Leverage ratio and risk-taking: theory and practice

We assess the impact of the leverage ratio capital requirements on the risk‑taking behaviour of banks both theoretically and empirically. We use a difference‑in‑differences (DiD) setup to compare the behaviour of UK banks subject to the leverage ratio requirements (LR banks) to otherwise similar banks (non‑LR banks). Conceptually, introducing binding leverage ratio requirements into a regulatory framework with risk-based capital requirements induces banks to reoptimise, shifting from safer to riskier assets (higher asset risk). Yet, this shift would not be one‑for‑one due to risk‑weight differences, meaning the shift would be associated with a lower level of leverage (lower insolvency risk). The interaction of these two changes determines the impact on the aggregate level of risk. Empirically, we show that LR banks did not increase asset risk, and slightly reduced leverage levels, compared to the control group after the introduction of leverage ratio in the UK. As expected, these two changes lead to a lower aggregate level of risk. Our results show that credit default swap spreads on the five‑year subordinated debt of LR banks dropped relative to non‑LR banks post leverage ratio introduction.