Effective Macroprudential Policy: Cross-Sector Substitution from Price and Quantity Measures

Abstract

Macroprudential policy is increasingly being implemented worldwide. Its effectiveness in influencing bank credit and its substitution effects beyond banking have been a key subject of discussion. Our empirical analysis confirms the expected effects of macroprudential policies on bank credit, both for advanced economies and emerging market economies. Yet we also find evidence of substitution effects towards non-bank credit, especially in advanced economies, reducing the policies' effect on total credit. Quantity restrictions are particularly potent in constraining bank credit but also cause the strongest substitution effects. Policy implications indicate a need to extend macroprudential policy beyond banking, especially in advanced economies.

Impacts of loan-to-value ceilings in the Brazilian mortgage loans

“…Results obtained through DID corroborated the study by Cizel et al (2016) and Araujo et al (2017), since they show statistically significant effects of the regulation on the quality of the regulated group (SFH) in comparison to the non-regulated one (SFI) in the short run (6 months after treatment). In the long run, the effect vanishes, which may be the result of substitution effect between regulated and non-regulated loans, which influences DID estimates but not ITS, which continues to show LTV significant effects after 12 months.…”

“…Borrowers may have migrated to non-regulated SFI loans with higher interest rates. Cizel et al (2016) investigate the effect of replacing the regulated sector (bank credit) with the non-regulated one (non-bank credit) in an empirical study based on data from 40 countries. According to their study, the regulated mortgage market decreased 10% in the two years after the implementation of macroprudential policies; whereas the non-regulated segment showed increased growth rate of 5%.…”

“…Results show the housing credit growth dynamics was not affected by the LTV enactment in Brazil. Although Lim et al (2011) and Cizel et al (2016) find decreasing mortgage loans after LTV ceiling implementations, this effect is not observed in Brazil. Competing hypothesis can explain this result.…”

The International Aspects of Macroprudential Policy

“…This is unlikely to fully address issues around endogeneity and selection bias. Several papers go further by instrumenting for the policy changes and/or creating some type of exogenous macroprudential policy shock (Cizel et al, 2019;Ahnert et al, 2020). This is more likely to lead to unbiased estimates, but the first-stage is often imprecisely estimated because it is difficult to predict when tools are adjusted.…”

“…One of the best approaches is propensity score matching, which compares the dependent variables in countries that have adjusted policy with the "nearest neighbor" or a synthetic construct of similar countries that do not adjust policy. This approach was first used in and Klein (2014, 2015), with more recent applications in Cizel et al (2019) and Frost et al (2020).…”