The Micro Impact of Macroprudential Policies: Firm-Level Evidence
Author: Meghana Ayyagari1, Thorsten Beck2, and Mr. Maria Soledad Martinez Peria3
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund
  • | 2 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund
  • | 3 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

Contributor Notes

Author’s E-Mail Address: ayyagari@gwu.edu; TBeck@city.ac.uk; mmartinezperia@imf.org;
Publication Date:
07 Dec 2018
eISBN:
9781484385654
ISBN:
9781484385654
Language:
English
Keywords:
WP; financing growth; firm assets; micro firm; balance sheet data; financing constraint; firm characteristic; return on assets; firm leverage; bank financing; size classification; firm heterogeneity; coverage ratio
Download PDF (2.1 MB)

Combining balance sheet data on 900,000 firms from 48 countries with information on the adoption of macroprudential policies during 2003-2011, we find that these policies are associated with lower credit growth. These effects are especially significant for micro, small and medium enterprises (MSMEs) and young firms that, according to the literature, are more financially constrained and bank dependent. Among MSMEs and young firms, those with weaker balance sheets exhibit lower credit growth in conjunction with the adoption of macroprudential policies, suggesting that these policies can enhance financial stability. Finally, our results show that macroprudential policies have real effects, as they are associated with lower investment and sales growth.

Abstract

Combining balance sheet data on 900,000 firms from 48 countries with information on the adoption of macroprudential policies during 2003-2011, we find that these policies are associated with lower credit growth. These effects are especially significant for micro, small and medium enterprises (MSMEs) and young firms that, according to the literature, are more financially constrained and bank dependent. Among MSMEs and young firms, those with weaker balance sheets exhibit lower credit growth in conjunction with the adoption of macroprudential policies, suggesting that these policies can enhance financial stability. Finally, our results show that macroprudential policies have real effects, as they are associated with lower investment and sales growth.

1. Introduction

Macroprudential policies have been the focus of increased attention in the post-Global Financial Crisis regulatory reform agenda. The objective of these policies is to increase the resilience of financial institutions and borrowers to aggregate shocks, contain the build-up of systemic vulnerabilities resulting from procyclical feedback between asset prices and credit, excessive leverage, and volatile funding, and control structural vulnerabilities in the financial system (IMF, 2013). Given its stated objectives, macroprudential policies would be expected to curb average firm credit growth and, potentially through the effect on credit, affect firm investment and sales growth. An interesting question, however, is whether macroprudential policies have distributional consequences with their impact varying by type of firm. Specifically, are smaller and younger firms that are often credit constrained and typically more dependent on bank finance than larger firms more affected by macroprudential policies? If so, is the impact uniform across these firms or are those that are financially weaker more affected in line with the notion that macroprudential policies should enhance financial stability? 5

This paper is a first attempt at assessing the effectiveness of macroprudential policies in reducing firm credit and their impact on firms’ investment and sales growth. In addition, we explore the distributional consequences of macroprudential policies by examining how they affect different firms. We combine firm-level data on more than 900,000 firms between 2003 and 2011 in 48 countries with detailed data on the use of macroprudential policy instruments in these countries. The micro data allow us to (i) investigate the effect of these policies on firms of different sizes and ages, (ii) separately examine borrower-targeted and financial institution-targeted instruments,6 (iii) differentiate between growth in short-term and long-term financing, and (iv) examine the behavior of real variables such as firm investment and sales growth.

There are several advantages to using micro data to examine the impact of macroprudential policies. First, using firm-level data and focusing on the differential effects of macroprudential policies across firm groups helps to mitigate endogeneity concerns regarding the adoption of macroprudential policies, as it is harder to argue that credit developments in individual firms or specific firm groups will drive the adoption of aggregate macroprudential policies. Second, by conducting the analysis at the firm-level we can include country-year fixed effects to control for the impact of other macroeconomic developments (e.g., monetary policy) that might also affect firm credit growth.

