor and holds about $23 million of it.
  3. Sep 29, 2025First Brands, an auto-parts maker that sells to the major parts retailers, files for Chapter 11.
  4. Oct 7, 2025Great Elm Capital, a BDC, details its First Brands loans; its net asset value per share later falls from $12.10 to $10.01, a decline its chief executive attributes mainly to First Brands.
  5. Oct 8, 2025Jefferies discloses that its Point Bonita fund has about $715 million invested in receivables First Brands was servicing.
  6. Oct 22, 2025Origin Bancorp raises its credit-loss provision by $29.5 million, mainly because of suspected borrower fraud at Tricolor.
  7. Oct 23, 2025First Citizens reports an $82 million individual client loss in its commercial services business; its filed presentation links it to First Brands.
  8. Dec 17, 2025Federal prosecutors unseal criminal charges against Tricolor executives. Tricolor's bankruptcy trustee later alleges about $683 million of fictitious loans in its records.
  9. Feb 2026Genuine Parts, which distributes auto parts First Brands made, reports a $151 million charge for credit losses on amounts First Brands owed it.

Source: Origin Bancorp 8-Ks, Sep 10 and Oct 22, 2025; Jefferies Financial Group 8-K, Oct 8, 2025; M&T Bank 10-Q, quarter ended June 30, 2026 (Tricolor bankruptcy, charges and trustee's allegation); Genuine Parts 10-K for 2025. Data as of 2026-08-04. Every entry restates a company's own filing; the full list of filings is in Sources.

Read the timeline for who was exposed, not for the fraud allegations, which are for the courts. Banks in Louisiana, Mississippi and Texas had lent to a subprime auto lender. A North Carolina bank took an $82 million loss through its commercial services business. A BDC lost about a sixth of its net asset value per share. An investment bank's fund held $715 million of receivables whose collection depended on the company that failed. And an auto-parts distributor, nobody's idea of a lender, took a $151 million charge on money First Brands owed it. The losses traveled in every direction the money had come from.

A thought experiment: if the lines get drawn

Suppose, as a thought experiment, that a stretch of weak borrower results makes private credit funds draw on their bank lines: to carry loans that have stopped paying cash, to meet investors who want their money back, or simply because the cash is cheaper than selling loans. Nobody has to predict that to ask what it would do to the banks' numbers, and the Call Reports answer it.

Banks have $489 billion of undrawn commitments to private credit and private equity funds. If a third of that were drawn, their loans to those funds would rise by about $163 billion, from $856 billion to just over $1 trillion, without a single new credit decision. If half were drawn, about $245 billion. That is a matter of arithmetic, not a forecast. It shows why a line that looks small on the drawn side can matter more than the loan beside it.

Exhibit 13Bank loans to private credit and private equity funds if their undrawn lines were partly drawn ($ billions)

Source: FFIEC Call Reports: drawn loans ($856.3 billion) and undrawn commitments ($489.2 billion) to business credit intermediaries and private equity funds; Finantrix arithmetic. Data as of 2026-06-30. An illustration, not a forecast: it assumes no repayments and no new commitments.

Show the data
Bank loans to private credit and private equity funds if their undrawn lines were partly drawn ($ billions)
Today, June 30, 2026$856B
If a third of the lines were drawn$1,019B
If half were drawn$1,101B
If all were drawn$1,345B

The order in which a loss would be felt is also visible in the record. A fund's own investors and a BDC's shareholders sit first: their net assets absorb markdowns, as Great Elm's did. The fund's lenders sit behind them, and banks are the largest of those lenders. A bank's loss, if one comes, reaches its own shareholders and, in the extreme, the deposit insurance fund. That is a long way down. But it is not "somebody else's problem," which is how the popular story ends.

Myth three: this is a problem for the giants

The largest banks dominate the numbers: the ten biggest lenders hold 73% of the total. It is tempting to conclude that a community or regional bank can read this as someone else's story.

The record says otherwise. On June 30, 2026, 633 insured banks reported loans to nonbank financial institutions. The first banks to disclose exposure to the 2025 failures in their SEC filings were not the largest. They were Origin, Renasant and Triumph, banks with between $6 billion and $27 billion in assets, which had lent to a nonbank lender directly. A small bank does not need a $50 billion private credit book to be exposed. It needs one finance-company relationship that looked well secured.

