Nonbank lending across the yield curve: When monetary tightening is cushioned or reinforced
Does nonbank lending cushion or reinforce monetary tightening depending on which part of the yield curve moves, and which types of nonbank institutions supply credit? This is the question that Peter Bednarek, Francesco Febbraro, Steven Ongena, and Natalja von Westernhagen analyze in this paper.
The authors use German supervisory credit-register data for 2019–2023, matched with bank balance sheets, insurer solvency measures, and borrower characteristics. The full dataset contains approximately 15.7 million lender–firm–quarter observations.
They construct three mutually orthogonal monetary-policy surprises from interest-rate movements around ECB announcements: short-end surprises based on one-month overnight index swap rates, medium-end surprises based on two-year rates after removing short-rate movements, and long-end surprises based on ten-year rates after removing shorter-rate movements. Quarterly surprises enter regressions with a one-quarter lag.
The baseline identification compares bank and nonbank credit exposures to the same firm in the same quarter, controlling for lender fixed effects. This separates relative credit-supply responses from firm-level demand shocks, assuming financing-demand shocks are common across the firm’s lenders. Additional regressions test mechanisms involving bank capitalization and insurer solvency.
Findings
- Nonbanks cushion short-end tightening. After a one-standard-deviation short-end surprise, nonbanks increase credit by 0.32% relative to banks, partly offsetting banks’ lending contraction.
- Nonbanks reinforce medium- and long-end tightening. Following equivalent surprises at the medium and long end, nonbank credit falls by 0.36% and 0.34% relative to banks, respectively.
- Leasing and factoring companies drive this reversal. After short-end tightening, capital-constrained banks redirect funding toward these relatively low-default-risk intermediaries, supporting their onward lending. This funding channel does not operate after medium- and long-end surprises.
- Insurers partly offset longer-end tightening. Higher medium- and long-term rates reduce the value of insurers’ longer-duration liabilities more than their assets, improving solvency and supporting lending.
The composition of the nonbank sector matters. The overall lending response depends on both the yield-curve segment affected and the relative importance of different nonbank institutions. - The findings do not establish a symmetric QE effect. The paper identifies surprises at different yield-curve maturities so its tightening results cannot automatically be reversed to describe QE easing.

When nonbanks dampen and when they amplify: A yield-curve view of the nonbank lending channel
Authors: Peter Bednarek, Francesco Febbraro, Steven Ongena, Natalja von Westernhagen
From: Deutsche Bundesbank, University of Groningen, University of Zurich
Nonbank finance across policy tools: When nonbank growth reinforces QE but leaves rate transmission unchanged
How does a growing role for nonbank financial institutions change the transmission of policy-rate changes versus quantitative easing, and does the source of nonbank growth matter? This is the question Nicolò Bandera, Kristina Bluwstein, Michael Kumhof, Mauricio Salgado-Moreno, and Jacob Stevens are analyzing in this paper.
The authors develop a dynamic stochastic general equilibrium (DSGE) model featuring banks, nonbanks, households, firms, and a central bank. Nonbanks hold government bonds, corporate bonds, and wholesale bank deposits; firms finance investment through bank loans and corporate bonds, which are imperfect substitutes.
They estimate the model using Bayesian methods and 19 UK macroeconomic and financial time series from 1997Q2–2025Q3, treating 2020–2021 as missing to avoid pandemic-specific distortions. They then simulate rate cuts and QE under two sources of nonbank expansion: firms preferring bond finance, or bank borrowing becoming more expensive.
Findings
- Conventional rate policy remains effective. A larger nonbank sector has little effect on the output and inflation responses to a policy-rate cut, because interest-rate changes continue to pass through banks and nonbanks.
- QE is stronger when firms prefer market finance. When nonbank growth reflects firms’ increased preference for bond funding, QE’s effects on output and inflation are 15.4% and 10.3% larger, respectively, than in the baseline simulation.
- Costlier bank finance does not produce the same amplification. When higher bank lending costs drive nonbank expansion, QE’s overall effects remain largely unchanged: expensive bank borrowing offsets the stronger bond-market channel.
- Portfolio rebalancing is the main mechanism. QE purchases remove government bonds from nonbanks’ portfolios, encouraging purchases of corporate bonds, lowering corporate yields, and supporting investment; wholesale deposits and bank reserves provide additional transmission channels.
- The baseline QE effects are positive but model-dependent. A simulated £100 billion purchase, with policy rates held at the effective lower bound for eight quarters, raises peak output by approximately 0.15% and peak annualized inflation by 0.45 percentage points.

Lighting the shadows: central bank policy transmission through banks and non-banks
Authors: Nicolò Bandera, Kristina Bluwstein, Michael Kumhof, Mauricio Salgado-Moreno, Jacob Stevens
From: Bank of England, Centre for Macroeconomics