Our results indicate that macroprudential policies are negatively associated with firm financing growth, but there are heterogenous effects depending on the type of policies and firms. In estimations combining all types of firms, we find that the index of borrower-targeted macroprudential policies is robustly and negatively associated with growth in long-term firm financing, while policies targeted at financial institutions do not appear to be significantly correlated with firm financing growth. This is consistent with the argument that avoidance or leakage (i.e., a situation where credit activity migrates to institutions that are not covered by the macroprudential instruments) is easier when policies target institutions rather than borrowers.7

We also find differential effects of macroprudential policies based on firm size and age. Financing growth for MSMEs (firms with fewer than 250 employees) and young firms (those three or less years of age) is more negatively correlated with macroprudential policies than for larger and older firms. This could be driven by the fact that MSMEs and young firms are typically opaque and dependent on bank relationship lending (Petersen and Rajan, 1994; Berger and Udell, 1995; Beck, Demirguc-Kunt and Maksimovic, 2008), whereas in the presence of macroprudential policies that limit bank credit, larger firms can substitute this type of financing with non-bank finance.

These results are robust to controlling for industry demand shocks through industry-time fixed effects, accounting for differences in firm creditworthiness, and allowing for a heterogenous impact of other macroeconomic policies (such as monetary policy) across firm size and age. We also confirm the results for borrower targeted measures, specifically loan-to-value ratio caps, using recent data on the intensity of macroprudential tools (Cerutti et al., 2017), where we show that the long-term financing growth of MSMEs is negatively associated with cumulative changes in the intensity of loan-to-value ratio caps.8

Given the above results, an interesting question is whether among MSMEs and young firms, those which could pose more severe financial stability concerns, due to poor creditworthiness, are most impacted by macroprudential policies. To address this question, we measure firms’ creditworthiness using the leverage ratio, interest coverage ratio, and an indicator of profitability (return on assets or ROA) and include interactions of these variables with the macroprudential measures. We find that among MSMEs and young firms, the negative association between credit growth and macroprudential policies is stronger for the least credit worthy and riskiest firms, in line with the stability-enhancing goal of such policies.

Finally, we also document important real effects of macroprudential policies on MSMEs and young firms. MSMEs in countries with borrower targeted macroprudential instruments have lower investment and sales growth; young firms in economies with borrower and/or financial institution targeted macroprudential instruments have lower investment and sales growth. Thus, we show that macroprudential policies have an effect not only on financial stability, but also the real economy.

Our paper relates to several strands of literature. First, it relates to the literature that has documented the importance of financing constraints for firm growth and shown that younger and smaller firms are generally more financially constrained and tend to be more dependent on bank financing and relationship lending than older and larger firms. Using either Tobin’s Q model or the Euler equation of investment, an extensive literature has documented the existence of financing constraints, by showing a higher investment-cash flow sensitivity for these firms (e.g., Abel, 1980; Fazzari et al., 2000 are among the earlier studies). While most of this earlier literature has used information on larger, listed firms, a more recent literature using firm-level surveys has shown that smaller firms are more likely to report financing obstacles and are more constrained in their growth by such obstacles (Beck et al. 2005, 2006). 9 At the same time several studies, starting with Petersen and Rajan (1994) and Berger and Udell (1995), document that smaller and younger firms are more dependent on bank relationship lending. Our analysis expands this literature by showing that MSMEs and young firms, known to be more financially constrained and dependent on bank financing than large firms, are more impacted by macroprudential regulations.

Second, our findings also relate to the macro literature on the differential sensitivity of small firms to policy shocks of various kinds. In a seminal paper, Gertler and Gilchrist (1994) present evidence that small firms are more sensitive to monetary policy shocks.10 Chodorow-Reich (2014) examines the impact of credit supply disruptions associated with the Global Financial Crisis and finds bigger effects among small firms. Forbes (2007) shows that taxes on short-term capital flows in Chile increased financing constraints for small, but not for large firms. Lilienfeld-Toal, Mookherjee and Visaria (2012) show that a judicial reform in India had important distributional consequences, resulting in lower (higher) bank financing for small (large) firms. Our paper contributes to this literature by being the first to focus on the distributional effects of macroprudential policies across firms.

