By Federico D’Amario, Sebastian de-Ramon and William B. Francis
Despite major changes to the UK banking system, including post-crisis reforms and the extraordinary conditions of the COVID-19 period, the central findings remain remarkably robust. Banks continue to pass the cost of higher capital predominantly to corporate borrowers rather than household (residential mortgage) borrowers. Updated estimates suggest pass-through rates of around 7 to 10 basis points on corporate lending spreads for every 1 percentage point increase in capital ratios, broadly consistent with the previous estimates of around 10 to 12 basis points.
Below we summarise the main findings, why they matter, and how they inform the wider debate on the economic cost of capital.
Why capital affects lending spreads
Banks finance lending with a mix of retail deposits, wholesale funding and equity capital. When capital requirements rise, banks typically adjust their balance sheets to maintain a voluntary buffer above the minimum – something we observe consistently in the data (eg, de-Ramon, Francis and Harris (2022)). They do this through one or a combination of the following actions: raising more capital, shrinking assets, changing the mix of risk-weighted assets, and altering the pricing and composition of lending.
Because corporate loans tend to be riskier, attracting a higher regulatory risk-weight, and shorter-term relative to household mortgages, banks tend to emphasize adjusting corporate lending spreads when seeking to optimize their capital position. This emphasis is consistent with evidence that corporate lending adjusts more strongly than household mortgage lending when banks face tighter constraints (eg, Noss and Taffano (2016); Kanngiesser et al (2019)). It is this focus that underpins the ‘corporate lending channel’ through which changes in prudential capital policy can affect the real economy.
Updating the VECM: What the new data say
We update the Vector Error Correction Model (VECM) used by de-Ramon and Straughan with data through 2024, incorporating periods of substantial regulatory reform (eg, Basel II in 2007 and Basel III in 2014), structural change and economic stress (eg, the COVID-19 pandemic). Briefly, our key findings are as follows:
- Overall, long-run relationships remain stable and broadly in line with those reported in the earlier study. A one percentage point permanent increase in the risk-based capital ratio is associated with an increase of around 8 to 10 basis points on lending spreads to Private Non-Financial Corporate (corporate) borrowers, which is consistent with 10 to 12 basis points reported in de-Ramon and Straughan. The long-run relationship between risk-based capital ratios and mortgage spreads remains statistically insignificant, consistent with the 2017 findings. Corporate spreads adjust more quickly toward equilibrium than mortgage spreads, consistent with the idea that banks focus on altering lending that attracts higher regulatory risk weights first when managing towards higher capital ratios.
- Short-run responses differ when COVID observations are included. Including COVID period data introduces short-run volatility, consistent with the extreme economic shock of 2020–21. As a result, and following Lenza and Primiceri (2022), for structural inference we rely primarily on estimations that either exclude observation from the COVID-19 period (our truncated sample) or use the full sample and explicitly account for the period (using time dummies). In both the truncated and full samples, the long-run results are robust and consistent with de-Ramon and Straughan (2017).
Impulse responses: How spreads react to a change in capital policy?
Using the same identification strategy as de-Ramon & Straughan, ie, a Cholesky decomposition, we estimate the response of lending spreads to a prudential capital change that raises risk-based capital ratios by 1 percentage point. Our impulse response functions, shown below with 68% confidence intervals (captured by the shaded bands) indicate that when excluding COVID-19 observations, corporate spreads rise sharply to above 8 basis points in the medium-term and then gradually settle between 7 to 8 basis points (Chart 1). Mortgage spreads also increase, though more modestly. This increase, however, is not statistically significantly different from zero in the long run (Chart 2). The capital ratio increases by more than 1 percentage point in the short run as banks adjust capital and balance sheets in response to increased requirements and then settles down to around 1 percentage point higher in the long run (Chart 3).
Chart 1: Corporate Spread
Chart 2: Mortgage Spread
Chart 3: Risk-based Capital Ratio
Footnotes
- Note: The solid line shows the central estimated response to a change in capital policy; the shaded area shows uncertainty (68% confidence interval).
Extending the model to include updated measures of competition
We expand the original model by incorporating additional measures of banking sector competition, including the Lerner index (mean and median), alongside the Boone Indicator and Herfindahl–Hirschman index. Stationarity, lag selection and cointegration tests all support maintaining two long-run equilibrium relationships. Weak exogeneity tests show that competition indices and insolvency rates, as used in de-Ramon and Straughan, can be treated as exogenous, allowing for a more efficient specification. The inclusion of the Lerner index does not materially alter our central results: the pass-through remains concentrated in corporate lending spreads.
What does all this mean for the economic cost of capital?
The evidence continues to point to modest, but nonzero economic costs, concentrated in corporate lending. Corporate spreads rise by 8 to 12 basis points per additional percentage point of capital. Mortgage spreads show no statistically significant long-run response. Because corporate borrowing is closely linked to investment, the short run effects of higher capital requirements on output operate through aggregate demand, while their persistence is consistent with a supply side effect in the long run.
Conclusion
Our updated analysis using more than 30 years of data and richer measures of competition reinforces the core findings of de-Ramon and Straughan (2017). The corporate lending channel remains the main pathway through which prudential capital requirements affect the real economy. While the UK banking system has evolved significantly since the original study, the fundamental economics of capital pass-through appear largely unchanged. The estimated pass-through range of 7 to 12 basis points per 1 percentage point increase in bank capital ratios remains a stable, evidence-based benchmark for assessing the economic cost of capital in the Prudential Regulation Authority’s approach to cost benefit analysis. More importantly, the update shows that the PRA’s approach to weighing the costs and benefits of capital requirements remains suitable, supporting its judgments about the appropriateness of banks’ capital levels and risk weights.
Share your thoughts with us at Bankinsights@bankofengland.co.uk.