The effect of bank capital requirements on the real economy and their interaction with monetary policy
The effect of bank capital requirements on the real economy and their interaction with monetary policy
Gabriele Cozzi, Matthieu Darracq Pariès, Peter Karadi, Jenny Körner, Christoffer Kok, Falk Mazelis, Kalin Nikolov, Elena Rancoita, Alejandro Van der Ghote, Julien Weber 03 March 2020
Since the financial crisis, central banks and regulatory authorities have been given new powers to set macroprudential bank capital requirements (BIS 2018: Chapter IV, Cerruti et al. 2019). It is therefore important to understand how these instruments affect the economy at large. In a new paper (Cozzi et al. 2020), we lay out the implications of a number of important macro models at the ECB for the real impact of capital requirements as well as their interactions with monetary policy. This work is part of a broader research effort by the ECB to analyse the interactions between monetary policy and macroprudential policy. A companion paper (Albertazzi et al. 2020) takes an empirical perspective and reviews the evidence linking monetary policy and bank stability.
The real impact of capital requirements: A model comparison exercise
We start by comparing how a 1% increase in the minimum bank capital requirement affects real activity in four models frequently used for policy analysis at the ECB (Figure 1). The models are referred to with the acronyms DJP, DKR, NAWM II, and 3D.1
Figure 1 GDP and inflation evolution in response to a 1% increase in the minimum bank capital requirement in the 3D, DJP, DKR, and NAWM II models
The conduct of monetary policy also matters. The stronger that monetary policy leans against the negative effects of a capital-requirements increase on aggregate demand, the smaller the impact on output and inflation over the transition. Yet again, this suggests that the costs of imposing higher capital requirements may be elevated more when monetary policy is constrained by a binding effective lower bound on nominal interest rates.
The comparative model exercise reveals considerable similarities across the different analytical frameworks in the way the economy responds to an increase in bank capital requirements. All models incorporate a strong link between bank capital and credit supply, and all generate some decline in lending on impact as banks restrict credit supply. This reduces aggregate demand and leads to a modest but significant fall in GDP (peak decline of 0.15-0.35%) despite the reaction of the monetary authority, which lowers nominal interest rates.
Several interesting differences in both the short- and long-run behaviour of the different models can be seen in Figure 1. In the long run, the probability of bank default plays an important role; this explains why output fully recovers in the 3D model, in which bank default plays an important role. Making banks safer using higher capital requirements has a positive effect on economic activity in the long run. In the short run, the impact on credit supply is considerably smoothed when banks reduce their voluntary buffers (as in the DJP and DKR models) or cut dividend pay-outs (in the 3D model). To the extent that the ability of banks to either reduce their leverage gradually or access the equity market may depend on the state of the economy, our results suggest that the real impact of capital measures may vary over time.
In an inter-dependent world, it is also important to analyse potential spillovers to other economies when one country increases its bank capital requirements. This is especially critical in a monetary union where monetary policy is decided centrally but macroprudential measures are taken at the country level. Simulations with a two-country version of the DKR model show that capital measures in one economy create spillovers through the operation of the single monetary policy and through financial and trade linkages. Increasing capital requirements depresses GDP growth and inflation in the affected country and, if the country is large enough, in the entire monetary union. Monetary policy reacts with a reduction in nominal interest rates, boosting economic activity in the rest of the monetary union. Financial linkages work through the cross-border lending of large banks. Tighter capital requirements lead to lending restrictions not only at home but also in foreign markets. Finally, lower aggregate demand in the affected country reduces the export demand for all other members of the monetary union.
The impact of capital requirements on the monetary transmission mechanism
We also examine how macroprudential policy affects the monetary transmission mechanism. A common lesson from all the contributions is that bank leverage and the default risk of the banking system have a significant impact on the way monetary and other shocks are transmitted to the real economy.
Simulations with the Gertler and Karadi (2011) model as well as work by Darracq-Paries et al. (2019) and Mazelis (2016) show that a high level of capital requirements and a safe banking system make the economy less responsive to conventional and unconventional monetary policy actions. Banks are exposed to economic shocks (including those resulting from monetary policy); their net worth becomes more sensitive to these shocks when leverage is high and assets are risky. Positive (and negative) shocks are then amplified because they also raise (or lower) bank net worth and expand (or contract) credit supply. High capital requirements stabilise banks’ net worth, weakening the bank capital-related amplification mechanism. This makes the economy less responsive to short term interest rates or unconventional measures such as the asset purchase programme (APP).
Because monetary policy is less powerful when bank leverage is low (and capital requirements are high), the economy is not necessarily less responsive to non-monetary (e.g. demand) shocks. Even though the shocks themselves have a smaller impact with better-capitalised banks, monetary policy has a smaller impact, leaving the net effect from the impact of non-monetary shocks broadly unchanged. However, if monetary policy becomes constrained by the effective lower bound on nominal interest rates, the impact of exogenous shocks will depend again on the leverage of the banking system, and a safer banking sector will reduce output volatility.
How can monetary and macroprudential policies work together to reduce the frequency and severity of financial crises?
We also study how monetary and cyclical macroprudential policies should interact in a model of endogenous financial crises due to Van der Ghote (2018). In this model, crises occur due to the strong two-way feedbacks between banks’ net worth and asset prices when financial intermediaries experience losses which tighten their capital constraints. When banks lose net worth and are forced to deleverage, they ‘fire-sell’ securities at depressed prices and cut their lending to the real economy, leading to a downward spiral of lower asset prices and lower bank net worth. The result is a financial crisis characterised by a deep recession and a significant reduction in financial intermediation.
In order to reduce the probability and severity of financial crises, macroprudential measures are activated in order to lean against excessive leverage during the boom. They are released following crises in order to support lending. Monetary policy in the model is also used to lean against imbalances and speed up the economy’s recovery from a crisis. This implies that real interest rates are somewhat higher during booms and somewhat lower during deep financial recessions than a pure focus on inflation stabilisation would call for. The potential role of monetary policy in stabilising the financial cycle is relatively minor in the model if macroprudential measures work well. If, in practice, macroprudential policy is less effective in delivering financial stability (due for example to the presence of a large shadow banking sector), then this minor role of monetary policy may become relatively more relevant. Still, this needs to be weighed against the cost of increased inflation variability.
Finally, we argue that to the extent that macroprudential policy is effective at moderating credit booms and busts, this should reduce the need for negative nominal interest rates. During boom times, macroprudential measures restrict lending and cool down the economy, engendering lower interest rates compared to a world without macroprudential regulation.
Our analysis carries a key lesson for policymakers: the health of banks and the nature of macroprudential policy matter a lot for monetary policy transmission. A weaker banking system amplifies the impact of monetary policy but also contributes to economic instability and the need for ultra-loose monetary policy in the aftermath of crises. Macroprudential measures that put banks on a sounder financial footing and tame the financial cycle bring an additional benefit by eliminating periods when the economy needs deeply negative real interest rates.
However, as our model comparison reveals, activating a macroprudential capital measure carries short-term costs to economic activity. These costs could be larger when banks deleverage quickly while paying dividends to shareholders, and when monetary policy cannot respond due to a binding effective lower bound on nominal interest rates, which underscores the need for truly countercyclical macroprudential policy. Regulators should build banks’ capital buffers during good times, when banks can issue new equity and when looser monetary policy can mitigate the costs of increasing capital requirements.
Authors’ note: This column represents the views of the authors and not those of the ECB or the Eurosystem.
Albertazzi, U, F Barbiero, D Marques-Ibanez, A Popov, C Rodriguez D´Acri, T Vlassopoulos (2020), “Monetary Policy and Bank Stability: The Toolbox Reviewed”, ECB Working Paper No. 2377
BIS (2018), Annual Economic Report 2018.
Cerruti, E, S Claessens and L Laeven (2019), “The use and effectiveness of macroprudential policies”, BIS Papers No. 86
Christoffel, K, G Coenen and A Warne (2008), “The New Area-Wide Model of the euro area: A micro-founded open-economy model for forecasting and policy analysis”, ECB Working Paper No. 944.
Coenen, G, P Karadi, S Schmidt and A Warne (2018), “The New Area-Wide Model II: an extended version of the ECB’s micro-founded model for forecasting and policy analysis with a financial sector”, ECB Working Paper No. 2200.
Cozzi, G, M Darracq-Paries, C Kok, J Koerner, F Mazelis, K Nikolov, E Rancoita, A Van der Ghote and J Weber (2020), “Macroprudential policy measures: macroeconomic impact and interactions with monetary policy”, ECB Working Paper No. 2376
Darracq Pariès, M, C Kok-Sorensen and D Rodriguez-Palenzuela (2011), “Macroeconomic Propagation under Different Regulatory Regimes: Evidence from an Estimated DSGE Model for the Euro Area”, International Journal of Central Banking 7(4): 49–113.
Darracq Pariès, M, P Jacquinot and N Papadopoulou (2016), “Parsing financial fragmentation in the euro area: a multi-country DSGE perspective”, ECB Working Paper No. 1891.
Darracq Pariès, M, J Koerner and N Papadopoulou (2019), “Empowering Central Bank Asset Purchases: The Role of Financial Policies”, ECB Working Paper No 2237.
Gertler, M and P Karadi (2011), “A Model of Unconventional Monetary Policy”, Journal of Monetary Economics 58(1): 17–34.
Mazelis, F (2016), “Non-Bank Financial Institutions and Monetary Policy”, Working Paper.
Mendicino, C, K Nikolov, J Suarez and D Supera (2018), “Optimal Dynamic Capital Requirements”, Journal of Money, Credit and Banking 50(6): 1271–1297.
Van der Ghote, A, (2018), “Coordinating Monetary and Financial Regulatory Policies”, ECB Working Paper 2155.
1 DKR: Darracq-Paries, Kok-Sorensen and Rodriguez-Palenzuela (2011) model; DJP: Darracq-Paries, Jacquinot and Papadopoulou (2016) model; 3D: Mendicino, Nikolov, Suarez and Supera (2018a) model (so-called because it models the defaults of banks, firms and households); NAWM II: Coennen, Karadi, Schmidt and Warne (2018) model (NAWM stands for new area-wide model with a financial sector, the version extends and re-estimates the original 2008 NAWM model.