Finally, our paper builds on and contributes to a small but rapidly expanding literature on the effects of macroprudential policies across countries.11 Cerutti et al (2015) document the use of various macroprudential policies in 119 countries over the period of 2000–13 and find that macroprudential policies are associated with lower aggregate (country-level) growth in credit. Claessens et al (2013) use balance sheet data of individual banks in 48 countries over 2000–10 to show that borrower-based measures (such as loan-to-value (LTV) and debt-service-to-income (DSTI) caps) along with credit growth and foreign currency lending limits are effective in reducing the growth in banks’ leverage, asset, and non-core to core liabilities. Akinci and Olmstead-Rumsey (2018) record the tightening and easing of macroprudential policies every quarter from 2000 onwards in 57 countries and show that these policies are used in tandem with bank reserve requirements, capital flow management measures, and monetary policy. Lim et al (2011) study a smaller subset of 49 countries and find that macroprudential policies are associated with reductions in the procyclicality of credit and leverage. To our knowledge, our paper is the first to consider the effect of macroprudential policies on firms’ financing growth and real activity across countries, using micro-level evidence to shed light on the distributional effects of such policies across firms.

There are two important data qualifications to consider. First, because we combine data across many countries and an array of different macroprudential tools, we cannot speak to the effectiveness of specific tools. Second, notwithstanding the advantages of using micro-data, there may still be some residual endogeneity concerns, as changes in aggregate debt growth might lead to the adoption of macroprudential tools and the implementation of these policies might result in changes both in the demand for and supply of credit. Focusing on within-firm variation and considering the differential effects across firm groups allows us to address the endogeneity to a certain degree. Hence, we are cautious in drawing causal inferences from our results.

The remainder of the paper is structured as follows. Section 2 discusses the data and the empirical methodology. Section 3 presents the results on the association between macroprudential policies and firm financing and growth. Section 4 concludes.

2. Data and methodology

To investigate the association between macroprudential policies and firm financing and real performance, we combine firm-level balance sheet data with country level information on macroprudential policies. We complement these data with other macroeconomic data. Appendix Table A lists the countries in our sample with the respective number of firms included in the analysis.

We use data from Orbis, a commercial database distributed by Bureau van Dijk containing basic firm-level information including data on external financing for over 900,000 companies across 48 countries over the period of 2003 to 2011. Compared with other databases, the unique advantage of using Orbis is that it includes data on large and small, listed and unlisted firms. We “clean” the data in a number of ways. First, we restrict our analysis to non-financial firms and drop all duplicate observations or double reports for the same firm.12 Second, we only include in our sample countries that have at least 25 firms over the entire period. Third, we drop all firms that were acquirers in an acquisition deal, post-acquisition, or that merged with others following the merger since such transactions can result in sharp changes in firms’ balance sheets. Fourth, we drop observations with zero values for total assets and employees and we remove outliers such as negative values for total assets and employees and listed MSMEs.

As shown in Appendix A, we have a wide variation in the number of firms across countries, ranging from 224,786 firms in France and over 110,000 firms in Italy and Spain, respectively, to fewer than 100 firms in Brazil, Chile, Kazakhstan, Mexico, Netherlands, and Philippines, respectively.13 To address the unbalanced nature of our data, we weight all our estimations with the inverse of the number of firms in each country.

We construct the following financing variables: Short-term financing which is the growth in short-term debt (with residual maturity of less than one year), Long-term financing which is growth in long-term debt (with residual maturity of one year or more) and Overall Financing Growth which is the growth in total financing (defined as the sum of short- and long-term debt), where growth is the annual growth rate, defined as the log-difference of the variable.14 To reduce the impact of outliers, we drop the top and bottom 5% of each of the financing variables. We also drop observations for which we do not have all three variables available to make results comparable across the three dependent variables. We then create a consistent sample across all three variables.

To examine the real implications of macroprudential policies, we look at Investment, which is the growth in fixed assets, and Sales Growth, which is the growth in operating turnover. As before, growth rates are computed as log-differences and we remove top and bottom 5% outliers.

We control for the log of total assets to account for changes in external financing due to firm growth. We define MSME firms as those with fewer than 250 employees, micro firms as those with one to nine employees, SME firms as firms with 10 to 249 employees.15 We categorize each firm into a firm size class according to the median employees across all observations available during the sample period. Thus, the firm size classifications are fixed avoiding the reclassification bias (e.g. see Moscarini and Postel-Vinay, 2012) where firms are classified into larger size bins as the economy grows. We define young firms as firms that are three years or younger (since incorporation).

To measure the creditworthiness of firms, we use three measures: Leverage ratio which is the ratio of total debt to total assets, Interest Coverage which is a dummy that takes the value 1 for financially distressed firms (i.e., those with interest coverage ratios less than 1) 16 and 0 otherwise, and Profitability as measured by the return on assets.

