Did the financial crisis affect the debt maturity of multinationals and local firms differently?

(1)

Did the financial crisis affect the debt maturity of

multinationals and local firms differently?

Jannick van IJzendoorn

Abstract: Using a sample of firms from 15 European countries, we demonstrate that the

2008-2009 financial crisis had a different effect on the debt maturity of multinationals compared to local firms. While we note a decrease in debt maturity amongst local firms in the period 2008-2011, we observe an increase amongst multinationals. More specifically, we uncover that multinationals reduced their short-term debt usage in that period, whereas local firms did not. This finding is consistent with the view that multinationals are better able to avoid contracting short-term debt during credit crises by reallocating funds internally. We also reveal that the weight of bank credit in the financing of the private sector has a less negative effect on the debt maturity of multinationals compared to local firms. Our main findings are robust against alternative proxies for multinationality and alternative measures of the financial crisis.

Key words: financial crisis, debt maturity, multinational firms [JEL: F23, G32]

Master Thesis | s2754835

Program: MSc International Financial Management

Supervisor: Dr. E. (Egle) Karmaziene

(2)

1

(3)

2

1. Introduction

Many papers document how the 2008-2009 financial crisis affected the health of the global economy. Particularly harmful to economic growth was the worldwide collapse of corporate investment. This collapse was especially persistent in Europe, where net investments had still not recovered to pre-crisis levels by 2018 (Kalemli-Özcan et al., 2019). Some studies link the investment implosion to a credit supply shock, meaning that companies invested less because it became increasingly difficult for them to obtain bank debt. If a credit crunch was indeed to blame, it follows that investment cutbacks should have been greatest for companies that relied on bank finance. In line with this prediction, Duchin et al. (2010) show that investment de-clined most for firms that had low cash reserves or high short-term debt, were financially con-strained, or operated in industries that depended on external funding.

Given that capital structure seems to account for (some of) the post-2007 investment behaviour of firms, it is surprising that the literature pays little attention to the maturity struc-ture of corporate debt. While longer debt is commonly associated with lower capital expendi-ture (Aivazian et al., 2005) and lower investment efficiency (Gomariz & Ballesta, 2014), evi-dence suggests that firms relied on existing long-term loans to sustain investment in the peri-od 2008-2012 (Almeida et al., 2011). This applies especially to companies that depended on external finance (Gonzalez, 2015). To continue investing during the crisis after their long-term loans matured, these firms had to revert to shorter debt structures. Therefore, the propor-tion of long-term debt to short-term debt on their balance sheets declined.

This paper sheds more light on the development of debt maturity during the recession by examining the role of firm internationalization. We ask ourselves the question, did the cri-sis affect the debt maturity of multinationals differently compared to local firms? Having an operation presence in a foreign market should give multinationals greater financial flexibility compared to local firms, both through methods of internal and external financing. Thus, we should observe multinationals holding higher proportions of long-term debt and/or lower pro-portions of short-term debt in the period 2008-2012. To test this hypothesis, we construct a sample consisting of 2071 firms and 23,434 firm-year observations for 15 European countries over the period 2003-2016. In the empirical analysis, our attention mainly centres on the combined effect of the financial crisis and firm internationalization on debt maturity.

(6)

5

effect of the crisis on the short-term debt usage of multinationals. Namely, despite the uni-form reduction in long-term debt usage across all stages of internationalization, we see that short-term debt usage decreased for multinationals, whereas it did not for local firms. Because the effect is concentrated in short-term debt, our results are consistent with the view that mul-tinationals were able to uphold investment by reallocating funds via their cross-border net-work. As such, internal financing could have allowed multinationals to avoid contracting short-term debt. It is unclear whether this differential effect on debt maturity becomes more pronounced as the firm’s international activities intensify. Evidence from the foreign assets ratio suggests that highly internationalized firms experienced a larger increase in debt maturi-ty than moderately internationalized firms did, while evidence from the foreign sales ratio suggests that their increase was smaller. Additional tests bear out that the positive influence of multinationality persisted from 2008 to 2011 and vanished in 2012. Moreover, multinationals from countries with a more severe financial crisis reduced their short-term debt usage further, which is consistent with credit being less available in those areas. Furthermore, we reveal that the weight of bank credit in the financing of the private sector affects the debt maturity of multinationals less negatively compared to local firms. That is, we consistently obtain a less negative effect of bank credit on the debt maturity of multinationals. In some regressions, we find that bank concentration affects the debt maturity of local firms less negatively compared to multinationals, which is consistent with lending relationships generally being stronger for domestic companies. However, due to a combination of reliability and endogeneity concerns, the link between lending relationships and multinationality requires further investigation.

(7)

6

2. Hypotheses

2.1 Background: the financial crisis, bank credit, and the role of debt maturity

Why did bank credit become less available during the crisis? One answer comes from Brun-nermeier (2009), who describes how large portfolio losses can lead banks to refuse new loans to corporations. As most banks are highly levered, they must either raise capital or sell assets to overcome equity depreciations. Not renewing existing loans and not providing novel loans are two methods banks use to shrink their asset base. These methods can explain how enor-mous losses amongst major banks could have led to a severe contraction in the supply of bank debt in the period 2008-2012. Although some scholars see little reason to associate the reduc-tion in capital expenditure with the bank lending channel (e.g. Kahle & Stulz, 2013), a num-ber of papers provide strong empirical evidence for a causal link between the two phenomena. Consistent with a direct connection, Duchin et al. (2010) find that US firms reduced invest-ment most if they relied on external funds, while Iyer et al. (2014) find that Portuguese banks reduced their credit supply most if they depended on interbank borrowing. Bucă and Ver-meulen (2017) further show that investment fell more in industries that depended on banks.

Given the contraction in bank credit, the inopportune maturity composition of existing debt became an important contributor to decreasing investment. Almeida et al. (2011) demon-strate that US companies with a large portion of their debt maturing at the onset of the crisis struggled to renew their loans and hence experienced heavier investment cutbacks. Specifical-ly, corporate investment fell by 2.5% more for firms holding long-term debt that matured in the last quarter of 2007 than for comparable firms whose debt matured well afterwards. The authors find no effect of maturity composition on investment for firms that did not depend on long-term debt for funding. Vermoesen and Deloof (2013) observe a similar relationship for Belgian SMEs: investment fell most sharply for firms whose long-term debt matured within a year after the crisis started. The authors establish that maturity composition did not reduce corporate investment in periods where no credit supply shock occurred.

(8)

con-7

ditions worsened. Evidence from Campello et al. (2010) also indicates that credit lines attenu-ated the negative effect of the financial crisis on corporate spending, and that conversely, when access to credit lines was limited, firms chose either to save or to invest. Campello et al. (2012) further demonstrate that credit lines were the main instruments responsible for provid-ing the liquidity European firms needed to cope with increased credit restrictions.

Since credit lines commonly have shorter maturities than bank loans (Jiménez et al., 2009), it follows that the maturity of corporate debt declined during the crisis. Gonzalez (2015) is the first to analyse specifically the development of debt maturity during the crisis on a global scale. The author demonstrates that corporate debt maturity decreased in the period 2008-2012, but only for firms that depended on external finance beforehand. Gonzalez (2015) evinces that the fall in debt maturity was due to a higher increase in short-term debt usage than in long-term debt usage.

2.2 First hypothesis: multinationality

From Gonzalez (2015), it becomes evident that heterogeneity in the dependence of firms on external finance resulted in a differential effect of the crisis on the maturity of corporate debt. In a comparable fashion, our analysis aims to uncover an additional feature that mitigates the effect of increasingly stringent credit conditions on corporate debt maturity, i.e. firm interna-tionalization. The question then becomes, did the crisis affect the debt maturity of multina-tionals differently compared to local firms?

(9)

8

decreased during the financial crisis (Gonzalez, 2015), we should expect that it declined less for multinationals due to their proportion of long-term debt falling less substantially and/or their proportion of short-term debt rising less substantially.

Nevertheless, without the force of empirical studies on the lending behaviour of non-US multinationals during the crisis, it is hard to disentangle the factors that are responsible for the (potential) positive relationship between multinationality and long-term debt usage. Cru-cially, we cannot ascertain whether it stems from multinationals having easier access to do-mestic capital, to foreign capital, or to both. Hypothetically, dodo-mestic banks could deem mul-tinationals more trustworthy solely in virtue of their being internationally diversified. In turn, this could make domestic banks more willing to extend long-term credit to multinationals. Although we might worry about the lack of empirical support for this particular explanation as well, it is at least consistent with a positive relationship between multinationality and ac-cess to debt. For instance, Park et al. (2013) show that the fraction of non-US multinationals that issue debt lies between 34.3% and 37.8%, whereas the corresponding fraction of domes-tic firms is only 29.8%. Importantly, the validity of such findings does not require multina-tionals to shift between domestic and international capital markets. As our data does not allow us to resolve this ambiguity, we will pre-emptively omit all references to the location of capi-tal markets when interpreting our results.

(10)

9

during currency depreciations and thereafter. Lastly, the authors bear out that affiliates rely on parent equity most heavily when domestic firms are maximally constrained. This last result in particular supports the contention that multinationals should have been less eager for short-term credit lines during the recession. We refer to the resulting hypothesis as the multination-ality hypothesis.

Multinationality hypothesis: Multinationality is positively associated with the relationship

between the financial crisis and corporate debt maturity.

2.3 Second hypothesis: the weight of banks

Received theory dictates that banks positively affect global prosperity (Levine & Zervos, 1998; Demirgüç-Kunt & Maksimovic, 1998; Beck & Levine, 2004). In many countries, banks constitute the primary source of debt financing. As such, a better functioning banking sector will generally lead to a rise in economic activity. Because banks are amongst the most domi-nant participants in domestic capital markets, the weight of banks in the economy should sig-nificantly affect the capital structure decisions of firms. The question for then becomes, does the weight of banks affect the debt maturity of multinationals and domestic firms differently?

Some studies suggest that the weight of banks has a negative impact on corporate debt maturity in general, mainly because banks prefer giving out shorter loans to issuing long-term debt. According to Titman (2002), much of the corporate finance literature erroneously pre-supposes that global markets for different securities are perfectly integrated. Consequently, many researchers tacitly assume that firms need not worry about investor preferences when designing their capital structure. After all, suppliers of capital are free to construct a portfolio with any cash flow pattern that they see fit when markets are perfectly integrated. Unsurpris-ingly, this assumption does not survive in the real world. As Titman (2002) argues, it is much more prudent to assume that there are some pre-existing market conditions influencing the capital structure choices of firms.

(11)

10

with these points, Fan et al. (2012) find that debt maturity relates strongly and negatively to the size of a country’s bank deposits. In addition, Gonzalez (2015) reports that the weight of banks accentuated the overall reduction in debt maturity during the financial crisis, suggesting that market conditions become more influential as capital markets break down.

By way of extending our multinationality hypothesis, we conjecture that the flexibility in external funding enjoyed by multinationals weakens the impact of bank weight on the ma-turity of corporate debt. Similarly, the internal financing channel of multinationals constitutes another reason why domestic markets for bank debt should matter less for the debt maturity of internationalized firms. We will refer to the next hypothesis as the banking hypothesis.

Banking hypothesis: Multinationality is positively associated with the relationship between

bank weight and corporate debt maturity.

3. Methodology

3.1 Data

In order to ensure a close fit between this paper and related studies, we retrieve the majority of our firm-level information from Worldscope1. Our data focuses on the European subset that Gonzalez (2015) uses for his analyses. Consequently, our sample comprises 15 European countries: Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, the Nether-lands, Norway, Portugal, Spain, Sweden, Switzerland, and the United Kingdom. Importantly, this list covers a multiplicity of institutional environments and thereby guarantees a sufficient measure of variety in the weight of banks in the economy. To minimize the influence of sur-vivorship bias, we include both active and inactive firms. As additional filters, we consider only securities and primary quotes, and we remove all firms that operate in the financial in-dustry (SIC codes 6000-6999). In the end, our sample consists of 2071 firms and 23,434 firm-year observations for 15 European countries over the period 2003-2016. Information to fill our country-level variables comes from several World Bank datasets. Please refer to appendix A for a comprehensive overview of all the variables and their original sources.

3.2 Models and variables

Our first model aims to capture the differential impact of the crisis on the debt maturity of multinationals and local firms. The regressions regarding the multinationality hypothesis are therefore variations upon the following benchmark:

(12)

11

DEBT MAT_it = a₀+ a₁DCRISIS + a₂MNC_it−1+ a₃(DCRISIS x MNC_it−1)

+ a4CONTROLS + ∑ INDUSTRY + ∑ COUNTRY + εit (1)

where CONTROLS is a vector that includes multiple firm-level and country-level control variables. The regressions regarding the banking hypothesis follow a comparable benchmark:

DEBT MAT_it = a₀+ a₁DCRISIS + a₂MNC_it−1+ a₃BANK_kt+ a₃(MNC_it−1 x BANK_kt)

+ a₄CONTROLS + ∑ INDUSTRY + ε_it (2)

In both models, debt maturity (DEBT MAT) is our independent variable. We define debt ma-turity as the percentage of a firm’s total debt that has a mama-turity of more than one year. Using this definition dovetails with prior literature (e.g. Fan et al., 2012; Antoniou et al, 2006; Gon-zalez, 2015). In testing the multinationality hypothesis, we mainly focus on two terms, the first being the crisis dummy DCRISIS. Given that long-term loans helped sustain corporate investment from 2008 until 2012 (Almeida et al., 2011), we assign DCRISIS the value one for the years 2008, 2009, 2010, 2011, and 2012, and zero otherwise. Second, we introduce into our model the multinationality dummies MNC10 and MNC50. Conforming to a strand of re-search that harkens back to Geyikdagi (1981) and Fatemi (1984), we use the ratio of foreign sales to consolidated sales as a proxy for multinationality. We categorize a firm as multina-tional if it reports a foreign sales ratio of 10% or more, which conforms to prior studies on firm internationalization (e.g. Gonenc, 2005; Doukas & Pantzalis, 2003) and to the official requirements of the Statement of Financial Accounting Standard No. 14 (FASB 1976). We add the MNC50 dummy to isolate the effect of the crisis on those firms in our sample that are most internationalized. Accordingly, MNC10 (MNC50) takes the value one if the firm’s for-eign sales ratio is at least 10% (50%), and zero otherwise.

(13)

12

annual fraction of bank assets held by the three largest commercial banks in the country (Demirgüç-Kunt et al., 2004; Beck et al., 2006). Although BANK CONC is more vulnerable to endogeneity charges compared to BANK CRED, we include it to isolate the impact of rela-tionship banking. Higher bank market concentration incentivizes banks to put effort into ac-quiring soft information by developing close relationships with borrowers over time (Boot, 2000). The resulting proximity between the bank and the borrower mitigates problems of asymmetric information and increases the availability of credit. In keeping with our banking hypothesis, local firms should be more eager to establish close ties with domestic banks to circumvent barriers to external credit and internal financing.

Next, we introduce a number of variables that control for the impact of various firm and country characteristics on debt maturity. We refer to appendix B for independent justifi-cations regarding the relevance of each variable. At the firm level, we control for asset maturi-ty (ASSET MAT), growth (GROWTH), size (SIZE), firm qualimaturi-ty (QUALITY), earnings vola-tility (VOL EBIT), leverage (LEV), and dependence on external finance (EXT FIN). Combin-ing these control variables puts our model in line with similar empirical research (Antoniou et al., 2006; Gonzalez, 2015).

On the country level, we first account for the influence of creditor rights (C RIGHTS). For this variable, we use a well-known index developed by Djankov et al. (2007). Their index aggregates four powers of secured lenders in case of a bankruptcy. Accordingly, C RIGHTS ranges from zero (poor creditor rights) to four (strong creditor rights). To control for the effi-ciency of the country’s legal system, we introduce a second variable, namely the rule of law (RULE LAW). Here, we use an index that originates in Kaufmann et al. (2009) and embodies the degree to which agents trust in and abide by the rules of society. The value of RULE LAW varies between -2.5 to 2.5. Low values indicate a less efficient legal system, and high values a more efficient one.

Lastly, we introduce country and industry dummies into our multinationality model and industry dummies into our banking model to mitigate the effect of unobserved variables2. We use panel data for all our estimations. Before testing, we winsorize continuous firm-level variables at the bottom 1% and the top 99% to subvert the impact of outliers. In addition, we lag all independent firm variables by one year to mitigate problems of endogeneity.

2_{In case of the banking model, country dummies would nullify a non-trivial part of the cross-sectional}

(14)

13

3.3 Descriptive statistics

Table 1 provides mean values for our firm variables. To gain an overview of how those values changed for different firm types, we divide table 1 into three panels. Panel A presents pre-crisis (2003-2007), mid-pre-crisis (2008-2012), and post-pre-crisis (2013-2016) means for local firms. Panels B and C do the same for firms with a foreign sales ratio of at least 10% and 50%, re-spectively. We note that local firms comprise 35.24% of our sample, while multinationals make up 64.76%. In turn, highly internationalized firms represent 34.14% of the observations, which amounts to 52.72% of all multinationals. In accord with our multinationality hypothe-sis, we observe that the average debt maturity during the recession was lower for local firms and higher for multinationals compared to the pre-crisis mean. The observed increase in debt

Table 1: Variable means

Obs. Pre-crisis Mid-crisis Post-crisis Overall

Panel A: local firms

DEBT MAT 8,257 0.56 0.55 0.60 0.57 ASS MAT 8,257 26.36 25.07 24.80 25.45 GROWTH 8,257 2.18 1.83 1.94 1.99 SIZE 8,257 11.42 11.68 11.86 11.64 FIRM QUAL 8,257 0.05 0.23 0.11 0.13 VOL EBIT 8,257 1.51 1.64 1.50 1.55 LEV 8,257 0.61 0.80 0.76 0.72 EXT FIN 8,257 0.24 0.23 0.40 0.28 Panel B: multinationals (10%) DEBT MAT 15,177 0.59 0.62 0.65 0.62 ASS MAT 15,177 26.01 24.61 23.94 24.93 GROWTH 15,177 1.95 1.77 1.91 1.87 SIZE 15,177 13.27 13.55 13.68 13.49 FIRM QUAL 15,177 0.15 0.24 0.12 0.17 VOL EBIT 15,177 1.23 1.25 1.19 1.23 LEV 15,177 0.62 0.73 0.67 0.67 EXT FIN 15,177 0.24 0.23 0.23 0.23 Panel C: multinationals (50%) DEBT MAT 8,001 0.63 0.65 0.67 0.65 ASS MAT 8,001 27.66 26.37 25.91 26.70 GROWTH 8,001 1.94 1.82 1.94 1.90 SIZE 8,001 13.42 13.69 13.85 13.64 FIRM QUAL 8,001 0.15 0.18 0.15 0.16 VOL EBIT 8,001 1.28 1.27 1.28 1.28 LEV 8,001 0.64 0.77 0.73 0.71 EXT FIN 8,001 0.25 0.24 0.24 0.25

(15)

14

maturity amongst multinationals is compatible with Gonzalez (2015), who also observes an increase in debt maturity amongst a subset of firms, i.e. amongst those that did not depend on external finance before 2008. Similarly, we observe an increase in debt maturity amongst the subset of firms that are multinationals. We conduct two independent-samples t-tests to ascer-tain the observed difference between the average debt maturity of local firms and multination-als for the pre-crisis and mid-crisis time intervmultination-als, respectively. We obtain significant differ-ences in pre-crisis scores (t(8368) = -5, p < 0.01) and mid-crisis scores (t(8440) = -10, p < 0.01). This bolsters the prediction that multinationality is positively associated with the rela-tion between the crisis and debt maturity. To proceed with the control variables, they align with Gonzalez (2015) in that earnings volatility and leverage show an increase during crisis years, whereas asset maturity and growth exhibit a decrease. Contrastingly, we observe a rise instead of a drop in size and firm quality across all panels.

Table 2 presents descriptive statistics per country. Countries that represent a signifi-cant part of the observations are the United Kingdom (23.16%), France (18.48%), and Ger-many (15.06%). With representations of 0.90% and 1.75% respectively, Ireland and Austria have the smallest presence. Furthermore, we see that the values for debt maturity tend to clus-ter around the 60% mark. For all country-level variables, we see substantial differences be-tween geographies. Table A1 provides the sample distribution across industries and table A2 provides the correlation matrix. Please note that all A-tables are included in appendix C.

Table 2: Country statistics

Obs. DEBT MAT RULE LAW C RIGHTS BANK CRED BANK CONC

Austria 410 0.60 1.86 3 90.59 71.62 Belgium 688 0.57 1.38 2 59.88 73.15 Denmark 831 0.59 1.96 3 177.23 82.18 Finland 1,040 0.62 1.98 1 80.66 93.40 France 4,331 0.58 1.45 0 88.70 61.29 Germany 3,528 0.59 1.70 3 91.98 72.90 Ireland 212 0.74 1.68 1 114.22 72.72 Italy 1,495 0.53 0.45 2 82.17 56.69 Netherlands 747 0.62 1.83 3 114.45 82.98 Norway 719 0.74 1.96 2 97.28 93.90 Portugal 427 0.58 1.10 1 136.68 86.18 Spain 836 0.59 1.11 2 141.99 64.65 Sweden 1,596 0.58 1.94 1 115.719 94.77 Switzerland 1,147 0.61 1.86 1 156.40 86.98 UK 5,427 0.63 1.72 4 155.32 54.84

(16)

15

4. Empirical analysis

4.1 Testing the multinationality hypothesis

Table 3 displays the results of our first panel regressions. Column (1) presents the effect of the financial crisis when controlling for the firm-level determinants of debt maturity. We see that DCRISIS is negative and significant, suggesting that the overall maturity of debt shortened during the crisis. For all but growth and leverage, our firm-level controls align with the litera-ture – again, we refer to appendix B for a lengthier discussion. Column (2) shows the regres-sion results when also considering country-level determinants. The firm controls remain un-changed, whereas DCRISIS retains its negative direction but loses significance. We can inter-pret this as a sign that our analysis tracks the paper by Gonzalez (2015). There, the author also obtains a negative and insignificant DCRISIS coefficient after accounting for country effects, while it was also negative and significant before. Bank concentration is negative and signifi-cant, suggesting that bank weight negatively influences the maturity of debt. The insignifi-cance of bank credit is not in line with the significant coefficient found by Gonzalez (2015).

We examine our multinationality hypothesis in columns (3) to (6) of table 3. To reiter-ate, we theorize that the financial crisis had a different effect on the debt maturity structure of multinationals compared to local firms. Specifically, we hypothesize that the maturity of cor-porate debt decreased less for multinationals than it did for domestic firms. We test this pre-diction by including the interaction between DCRISIS and MNC10 into our estimations. Cur-rently, we obtain two central results. First, DCRISIS turns negative and significant in both specifications, revealing strong evidence that local firms suffered a reduction in corporate debt maturity during the recession. Second, we see that DCRISIS x MNC10 shows up posi-tive and significant in both specifications, indicating that the crisis had a less detrimental ef-fect on the debt maturity of multinationals compared to local firms. Because the positive in-teraction coefficient is everywhere larger than the negative DCRISIS coefficient, our results suggest that the debt maturity of multinationals did not decline in the wake of the recession. Instead, they show an increase in debt maturity for internationalized firms.

(17)

16

highly internationalized firms still decreased their debt maturity less than local firms did, they were seemingly unable to increase their debt maturity in the way moderately internationalized firms could. We rectify this apparent contradiction with our previous findings by adding country controls in column (6). The resulting interaction coefficient is again larger in magni-tude, again suggesting that the debt maturity of multinationals generally increased.

Table 3: Debt maturity and foreign sales ratio

(1) (2) (3) (4) (5) (6) DCRISIS -0.0108*** -0.0074 -0.0329*** -0.0295*** -0.0330*** -0.0269*** (0.0042) (0.0049) (0.0077) (0.0082) (0.0077) (0.0084) MNC10 -0.0033 -0.0027 (0.0062) (0.0062) DCRISIS x MNC10 0.0342*** 0.0337*** (0.0091) (0.0091) MNC50 0.0275*** 0.0281*** (0.0072) (0.0072) DCRISIS x MNC50 0.0293*** 0.0286*** (0.0100) (0.0100) ASS MAT 0.0022*** 0.0022*** 0.0022*** 0.0022*** 0.0022*** 0.0022*** (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) GROWTH -0.0013 -0.0014 -0.0013 -0.0014 -0.0013 -0.0013 (0.0010) (0.0011) (0.0010) (0.0010) (0.0012) (0.0012) SIZE 0.0298*** 0.0297*** 0.0290*** 0.0288*** 0.0253*** 0.0252*** (0.0010) (0.0010) (0.0011) (0.0011) (0.0013) (0.0013) FIRM QUAL 0.0028*** 0.0028*** 0.0028*** 0.0028*** 0.0028*** 0.0028*** (0.0007) (0.0007) (0.0007) (0.0007) (0.0008) (0.0008) VOL EBIT -0.0005 -0.0005 -0.0005 -0.0005 -0.0004 -0.0004 (0.0006) (0.0006) (0.0006) (0.0006) (0.0008) (0.0008) LEV 0.0021 0.0023 0.0022 0.0023 0.0003 0.0005 (0.0019) (0.0019) (0.0019) (0.0019) (0.0023) (0.0023) EXT FIN 0.1365*** 0.1364*** 0.1364*** 0.1362*** 0.1171*** 0.1169*** (0.0130) (0.0130) (0.0130) (0.0129) (0.0132) (0.0132) RULE LAW 0.0193 0.0203 0.0320 (0.0306) (0.0306) (0.0370) C RIGHTS -0.0204** 0.0852*** 0.1024*** (0.0086) (0.0143) (0.0155) BANK CRED -0.0001 -0.0000 -0.0002 (0.0002) (0.0002) (0.0002) BANK CONC -0.0013*** -0.0013*** -0.0010** (0.0004) (0.0004) (0.0004) cons 0.0331 0.1533*** 0.1017*** 0.1018 0.1583*** 0.1109 (0.0246) (0.0562) (0.0203) (0.0776) (0.0233) (0.0922) Obs. 21404 21404 21404 21404 14819 14819 R-squared 0.1295 0.1300 0.1303 0.1308 0.1421 0.1425

(18)

17

Our multinationality hypothesis offers two explanations for the different effect of the crisis on the debt maturity of multinationals and local firms. One, multinationals have easier access to long-term credit. Two, multinationals require less short-term debt due to the availability of internal funds and/or long-term credit. Because a ‘simple’ measure of debt maturity does not distinguish between changes in short-term debt usage and changes in long-term debt usage, we employ alternative dependent variables in table 4. The dependent variable in columns (1), (3) and (5) is the ratio between long-term debt and lagged total assets (LTD). In columns (2), (4) and (6), it is the ratio between short-term debt and lagged total assets (STD). Since these variables strictly do not constitute measures of debt maturity, we change the control variables to the traditional determinants of capital structure found in Rajan and Zingales (1995). These are profitability (PROF), growth (GROWTH), asset tangibility (TANG), and size (SIZE). In keeping with papers that show a positive association between creditor rights and debt usage (e.g. Demirgüç-Kunt & Maksimovic, 1999), we also control for creditor rights.

Table 4: Debt-to-assets ratios and foreign sales ratio

(1) LTD (2) STD (3) LTD (4) STD (5) LTD (6) STD DCRISIS -0.0110*** -0.0006 -0.0091 0.0048 -0.0089 0.0048 (0.0041) (0.0019) (0.0080) (0.0040) (0.0081) (0.0040) MNC10 0.0200** 0.0067* (0.0091) (0.0036) DCRISIS x MNC10 -0.0025 -0.0084* (0.0093) (0.0043) MNC50 0.0515*** 0.0059 (0.0132) (0.0047) DCRISIS x MNC50 -0.0156 -0.0090* (0.0115) (0.0048) PROF -0.2764*** -0.1221*** -0.2739*** -0.1216*** -0.3025*** -0.1366*** (0.0729) (0.0279) (0.0725) (0.0279) (0.0897) (0.0341) GROWTH 0.0042* 0.0032** 0.0042* 0.0032** 0.0037 0.0029 (0.0025) (0.0016) (0.0025) (0.0016) (0.0032) (0.0019) TANG 0.0020*** 0.0001** 0.0020*** 0.0001** 0.0020*** 0.0001* (0.0002) (0.0001) (0.0002) (0.0001) (0.0002) (0.0001) SIZE 0.0042*** -0.0068*** 0.0026 -0.0071*** -0.0008 -0.0062*** (0.0014) (0.0006) (0.0018) (0.0007) (0.0023) (0.0008) C RIGHTS -0.0136*** 0.0064*** 0.0852*** 0.0031 0.1181*** -0.0031 (0.0044) (0.0016) (0.0167) (0.0106) (0.0189) (0.0129) cons 0.1304*** 0.1099*** 0.0068 0.1838*** 0.0076 0.1808*** (0.0305) (0.0102) (0.0312) (0.0170) (0.0288) (0.0203) Obs. 23106 23106 23106 23106 15992 15992 R-squared 0.0386 0.0589 0.0391 0.0592 0.0392 0.0624

(19)

18

Contrary to Gonzalez (2015), columns (1) and (2) of table 4 suggest that a decrease in long-term debt usage (as opposed to an increase in short-long-term debt usage) was responsible for the overall fall in corporate debt maturity during the recession. Namely, DCRISIS has a negative and significant effect on long-term debt usage, and a negative and insignificant effect on short-term debt usage. As we divert from Gonzalez (2015) by focusing solely on firms from Western Europe, this contradiction might result from Western European firms being more liquid than the global average and hence having to resort less to short-term credit in times of crisis. If this were true, then the increasing difficulty of accessing long-term loans would still hold as an explanation for the negative effect on long-term debt usage we observe here. In accord with this explanation, Albouy et al. (2017) evince that European firms hold relatively high amounts of cash compared to US firms. However, their comparison does not include the rest of the world, so it is unclear whether higher liquidity actually drives our diverging from Gonzalez’ (2015) global study. Furthermore, cash-holdings tend to go down with better inves-tor protection (Brockman & Chung, 2003), higher capital market development (Kim et al., 1998), and improved banking relationships (Ferreira & Vilela, 2004). In addition, European firms reverted to external sources of liquidity to obtain the financial resources they needed to cope with increased credit restrictions during the recession (Campello et al., 2012). These reasons seem to speak against the claim that Western European firms hold more cash than the global average. To proceed with the analysis, we see no significant effect of the crisis on the debt ratios of local firms in columns (3) and (4). We do note a negative and significant effect of the crisis on the short-term debt ratio of multinationals, suggesting that the differential im-pact on debt maturity lies in the fact that multinationals used less short-term debt than local firms did. This result persists when we substitute MNC50 for MNC10 in columns (5) and (6). Overall, the current regressions align with the view that multinationals can avoid contracting short-term debt by reallocating funds internally. For now, we find no evidence that is con-sistent with multinationals being better able to access external long-term debt during the re-cession, or with multinationals offsetting short-term debt with long-term debt.

4.2 Robustness of the multinationality hypothesis

(20)

19

operations being present in multiple countries. If having foreign sales does not establish an operation presence per se, using the foreign sales ratio to proxy for multinationality will sig-nificantly hurt the validity of our results. Because the empirical analysis exclusively relies on the foreign sales ratio, we will repeat the regressions with an alternative measure of multina-tionality, namely the foreign assets ratio. This ratio partly addresses Burgman’s (1996) worry by looking at the internationalization of assets instead of the internationalization of sales, the first potentially being a superior approximation of the firm’s actual presence in other geogra-phies. We display the results of our first robustness test in table 5. Similar to our previous regressions, we assign MNC10* (MNC50*) the value one if the firm’s foreign assets ratio is at least 10% (50%), and zero otherwise.

To begin, we note that DCRISIS exhibits a similar pattern as it does in table 3: it starts negative and significant in column (1) and turns negative and insignificant in column (2). The firm and country controls behave in a comparable fashion as well. As becomes evident in col-umns (3) to (6), the current estimations strongly confirm our central finding that the crisis affected the debt maturity of multinationals differently compared to local firms. Again, the positive interaction terms are larger in magnitude than the negative DCRISIS coefficients, implying that multinationals increased their debt maturity when local firms decreased theirs. We see one notable change in table 5: the positive interaction terms between multinationality and the crisis are larger in magnitude for MNC50 than for MNC10. As opposed to what we found before, this suggests that the crisis had a less detrimental impact on the debt maturity of highly internationalized firms compared to moderately internationalized firms. This is more consistent with our hypothesis than the notion that firms become less resilient to negative in-fluences as internationalization intensifies. Therefore, we may reasonably contrive that the foreign assets ratio is a superior tool for measuring the local operation presence highlighted by Jang (2017). We repeat our analysis of the short- and long-term debt-to-assets ratios in table A3. In line with our previous results, we note a negative and significant effect of the crisis on overall long-term debt usage and a negative and insignificant effect on overall short-term debt usage. We also confirm that the differential effect of the recession on the debt ma-turity of multinationals is due to multinationals reducing their usage of short-term debt.

(21)

20

RA TIO), total foreign sales (FOR SAL), and total foreign assets (FOR ASS). Looking at the interaction effects between these proxies and the crisis dummy, we find little evidence for a linear relationship. The only measure that results in a positive and significant – though excess-

Table 5: Debt maturity and foreign assets ratio

(1) (2) (3) (4) (5) (6) DCRISIS -0.0108*** -0.0074 -0.0208*** -0.0174*** -0.0206*** -0.0153** (0.0042) (0.0049) (0.0060) (0.0066) (0.0060) (0.0069) MNC10* 0.0058 0.0062 (0.0057) (0.0057) DCRISIS x MNC10* 0.0217*** 0.0217*** (0.0083) (0.0083) MNC50* 0.0343*** 0.0347*** (0.0078) (0.0078) DCRISIS x MNC50* 0.0333*** 0.0328*** (0.0110) (0.0110) ASS MAT 0.0022*** 0.0022*** 0.0022*** 0.0022*** 0.0024*** 0.0024*** (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) GROWTH -0.0013 -0.0014 -0.0014 -0.0015 0.0006 0.0005 (0.0010) (0.0011) (0.0010) (0.0011) (0.0012) (0.0012) SIZE 0.0298*** 0.0297*** 0.0284*** 0.0283*** 0.0268*** 0.0267*** (0.0010) (0.0010) (0.0011) (0.0011) (0.0014) (0.0014) FIRM QUAL 0.0028*** 0.0028*** 0.0028*** 0.0028*** 0.0022*** 0.0022*** (0.0007) (0.0007) (0.0007) (0.0007) (0.0008) (0.0008) VOL EBIT -0.0005 -0.0005 -0.0006 -0.0006 -0.0006 -0.0005 (0.0006) (0.0006) (0.0006) (0.0006) (0.0008) (0.0008) LEV 0.0021 0.0023 0.0023 0.0024 0.0039* 0.0040* (0.0019) (0.0019) (0.0019) (0.0019) (0.0023) (0.0023) EXT FIN 0.1365*** 0.1364*** 0.1363*** 0.1361*** 0.1070*** 0.1068*** (0.0130) (0.0130) (0.0130) (0.0130) (0.0125) (0.0125) RULE LAW 0.0193 0.0229 -0.0100 (0.0306) (0.0306) (0.0377) C RIGHTS -0.0204** 0.0857*** 0.0884*** (0.0086) (0.0143) (0.0165) BANK CRED -0.0001 -0.0001 -0.0002 (0.0002) (0.0002) (0.0002) BANK CONC -0.0013*** -0.0013*** -0.0009** (0.0004) (0.0004) (0.0004) cons 0.0331 0.1533*** 0.1037*** 0.1020 0.0879*** 0.1293 (0.0246) (0.0562) (0.0204) (0.0776) (0.0247) (0.0950) Obs. 21404 21404 21404 21404 14614 14614 R-squared 0.1295 0.1300 0.1302 0.1307 0.1467 0.1470

(22)

21

ively small – interaction coefficient is total foreign sales. All of the other interaction coeffi-cients are insignificant. Considering what we saw in table 5, it remains possible that financing benefits increase as firms expand their activities abroad. What our current regressions show is that most likely, those benefits do not move in tandem with foreign sales or foreign assets. This being so, we may say that the local operation presence in itself determines whether mul-tinationals enjoy financing benefits. There may exist some co-movement between the ‘usabil-ity’ of this operation presence for internal financing and the firm’s foreign assets, but that co-movement is not direct enough to observe a significant linear relationship here.

(23)

22

Table 6: Debt maturity, foreign sales ratio, and crisis periods

(1) (2) (3) (4) (5) (6) DCRISIS1 -0.0138** -0.0100 -0.0359*** -0.0323*** -0.0365*** -0.0302*** (0.0057) (0.0064) (0.0105) (0.0109) (0.0104) (0.0111) DCRISIS2 -0.0084 -0.0047 -0.0348*** -0.0311*** -0.0345*** -0.0283** (0.0058) (0.0063) (0.0106) (0.0110) (0.0106) (0.0112) DCRISIS3 -0.0093 -0.0078 -0.0226 -0.0209 -0.0225 -0.0187 (0.0080) (0.0083) (0.0147) (0.0149) (0.0146) (0.0149) MNC10 -0.0033 -0.0027 (0.0062) (0.0062) DCRISIS1 x MNC10 0.0342*** 0.0339*** (0.0124) (0.0124) DCRISIS2 x MNC10 0.0407*** 0.0402*** (0.0126) (0.0126) DCRISIS3 x MNC10 0.0207 0.0200 (0.0175) (0.0175) MNC50 0.0275*** 0.0280*** (0.0072) (0.0072) DCRISIS1 x MNC50 0.0277** 0.0270** (0.0137) (0.0137) DCRISIS2 x MNC50 0.0368*** 0.0361*** (0.0138) (0.0139) DCRISIS3 x MNC50 0.0169 0.0165 (0.0192) (0.0192) ASS MAT 0.0022*** 0.0022*** 0.0022*** 0.0022*** 0.0022*** 0.0022*** (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) GROWTH -0.0013 -0.0013 -0.0013 -0.0013 -0.0012 -0.0013 (0.0011) (0.0011) (0.0011) (0.0011) (0.0012) (0.0012) SIZE 0.0298*** 0.0297*** 0.0290*** 0.0288*** 0.0253*** 0.0252*** (0.0010) (0.0010) (0.0011) (0.0011) (0.0013) (0.0013) FIRM QUAL 0.0028*** 0.0028*** 0.0028*** 0.0028*** 0.0028*** 0.0028*** (0.0007) (0.0007) (0.0007) (0.0007) (0.0008) (0.0008) VOL EBIT -0.0005 -0.0005 -0.0005 -0.0005 -0.0004 -0.0004 (0.0006) (0.0006) (0.0006) (0.0006) (0.0008) (0.0008) LEV 0.0020 0.0022 0.0020 0.0022 0.0001 0.0003 (0.0020) (0.0020) (0.0020) (0.0020) (0.0023) (0.0023) EXT FIN 0.1368*** 0.1366*** 0.1366*** 0.1365*** 0.1174*** 0.1172*** (0.0130) (0.0130) (0.0130) (0.0130) (0.0132) (0.0132) RULE LAW 0.0205 0.0215 0.0335 (0.0306) (0.0307) (0.0370) C RIGHTS -0.0208** 0.0854*** 0.1027*** (0.0086) (0.0143) (0.0155) BANK CRED -0.0001 -0.0000 -0.0002 (0.0002) (0.0002) (0.0002) BANK CONC -0.0013*** -0.0013*** -0.0010** (0.0004) (0.0004) (0.0004) cons 0.0332 0.1507*** 0.1015*** 0.0978 0.1582*** 0.1044 (0.0246) (0.0564) (0.0203) (0.0778) (0.0233) (0.0923) Obs. 21404 21404 21404 21404 14819 14819 R-squared 0.1296 0.1301 0.1304 0.1309 0.1422 0.1426

(24)

23

insignificant. We might ascribe this lack of significance to the foreign sales ratio being an inferior proxy for multinationality. Alternatively, the effect may only be strong enough if we consider the crisis as a whole and do not divide it into smaller subparts. The last interpretation is less tenable, however, because the foreign assets ratio does deliver significant sub-effects.

A final worry is that crisis dummies do not control for differences in the effect of the recession on the economies of individual countries. Consequently, we repeat our regressions with two approximations of the extensiveness of the crisis, which are crisis intensity (CR INT) and economic growth (EC GROWTH). To start, CR INT it is the absolute difference between the average GDP growth rates for the periods 2003-2007 and 2008-2012. In non-crisis years, CR INT takes the value of zero. We divide the output regarding the foreign sales ratio and the foreign assets ratio between tables A8 and A9. Columns (3) to (6) of both tables present a negative and highly significant coefficient for CR INT, suggesting that local firms from countries that suffered a more intense financial crisis experienced a starker decline in debt maturity. We obtain positive and highly significant interactions between multinationality and crisis intensity that are larger in magnitude across all estimations, implying that multina-tionals from countries that suffered more actually experienced a higher increase in debt ma-turity than did multinationals from countries that suffered less. Additional analyses in tables A10 and A11 suggest that this is due to a higher decrease in short-term debt usage amongst multinationals from countries with a more intense crisis. This observation is consistent with credit becoming less available and/or lending conditions deteriorating further in countries that were hit hardest.

(25)

24

EC GROWTH may yield less relevant insights into firm debt usage during the crisis period specifically. To proceed, we again divide the output regarding the foreign sales ratio and the foreign assets ratio between tables A12 and A13. EC GROWTH starts positive and significant in three out of four regressions using only firm-level control variables, but turns insignificant when including country controls as well. Hence, there is no reliable effect of economic growth on the debt maturity of local firms. The multinationality interactions are negative and significant in all estimations, meaning that higher economic growth rates are associated with lower debt maturity amongst multinationals. Additional analyses in tables A14 and A15 bear out that this is likely due to multinationals using more short-term debt as the economy strengthens, although the results suggesting this are relatively weak. Still, multinationals using more short-term debt is consistent with credit becoming increasingly available and/or lending conditions becoming increasingly attractive as the economy strengthens, potentially inducing multinationals to lean less on internal funding and more on external borrowing.

4.3 Testing the banking hypothesis

We test our banking hypothesis in table 7. We expect the weight of bank credit in the financ-ing of the private sector to have a less negative impact on the debt maturity of multinationals compared to local firms. Since having an international network gives firms more funding flex-ibility, local market conditions will shape the capital structure of multinationals less than the capital structure of firms that rely for a larger part on domestic capital. In all specifications, we observe a negative and significant bank credit coefficient, confirming that the weight of banks negatively affects the debt maturity of local firms. In columns (2) and (3), the interac-tions between multinationality and bank credit are positive and significant. This suggests that indeed, the weight of banks affects multinationals less. In addition, we obtain negative and significant BANK CONC coefficients across all specifications. Here, however, multination-ality does not visibly mitigate the relationship. In the last two columns, we add DCRISIS to the interactions to see whether the results are different for crisis years. While the interaction between MNC50 and bank credit remains positive and significant in column (5), the interact- Table 7: Debt maturity, foreign sales ratio, and bank weight

(26)

25 Table 7 continued (1) (2) (3) (4) (5) MNC10 -0.0204 -0.0008 (0.0275) (0.0352) MNC10 x B CRED 0.0004*** 0.0003 (0.0001) (0.0002) MNC10 x B CONC -0.0001 -0.0005 (0.0003) (0.0004) MNC50 -0.0307 -0.0107 (0.0300) (0.0382) MNC50 x B CRED 0.0005*** 0.0004* (0.0001) (0.0002) MNC50 x B CONC 0.0002 -0.0001 (0.0003) (0.0004) DC x MNC10 x B CR 0.0000 (0.0003) DC x MNC10 x B CO 0.0006 (0.0006) DC x MNC50 x B CR 0.0001 (0.0003) DC x MNC50 x B CO 0.0005 (0.0007) ASS MAT 0.0022*** 0.0022*** 0.0022*** 0.0022*** 0.0022*** (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) GROWTH -0.0015 -0.0015 -0.0013 -0.0015 -0.0013 (0.0011) (0.0011) (0.0012) (0.0011) (0.0012) SIZE 0.0277*** 0.0268*** 0.0237*** 0.0268*** 0.0238*** (0.0009) (0.0010) (0.0012) (0.0010) (0.0012) FIRM QUAL 0.0028*** 0.0028*** 0.0027*** 0.0028*** 0.0027*** (0.0007) (0.0007) (0.0008) (0.0007) (0.0008) VOL EBIT -0.0006 -0.0006 -0.0005 -0.0006 -0.0004 (0.0006) (0.0006) (0.0008) (0.0006) (0.0008) LEV 0.0031 0.0033* 0.0022 0.0033* 0.0021 (0.0019) (0.0019) (0.0023) (0.0019) (0.0023) EXT FIN 0.1358*** 0.1364*** 0.1194*** 0.1361*** 0.1194*** (0.0136) (0.0135) (0.0137) (0.0135) (0.0137) RULE LAW 0.1030*** 0.1038*** 0.1211*** 0.1042*** 0.1223*** (0.0062) (0.0062) (0.0072) (0.0062) (0.0073) C RIGHTS 0.0023 0.0027 0.0043** 0.0027 0.0044** (0.0017) (0.0017) (0.0021) (0.0017) (0.0021) cons 0.2057*** 0.2018*** 0.2290*** 0.2066*** 0.2362*** (0.0212) (0.0275) (0.0285) (0.0334) (0.0342) Obs. 21353 21353 14776 21353 14776 R-squared 0.1213 0.1221 0.1355 0.1229 0.1362

(27)

26

ion between MNC10 and bank credit turns insignificant in column (4). The interactions be-tween multinationality and bank concentration stay insignificant. Moreover, we report insig-nificant coefficients for all of the three-way interactions between DCRISIS, the banking dummies, and the multinationality dummies. Based on columns (2) and (3), we conclude that our current findings align with our prediction that bank weight affects the debt maturity of multinationals less negatively compared to local firms.

4.4 Robustness of the banking hypothesis

Rerunning our banking regressions with the foreign assets ratio in table A16 again results in positive and significant interactions between bank credit and multinationality, which corrobo-rates our main finding. In contrast to our earlier results, we obtain negative and significant coefficients for MNC10* x B CONC, suggesting that bank concentration affects the debt ma-turity of multinationals more negatively compared to local firms. This finding is consistent with the claim that local firms have more incentive to develop close lending relationships with domestic banks. If it were true that ceteris paribus, the negative effect of bank concentration on corporate debt maturity weakens as firms become increasingly domestic, we would expect the negative influence of multinationality to intensify – or at least remain significant – as we consider firms that are most internationalized. Contrary to this expectation, the BANK CONC interaction terms become insignificant when substituting MNC10* for MNC50*. In turn, this can be construed as a lack of evidence for the view that lending relationships become stronger as the firm’s domestic focus increases. Noting this duality, it is possible that endogenous sample characteristics such as the capital structure preferences of corporations are responsible for the negative interactions between MNC10* and BANK CONC observed before. The com-bined problems of unreliability and endogeneity provide sufficient ground for future studies to investigate the relationship between bank concentration and multinationality more carefully.

5. Conclusion

(28)

differ-27

ent database, it is conceivable that the calculations behind it do not coincide with the underly-ing makeup of our other (dependent and independent) variables. In turn, this would explain the observed lack of association between leverage and growth opportunities on the one hand and debt maturity on the other, with a possible spill-over into earnings volatility. The ultimate consequences of this abnormality for our central finding are unclear. A second limitation is our inability to approximate firm internationalization via the number of foreign subsidiaries. Complementing our model with this proxy presents a promising strategy for addressing Burgman’s (1996) concern that traditional measures of multinationality do not tell us whether foreign income comes from foreign sources. Some authors obtain foreign subsidiary data from Osiris (e.g. Park et al., 2013), but we could not access that database when writing this paper.

(29)

28

6. References

Aivazian, V., Ge, Y., Qiu, J., 2005. Debt maturity structure and firm investment. Financial Management 34, 107-119.

Albouy, M., Bonnet, C., Dupuy, P., Mchawrab, S., 2017. Why so much cash in European firms? An empirical study 1986-2015. Unpublished working paper.

Almeida, H., Campello, M., Laranjeira, B., Weisbenner, S., 2012. Corporate debt maturity and the real effects of the 2007 credit crisis. Critical Finance Review 1, 3-58.

Antoniou, A., Guney, Y., Paudyal, K., 2006. The determinants of debt maturity structure: evi-dence from France, Germany and the UK. European Financial Management 12, 161-194. Auken, H.E., Holman, T., 1995. Financial strategies of small, public firms: a comparative analysis with small, private firms and large, public firms. Entrepreneurship Theory and Prac-tice 20, 29-41.

Bae, K. H., Goyal, V. K., 2009. Creditor rights, enforcement, and bank loans. Journal of Fi-nance 64, 823–860.

Barclay, M.J., Smith, C.W., 1995. The maturity structure of corporate debt. Journal of Fi-nance 50, 609-631.

Beck, T., Demirgüç-Kunt, A., Levine, R., 2006. Bank concentration, competition, and crises: first results. Journal of Banking & Finance 30, 1581-1603.

Beck, T., Levine, R., 2004. Stock markets, banks, and growth: panel evidence. Journal of Banking & Finance 28, 423-442.

Bodnar, G.M., Tang, C., Weintrop, J., 2003. The value of corporate international diversifica-tion. Unpublished working paper.

Boot, W., 2000. Relationship banking: what do we know? Journal of Financial Intermediaries 9, 7-25.

Brockman, P., Chung, D.Y., 2003. Investor protection and firm liquidity. Journal of Finance 58, 921-938.

Brunnermeier, M., 2009. Deciphering the liquidity and credit crunch 2007–2008. Journal of Economic Perspectives 23, 77-100.

Bucă, A., Vermeulen, P., 2017. Corporate investment and bank-dependent borrowers during the recent financial crisis. Journal of Banking & Finance 78, 164-180.

Burgman, T.A., 1996. An empirical examination of multinational corporate capital structure. Journal of International Business Studies 27, 553-570.

(30)

29

Campello, M., Graham, J.R., Harvey, C.R., 2010. The real effects of financial constraints: evidence from a financial crisis. Journal of Financial Economics 97, 470-487.

Christophe, S.E., 1997. Hysteresis and the value of the U.S. multinational corporation. The Journal of Business 70, 435-462.

Creal, D.D., Robinson, L.A., Rogers, J.L., 2014. The multinational advantage. Unpublished working paper.

Demirgüç-Kunt, A., Laeven, L., Levine, R., 2004. Regulations, market structure, institutions, and the cost of financial intermediation. Journal of Money, Credit, and Banking 36, 593–622. Demirgüç-Kunt, A., Maksimovic, V., 1998. Law, finance, and firm growth. Journal of Fi-nance 53, 2107-2137.

Demirgüç-Kunt, A., Maksimovic, V., 1999. Institutions, financial markets, and firm debt ma-turity. Journal of Financial Economics 54, 295–336.

Denis, D.J., Denis, D.K., Yost, K., 2002. Global diversification, industrial diversification, and firm value. Journal of Finance 57, 1951-1979.

Desai, M.A, Forbes, K.J., 2008. Financial constraints and growth: multinational and local firm responses to currency depreciations. Review of Financial Studies 21, 2857-2888.

Desai, M.A., Foley, C.F., Hines Jr., J.R., 2004. A multinational perspective on capital struc-ture choice and internal capital markets. Journal of Finance 59, 2451–2481.

Diamond, D., 2004. Presidential address, committing to commit: short-term debt when en-forcement is costly. Journal of Finance 59, 1447-1479.

Djankov, S., McLiesh, C., Shleifer, A., 2007. Private credit in 129 countries. Journal of Fi-nancial Economics 84, 299–329.

Doukas, J.A., Pantzalis, C., 2003. Geographical diversification and agency costs of debt of multinational firms. Journal of Corporate Finance 9, 59–92.

Duchin, R., Ozbas, O., Sensoy, B.A., 2010. Costly external finance, corporate investment, and the subprime mortgage credit crisis. Journal of Financial Economics 97, 418–435.

Fan, J.P.H., Titman, S., Twite, G., 2012. An international comparison of capital structure and debt maturity choices. Journal of Financial and Quantitative Analysis 47, 23–56.

Fatemi, A.M, 1984. Shareholder benefits from corporate international diversification. Journal of Finance 39, 1325-1344.

Ferreira, M.A., Vilela, A.S., 2004. Why do firms hold cash? Evidence from EMU countries. European Financial Management 10, 295-319.

(31)

30

Geyikdagi, M., 1981. The cost of equity capital and risk of 28 US multinational corporations vs. 28 US domestic corporations: 1965-1978. Management International Review 1, 89-94. Gilchrist, S., Zakrajšek, E., 2012. Bank lending and credit supply shocks. In: Allen, F., Aoki, M., Fitoussi, J.P., Kiyotaki, N., Gordon, R., Stiglitz, J.E. (eds.) The global macro economy and finance. International Economic Association Series. Palgrave Macmillan, London.

Gomariz, M.F.C., Ballesta, J.P.S., 2014. Financial reporting quality, debt maturity and in-vestment efficiency. Journal of Banking & Finance 40, 494-506.

Gonenc, H., 2005. Comparison of debt financing between international and domestic firms: evidence from Turkey, Germany and UK. International Journal of Managerial Finance 1, 49-68.

Gonzalez, V.M., 2015. The financial crisis and corporate debt maturity: the role of banking structure. Journal of Corporate Finance 35, 310-328.

Ivashina, V., Scharfstein, D., 2010. Bank lending during the financial crisis of 2008. Journal of Financial Economics 97, 319–338.

Iyer, R., Da-Rocha-Lopes, S., Schoar, A., 2014. Interbank liquidity crunch and the firm credit crunch: evidence from the 2007-2009 crisis. Review of Financial Studies 27, 347-372.

Jang, Y., 2017. International corporate diversification and financial flexibility. Review of Fi-nancial Studies 30, 4133-4178.

Jiménez, G., López, J.A., Saurina, J., 2007. Empirical analysis of corporate credit lines. Re-view of Financial Studies 22, 5069-5098.

Kahle, K.M., Stulz, R.M., 2013. Access to capital, investment, and the financial crisis. Journal of Financial Economics 110, 280–299.

Kale, J.R., Noe, T., 1990. Risky debt maturity choice in a sequential game equilibrium. Jour-nal of Financial Research 13, 155-166.

Kalemli-Özcan, S., Laeven, L., Moreno, D., 2019. Debt overhang, rollover risk, and corporate investment: evidence from the European crisis. Unpublished working paper.

Kaufmann, D., Kraay, A., Mastruzzi, M., 2009. Governance matters VIII: aggregate and indi-vidual governance indicators, 1996–2008. Unpublished working paper.

Kim, C.S., Mauer, D.C., Sherman, A.E., 1998. The determinants of corporate liquidity: theory and evidence. Journal of Financial and Quantitative Analysis 33, 335-359.

Levine, R., Zervos, S., 1998. Stock markets, banks, and economic growth. American Eco-nomic Review 88, 537-558.

(32)

31

Malatesta, P., Dewenter, K., 2001. State-owned and privately owned firms: an empirical anal-ysis of profitability, leverage, and labor intensity. American Economic Review 91, 320-334. Mansi, S.A., Reeb, D., 2002. Corporate international activity and debt financing. Journal of International Business Studies 33, 129-147.

Morris, J., 1992. Factors affecting the maturity structure of corporate debt. Unpublished work-ing paper.

