Higher Bank Capital Requirements and Mortgage Pricing: Evidence from the Counter-Cyclical Capital Buffer
Christoph Basten
Review of Finance, Volume 24, Issue 2, March 2020, Pages 453–495, https://doi.org/10.1093/rof/rfz009
Basel III introduced the macroprudential tool Counter-Cyclical Capital Buffer (CCyB) as capital requirements that policy-makers can temporarily increase when lending appears excessive. It is intended firstly to strengthen bank resilience when credit risks are increasing, and secondly to slow down lending then by increasing its costs. As globally first country, Switzerland activated the CCyB in 2013 focused on residential mortgages.
We evaluate its effects with a unique dataset that contains responses from multiple banks to each household mortgage application. This allows us to keep the borrower side fixed and so rule out endogenous selection of different borrowers to different banks. Doing so has become common in analyses on corporate credit but has so far not been achieved for lending to households.
We exploit that banks were differentially CCyB affected along two dimensions. First, the CCyB increased capital requirements for mortgages but not for other loans. It did so not only for mortgages granted henceforth, but also for those already on bank balance sheets. As increased costs could not be passed on to existing borrowers, banks with larger pre-existing mortgage intensity had to recoup larger costs elsewhere. Second, banks with smaller cushions between their pre- existing capital ratios and regulatory requirements were brought closer to violating regulation and triggering supervisory intervention. So they had to increase their capital, reduce risk-weighted mortgages, or both. This allows us to identify the effects of the CCyB on banks’ mortgage offers with a Difference-in-Difference estimation, comparing how mortgage offers change from the pre- to the post-activation period for more vs. less CCyB affected banks.
We find no evidence of explicit rationing. But banks with higher mortgage specialization or lower capital cushions raise prices by an extra eight basis points. To the extent to which these differential price increases cause fewer offers to be accepted, they should shift new mortgage origination from more to less mortgage exposed banks. To the extent to which banks with higher price increases still have their offers accepted, they can earn larger margins, thus increase retained earnings, and thus rebuild their capital cushions. In bank level data we find that the differential price increases do indeed translate into both lower mortgage growth and higher capital increases, both of which contribute to rebuilding capital cushions.
The bank level results show that our response level results are not specific to the online context but carry over to the population of all Swiss commercial banks. While effects vary with banks’ mortgage exposures and capitalization, they are not found to be amplified for applications with higher loan-to-value ratios. The link between LTV ratios and mortgage risk weights in the Swiss Standardized Approach to capital requirements appears too weak. Overall our results show financial stability benefits of the CCyB that remain hidden in more aggregate data.