The Credit Bridge: Banks May Have Moved Risk Into the Shadow System—Not Removed It
Zeeshan · 2026-09-15
Bank lending to nondepository financial institutions is growing and concentrated. The credit risk may be moving through a larger, less visible funding bridge.
By Zeeshan | YouYaa Intelligence | 15 September 2026
A loan can leave a bank’s old reporting category without leaving the financial system.
The controversial thesis
Banks and nondepository financial institutions are often described as separate worlds. Banks take deposits and lend. Nonbanks provide mortgage finance, business credit, consumer loans, and investment products without the same deposit model.
That separation is becoming harder to see.
The FDIC says bank lending to nondepository financial institutions, or NDFIs, was among the fastest-growing bank-lending segments in recent years. It was also heavily concentrated at the largest banks.[1]
The controversial question is simple:
Did banks reduce credit risk—or move it into a more complex chain that is harder to see?
This is not a claim that every NDFI is unsafe. The FDIC says credit-quality measures for NDFI loans remained favorable in 2025.[1] The issue is structure: concentration, opacity, and the possibility that many nonbank lenders depend on the same bank funding lines when markets become difficult.
The bridge is growing
NDFIs include mortgage lenders, business-credit providers, consumer-finance companies, finance companies, broker-dealers, private funds, and other credit intermediaries. They can reach borrowers quickly and serve segments that banks may not serve directly.
But banks still sit behind parts of the chain. They may provide warehouse lines, revolving credit, secured loans, liquidity facilities, and other financing to NDFIs.
The FDIC’s 2026 Risk Review says that more than half of NDFI loans were to credit intermediaries for mortgages, business loans, and consumer loans.[1]
| FDIC finding | Why it matters |
|---|---|
| NDFI lending was among the fastest-growing bank-lending segments | It is becoming strategically important to bank balance sheets |
| Lending was heavily concentrated at the largest banks | A small group of institutions may carry a large share of the bridge |
| More than half went to mortgage, business, and consumer credit intermediaries | Risk can travel through several layers before reaching the end borrower |
| 2025 credit-quality measures remained favorable | This is a structural warning, not proof of current distress |
| Reporting changes created more granular data | Better visibility can reveal risk that was previously grouped elsewhere |
The same facts support two views. NDFI lending can diversify access to credit. It can also create a concentrated dependency that is invisible until funding is withdrawn.
The number is large—but definitions matter
A Federal Reserve/FRED series for loans to nondepository financial institutions showed approximately $2.028 trillion on 2 September 2026.[2] The series is subject to its own definitions, revisions, and reporting changes, so the number should not be treated as a perfect measure of total shadow-bank risk.
It is still a useful signal.
A balance above $2 trillion means treasury teams should not treat bank-to-NDFI lending as a small specialist activity. It is part of the credit plumbing.
The right question is not only “How much do banks lend to nonbanks?” It is:
How many borrowers, lenders, warehouses, and funds depend on the same refinancing window?
Why concentration matters
If many NDFIs borrow from a small group of large banks, a bank may face a portfolio problem even when each individual loan looks acceptable. A common funding source can turn separate companies into one correlated exposure.
A mortgage lender may be healthy. A consumer lender may be healthy. A business-credit platform may be healthy. If all three depend on the same warehouse market, collateral rules, or bank credit committee, stress can arrive through the funding channel rather than the borrower’s income statement.
| Layer | Typical role | Possible stress point |
|---|---|---|
| Bank | Provides warehouse or secured financing | Reduces limits or raises margins |
| NDFI | Originates or purchases loans | Cannot fund new originations |
| Credit intermediary | Packages or distributes exposure | Investors demand faster repayment |
| End borrower | Receives mortgage, business, or consumer credit | Faces tighter terms or delayed funding |
| Market | Prices collateral and liquidity | Haircuts and spreads rise together |
This is the credit bridge. A problem at one layer can change behaviour at every other layer.
The opacity problem
The FDIC notes that reporting changes in 2025 provided more granular data on bank exposures to NDFIs.[1] Better data is good news. It also suggests that earlier totals did not show the full shape of the exposure.