Myers, S.C., 1977. Determinants of corporate borrowing. Journal of Financial Economics 5, 147–175.

Myers, S.C., Majluf, N.S., 1984. Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics 13, 187-221. Ozkan, A., 2000. Determinants of capital structure and adjustment to long run target: evidence from UK company panel data. Journal of Business Finance & Accounting 28, 175–198. Park, S.H., Suh, J., Yeung, B., 2013. Do multinational and domestic firms differ in their lev-erage policies? Journal of Corporate Finance 20, 115-139.

Pöyry, S., Maury, B., 2010. Influential ownership and capital structure. Managerial and Deci-sion Economics 31, 311-324.

Qian, J., Strahan, P.E., 2007. How law and institutions shape financial contracts: the case of bank loans. Journal of Finance 62, 2803–2834.

Rajan, R., 1992. Insiders and outsiders: the choice between informed and arm's length debt. Journal of Finance 47, 1367–1400.

Rajan, R.G., Zingales, L., 1995. What do we know about capital structure? Some evidence from international data. Journal of Finance 50, 1421–1460.

Sarkar, S., 1999. Illiquidity risk, project characteristics, and the optimal maturity of corporate debt. Journal of Financial Research 22, 353-370.

Scherr, F.C., Hulburt, H.M., 2001. The debt maturity structure of small firms. Financial Man-agement 30, 85-111.

Titman, S., 2002. The Modigliani and Miller theorem and the integration of financial markets. Financial Management 31, 101-115.

(33)

32

7. Appendices

7.1 Appendix A: Variables

Name Definition Source

Crisis variables

DCRISIS Dummy variable that takes the value of one for the years 2008, 2009, 2010, 2011, and 2012, and zero otherwise.

DCRISIS1 Dummy variable that takes the value of one for the years 2008 and 2009, and zero otherwise.

DCRISIS2 Dummy variable that takes the value of one for the years 2010 and 2011, and zero otherwise.

DCRISIS3 Dummy that takes the value of one for the year 2012, and zero other-wise.

CR INT For the period 2008-2012, CR INT measures the absolute difference between the mean GDP growth rates for each year and the period 2003-2007. It takes the value of zero for non-2008-2012 years.

World Bank

EC GROWTH The GDP growth rate per annum. World Bank

Firm variables

DEBT MAT The percentage of total debt that has a maturity of more than one year. Worldscope

LT DEBT The ratio between long-term debt and lagged total assets. Worldscope

ST DEBT The ratio between short-term debt and lagged total assets. Worldscope MNC10 Dummy variable that takes the value of one if the firm’s foreign sales

ratio is at least 10%. Worldscope

MNC50 Dummy variable that takes the value of one if the firm’s foreign sales

MNC10* Dummy variable that takes the value of one if the firm’s foreign assets

MNC50* Dummy variable that takes the value of one if the firm’s foreign assets

FOR SAL The firm’s total foreign sales. Worldscope

FOR ASS The firm’s total foreign assets. Worldscope

FS RATIO The ratio between foreign sales and total sales. Worldscope

FA RATIO The ratio between foreign assets and total assets. Worldscope

ASSET MAT The ratio between fixed assets and total assets. Worldscope

GROWTH The ratio between common shareholder’s equity and market value. Worldscope/Datastream

SIZE The natural logarithm of sales. Worldscope

FIRM QUAL The ratio between net income plus depreciation and net debt. Worldscope VOL EBIT The absolute value of change in earnings before interest and taxes. Worldscope

LEV The ratio between total debt and the firm’s market value. Worldscope/Datastream

EXT FIN The ratio between total debt and total assets in the year 2006. Worldscope PROF Earnings before interest and taxes plus depreciation divided by total

assets. Worldscope

TANG The ratio between ‘property, plant, and equipment’ and total assets. Worldscope

Country variables

RULE LAW The rule of law constitutes one of the six dimensions of the Worldwide Governance Indicators and captures the extent to which agents have confidence in and abide by the rules of society. It compiles four ingre-dients of legal efficiency: the quality of contract enforcement, property rights, the police and the courts, and the likelihood of crime and vio-lence. The variable ranges from -2.5 to 2.5; low values indicate a less efficient legal system, and high values a more efficient one.

Kaufmann et al. (2009)

C RIGHTS An index that aggregates four powers of secured lenders in case of a bankruptcy. The first power is whether there are restrictions when a debtor files a petition for reorganization. Second, whether secured creditors can acquire their collateral after the petition for reorganiza-tion receives approval. Third, whether a bankrupt firm first pays secur-

(34)

33

7.2 Appendix B: Control variables

The following section presents independent justifications for the control variables used in our model. After each justification, we briefly discuss the results in table 4 in light of prior litera-ture.

Asset maturity

Justification: One reason firms use debt is to invest in novel projects that involve new assets.

However, firms also have existing projects with an existing asset base. As these assets mature, firms must eventually decide on how best to refinance them. According to Myers (1977), firms can minimize the agency costs of debt by matching refinancing decisions to the maturity structure of their existing assets. In other words, firms with longer-maturity assets in their portfolios should refinance their projects with longer-maturity debt, whilst the reverse should hold for firms that hold shorter-lived assets (Auken & Holman, 1995). Ozkan (2000) finds strong empirical support for this expectation. In a different study involving a sample of small enterprises, Scherr and Hulburt (2001) find further evidence that small firms – like larger ones – tend to match the maturity of liabilities with the maturity of assets.

Result: Asset maturity is positive and significant, which is consistent with Ozkan

(2002) and Scherr and Hulbert (2001). In addition, it supports the argument that firms can minimize the agency costs of debt financing by matching the maturity of novel debt to the maturity of existing assets (Myers, 1977; Auken & Holman, 1995).

Growth

Justification: Myers (1977) introduces the underinvestment problem, which is an agency

problem between shareholders and debt holders where managers erroneously forego a pro-spective investment that would make a positive contribution to the market value of the firm.

ed creditors out of the proceeds of its liquidation. Fourth, whether an administrator – and not management itself – is responsible for running the business during the reorganization. Djankov et al. (2007) add a value of one to the index when a country assures each of the foregoing powers to secured lenders. Hence, the variable ranges from zero (poor creditor rights) to four (strong creditor rights).

BANK CRED The ratio between private credit by deposit money banks and GDP. Fin. Dev. and Structure Dataset (World Bank). Beck et al. (2006) BANK CONC The percentage of bank assets held by the three largest commercial

(35)

34

The underinvestment problem occurs in situations where debt holders apprehend a large part of the cash flows of an investment opportunity, leaving insufficient returns to satisfy share-holders. From this, it follows that shareholders have insufficient incentive to proceed with the investment. One way for firms to mitigate this agency problem is by shortening the effective maturity of debt. That is, if debt matures before its holders are able to accrue the benefits of a given project, the underinvestment problem will vanish. Thus, Myers (1977) predicts that firms with more growth opportunities should hold shorter-maturity debt. Both Barclay and Smith (1995) and Ozkan (2000) confirm this prediction.

Result: In contrast to previous studies (Barclay & Smith, 1995; Ozkan, 2000), we

ob-serve no significant effect of growth opportunities. Hence, our analysis fails to provide evi-dence in favor or against the underinvestment problem conceptualized by Myers (1977).

Size

Justification: Barclay & Smith (1995) demonstrate that larger firms hold more long-term debt

than smaller firms do. The underpinning rationale is that larger firms are better able to derive economies of scale from the steep fixed cost component associated with issuing public debt. Since it is harder for smaller firms to take advantage of these scale economies, they will in-cline more toward bank debt that involves lower fixed costs. Small firms that prefer the bank-ing channel for this reason will hold shorter-maturity debt. We refer to Ozkan (2000) for con-firmatory evidence.

Result: Consistent with Ozkan (2000), firm size is positive and significant. This

but-tresses the idea that larger firms are in a better position to derive economies of scale from the fixed costs associated with issuing public debt, and that some smaller firms will revert to short-term bank debt because those scale economies are less available to them (Barclay & Smith, 1995).

Firm quality

Justification: Flannery (1986) posits that structural information asymmetries between firm

Did the financial crisis affect the debt maturity of multinationals and local firms differently?