|
|
They also rebalanced portfolios towards safer borrowers. However, surprising outcomes surface for banks in the lowest tail of capitalisation. The regulatory impact on lending eventually affects firms’ investment decisions.
Regulators and supervisors impose requirements on banks to guide their behaviour. Still, the effect of higher capital requirements on lending is debated, both theoretically and empirically. The issue is policy relevant, as shown by the attention that international bodies, including the Financial Stability Board, have devoted to this topic (Financial Stability Board 2018, 2019). Acharya et al. (2021) also suggest that regulators’ aggressive loosening can end up directing credit to suboptimal targets.
In the long run, the benefits of capital strength are hardly contended: large capital buffers enhance bank resilience, including against black swans (i.e. unforeseeable events) (Mendicino et al. 2021), and this should secure more stable lending flows.
By contrast, in the short run, some scholars suggest that stricter risk-based capital requirements increase banks’ funding costs; this makes lending less attractive, at least during a transition phase (e.g. Aiyar et al. 2014, Aiyar et al. 2014, Acharya et al. 2018, Gropp et al. 2019). Others argue that an increase in capital requirements might, under some conditions, reduce average banks’ funding costs and thus create the conditions to increase bank lending (Begenau 2020, Admati et al. 2013, Bassett and Berrospide 2018).
Finally, to complete the options on the shelf, recent papers open the door to a half-way conclusion: capital requirements depress lending up to a certain bindingness, but this impact becomes comparatively milder for banks that get particularly constrained; it could even be reverted for some lenders, for which stricter rules would eventually produce an increase in lending (Bahaj and Malherbe 2020).
To shed light on the issue, one can exploit quasi-natural experiments, coupled with granular data on firms, banks, and their relationships (Galardo and Vacca 2022). To do this, we need a few ingredients. First, an exogenous shock should break the timeline in a period before and a period after the shock. Second, it should be possible to separate the sample of banks into two subsamples, in order to compare their after-shock reactions: more affected banks versus less or unaffected banks...
more at CEPR