How we did this

Bank figures come from the FFIEC Call Report bulk files for every quarter from December 2024 to June 2026, and every year-end from 2010 to 2023: Schedule RC-C, item 9.a (loans to nondepository financial institutions), its Memorandum items by kind of borrower, and Schedule RC-L unused commitments by the same kinds. Each bank is counted once, using its consolidated figure where it files one. The breakdown by kind of borrower is collected only from banks with $10 billion or more of assets. The reporting instructions changed in December 2024, and the total jumped by $261 billion that quarter while the same banks' commercial and industrial loans fell by $131 billion, so every growth figure here starts in December 2024 or compares the same banks. Some banks re-sorted loans between categories in 2025, so category growth is reported only alongside that caution.

BDC figures come from the SEC's structured company facts for the 172 BDCs (file numbers beginning 814-) that filed a 10-Q or 10-K between April and October 2026. Portfolio values, liabilities, net assets and payment-in-kind income are taken from the face of each fund's statements. Fourteen BDCs' structured data contains net-asset figures inconsistent with their own balance sheets in at least one period; for those periods we use total assets minus total liabilities. Non-accrual figures are not in the structured data. They were read from the text of each fund's 10-Q, and the sentence each figure comes from is kept in the research file.

Event figures are quoted from the companies' own filings. Where a widely reported statement was made only on an earnings call and we did not capture a primary transcript, we left it out. What we cannot see, we have not estimated: most private credit funds do not file with the SEC, and their size is not in these numbers.

What it means if you are deciding something

For a bank board or credit committee, the useful question is not whether private credit is a bubble. Nobody in the room can settle that. The question that can be settled is how much of the bank's exposure to private credit sits in a line most reports skip: loans and undrawn commitments to nonbank financial institutions. Ask for it drawn and undrawn, by kind of borrower, by fund sponsor, and with the advance rate and collateral for each facility. Ask what happens to each line in a month when every fund draws at once. The worksheet that comes with this report sets those questions out on one page.

For an insurer, pension fund or wealth manager allocating to private credit, the question is who sits beside you in the capital structure. A fund with a bank line has a senior lender that will be paid before you and may tighten terms when you most need it to wait. Ask each manager how much of the fund's assets are financed with bank borrowing and on what terms.

For a vendor selling into either side, this is where budgets are moving. Banks are being asked to report exposure to nonbank lenders in more detail than ever before, and the 2025 failures turned fraud in pledged collateral and double-pledged receivables into board topics. Monitoring, collateral verification and exposure reporting for lender finance are no longer niche.

The rule

When risk leaves a balance sheet, find who lent to the buyer.

The next time a headline says that private credit has taken lending risk off the banks, ask three questions before sharing it. How much do banks lend to the lenders, drawn and undrawn? How concentrated is that lending, and in whose funds? And if a borrower fails, whose losses come first and whose come next? A market where the answers are small, diverse and junior to plenty of equity has moved risk. One where the answers are a trillion dollars, ten banks and a credit line waiting to be drawn has mostly moved where the risk is written down.

