Given the secular decline of real interest rates around the world, negative policy rates are likely to become a more prevalent feature of the financial landscape. Hence, whether and how negative rates affect the transmission of monetary policy is a pertinent question for policy-makers. My paper attempts to address this question by examining the impact of the European Central Bank’s (ECB’s) negative policy rates on the lending behaviour of euro area banks.

The relative exposure of banks to negative rates can be differentiated by their reliance on retail deposit funding. In normal times, market rates and retail deposit rates decrease when the policy rate decreases. However, banks are reluctant to pass on negative rates to their retail depositors for fear that their customers will simply withdraw and hold their savings as cash. Hence, high deposit banks (banks that are heavily reliant on retail deposit funding) are more affected by negative rates than low deposit banks because they no longer benefit from a lower cost of funding. In contrast, low deposit (banks which are more dependent on non-deposit funding) still benefit since the pass-through to market rates is relatively unimpeded. Using standard panel econometric techniques, I exploit differences in bank deposit ratios in order to measure the impact of negative policy rates on bank lending behaviour. I also conduct a variety of tests to check that my results are robust and not driven by potential confounding factors such as the ECB’s public sector purchase programme or changes in the regulatory environment.

I find that in the 18 months following the ECB’s first foray into negative rates in June 2014, banks that were more affected by negative rates increased their supply of credit relative to banks that were less affected. The increase in credit was limited to lending to mortgage lending rather than lending to non-financial corporations (NFCs). While the effects on lending volumes are highly significant, I find only limited evidence of a price effect—that is, some banks reduced their lending spreads in response to negative rates. On bank profitability, I find that the net interest margins of high deposit banks were unaffected relative to low deposit banks. This suggests that the positive effect on lending is large enough to offset the adverse cost of funding impact.

Motivated by the main results, I perform two small extensions. First, I uncover an important role for bank capitalisation in banks’ response to negative rates. Whilst I cannot detect significant effects on net interest margins for high deposit banks relative to low deposit banks in the overall sample, I do find significant negative effects when the sample is restricted to banks with relatively high capitalisation. In contrast, the insignificant results remain in the sample of banks with relatively low capitalisation. These findings can be explained by differences in shock absorbency. Highly capitalised have more capacity to absorb shocks, but less capitalised banks need to find other avenues to maintain their profitability in order to prevent erosion of relatively thinner capital cushions.

In the second extension, I extrapolate the time period under investigation to the end of 2018. This opens the econometric specification up to greater threat from confounding factors and incorrect identification of high and low deposit banks. The advantage is that it allows for analysis of how bank lending behaviour changes over time in an environment of persistently negative rates. I find that the positive effect on lending dissipates as the negative rate environment persists, such that by the end of 2018, there are no significant differences in lending between high and low deposit banks. This suggests that the strategy of increasing lending volumes may not have been sustainable. This does not necessarily mean that banks are completely out of options—bank fees for both high and low deposit banks have been rising recently, which is at least suggestive that banks are not yet completely stymied in how they respond to negative rates. Importantly, there is (still) no evidence of contractionary effects due to negative rates.

Overall, the results are somewhat counter to prevailing theories that hold that negative rates could have adverse effects on bank lending volumes. The most well-known of these is the theory of the ‘reversal rate’ by Brunnermeier and Koby (2019), which proposes a time-varying, state-dependent rate below which further rate cuts induce contractionary rather than expansionary effects. The reversal rate is breached when the adverse effects of negative rates on profitability are such that banks hit their capital constraints, resulting in a reduction in credit supply. Viewed through this framework, my results would indicate that the reversal rate was not breached. Moreover, they hint at something missing in the heretofore discussion—rather than take the hit on profitability ‘on the chin’, banks seek other avenues to maintain their profit margins. From a theoretical perspective, this has thorny and uncomfortable connotations—for example, that banks were not behaving optimally to begin with. On the other hand, given the heightened level of uncertainty and risk aversion at the time, this is perhaps not an implausible assertion. Importantly, the adverse effects of negative rates may become more evident as banks become more limited in their options to maintain profit margins.

The findings of this paper have several policy implications. First, no contractionary effects on lending were detected, which implies that as at the end of 2018, there was still space to cut rates further (the ECB has since cut the deposit facility rate by 10 basis points to -0.5%). Second, negative rates have heterogeneous impacts. Most prominently, high deposit banks are more sensitive to the negative rate environment than low deposit banks. This implies additional heterogeneity in monetary transmission through banks when policy rates are negative. Third, negative rates could affect risk in the financial system in different ways. For example, I find (limited) evidence that some banks respond to negative rates by lowering lending terms. Another potential concern is that the positive effects on lending are limited to the mortgage sector, which may point to a lack of productive investment opportunities in the NFC sector. Rapid growth in mortgage lending could also fuel asset price inflation and the procyclicality of the financial cycle, which would have adverse implications for financial stability. 

Garyn Tan is a researcher at De Nederlandsche Bank.