How do macroprudential policies interact with the core objectives of monetary policy, namely, output and price stability? As a response to the Global Crisis, central banks and regulators across the world have increasingly relied on macroprudential policies to maintain financial stability. A recent literature has shown that policymakers can moderate credit and asset price cycles using macroprudential instruments (e.g. Akinci and Olmstead-Rumsey 2018, Bruno et al. 2017, Kuttner and Shim 2016). This may allow negative output tail risks emanating from the link between excess credit and costly financial crises to be reduced (Schularick and Taylor 2012, Jordà et al. 2013). However, there is very little empirical evidence on how the use of such instruments affects the traditional objectives of monetary policy. This lack of evidence is in part a result of two empirical challenges that make identification and measurement difficult. In new research, we make inroads into addressing these two challenges (Richter et al. 2018).

## Narrative identification and intensity adjustment of loan-to-value policy actions

First, macroprudential measures are not randomly assigned. When we want to assess the output and price effects of macroprudential policies, we need to ensure that these policies were not taken in response to current or expected developments in output and inflation. We therefore focus on one specific tool that is frequently used to tackle boom-bust cycles in credit and housing markets, but rarely to target real economic variables – changes in regulatory loan-to-value (LTV) limits. We rely on a novel, hand-collected dataset detailing the intentions or stated objectives of policymakers when they change LTV limits. In a similar spirit to the narrative identification of monetary policy shocks, we argue that macroprudential actions taken without reference to current or expected trends in output and inflation can be seen as exogenous with respect to the objectives of monetary policy. We compiled a comprehensive new panel dataset consisting of quarterly observations on 89 LTV actions which do not have output or inflation as their stated objectives, building on the database developed by Shim et al. (2013).

Second, measuring the intensity of macroprudential policy actions is difficult and existing databases usually use dummy or indicator variables for loosening (−1), tightening (+1), or no changes (0) in LTV limits. When we use a dummy variable, a decrease in the maximum LTV ratio from 80% to 60% and a decrease from 80% to 75% are treated equally. Clearly, the effects of the two policies might differ and the size of the adjustment matters. We therefore use the size of changes in LTV limits to estimate the effects of a one percentage point change in LTV limits on output and inflation. We also adjust for the scope of loans affected by the LTV policy as changes in LTV limits often apply only to a specific segment of mortgage markets. We can only assign a value to this change if there was previously an LTV limit in place. As a result, our sample is reduced to 53 actions that are both exogenous to the real economy and quantifiable in the above sense. In line with the literature cited above, we assign a positive value to a tightening in LTV limits, for example, +10 for a change from 80% to 70%.

## Effects on output and inflation

Using this new dataset, we employ local projections (Jordà 2005) to estimate impulse responses to a change in LTV limits. We find that a 10 percentage point decrease in the maximum LTV ratio leads to a 1.1% reduction in output after 16 quarters (see Figure 1). A back-of-the-envelope calculation suggests that the two-year impact of such a 10 percentage point LTV tightening can be viewed as roughly comparable to that of a 25 basis point increase in the policy rate. However, the effects are imprecisely estimated and only present in emerging market economies, but not in advanced economies (see Figure 2). We also find that tightening LTV limits has larger output effects than loosening them. The price response is initially slightly positive, but in most specifications even smaller than the GDP response. The response of consumption is similar to that of GDP.

To rule out the possibility that our results are driven by LTV actions taken after the Global Crisis, we only consider LTV changes implemented until 2006 and obtain very similar results. Finally, to ensure that our results are not driven by a single country, we drop from the sample China, Hong Kong SAR, and Iceland one by one, which have used LTV policies most frequently, and obtain consistent results.

**Figure 1** Output and inflation effects of macroprudential policy

*Notes*: The blue lines display the coefficients of cumulative responses of real GDP and the price level over the 16 quarters following a 1 percentage point decrease in maximum LTV ratios. Shaded areas refer to 1 standard deviation (dark) and 1.96 standard deviations.

**Figure 2** Output effects in subsamples

*Notes*: The blue lines display the coefficients of cumulative responses of real GDP over the 16 quarters following a 1 percentage point decrease in maximum LTV ratios. Shaded areas refer to 1 standard deviation (dark) and 1.96 standard deviations.

## Effects on the financial cycle

These results show that the effects of LTV changes on output and inflation seem rather small. Next, we ask whether changes in LTV limits achieve the desired goal of moderating financial cycles. In order to address this question, we cannot apply our previous approach; most LTV changes are clearly implemented as a reaction to developments in credit and housing markets. We therefore use a two-step inverse propensity weighting strategy to re-randomise LTV actions. In the first stage, we estimate the probability of a treatment, in our case, a tightening in LTV limits (we use a dummy variable here). In the second-stage regression, observations are weighted inversely to the estimated probability of receiving treatment, thus giving a greater weight to those actions that come closer to the random allocation ideal.

Using this procedure, we indeed find that tightening LTV limits reduces credit and asset price growth. In particular, household credit and mortgage credit are significantly reduced by around 5% after 16 quarters (Figure 3), while house prices fall significantly by 8% over the four years (Figure 4). Tightening LTV limits also has a negative effect on stock prices which remains however insignificant.

**Figure 3 **Credit variables

*Notes*: The blue lines display the coefficients of cumulative responses of real household credit and real mortgage credit over the 16 quarters following a tightening in maximum LTV ratios. Shaded areas refer to 1 standard deviation (dark) and 1.96 standard deviations.

**Figure 4 **Asset prices

*Notes*: The blue lines display the coefficients of cumulative responses of real stock price indices and real house prices over the 16 quarters following a tightening in maximum LTV ratios. Shaded areas refer to 1 standard deviation (dark) and 1.96 standard deviations.

## Policy implications

Combining the above results, we conclude that central banks could be in a position to use macroprudential instruments to manage financial booms without affecting monetary policy objectives in a major way. Furthermore, our use of a scope-adjusted quantified LTV change variable may provide guidance to policymakers on how to calibrate macroprudential tools. Finally, the evidence demonstrates that changes in maximum LTV ratios introduced under financial objectives tend to have rather substantial effects on credit and house prices as intended.

## References

Akinci, O, and J Olmstead-Rumsey (2018), “How Effective are Macroprudential Policies? An Empirical Investigation”, *Journal of Financial Intermediation* 33: 33–57.

Bruno, V, I Shim, Ilhyock and H S Shin (2017), “Comparative Assessment of Macroprudential Policies”, *Journal of Financial Stability* 28: 183–202.

Kuttner, K N and I Shim (2016), “Can Non-interest Rate Policies Stabilise Housing Markets? Evidence from a Panel of 57 Economies”, *Journal of Financial Stability* 26: 31–44.

Jordà, O (2005), “Estimation and Inference of Impulse Responses by Local Projections”, *American Economic Review *95(1): 161–82.

Jordà, O, M Schularick and A M Taylor (2013), “When Credit Bites Back”, *Journal of Money, Credit and Banking* 45(2): 3–28.

Richter, B, M Schularick and I Shim (2018), “The Costs of Macroprudential Policy”, CEPR Discussion Paper 13124.

Schularick, M and A M Taylor (2012), “Credit Booms Gone Bust: Monetary Policy, Leverage Cycles, and Financial Crises, 1870–2008”, *American Economic Review* 102(2): 1029–61.

Shim, I, B Bogdanova, J Shek and A Subelyte (2013), “A Database for Policy Actions on Housing Markets”, *BIS Quarterly Review*, September: 83–95.