Household Debt, Corporate Debt, and the Real Economy: Some Empirical Evidence

ABSTRACT

The rapid accumulation of private debt is widely viewed as a major risk to financial and economic stability. This article systematically and comprehensively assesses the effect of private debt buildup on economic growth. In the spirit of the existing study that separately examines the effects of two types of private debt – household debt and corporate debt – on growth in advanced economies, we specifically provide new evidence on the growth-private debt nexus in both advanced and emerging market economies (EMEs). Moreover, we construct financial peaks in terms of the speed of debt accumulation rather than crisis dates and find that in both advanced and EMEs, corporate debt buildups cause more financial peaks than household debt buildups. Furthermore, corporate debt-induced financial recessions inflict a bigger damage on output than household debt-induced financial recessions in EMEs. Overall, our evidence suggests that policymakers would do well to closely monitor not only household debt but also corporate debt.

KEYWORDS:

• Private debt
• household debt
• corporate debt
• output
• economic growth
• business cycle
• crisis

JEL CLLASSIFICATION:

• E32
• E44
• G01

Supplemental data

Supplemental data for this article can be accessed on the publisher’s website.

Acknowledgments

This article was initially prepared for the Asian Development Bank Institute’s 21st Annual Conference: Managing Private and Local Government Debt in Asia, 30 November–1 December 2017. This was also used as a background article for the Asian Development Outlook 2018 and a preliminary version was published as ADB working paper No. 567. We thank Jaeyoung Yoo for his excellent research assistance, Hyein Han and Cynthia Castillejos-Petalcorin for their superb editorial work, and the Asian Development Bank for its financial support.

Notes

1. For example, see Glick and Lansing (2010) and Mian and Sufi (2014), and the literaturereviewed in the next section.

2. Household debt has received more attention than corporate debt. For example, Glick and Lansing (2010) shows that many advanced economies experienced rapid increases in household leverage and countries with the largest increase in household leverage experienced the fastest rise in house prices and the largest decline in subsequent household consumption. Based on United States county data, Mian and Sufi (2014) also find that the increase in household debt before the GFC predicts the severity of the downturn during the Great Recession.

3. See Bernardini and Forni (2017) and Figure 1 therein. See also Section II.

4. Mian, Sufi, and Verner (2017) also show that, if the underlying credit shock is demand-driven, even models of agents with flawed expectations are not consistent with empirical facts because these models imply increases in the interest rate, which is counterfactual.

5. See Mian, Sufi, and Verner (2017) for the references that explain sources of credit supply shocks. For example, as argued by Justiniano, Primiceri, and Tambalotti (2015) and Schmitt-Grohé and Uribe (2016), credit supply expansion may originate from foreign capital inflows as well.

6. The nonlinear nexus between public debt and economic growth is well established in the literature. Relevant studies include Baum, Checherita, and Rother (2013), Checherita and Rother (2012), Égert (2015), Kumar and Woo (2010), and Reinhart and Rogoff (2010). However, less is known about the impact of private debt accumulation on growth.

7. While Mian, Sufi, and Verner (2017) also report some differences between the experiences of advanced and EMEs, the comparison was not a main objective of their paper.

8. In Section 3, we find that most advanced countries reached financial peaks only around GFC. See Section 3 for details.

9. In Section 3, we find that only two EMEs experienced rapid increase in private debts before the GFC.

10. We follow the approach by Jordà, Schularick, and Taylor (2013) in defining financial peaks solely based on actual financial crisis dates and compare these results with ours.

11. This different timings of the effects of household and corporate debts is also highlighted by Mian, Sufi, and Verner (2017).

12. Following Mian, Sufi, and Verner (2017), we exclude the People’s Republic of China, India, and South Africa, for which the data for private debt start from 2006 to 2007, as well as Luxembourg, for which the private debt data are too volatile. For most countries, the amount of private debt of nonfinancial sector is exactly the same as the sum of household debt and nonfinancial corporate debt, but there are small discrepancies in some cases.

13. We calculate real consumption by multiplying consumption share to output-side real GDP at chained PPPs because consumption share is reported using current PPPs. However, our findings seldom change if we use GDP at constant national prices instead.

14. Asian EMEs refer to the four countries hit hardest during the Asian financial crisis, namely Indonesia, the Republic of Korea, Malaysia, and Thailand.

15. While Mian, Sufi, and Verner (2017) adopted a proxy VAR where the mortgage sovereign spread is used as an instrument, we cannot do so because the data are not available for many EMEs. Mian, Sufi, and Verner (2017) also report figures based on VAR for which Cholesky decomposition is used to identify shocks. However, we believe that estimating Equation (1)(1) ${\mathrm{\Delta }}_3{y}_{it+k}={\beta }_0+{\beta }_H{\mathrm{\Delta }}_3{d}_{it-1}^{HH}+{\beta }_C{\mathrm{\Delta }}_3{d}_{it-1}^{Corp}+u_{it+k}$(1) generates essentially the same dynamics as VAR identified from Cholesky decomposition since it essentially imposes restrictions between current variables only. In Equation (1)(1) ${\mathrm{\Delta }}_3{y}_{it+k}={\beta }_0+{\beta }_H{\mathrm{\Delta }}_3{d}_{it-1}^{HH}+{\beta }_C{\mathrm{\Delta }}_3{d}_{it-1}^{Corp}+u_{it+k}$(1) , the dependent variables are mostly future values relative to the explanatory variables which can thus be considered as exogenous. In this sense, both methods show the same dynamic relation between three variables – real GDP, household debt and corporate debt. Further, the dependent variable is the three-year change in logarithm as in Mian, Sufi, and Verner (2017) since household debt increases for three to four years after a shock (p. 1765 in Mian, Sufi, and Verner (2017)). Hence, it is difficult to apply GMM methods of Arellano and Bond.

16. We thank an anonymous referee for pointing out the problem.

17. Results are not reported for brevity purpose but are available upon request.

18. In fact, the average duration of expansions before FCPs is much longer than that before normal peaks. Hence, the amplitude of the variables of the FCPs is higher than that of normal peaks. However, as reported in Appendix Table 6, the annual growth rate, the amplitude divided by duration, is actually lower during expansions before FCPS than before normal peaks.

19. The results, based on FCPs, are quite similar and are available upon request.

20. We gratefully appreciate JST for sharing their Stata program for generating the regression results. Stata is a general-purpose statistical software package created in 1985 by StataCorp.

21. The number of FCPs and household and corporate FCPs is small for EMEs. So, instead of reporting the estimates separately, we combine advanced economies and EMEs together to generate estimates for the full sample.

22. This finding is also emphasized by JST.

23. The interactions terms with household excess credit are generally more statisticallysignificant.

Additional information

Funding

This work was supported by the Asian Development Bank.