Opacity does not mean fraud. It means a board may know the counterparty but not the economic chain behind the counterparty.
A bank can report an exposure to an NDFI while the true risk depends on mortgages, small businesses, auto loans, or consumer balances originated elsewhere. A fintech can report strong loan performance while depending on a short-term warehouse line that can be repriced or cancelled.
The risk map must therefore include both the direct borrower and the funding ecosystem.
What CFOs and fintech operators should test
For a CFO, the key concern is not whether a lender is regulated. It is whether the lender can keep funding the company through a stressed market.
For a fintech operator, the critical document may be the warehouse agreement, not the customer loan book. Review advance rates, eligibility rules, borrowing-base tests, margin triggers, concentration limits, early-amortisation clauses, and lender discretion.
For HNWIs and family offices, the issue appears in private credit vehicles, specialty finance funds, mortgage products, and structured cash strategies. Strong recent performance does not prove that liquidity will remain available when everyone wants to exit.
| Test | What management should know |
|---|---|
| Funding concentration | Which bank or facility provides the largest share of cash? |
| Collateral eligibility | Which loans stop qualifying after a downgrade, delay, or data problem? |
| Advance-rate sensitivity | How much cash disappears if the advance rate falls by 5 or 10 points? |
| Renewal risk | What happens if the line is not renewed on schedule? |
| Correlated borrowers | Which portfolios depend on the same employer, property market, platform, or bank? |
| Data transfer | Can the lender verify collateral quickly if systems are disrupted? |
A stress test that only changes borrower defaults misses funding-chain risk.
The “risk left the bank” illusion
Regulation and capital requirements can encourage banks to use partnerships, facilities, and distribution channels. That can be efficient. It can also create the illusion that risk has left the regulated system.
Risk may instead have changed form. It may become a bank commitment, a warehouse line, a collateral call, a liquidity guarantee, or a portfolio of loans to intermediaries.
The FDIC’s evidence does not prove that this bridge will fail. It shows that the bridge is growing and that the largest banks are important providers of its funding.[1]
That distinction matters. The right response is not to shut down nonbank credit. It is to measure the bridge honestly.
Conclusion
The most important credit risk may not sit in the final borrower’s loan. It may sit in the refinancing relationship between a bank and the nonbank that delivered the loan.
The FDIC identifies NDFI lending as one of the fastest-growing bank-lending segments, concentrated at the largest banks, with more than half of exposure going to credit intermediaries for mortgages, business loans, and consumer loans.[1]
A Federal Reserve/FRED series places loans to NDFIs at about $2.028 trillion on 2 September 2026.[2]
The controversial conclusion is this:
Banks may not have removed credit risk. They may have built a larger bridge into a less visible part of the system.
FAQ
What are NDFIs?
NDFIs are nondepository financial institutions. They provide financial services or credit without operating as traditional deposit-taking banks. Examples include finance companies, mortgage lenders, credit intermediaries, broker-dealers, and some private funds.
Why do banks lend to NDFIs?
Banks can earn interest, support loan origination, provide warehouse funding, and serve credit channels indirectly through NDFIs.
How large is bank lending to NDFIs?
A Federal Reserve/FRED series showed approximately $2.028 trillion on 2 September 2026. The series has specific definitions and can be revised.[2]
Is NDFI lending currently in crisis?
The FDIC said NDFI credit-quality measures remained favorable in 2025.[1] The article identifies structural concentration and funding-chain risk, not an inevitable crisis.
Why does concentration at the largest banks matter?
If many NDFIs depend on a small group of banks, one bank’s change in limits, margins, or underwriting can affect many lenders and end borrowers at once.
What should a fintech check first?
Review warehouse and financing agreements, including advance rates, collateral rules, margin triggers, renewal terms, concentration limits, and lender discretion.
Is this investment advice?
No. This is general analysis of bank-to-NDFI lending, credit intermediation, funding concentration, and financial stability. Obtain appropriate legal, accounting, compliance, and regulated financial advice for individual circumstances.
References
[1] FDIC, “Risk Review 2026,” sections on NDFI lending
[2] Federal Reserve / FRED, “Loans to Nondepository Financial Institutions,” series LNFACBW027SBOG
