Financial frictions in the US: asymmetric effects per industry

ABSTRACT

The article explores the link between lending rates and real output asymmetries and bankruptcy costs across US industries. The results document that bankruptcy costs impact such asymmetries, indicating stronger financial frictions in those industries, which are expected to delay the recovery of firms in stressful events.

KEYWORDS:

• Lending rates asymmetries
• bankruptcy costs
• US industrial sectors
• Skewness

JEL CLASSIFICATION:

• O40
• O30
• E44

I. Introduction

The presence of financial frictions in the corporate sector might be due to either the inability of firms to issue new equity and debt instruments, their inability to borrow from the banking sector, their greater dependence on bank loans, their prevalence of credit constraints, or the illiquidity of their assets (Graham and Harvey 2001). The exploration of the role of financial frictions lies within the boundaries of the pecking order theory, according to which asymmetric information costs play the primary role in determining the capital structure choice of firms. Accordingly, financially unconstrained firms are likely to depend less on internal funds than their financially constrained peers (Myers and Majluf 1984). Moreover, given the presence of informational asymmetries, firms prefer debt to equity financing when go for external funds and issue equity only as a last resort. Almeida and Campello (2010) and Gracia and Mira (2014) provide strong evidence on the role of financial frictions in determining the capital structure of corporate firms.

The goal of this article is to explore for the first time, to our knowledge, the role of asymmetries in financial frictions across US industries. Although the US is a highly developed country characterized by low financial frictions, there are certain industries that suffer more from such frictions, captured by higher bankruptcy costs. Such asymmetric frictions magnify the reactions of lending rates to shocks and delay their recovery.

Ordoñez (2012) investigates how business cycle asymmetries differ across countries. His findings document that the asymmetric movements of lending rates and output are stronger in less-developed countries, especially, those with weaker financial systems. He introduces financial frictions into a learning model with endogenous flows of information and generates asymmetric lending rates as in Veldkamp (2005) and asymmetric economic activity as in Van Nieuwerburgh and Veldkamp (2006). His environment with asymmetric information induces borrowers to falsely renege on their loans, while bankruptcy costs represent measures of how costly it is to overcome financial frictions, leading to higher lending rates and depressing economic activity. His results illustrate that uncertainty is asymmetric: it increases during crises and declines gradually thereafter. Gilchrist, Yankov, and Zakrajsek (2009) use corporate bonds trading and find that credit market shocks have contributed significantly to US economic fluctuations, while Gilchrist and Zakrajsek (2012) employ credit spreads and document that innovations to the component that is orthogonal to the current state of the economy lead to declines in economic activity. Finally, Christiano, Motto, and Rostagno (2012) highlight the relevance of risk shocks to capture crises and recoveries and match moments of financial variables, such as the counter-cyclicality of lending rates. While all these studies focus on national economic variables, our study extends them by considering variables across US industrial sectors.

II. Data

Annual data on real lending rates and real GDP per capita are used, spanning the period 1971–2018, across US industrial sectors. Data on nominal lending rates, real GDP per capita and US inflation are obtained from Bloomberg, while the remaining data on lending rates are obtained from the Global Economy.Com company (https://www.theglobaleconomy.com/USA). Real lending rates are measured by subtracting the Hodrick-Prescott (HP) trend of US inflation from nominal rates.

Moreover, financial frictions are defined as the asymmetric ex-post information about repayment possibilities, i.e. borrowers know more about their income than lenders. A measure of this friction is the cost for lenders to confirm such information, which is provided by the cost of taking a defaulting borrower to bankruptcy. Such (annual) data on bankruptcy costs are obtained from Moody’s Investor reports and they compute estimates of how difficult it is for lenders to go through bankruptcy procedures. These measures of costs include court/bankruptcy authority costs, attorney fees, bankruptcy administrator fees, accountant fees, notification and publication fees, assessor or inspector fees, asset storage and preservation costs, auctioneer fees, government levies, and other associated insolvency costs. Such (proprietary) data are made available from Bankruptcy Data by the New Generation Research company (https://www.bankruptcydata.com). In addition, data on time for bankruptcy are also obtained from the same source and are measured as estimated duration, in years, of the time to resolve the insolvency incident; it measures the duration from the moment of default to the point at which the fate of the incident is determined.

We measure unconditional asymmetry as the skewness of changes over time. Such a measure shows that if lending rates in an industry are more likely to experience large jumps rather than large reductions of the same magnitude, the skewness of their changes is positive, while a stronger asymmetry is captured by a larger positive skewness. Similarly, we expect negative skewness for output per capita. The analysis constructs the distribution of changes for lending rates and GDP per capita in each industry. Then, the unconditional skewness for each one of these distributions is measured as:

$\begin{array}{rl}& Skewness=\left[T\sqrt{\left(T-1\right)}\sum _{t=1}^T{\left(x_t-\bar{x}\right)}^3\right]\\ & /\left[T\sqrt{\left(T-1\right)}\sum _{t=1}^T{\left(x_t-\bar{x}\right)}^3\right]\\ & \left[\left(T-2\right)\sum _{t=1}^T{\left(x_t-\bar{x}\right)}^2\right]\\ & -{\left[\left(T-2\right)\sum _{t=1}^T{\left(x_t-\bar{x}\right)}^2\right]}^{3/32-2}\end{array}$

where T is the number of observations (number of years), x_t = X_t−X_t-1, X_t is the variable measured in period t and x bar is the sample mean of the time series.

The industries included are: Automotive, Banking, Beverage-Food-Tobacco, Capital Equipment, Chemicals-Plastics-Rubber, Construction-Building, Consumer Goods-Durables, Consumer Goods-Nondurables, Energy-Electricity, Finance-RealEstate, Forest Products-Paper, Healthcare-Pharmaceuticals, High Tech Industries, Hotel-Gaming-Leisure, Media, Metals-Mining, Services-Business, Services-Consumer, Sovereign-Public Finance, Telecommunications, Transportation, Utilities-Electric.

III. Empirical evidence

Table 1 displays the estimates between the skewness of real lending rates and real output per capita (both as dependent variables) and bankruptcy costs and the time of bankruptcy. The findings document the presence of a positive link between bankruptcy costs and the asymmetry of lending rates and a negative link between those bankruptcy costs and the asymmetry of real output per capita. The estimates turn out to be statistically significant at various significance levels for certain industries (i.e. Automotive, Banking, Capital Equipment, Chemicals-Plastics-Rubber, Construction-Building, Consumer Goods-Durables, Finance-Real Estate, Hotel-Gaming-Leisure, Media, Metals-Mining, Services-Business, Services-Consumer, Sovereign-Public Finance). The strongest estimates come from the Services-Business sector, followed by the sector of Real Estate. The findings indicate that these industrial sectors face stronger financial distress and financial asymmetric information, with the statistical presence/significance of bankruptcy costs restricting potential recoveries of bankruptcy in case of an emergency. The results remain consistently similar with respect to the link between the skewness of real lending rates and real output per capita and the time of bankruptcy, with the association being negative now.

Table 1. Asymmetries and bankruptcy costs

Dependent variable: Skewness of real lending rates
	Automotive	Banking	Beverage-food-tobacco	Capital equipment	Chemicals-plastics-rubber	Construction-building
Constant	0.674	0.471	1.348**	0.529	0.493	0.386
	[0.15]	[0.23]	[0.03]	[0.19]	[0.21]	[0.38]
Bankruptcy costs	0.058***	0.083***	0.029	0.073***	0.079***	0.092***
	[0.00]	[0.00]	[0.18]	[0.00]	[0.00]	[0.00]
Time for bankruptcy	0.258**	0.316***	0.084	0.295***	0.316***	0.288***
	[0.02]	[0.00]	[0.17]	[0.00]	[0.00]	[0.01]
R^2-adjusted	0.52	0.55	0.27	0.59	0.57	0.61
		Consumer goods-durables	Consumer goods-nondurables	Energy-electricity	Finance-real estate	Forest products-paper
Constant		0.393	1.296**	1.335**	0.228	1.408***
		[0.42]	[0.04]	[0.03]	[0.59]	[0.01]
Bankruptcy costs		0.091***	0.034*	0.025	0.118***	0.021
		[0.00]	[0.10]	[0.16]	[0.00]	[0.22]
Time for bankruptcy		0.275***	0.096*	0.058	0.287***	0.042
		[0.01]	[0.09]	[0.36]	[0.00]	[0.36]
R^2-adjusted		0.56	0.22	0.15	0.52	0.13
	Healthcare-pharmaceuticals	High tech industries	Hotel-gaming-leisure	Media	Metals-mining	Services-business
Constant	1.359**	1.326**	0.616	0.543	0.522	0.493
	[0.02]	[0.02]	[0.18]	[0.22]	[0.26]	[0.31]
Bankruptcy costs	0.032	0.028	0.085***	0.093*** 0.102***	0.123***
	[0.14]	[0.19]	[0.00]	[0.00]	[0.00]	[0.00]
Time for bankruptcy	0.041	0.037	0.268***	0.285*** 0.302***	0.319***
	[0.30]	[0.32]	[0.01]	[0.00]	[0.00]	[0.00]
R^2-adjusted	0.19	0.15	0.53	0.57`	0.63	0.66
	Services-consumer	Sovereign-public finance	Telecommunications	Transportation	Utilities-electric
Constant	0.488	1.326**	1.405***	1.349**	0.783*
	[0.30]	[0.02]	[0.01]	[0.02]	[0.09]
Bankruptcy costs	0.069***	0.026	0.019	0.023	0.072***
	[0.00]	[0.17]	[0.30]	[0.26]	[0.00]
Time for bankruptcy	0.274***	0.082*	0.065	0.069	0.274***
	[0.00]	[0.10]	[0.15]	[0.14]	[0.01]
R^2-adjusted	0.52	0.2	0.13	0.15	0.58
Dependent variable: Skewness of real output per capita
	Automotive	Banking	Beverage-food-tobacco	Capital equipment	Chemicals-plastics-rubber	Construction-building
Constant	−0.051	−0.038	−0.501***	−0.042	−0.038	−0.026
	[0.21]	[0.34]	[0.00]	[0.31]	[0.35]	[0.48]
Bankruptcy costs	−0.082***	−0.096***	−0.033	−0.094***	−0.086***	−0.104***
	[0.00]	[0.00]	[0.17]	[0.00]	[0.00]	[0.00]
Time for bankruptcy	−0.203***	−0.264***	−0.059	−0.289***	−0.294***	−0.301***
	[0.00]	[0.00]	[0.38]	[0.00]	[0.00]	[0.00]
R^2-adjusted	0.49	0.53	0.22	0.63	0.59	0.6
		Consumer goods-durables	Consumer goods-nondurables	Energy-electricity	Finance-real estate	Forest products-paper
Constant		−0.031	−0.603***	−0.575***	−0.029	−0.493***
		[0.47]	[0.00]	[0.00]	[0.54]	[0.00]
Bankruptcy costs		−0.098***	−0.029	−0.023	−0.140***	−0.017
		[0.00]	[0.26]	[0.34]	[0.00]	[0.41]
Time for bankruptcy		−0.291***	−0.133**	−0.047	−0.296***	−0.054
		[0.00]	[0.05]	[0.50]	[0.00]	[0.45]
R^2-adjusted		0.58	0.26	0.17	0.55	0.18
	Healthcare-pharmaceuticals	High tech industries	Hotel-gaming-leisure	Media	Metals-mining	Services-business
Constant	−0.514***	−0.487***	−0.047	−0.052	−0.027	−0.023
	[0.00]	[0.00]	[0.44]	[0.40]	[0.56]	[0.62]
Bankruptcy costs	−0.026	−0.034	−0.097***	−0.104***	−0.126***	−0.132***
	[0.41]	[0.33]	[0.00]	[0.00]	[0.00]	[0.00]
Time for bankruptcy	−0.049	−0.032	−0.281***	−0.294***	−0.318***	−0.334***
	[0.51]	[0.59]	[0.00]	[0.00]	[0.00]	[0.00]
R^2-adjusted	0.23	0.18	0.59	0.61`	0.67	0.69
	Services-consumer	Sovereign-public finance	Telecommunications	Transportation	Utilities-electric
Constant	−0.048	−0.514***	−0.526***	−0.493***	−0.044
	[0.39]	[0.00]	[0.00]	[0.00]	[0.45]
Bankruptcy costs	−0.081***	−0.031	−0.024	−0.029	−0.089***
	[0.00]	[0.35]	[0.43]	[0.34]	[0.00]
Time for bankruptcy	−0.293***	−0.157**	−0.054	−0.065	−0.288***
	[0.00]	[0.05]	[0.57]	[0.54]	[0.00]
R^2-adjusted	0.54	0.23	0.17	0.19	0.63

Note: Figures in brackets denote p-values. ***: ≤0.01; **: ≤0.05; *:≤ 0.10.

