German banks have long grumbled about the squeeze to their net interest margins as a result of the European Central Bank’s (ECB’s) policy of paying a negative interest rates on reserves held at the ECB (banks that deposit at the ECB now pay 0.4 per cent on those deposits, rather than earn a positive interest rate as they would typically do in normal times). As a result, banks warn, their profitability will fall, likely curbing credit creation and hampering the economic recovery. Yet, micro-empirical evidence suggests that bank profits and banks’ balance sheet health has improved since the implementation of negative interest rates and that lending rates have fallen. What explains the difference?

In our paper (mimeo), we build a model that is consistent with both the concerns of bank CEOs and the empirical evidence. The model is thus able to explain, in a causal sense, the channels through which negative interest rate policy works and the costs and benefits of such a policy.

The model we built is consistent with a squeeze on net interest margins. Empirical evidence shows that retail banks have been reluctant to pass on negative central bank interest rates to customers, choosing instead to leave households earning at least zero per cent on their deposits. The fact that banks haven’t passed these negative interest rates on to their customers, however, means that banks’ net interest margins have shrunk. And, without a fall in deposit rates — the ultimate source of funding for banks — there is no reason why lending rates should fall as well. (A side note is in order here: For the banking system as a whole, reserves are an asset of the banking system and not a funding source.) In fact, as net interest margins shrink, bank profitability falls, and banks are forced to increase lending rates and cut credit to the private sector.

However, this explanation is incomplete because it ignores the role of expectations. Negative interest rates can also boost the value of bank assets. The problem is that for an individual bank CEO, it is not obvious that any rise in asset values is caused by the implementation of negative interest rates, while the fall in net interest margins is. By building a dynamic, general equilibrium macroeconomic model of the economy, we show that the effectiveness of negative interest rates rests on what current policy signals about future policy.

An extremely robust finding in decades of academic monetary policy research is that central banks adjust policy gradually. This implies that if policy rates are lowered today then policy rates will also be lower in the future. And even though deposit rates, already at zero, will not be lowered today by banks, the negative interest rates signal to households that deposit rates will rise more gradually in the future, inducing consumers and households to bring forward spending decisions to today and boost consumption demand.

As a result of households choosing to spend more today, banks experience a rise in the value of their assets. That is, bank profits can rise as a result of capital gains. Via the financial accelerator theory popularised by former Fed chairman Ben Bernanke and co-authors, the rise in bank profits will lower the spread between deposit rates and lending rates. With lending rates lower, firms’ investment demand increases, positively amplifying the effect of negative interest rates on economic activity.

Negative interest rate policy therefore has costs and benefits and the critical question is which effect dominates — the contractionary effect of shrinking net interest margins, or the expansionary effect from the signalling channel. In our paper, we conclude that quantitatively, the signalling channel dominates. However, this results in not universal. We find that negative interest rates are less effective when 1) the zero-lower bound period is expected to last for longer, 2) the banking system has more excess reserves, and 3) when the central bank is not able to signal (i.e. it cannot credibly commit to adjusting future policy gradually).

Not all central banks chose to set negative interest rates in the aftermath of the financial crisis despite many of them requiring additional monetary stimulus beyond standard interest rate cuts. While the ECB, the Bank of Japan, the Swiss National Bank and the Swedish Riksbank all adopted negative interest rates, the US Federal Reserve and the Bank of England notably did not, despite using other unconventional policy instruments such as quantitative easing.

An intriguing open question is whether the three factors identified above that alter the effectiveness of negative interest rates can account for the Fed and Bank of England’s choices.

Our paper ends by assessing the role of negative rates when a central bank can set policy “optimally”. We prove two results. Setting negative rates would either be redundant or harmful in the following scenarios: 1) when a central bank can perfectly and credibly commit to future policy and ii) when it can’t commit to any future actions at all. Therefore, negative rates are only a part of the optimal monetary policy toolkit in the intermediate case when a central bank has imperfect credibility. This intermediate case, however, is a pretty reasonable description of central banks in the real world.

A final point is worth sharing. The signalling channel of negative rates operates in a similar fashion to “forward guidance” policies that many central banks have also adopted. The crucial difference is that the effectiveness of forward guidance relies on a verbal promise (e.g. the central bank will slow down future interest rate hikes) that the public may not believe. Negative rates, however, works on a near universal truth of monetary policy, that when a central bank actually lowers its policy rate, it is usually slow in raising it again.

Authors’ disclaimer: Both authors were employed by the ECB during part of the writing of this paper. However, the views in this post and in the paper are those of the authors and does not necessarily reflect the views of the ECB’s Governing Council or its staff.



  • This blog is based on the authors’ paper “The Signalling Channel of Negative Interest Rates” by Oliver de Groot and Alexander Haas, Mimeo 2019, presented at the Royal Economic Society’s Annual Conference, April 2019, University of Warwick.
  • The post gives the views of its author, not the position of LSE Business Review or the London School of Economics.
  • Featured image by Beesmurf, under a Pixabay licence
  • Before commenting, please read our Comment Policy

Oliver de Groot is a full professor at the University of Liverpool Management School and holds the chair in macroeconomics. He is a macroeconomist with research interests in monetary economics, macro-finance, asset pricing and computational methods. Oliver has held positions at the European Central Bank in Frankfurt, the University of St Andrews in Scotland, and the Board of Governors of the Federal Reserve System in Washington DC. He holds an MA and a PhD from the University of Cambridge. E-mail:

Alexander Haas is a PhD candidate at the German Institute for Economic Research, DIW Berlin. He graduated from the University of Oxford with an MPhil in economics and spent several months as a graduate trainee at the European Central Bank. His research examines monetary policy, financial intermediation, and liquidity since the financial crisis. In fall 2019, Alex will be joining the University of Oxford’s DPhil in economics programme. E-mail: