VoxEU Column EU policies Financial Regulation and Banking

Waterbed effects of macroprudential policies: Evidence on cross-sector substitution

Macroprudential policies are being implemented around the globe. A key question is whether these policies prompt substitution toward the non-bank financial sector. This column presents compelling evidence of such ‘waterbed effects’ after macroprudential policy action. Substitution towards non-bank credit is stronger when policy measures applied to banks are binding and are implemented in countries with well-developed financial markets. While systemic risks may nonetheless decline, waterbed effects highlight the importance of developing macroprudential policies beyond banking. 

In the wake of the Global Crisis, macroprudential policies have come of age. Both advanced and emerging market economies have made substantial progress in implementing macroprudential frameworks (BIS 2018). Reflecting this, there is evidence of the growing use of macroprudential instruments worldwide (IMF 2013, 2018). These instruments seek to raise the resilience of the financial system and to dampen the amplitude of the financial cycle. Recently, calls for active macroprudential policies have emphasised their contribution to safeguarding financial stability in circumstances of prolonged easy monetary conditions (Praet 2018).

The implementation of macroprudential policies has, however, raised concerns about regulatory arbitrage. As noted by Goodhart (2008), “the more effective regulation is, the greater the incentive to find ways around it”. In this context, the implementation of macroprudential policy measures has been found to prompt circumvention through foreign banks (e.g. Aiyar et al. 2014, Frost et al. 2017, Cerutti and Zhou 2018). In turn, this has led to the development of reciprocation arrangements within the EU (ESRB 2015). Such arrangements may be expected to limit cross-border leakages. But there is another, perhaps greater concern that macroprudential policies stimulate arbitrage within the domestic financial sector and that this cross-sector arbitrage undermines the policies’ effectiveness. 

Cross-sector leakages

Most macroprudential measures only relate to the banking sector. This includes measures applied to lenders, such as countercyclical capital buffers and capital surcharges, as well as measures applied to borrowers, such as loan-to-value (LTV) and debt-to-income (DTI) caps. Research has established that these instruments have the intended effect on bank credit (Cerutti et al. 2017a). But what about leakages to non-bank credit? Such substitution effects across sectors have not yet been tested empirically in an international setting. 

To fill this gap, we have estimated the impact of macroprudential policies on non-bank credit to the non-financial private sector, using quarterly data from 37 countries over 1997-2014 (Cizel et al. 2019). To address endogeneity concerns, we apply three different methodologies – detrending, generalised method of moments (GMM) panel regressions, and propensity score matching (PSM) – to two different cross-country datasets on (macro)prudential policy measures. We also test whether the results differ between advanced economies and emerging market economies, given the differences in the degree of financial development and especially in the depth of non-bank credit markets.

Our results consistently indicate substantial substitution effects after macroprudential measures are activated. Waterbed effects are largest when policy constraints are binding on banks and when countries have well-developed non-bank credit markets, as in most advanced economies. Specifically, both using the simple detrending method and comparing similar advanced countries with and without macroprudential measures (PSM), we find that non-bank credit growth is roughly 11–14% higher than the baseline in the two years following the activation of macroprudential instruments (Figure 1). Our panel regressions, which control for the impact of other variables, also point to stronger non-bank credit growth after macroprudential measures are activated, but only in advanced economies when quantitative policy measures are binding. Additional robustness checks involving different data definitions and frequencies, and distinguishing between pre-crisis and post-crisis periods, confirm these findings.

Figure 1 Non-bank credit growth around macroprudential policy measures, PSM method

Source: Cizel et al. (2019) 
Note: The figure shows the effects of macroprudential policy events on the average cumulative credit growth rates during the period before and after the activation of macroprudential policies. The CCL data set refers to Cerutti et al. (2017a). CCFS refers to Cerutti et al. (2017b). The actual post-event growth rates are adjusted by the linearly extrapolated growth rates from the 2-year pre-event period. * p < .1, ** p < .05, *** p < .01. Standard errors in parentheses. 