The summary statistics in Table 1 shows high variation in external financing growth among firms in our sample, ranging from –190% to 200% for short-term financing and –140% to 150% for long-term financing growth. The median firm experienced a positive short-term financing growth (2.6%), but a decline in long-term financing growth (-7.7%). Overall financing growth was negative in the median firm in our sample (-3.4%) for the period we consider.

Table 1.

Summary statistics

This table presents summary statistics for all the firm-level variables in Panel A and country-level variables in Panel B. The number of observations for firm level variables, N, corresponds to the number of firms times the number of years of data available for each firm. The number of observations for country-level variables corresponds to the number of countries times the number of years of data for each country. All variables are defined in Appendix B.

article image

We combine the firm-level data with country panel data on the use of different macroprudential tools from the Global Macroprudential Policy Instruments (GMPI), as described in Cerutti et al. (2015). The GMPI survey is very detailed and covers 12 different instruments. The database distinguishes between (i) tools targeted at borrowers’ leverage and financial positions (BOR) and (ii) tools targeted at financial institutions (FIN).17 The former includes LTV and DSTI ratios, while the latter includes the following 10 instruments: dynamic loan-loss provisioning; countercyclical capital buffers; bank leverage ratio; capital surcharge for systemically important financial institutions; limits on interbank exposures; concentration limits; limits on foreign currency loans; limits on domestic currency loans; reserve requirement ratios; and taxes or levies on financial institutions. Each instrument is coded as 1 or 0 for each country-year depending on whether it was in use or not. Thus, the BOR index could range from 0 (no borrower-targeted instrument in place) to 2 (both borrower-targeted instruments in place) and the FIN index could range from 0 (no financial institution-targeted instrument in place) to 10 (all 10 financial institution-targeted instruments in place). Our third index (MPI) is the sum of BOR and FIN. Instruments are each coded for the period they were actually in place, i.e., from the date that they were introduced until the day that they were discontinued. While the survey captures the breadth of macroprudential policy across an array of tools and for a large cross-section of countries, it does not capture the intensity of the tools or the extent to which they were binding.18 The descriptive statistics in Table 1 show a high variation in the use of macroprudential tools across countries and over the sample period, ranging from zero to two instruments targeted at borrowers (out of two possible tools) and zero to six tools targeted at financial institutions (out of a possible maximum of 10 possible tools).

We control for several country-level time-varying factors to ensure that we do not confound the effect of macroprudential tools with other policies or macro factors.19 We control for the log change of GDP, GDP Growth, thus effectively controlling for economic growth, and the real monetary policy rate, Real Policy Rate, defined as the discount rate minus the inflation rate to control for the tightness of monetary policy. Finally, we control for the effect of the Global Financial Crisis by including a dummy, GFC, for the years 2008 and 2009, to account for the fact that firm credit could have declined as a result of the crisis.

To assess the relationship between macroprudential policies and growth in firms’ short-term, long-term, and overall financing, we estimate the following specification:

Financing growthijt=β1Macroprujt-1+β2Firm Assetsijt+β3Macrojt-1+β4GFCt+ηi+εijt(1)

where i denotes the firm, j the country and t the year. The dependent variable is one of the following three measures of financing growth: Short-term financing, Long-term financing, and Overall Financing Growth. Macropru is an indicator of macroprudential policies; Firm Assets refers to the log of total assets; Macro is a vector of macroeconomic variables including the real monetary policy rate and the log change of GDP. GFC is the Global Financial Crisis dummy variable for 2008 and 2009 to control for the generally lower growth during this period; ηi are firm fixed effects.

We lag the macroprudential and macroeconomic variables to reduce any bias that might come from reverse causation and allow for the time lag it takes for policy to affect firms’ financing growth. We include firm-fixed effects to control for time-invariant firm characteristics such as their sector and business model that could affect financing growth. We weight observations by the inverse of the number of firms per country and year so that each country has the same weight in our estimations. Finally, we cluster standard errors at the country-level, thus allowing error terms to be correlated across firms within a country.