Sources

  1. FFIEC Central Data Repository, Call Reports, bulk data, all schedules, December 31, 2010 to June 30, 2026. https://cdr.ffiec.gov/public/PWS/DownloadBulkData.aspx
  2. Board of Governors of the Federal Reserve System, FDIC and OCC, "Proposed Agency Information Collection Activities; Comment Request," Federal Register, December 27, 2023 (document 2023-28473). https://www.federalregister.gov/documents/2023/12/27/2023-28473/proposed-agency-information-collection-activities-comment-request
  3. U.S. Securities and Exchange Commission, XBRL company facts, 172 business development companies. https://data.sec.gov/api/xbrl/companyfacts/
  4. Ares Capital Corporation, Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/1287750/000162828026050307/arcc-20260630.htm
  5. Golub Capital BDC, Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/1476765/000147676526000048/gbdc-20260630.htm
  6. Blue Owl Capital Corporation, Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/1655888/000165588826000056/obdc-20260630.htm
  7. FS KKR Capital Corp., Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/1422183/000162828026053783/fsk-20260630.htm
  8. Apollo Debt Solutions BDC, Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/1837532/000119312526341358/ck0001837532-20260630.htm
  9. HPS Corporate Lending Fund, Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/1838126/000162828026056772/hps-20260630.htm
  10. Ares Strategic Income Fund, Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/1918712/000162828026054923/asif-20260630.htm
  11. Goldman Sachs Private Credit Corp., Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/1920145/000119312526340582/ck0001920145-20260630.htm
  12. Golub Capital Private Credit Fund, Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/1930087/000193008726000099/gcred-20260630.htm
  13. Ares Core Infrastructure Fund, Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/2031750/000203175026000064/aci-20260630.htm
  14. Jefferies Financial Group, Forms 8-K, October 8 and October 14, 2025. https://www.sec.gov/Archives/edgar/data/96223/000114036125037595/ef20056799_ex99-1.htm and https://www.sec.gov/Archives/edgar/data/96223/000114036125038023/ef20057006_ex99-1.htm
  15. Origin Bancorp, Forms 8-K, September 10 and October 22, 2025. https://www.sec.gov/Archives/edgar/data/1516912/000151691225000135/obnk-20250907.htm and https://www.sec.gov/Archives/edgar/data/1516912/000151691225000141/a09302025obkexhibit991er.htm
  16. Renasant Corporation, Form 8-K, September 11, 2025. https://www.sec.gov/Archives/edgar/data/715072/000071507225000219/rnst-20250911.htm
  17. Triumph Financial, Form 8-K, September 11, 2025. https://www.sec.gov/Archives/edgar/data/1539638/000153963825000011/tfin-20250911.htm
  18. Great Elm Capital Corp., Forms 8-K, October 7 and November 4, 2025. https://www.sec.gov/Archives/edgar/data/1675033/000119312525232655/d31018dex991.htm and https://www.sec.gov/Archives/edgar/data/1675033/000119312525264872/gecc-ex99_1.htm
  19. First Citizens BancShares, Form 8-K (third-quarter 2025 investor presentation), October 23, 2025. https://www.sec.gov/Archives/edgar/data/798941/000079894125000047/a3q2025investorpresentat.htm
  20. M&T Bank Corporation, Form 10-Q for the quarter ended June 30, 2026. https://www.sec.gov/Archives/edgar/data/36270/000003627026000050/mtb-20260630.htm
  21. Genuine Parts Company, Form 10-K for 2025. https://www.sec.gov/Archives/edgar/data/40987/000004098726000003/gpc-20251231.htm

This report is research, not investment advice. It does not recommend buying or selling any security. Figures come from regulatory filings and company disclosures as cited.

Finantrix deep dive

Take the report into your next meeting

Everything below, free. We email you a link to confirm the address.

  • The data workbook · Excel workbook, 82 KB
    Every exhibit's numbers and the full underlying dataset, with sources.
  • The slides · PowerPoint deck, 511 KB
    Editable charts for your own deck, each with its source line.
  • Lender-finance exposure review · Word document, 62 KB
    A one-page template for a credit or board risk committee.
  • The full report · PDF, 477 KB
    Every exhibit, with its source, plus the method.

Protected by reCAPTCHA. Cite and share with credit to Finantrix and a link to this page.

Frequently Asked Questions

How much do US banks lend to private credit funds?

On June 30, 2026, insured banks reported $446 billion of loans to business credit intermediaries, the Call Report category that holds private credit funds, BDCs and direct lenders, plus $410 billion to private equity funds. They had promised those two groups a further $489 billion in undrawn credit lines.

What are loans to nondepository financial institutions?

They are bank loans to lenders and funds that do not take deposits, such as private credit funds, BDCs, private equity funds, mortgage companies and consumer finance companies. Banks report them on Schedule RC-C, item 9.a of the Call Report; since December 2024 banks with $10 billion or more of assets also report them by kind of borrower.

Is private credit risk on bank balance sheets?

Partly. The loans to companies sit in funds, but banks lend to many of those funds, and had $856 billion of such loans outstanding and $489 billion more committed in mid-2026. A bank's exposure is to the fund, one step removed from the borrowing companies.

Are private credit loans going bad?

Not on average, by the measures BDCs publish: portfolios were marked at 98.1% of cost in June 2026 and interest paid in kind was 6.2% of investment income. But problems are concentrated: non-accruals at Golub Capital BDC rose from 0.6% to 2.9% of cost in nine months, and FS KKR reported 3.8% of its portfolio at fair value on non-accrual.

Deep DivePrivate CreditBank LendingCredit RiskBDCs
Share: