20 May 2019

ECB: Negative interest rates, excess liquidity and retail deposits: banks’ reaction to unconventional monetary policy in the euro area

Negative monetary policy rates are associated with a particular friction because the remuneration of retail deposits tends to be floored at zero. Authors investigate whether this friction affects banks’ reactions when the policy rate is lowered to negative levels, compared to a standard rate cut in the euro area.

The existing theoretical and empirical literature on banks’ role in the monetary policy transmission mechanism is inconclusive on bank reactions to changes in policy rates when these changes take place in negative territory.

Using confidential bank-level data for the euro area, covering a representative share of total euro area bank loans and a novel identification for the effect of negative interest rate policy (NIRP) on banks’ balance sheets, authors approach this question empirically.

They jointly consider banks’ exposure to the charge on excess liquidity (EL) and their reliance on retail deposit funding, as an essential identification mechanism for the impact of NIRP on banks. They find evidence that banks indeed operate differently under NIRP.

Banks that are highly exposed to NIRP (i.e. funded by large amounts of retail deposits and are holders of EL) extend significantly more loans to the NFPS during NIRP compared to the pre-NIRP period. These results suggest that NIRP has acted as an empowerment to the ECB’s large-scale asset purchases that were also launched during this period and injected large amounts of EL into the banking system.

The charge on EL appears to encourage banks to take action to mitigate it, thereby catalysing more active portfolio rebalancing.

Their results are coherent with results in the literature on the impact of NIRP in that they do find evidence of higher risk taking by banks, as risk-free EL is converted into loans. However, in contrast to some of these contributions, authors find that high retail deposit banks increase their lending during NIRP. The difference in the results stems from their use of a broader dataset for bank loans and the incorporation of the role of EL during NIRP.

Working paper

© ECB - European Central Bank