While regression (1) allows us to mitigate concerns about reverse causation and unobserved firm-level factors driving financing growth by using firm-level data and including firm fixed effects, the worry that our estimates could be biased due to time-varying omitted variables remains. In a second step, we therefore focus on within country-year variation in financing growth across different firm groups that the literature has identified as facing different degrees of financing constraints. Specifically, we distinguish between firms of different sizes and of different ages, as an extensive literature has shown that financing constraints are inversely related to the size and age of enterprises (e.g., Hadlock and Pierce, 2010). We estimate the following specification including country-year fixed effects in addition to firm fixed effects:

Financing growthijt=β1Macroprujt-1*Firm Characteristici+β2FirmAssetsjit+μjt+ηi+εijt(2)

Adding country-year fixed effects (μjt) allows us to control for any time-varying country factor that might affect financing growth for the average firm in the country. Moreover, we consider whether firms of different sizes and ages respond differently to the implementation of macroprudential tools than the average firm by including interactions of macroprudential policies with firm size and age.20 Given that macroprudential tools are implemented for aggregate and systemic stability considerations, rather than targeting specific firm groups, this also allows us to partly control for the confounding influence of credit growth and policy measures. Specifically, we focus on firms with one to nine employees (micro), 10 to 249 employees (SME), as well as firms that are three years or younger (since incorporation). As these firms are typically more bank-dependent and have less diversified external financing sources, we expect the effect of macroprudential policies to be stronger among them.

To examine whether macroprudential policies have real consequences we estimate the following specifications:

Firm growthijt=β1Macroprujt-1+β2Firm Assetsjit+β3Macrojt-1+β4GFCt+ηi+εijt(3)
Firm growthijt=β1Macroprujt-1*Firm Characteristics+β2Firm Assetsit+μjt+ηi+εijt(4)

where firm growth is measured by Investment and Sales growth and the other variables are as defined above. In equation (4), as in (2), explaining the differential effects of macroprudential policies across different firm groups allows us to mitigate endogeneity concerns.

3. Macroprudential policies and firm financing and growth

In Table 2 we explore the relationship between macroprudential policies and firms’ financing growth. We regress firms’ short-term, long-term and overall financing growth on different macroprudential policies, controlling for firm assets and a number of country-level variables. We include firm-fixed effects, thus controlling for other time-invariant firm-level characteristics. Hence, we exploit within-firm financing growth and its relationship with macroprudential policies.

Table 2.

Financing growth and macroprudential policies

This table estimates the following specification: Financing growthijt = β1Macroprujt-1 + β2Firm Assetsit + β3Macrojt-1 + β4GFCt + ηi + εijt. The dependent variable is one of the following three measures of financing growth: Short-term financing growth, Long-term financing growth, and Overall financing growth. Macropru is an indicator of macroprudential policies: MPI (overall index), BOR (borrower targeted measures), FIN (financial institution targeted measures). Firm Assets is the log of total assets. Macro is a vector of macroeconomic variables including the real monetary policy rate and the log change of GDP. GFC is the Global Financial Crisis dummy variable for 2008 and 2009 to control for the generally lower growth during this period. ηi are firm fixed effects. All regressions are estimated using ordinary least squares weighted by the number of observations in each country. Standard errors are clustered at the country level. All variables are defined in Appendix B. *, **, and *** represent significance at 10%, 5%, and 1% levels, respectively.

article image

While all nine coefficients on the macroprudential policies enter negatively, only the coefficient on BOR in the long-term financing growth regression enters significantly. The coefficient size suggests that applying one additional borrower-related macroprudential policy is associated with a 4.8 percentage points lower long-term financing growth. We find that firms reduce their financing growth as they grow larger, while GDP growth is positively and significantly associated with firm financing growth. Financing growth was significantly lower during the Global Financial Crisis, while there is no significant relationship between the real interest rate and firms’ financing growth over our period of analysis.

Our results are robust to a number of checks. First, in Appendix Table C, we include FIN and BOR in the same regression. The results are consistent with the results in Table 2, with only BOR in the long-term financing regression entering negatively and significantly. In unreported robustness tests, we also split FIN into two components: cyclical tools targeted at financial institutions (dynamic Provisions, counter-cyclical buffers, reserve requirement ratios, limits on domestic currency loans, FX and countercyclical reserve requirements and leverage) and structural tools (capital surcharges on SIFIs, limits on interbank exposures, concentration limits, limits on foreign currency loans, and tax on financial institutions). These two separate indicators do not enter significantly in any of the regressions. In Appendix D, we control for three measures of firm financial strength or creditworthiness: leverage, interest coverage ratio (< 1) dummy and return on assets. While our sample reduces due to the more limited availability of these data, the results are robust to including these additional firm characteristics.

