Bank balance sheets on both sides of the Atlantic have undergone a profound transformation. While traditional corporate lending has slowed, lending to non-bank financial institutions (NBFIs) has expanded rapidly (e.g. Acharya et al. 2024, 2025). This raises a key policy question: is bank lending to NBFIs simply lengthening the intermediation chain, or is it diverting credit away from the real economy?

We answer this question in a recent paper (Li et al. 2025) using new granular evidence from the euro area. Drawing on a comprehensive dataset combining supervisory information with loan-level data on banks’ exposures to both firms and NBFIs, we show that the answer is unambiguous: the rise of bank lending to NBFIs is not merely an evolution of the intermediation chain. Banks in the euro area are increasingly reallocating balance-sheet capacity toward safer, short-term, collateralised lending to NBFIs and away from firms. The fastest-growing category – reverse repurchase agreements (reverse repos) – finances government securities positions, not corporate lending. As this activity expands, firm credit contracts.

Reverse repos now dominate bank–NBFI lending

Since 2019, bank lending to NBFIs has grown by nearly 60%, compared with only about 20% growth in lending to non-financial corporations. The divergence has widened since mid-2022, when corporate lending stagnated while NBFI lending continued to rise steadily. As of 2024, bank credit to NBFIs amounts to roughly two-thirds of the volume of corporate loans. Figure 1 (Panel A) depicts the time-evolution of bank loans to firms and NBFIs over 2019–2024.

Figure 1 Bank loan volumes to firms and non-bank financial institutions (NBFIs)

A) Volume of loans to NBFIs and non-financial corporations (NFCs)

B) Volume of bank loans to non-bank financial institutions by instrument

Note: The top figure shows the volume of loans lent to non-bank financial institutions (NBFIs) versus non-financial corporations. The volumes are normalised by their value in January 2019. The bottom figure shows the volume of bank loans lent to NBFIs by instrument, which includes credit lines, term loans, revolving credit, reverse repos, and others.
Source: Li et al. (2025).

A striking feature of this shift is its composition. Reverse repos now account for around 45% of all bank–NBFI exposures.
  These are short-term, collateralised loans – typically backed by government securities – used predominantly to support securities-financing trades. Investment funds, particularly hedge funds, are major recipients. By contrast, NBFIs that actually lend to firms (e.g. leasing companies, factoring firms, or private-debt funds) remain a small share of recent growth. Hence, the expansion of bank–NBFI lending is largely associated with the growth in securities-financing rather than credit intermediation to firms.

What is driving the surge in bank lending to NBFIs?

Two forces underpin the rise of NBFI borrowing: strong demand from securities-investing NBFIs and bank balance-sheet constraints that make lending to them comparatively attractive.

On the demand side, borrowing by investment funds – especially hedge funds – has grown sharply, particularly through reverse repos in non-euro currencies backed by low-haircut collateral. These features align closely with the mechanics of leveraged trades in government bond markets, including the Treasury basis trade. Such trades have been fuelled by a surge in sovereign debt issuance and the unwinding of quantitative easing programmes by central banks.

Although NBFI demand is central, bank balance-sheet constraints amplify this pattern. Loans to NBFIs –  especially reverse repos – are safer, more liquid, and carry lower regulatory risk weights than loans to firms. Banks with weaker capital ratios expand NBFI lending more strongly, substituting away from higher-risk corporate exposures. Our evidence on the importance of bank capital constraints is consistent with prior findings in the literature (Acharya et al. 2024, Buchak et al. 2024, Chernenko et al. 2025, Xu 2025). Banks facing larger deposit outflows or greater reliance on targeted longer-term refinancing operation (TLTRO) funding also shifted toward reverse repos, which require less stable funding.

Bank-to-NBFI lending crowds out firm lending

These forces jointly shape a systematic reallocation of bank lending away from firms. Figure 2 shows a strong negative relationship between changes in banks’ NBFI lending and their corporate lending between 2019 and 2023. A formal empirical investigation shows that a one-percentage-point increase in the share of NBFI lending is associated with a 0.55-percentage-point decline in lending to firms. Substitution away from other liquid assets – such as reserves or government bonds – is far weaker, indicating that the reallocation comes directly at the expense of corporate credit.

Figure 2 Substitution between corporate and NBFI lending

Note: This figure presents a binscatter plot showing the relationship between changes in bank lending to non-financial corporations and to non-bank financial intermediaries (NBFIs) between December 2019 and December 2023. The variables on both axes are expressed as changes in the ratio of each loan type to total loans and securities. Each point represents the mean value within a bin of banks, and the fitted line shows the linear relationship between the two variables.
Source: Li et al. (2025).

Moreover, the reallocation is not offset elsewhere in the financial system. The reason is simple: the fastest-growing NBFI borrowers from banks are not lenders to firms – they are mostly securities investors. Firm-level data confirm that firms connected to banks that expand NBFI exposures face larger reductions in total borrowing, including from non-bank sources. The effects are concentrated among small and risky firms with limited access to market-based finance, producing declines in both borrowing and debt. This reallocation therefore has meaningful distributional effects within the corporate sector.

A simple framework explains why aggregate firm credit contracts

To interpret these findings, we develop a model in which a bank allocates its balance sheet across corporate loans, securities, and NBFI lending. Two insights emerge:

1. When NBFIs increase borrowing to finance securities positions, banks reallocate balance-sheet space to meet that demand, reducing corporate lending.
2. When regulatory costs rise, banks substitute even more strongly toward low-risk NBFI exposures. While NBFIs expand their own lending, the effect is too small to compensate for reduced bank lending.

In both cases, aggregate credit to firms declines, particularly when the marginal NBFI borrower is a securities investor rather than a credit provider.

Conclusion

The rapid rise in bank lending to NBFIs represents a structural shift in European financial intermediation. Far from lengthening the credit chain, this trend is making banks narrower, reallocating balance-sheet capacity toward securities-financing and away from the real economy. Understanding this shift – and its interaction with regulation, monetary policy, and market-based finance – is crucial for assessing the allocation of credit to the real economy.

References

Acharya, V V, N Cetorelli and B Tuckman (2024), “Where Do Banks End and NBFIs Begin?”, NBER Working Paper 32316.

Acharya, V V, M Gopal, M Jager and S Steffen (2025), “Shadow Always Touches the Feet: Implications of Bank Credit Lines to Non-Bank Financial Intermediaries”, NBER Working Paper 33590.

Buchak, G, G Matvos, T Piskorski and A Seru (2024), “The secular decline of bank balance sheet lending”, NBER Working Paper 32176.

Chernenko, S, R Ialenti and D S Scharfstein (2025), “Bank Capital and the Growth of Private Credit”, SSRN Working Paper.

Li, J, Y Ma, C Mendicino and D Supera (2025), “Bank to Non-Bank Lending and the Reallocation of Credit”, Columbia Business School Research Paper No. 5732322.

Xu, C (2025), “The Value of Contingent Liquidity from Banks to Nonbank Financiers”, SSRN 5084953.