Our results have implications for macroprudential policymaking. Overall, it may be argued that these waterbed effects reduce systemic risks. After all, non-bank finance is generally less leveraged and involves fewer liquidity mismatches than banking finance. Non-bank finance is also separated from systemic functions related to the payments infrastructure. Non-bank financial institutions generally do not have access to public safety nets, such as deposit insurance and lender-of-last-resort support, reducing any concerns about moral hazard. Indeed, Bats and Houben (2017) provide evidence that a greater share of market-based finance increases resilience to systemic risk. From this perspective, the substitution effects may be seen as reducing systemic risks. These considerations underscore the importance of reinvigorating initiatives such as the European Capital Markets Union (European Commission 2015).

But there is another, less comforting consideration. When credit growth shifts away from banks, but households and corporates continue to accumulate debt, the vulnerabilities that the macroprudential policies seek to address continue to rise. Excessive debt may eventually prompt a crisis, even if the debt is owed to investment funds or other capital market players. While contagion effects may be smaller than in the case of banking crises, macroeconomic costs may still be large on account of adverse wealth and confidence effects. Moreover, financing outside the banking system is typically subject to fewer reporting and regulatory requirements, making this harder to supervise. New, emerging risks are more difficult to monitor. These concerns highlight the need to develop macroprudential policy instruments beyond banking and to implement reporting requirements that fully capture non-bank credit activities. In all, waterbed effects should be taken into account in the design of macroprudential policy. 

Authors’ note: The views expressed here are those of the authors and not necessarily of De Nederlandsche Bank (DNB), the Bank for International Settlements (BIS), or any other organisation that the authors are affiliated with.


Aiyar, S, C W Calomiris, and T Wieladek (2014), “Does Macro-Prudential Regulation Leak? Evidence from a UK Policy Experiment”, Journal of Money, Credit and Banking, 46, 181–214.

Bats, J, and A Houben (2017), “Bank-based versus market-base financing: Implications for systemic risk”, DNP Working Paper no. 577. 

Bank for International Settlements (2018), Moving forward with macroprudential frameworks, BIS Annual Report, June.

Cerutti, E, and H Zhou (2018), “Cross-border Banking and the Circumvention of Macroprudential and Capital Control Measures”, IMF Working Paper no. 18/217.

Cerutti, E, S Claessens, and L Laeven (2017a), “The Use and Effectiveness of Macroprudential Policies”, Journal of Financial Stability, 28, 203–24.

Cerutti, E, R Correa, E Fiorentino, and E Segalla (2017b), “Changes in Prudential Policy Instruments – A New Cross-Country Database”, International Journal of Central Banking, 13, 477–503.

Cizel, J, J Frost, A Houben, and P Wierts (2019), “Effective Macroprudential Policy: Cross-Sector Substitution from Price and Quantity Measures”, Journal of Money, Credit and Banking, 51, 1209-1235.

European Commission (2015), “Action Plan on Building a Capital Markets Union”, Brussels.

European Systemic Risk Board (2015), “Recommendation on the assessment of cross-border effects of and voluntary reciprocity for macroprudential policy measures”, Report.

Frost, J, N van Horen, and J de Haan (2017), “International Banking and Cross-Border Effects of Regulation: Lessons from the Netherlands”, International Journal of Central Banking, 13, 293–313.

Goodhart, C (2008), “The Boundary Problem in Financial Regulation”, National Institute Economic Review, 206, 48–55.

International Monetary Fund (2013), “Key Aspects of Macroprudential Policy”, IMF Policy Paper, Washington.

International Monetary Fund (2018), “The IMF’s Annual Macroprudential Policy Survey—Objectives, Design and Policy Responses”, IMF Policy Paper and Note to G20, April, Washington.

Praet, P (2018), “The interaction between monetary policy and macroprudential policy”, speech, London.

1,365 Reads