Table 2 reports the results after computing the measure of skewness as the distribution of deviations from a real lending rate trend. In particular, the analysis obtains the difference between the real lending rates and their Hodrick-Prescott trend, and compute the skewness of such distribution. The findings document similar results to those presented in Table 1.

Table 2. Real lending rates asymmetries and bankruptcy costs (skewness is measured as a deviation from trend)

	Automotive	Banking	Beverage-food-tobacco	Capital equipment	Chemicals-plastics-rubber	Construction-building
Constant	0.621	0.396	1.274**	0.479	0.426	0.345
	[0.19]	[0.48]	[0.05]	[0.23]	[0.29]	[0.44]
Bankruptcy costs	0.064***	0.089***	0.037	0.082***	0.086***	0.101***
	[0.00]	[0.00]	[0.15]	[0.00]	[0.00]	[0.00]
Time for bankruptcy	0.271***	0.335***	0.089	0.314***	0.325***	0.296***
	[0.01]	[0.00]	[0.16]	[0.00]	[0.00]	[0.00]
R^2-adjusted	0.5	0.53	0.28	0.62	0.55	0.63
		Consumer goods-durables	Consumer goods-nondurables	Energy-electricity	Finance-real estate	Forest products-paper
Constant		0.365	1.233**	1.281**	0.214	1.362**
		[0.46]	[0.05]	[0.04]	[0.62]	[0.03]
Bankruptcy costs		0.103***	0.039*	0.02	0.127***	0.026
		[0.00]	[0.09]	[0.24]	[0.00]	[0.21]
Time for bankruptcy		0.293***	0.105*	0.064	0.296***	0.047
		[0.00]	[0.07]	[0.32]	[0.00]	[0.34]
R^2-adjusted		0.54	0.23	0.17	0.55	0.15
	Healthcare-pharmaceuticals	High tech industries	Hotel-gaming-leisure	Media	Metals-mining	Services-business
Constant	1.304**	1.295**	0.584	0.512	0.489	0.435
	[0.03]	[0.04]	[0.27]	[0.34]	[0.36]	[0.43]
Bankruptcy costs	0.037	0.034	0.099***	0.107***	0.118***	0.131***
	[0.13]	[0.17]	[0.00]	[0.00]	[0.00]	[0.00]
Time for bankruptcy	0.047	0.042	0.289***	0.296***	0.321***	0.338***
	[0.27]	[0.30]	[0.00]	[0.00]	[0.00]	[0.00]
R^2-adjusted	0.21	0.16	0.58	0.60`	0.67	0.71
	Services-consumer	Sovereign-public finance	Telecommunications	Transportation	Utilities-electric
Constant	0.447	1.254**	1.337**	1.296**	0.658
	[0.35]	[0.04]	[0.02]	[0.03]	[0.14]
Bankruptcy costs	0.077***	0.031	0.024	0.031	0.088***
	[0.00]	[0.16]	[0.26]	[0.23]	[0.00]
Time for bankruptcy	0.292***	0.097*	0.054	0.073	0.297***
	[0.00]	[0.08]	[0.19]	[0.13]	[0.00]
R^2-adjusted	0.54	0.23	0.14	0.17	0.61

Note: Similar to those in Table 1.

IV. Conclusion

The analysis documented the link between lending rates and real output asymmetries and bankruptcy costs across US industrial sectors. The evidence displayed that asymmetry was stronger in certain sectors, indicating stronger financial frictions and the important role of bankruptcy costs. Such frictions are expected to delay their recovery in case of crisis events. Potential implications have to do with the need of improving mechanisms that reduce financial frictions in these sectors, such as reductions in the complexity of bankruptcy procedures, which improve the efficiency in bankruptcy courts and codes.

Disclosure statement

No potential conflict of interest was reported by the authors.