3.1. Firm heterogeneity

In Table 3, we include interactions of macroprudential policies with firm size (Panels A and B) and age (Panel C) to examine their differential effect on firms of different sizes and ages. This along with the introduction of country-year fixed effects allows us to improve our identification further. While macroprudential tools might be adjusted in reaction to country-level developments, it is less likely that they are adjusted in response to developments within specific firm groups.

Table 3.

Financing growth and macroprudential policies: Allowing for firm size/age heterogeneity

This table estimates the following specification: Financing growthijt = β1Macroprujt-1*Firm Characteristici + β2Firm Assetsit + μjt + ηi + εijt Financing Growth is one of the following three variables: Short-term financing growth, Long-term financing growth or Overall financing growth. Macropru is an indicator of macroprudential policies: MPI (overall index), BOR (borrower targeted measures), FIN (financial institution targeted measures). Firm characteristics include MSME dummy (1 for firms with fewer than 250 employees and 0 otherwise) in panel A, and Micro (1-9 employees) and SME dummies (10-249 employees) in panel B and a Young dummy taking the value 1 for firms < 3 years since incorporation in panel C). ηi are firm fixed effects and μjt are country-year fixed effects. All estimations control for the log of total assets (Firm Assets). Regressions are estimated using ordinary least squares weighted by the number of observations in each country. Standard errors are clustered at the country level. All variables are defined in Appendix B. *, **, and *** represent significance at 10%, 5%, and 1% levels, respectively.

article image
article image
article image

The results in Panel A show that MSMEs’ short-term, long-term, and overall financing growth are negatively associated with the additional implementation of borrower-related macroprudential tools. While most of the nine interaction terms enter negatively, only the regressions coefficients for BOR enter significantly. The implementation of one additional borrower-related tool is associated with 8.6 percentage points lower short-term, 3.8 percentage points lower long-term, and 5.1 percentage points lower overall financing growth for MSMEs compared to large firms in the country. In Panel B, we distinguish further between microenterprises and SMEs, with the effects for large firms again captured by the country-year fixed effects. All six interactions of BOR with the micro- and the SME dummies enter negatively and significantly at least at the 10 percent significance level, with economic effects being somewhat higher for microenterprises than for SMEs. We also find that macroprudential tools targeted at financial institutions are associated with relatively lower financing growth for micro-enterprises, but not for SMEs.21 There is also a negative relationship between the overall index of macroprudential tools and the relative financing growth of both microenterprises and SMEs, though the relationship is only significant for microenterprises, but not for SMEs.

The results in Panel C show that the relationship between macroprudential tools and financing growth is relatively stronger and more negative for younger firms, especially for macroprudential tools targeted at borrowers and for long-term financing growth. Specifically, while all the interaction terms between macroprudential tools and the young firm dummy enter negatively, they only enter significantly for the BOR measure across all measures of financing growth and in the other two regressions of long-term financing growth. The relative economic effects are not as strong for young firms as they are for MSMEs; both short- and long-term financing growth is 2.3 percentage points lower for younger than for older firms after one additional borrower-targeted tool is adopted, while overall financing growth is 1.4 percentage points lower. Long-term financing growth also decreases by 4 percentage points for younger firms relative to older firms after one additional macroprudential tool targeted at financial institutions is adopted and by 2.5 percentage points for any additional macroprudential tool. While the negative coefficient estimates in Table 3 only tell us about the relative financing growth of specific firm groups, the combination of results in Tables 2 and 3 give us confidence in having identified a negative association of macroprudential policies (especially the ones targeted at borrowers) with financing growth of MSMEs and young firms.

We undertake a number of robustness tests of our results in Table 3. First, to address further endogeneity concerns we repeat the specifications in Table 3, but this time we control for the differential impact of other macro shocks on different firms by interacting the size and age dummies with the macro variables. Table 4 shows that the results from Table 3 hold when we allow for these additional interactions.

Table 4.

Financing growth and macroprudential policies: Interacting all country-level variables with firm size/age

This table estimates the following specification: Financing growthijt = β1Macroprujt-l*Firm Characteristici + β2Macrojt-l*Firm Characteristici +β3GFCt-l*Firm Characteristic+ β4Firm Assetsit + μjt + ηi + εijt. Financing Growth is one of the following three variables: Short-term financing growth, Long-term financing growth or Overall financing growth. Macropru is an indicator of macroprudential policies: MPI (overall index), BOR (borrower targeted measures), FIN (financial institution targeted measures). Firm characteristics include MSME dummy (1 for firms with fewer than 250 employees and 0 otherwise) in panel A, and Micro (1-9 employees) and SME dummies (10-249 employees) in panel B and a Young dummy taking the value 1 for firms ≤ 3 years since incorporation in panel C. Macro is a vector of macroeconomic variables including the real monetary policy rate and the log change of GDP. GFC is the Global Financial Crisis dummy variable for 2008 and 2009 to control for the generally lower growth during this period. ηi are firm fixed effects and μjt are country-year fixed effects. All estimations control for the log of total assets (Firm Assets). Regressions are estimated using ordinary least squares weighted by the number of observations in each country. Standard errors are clustered at the country level. All variables are defined in Appendix B. *, **, and *** represent significance at 10%, 5%, and 1% levels, respectively.

article image
article image
article image

Next, in Appendix E, we repeat our specifications in Table 3 but controlling for industry-year fixed effects in addition to firm and country-year fixed effects to better control for different industries experiencing different business cycle effects. Once again, all our findings from Table 3 hold when we allow for different industries facing different cyclicality. In Appendix F, we report estimations controlling for three measures of firm financial strength or creditworthiness: leverage, interest coverage ratio (<1) dummy, and return on assets. Again, the results are robust to including these additional firm characteristics.

Thus far, we have assessed the relationship between the implementation of macroprudential policies and financing of firms. We now turn our attention to indicators of the intensity of macroprudential measures, using data from Cerutti et al. (2017). Specifically, we focus on the loan-to-value cap for residential borrowers since most of our significant results come from the implementation of borrower macroprudential measures. While we would have liked to have information on the intensity of the other macroprudential measures, none of the existing databases provide that level of granularity for our sample of countries.

The underlying intensity measure is reported at the quarterly frequency and records regulatory changes in the loan-to-value ratio limits to real estate transactions. The index can take on values of 1 and -1 depending on whether the macroprudential tool was tightened or loosened in each quarter. A value of 0 indicates no policy change. We time-aggregate this indicator to the yearly frequency by taking the cumulative changes in the loan-to-value ratio cap starting from quarter one and keeping the cumulative index at the fourth quarter of every year. These data are, unfortunately, available for only 16 countries.

The results in Table 5 show that both the level and the change in loan-to-value ratio are associated with a relatively lower long-term financing growth of MSMEs, while neither short-term financing growth nor overall financing growth seem to be impacted (Panel A). On average, a tightening of the cumulative loan to value cap is associated with a 10.1 percentage point lower long-term financing growth for MSMEs relative to large firms. Similarly, on average, an increase in the cumulative tightening of the loan-value cap is associated with a 6.2 percentage point lower long-term financing growth for MSMEs relative to large firms.

Table 5.

Financing growth and the intensity of macroprudential policies

This table estimates the following specification: Financing growthijt = β1Intensity Macroprujt-l*Firm Characteristici + β2Firm Assetsit + μjt+ ηi + εijt Financing Growth is one of the following three variables: Short-term financing growth, Long-term financing growth or Overall financing growth. Intensity of macropru is Cumulative Intensity or ΔCumulative Intensity. Firm characteristics include MSME dummy (1 for firms with fewer than 250 employees and 0 otherwise) in panel A, and Micro (1-9 employees) and SME dummies (10-249 employees) in panel B and a Young dummy taking the value 1 for firms ≤ 3 years since incorporation in panel C. ηi are firm fixed effects and μjt are country-year fixed effects. All estimations include the log of asset (Firm Assets). Regressions are estimated using ordinary least squares and weighted by the number of observations in each country. Standard errors are clustered at the country level. All variables are defined in Appendix B. *, **, and *** represent significance at 10%, 5%, and 1% levels, respectively.

article image
article image
article image

These findings hold both for microenterprises and SMEs (Panel B). On the other hand, we do not find any significantly different effect of the level or change of macroprudential tools for the financing growth of young vs. older firms (Panel C). Together, these results provide some evidence that not only the implementation but also tightening of borrower-targeted macroprudential measures is negatively associated with firms’ financing growth.

3.2. Differential impact by financial strength

The findings above show that MSMEs and young firms are differentially more sensitive to macroprudential policies than larger and older firms presumably because they are more financially constrained and are more bank dependent. While this may not be the intended impact of these regulations, it begs the question whether the policies actually work to suppress credit to the riskiest firms at whom these policies are targeted. In other words, we would like to assess whether macroprudential policies hurt financial inclusion by limiting access to typically credit constrained firms across the board or whether they enhance financial stability and efficiency by restricting credit to poorly performing and risky firms. To try to get at this, we look within the sub-samples of micro, small, and young firms and interact the macroprudential variables with the financial strength or creditworthiness of the firm. As before, we use three measures of financial strength/creditworthiness – Leverage in Table 6, Interest Coverage in Table 7, and Profitability(ROA) in Table 8. At the outset, panel A-C of Appendix F show that controlling for the different measures of creditworthiness does not alter the findings in Table 3.

Table 6.

Financing growth and macroprudential policies: Allowing for firm size/age interactions with leverage

This table estimates the following specification: Financing growthijt = β1Macroprujt-1*Firm Leverage it + β2Firm Leverageit + β3Firm Assetsit + μjt+ ηi + εijt Financing Growth is one of the following three variables: Short-term financing growth, Long-term financing growth or Overall financing growth. Macropru is an indicator of macroprudential policies: MPI (overall index), BOR (borrower targeted measures), FIN (financial institution targeted measures). Firm leverage is the ratio of debt to assets. Panel A is estimated using a sub-sample of micro firms (1-9 employees), panel B using a sub-sample of SMEs (10-249 employees), and panel C using a sub-sample of young firms (≤ 3 years since incorporation). ηi are firm fixed effects and μjt are country-year fixed effects. All estimations include the log of assets (Firm Assets). Regressions are estimated using ordinary least squares weighted by the number of observations in each country. Standard errors are clustered at the country level. All variables are defined in Appendix B. *, **, and *** represent significance at 10%, 5%, and 1% levels, respectively.

article image
article image
article image
Table 7.

Financing growth and macroprudential policies: Allowing for firm size/age interactions with profitability

This table estimates the following specification: Financing growthijt = β1Macroprujt-1*Firm ROA it + β2Firm ROAit + β3Firm Assetsit + μjt+ ηi + εijt Financing Growth is one of the following three variables: Short-term financing growth, Long-term financing growth or Overall financing growth. Macropru is an indicator of macroprudential policies: MPI (overall index), BOR (borrower targeted measures), FIN (financial institution targeted measures). ROA refers to return on assets and is a measure of firm profitability. Panel A is estimated using a sub-sample of micro firms (1-9 employees), panel B using a sub-sample of SMEs (10-249 employees), and panel C using a sub-sample of young firms (≤ 3 years since incorporation). ηi are firm fixed effects and μjt are country-year fixed effects. All estimations include the log of total assets (Firm Assets). Regressions are estimated using ordinary least squares and weighted by the number of observations in each country. Standard errors are clustered at the country level. All variables are defined in Appendix B. *, **, and *** represent significance at 10%, 5%, and 1% levels, respectively.

article image
article image
article image
Table 8.

Financing growth and macroprudential policies: Allowing for firm size/age interactions with profitability

This table estimates the following specification: Financing growthijt = α1 + β1Macroprujt-1*Interest coverage it + β2Interest coverageit + β3Firm Assetsit + μjt+ ηi + εijt Financing Growth is one of the following three variables: Short-term financing growth, Long-term financing growth or Overall financing growth. Macropru is an indicator of macroprudential policies: MPI (overall index), BOR (borrower targeted measures), FIN (financial institution targeted measures). Interest Coverage is a dummy that takes the value 1 if interest coverage is less than 1 and 0 otherwise. Panel A is estimated using a sub-sample of micro firms (1-9 employees), panel B using a sub-sample of SMEs (10-249 employees), and panel C using a sub-sample of young firms (≤ 3 years since incorporation). ηi are firm fixed effects and μjt are country-year fixed effects. All estimations include the log of total assets (Firm Assets). All regressions are estimated using ordinary least squares weighted by the number of observations in each country. Standard errors are clustered at the country level. All variables are defined in Appendix B. *, **, and *** represent significance at 10%, 5%, and 1% levels, respectively.

article image
article image
article image