Multinationals, Asset Tangibility and Leverage
Lars van der Wal
MSc. Thesis International Financial Management Faculty of Economics and Business, University of Groningen
Student number: S3203255 Supervisor: dr. P.P.M. Smid
Date: 07-01-2021
Abstract
This paper examines the relationship between (in)tangible assets and leverage policies and how this relationship differs between multinational and domestic firms for a sample of 6,167 U.S. firms over the 2002-2019 period. Tangible assets are easy to collateralize, thereby providing firms with more tangible assets more borrowing power which will result in higher leverage ratios. However, in the last decades asset tangibility has decreased while leverage ratios have increased. Besides these trends, intangible assets have become more important and collateralizable. Since multinationals possess relatively more intangible assets than domestic firms it might suggest that the relationship between tangible assets and leverage is less pronounced (i.e., less positive) in multinational firms. Consequently, it might also suggest that the relationship between intangible assets and leverage is more pronounced (i.e., more positive) in multinational firms. The results reveal that tangible assets are positively related to both book and market leverage ratios. This relationship is less pronounced in multinational firms with a physical operational presence abroad, suggesting that they enjoy reduced frictions in international debt markets. However, no evidence is found that this is because these multinationals are able to collateralize their intangible assets relatively more than otherwise similar domestic firms.
2
1. Introduction
When firms want to expand their operations, they need financing. Often external borrowing is necessary in order to do so. Building on the trade-off theory of capital structure tangible assets are easy to value by outsiders (Frank and Goyal, 2008; Frank and Goyal, 2009). These assets of a firm can be posted as collateral to ensure lenders against the consequences of possible default of the borrower (Liberti and Sturgess, 2018). Firms that have more tangible assets will therefore have more borrowing power. This suggests a positive relationship between a firm’s tangibility and its leverage. However, asset tangibility ratios have decreased while leverage ratios have increased (Graham, Leary and Roberts, 2015). Along with these trends there has been an increase in multinational firms that participate in global economic activity (OECD, 2018). Multinationals possess valuable intangible assets, which have become more significant for firms over the last decades (Syverson, 2011). Intangible assets have become more collateralizable due to the development of sophisticated valuation methods (Loumioti, 2012). In addition, since international debt markets differ in their collateral requirements (Hall, 2012) multinationals might exploit this as they have access to these markets (Jang, 2017). This might suggest that multinational firms rely less on their tangible assets when it comes to their leverage policies because of the complementary use of intangible assets. Therefore, the following three research questions are addressed: (1) Does higher asset tangibility mean that firms hold more leverage? (2) Do multinational firms rely less on tangible assets compared to domestic firms when it comes to their leverage policy?(3) Do multinational firms rely more on intangible assets compared to domestic firms when it comes to their leverage policy?
3
4
below that of competitors in order to attract their customers. This can be achieved by building relationships with foreign lenders or switching from U.S. to foreign lenders.
This paper contributes to the literature regarding capital structure and more specifically the leverage policies of multinational firms. First, prior literature primarily focuses on the determinants of a firm’s capital structure, while this paper examines how the relationship between two of these determinants and leverage differs between multinational and domestic firms. Second, while prior literature primarily explored how these two types of firms differ in their leverage policies, this research explores the role of (in)tangible assets in order to explain these differences.
The remainder of this paper is organized as follows. Section 2 provides an overview of relevant literature and theories from which the hypotheses are derived; Section 3 describes the sample, methodology and variables; Section 4 presents the empirical results; and Section 5 summarizes and concludes.
2. Literature review and hypothesis development
2.1 Asset tangibility and leverage
Before discussing the effect of asset tangibility on leverage it needs to be established what the asset tangibility of a firm is. In general, this is measured as the ratio of fixed assets to the total book value of assets (Desai, Foley, and Hines, 2004; de Jong, Kabir, and Nguyen, 2008; Lemmon, Roberts, and Zender, 2008; Mittoo and Zhang, 2008; Bae and Goyal, 2009; Frank and Goyal, 2009; Park, Suh, and Yeung, 2013; Graham, Leary, and Roberts, 2015; Horsch, Longoni, and Oesch, 2020) where the fixed assets consists out of the book value of property, plant and equipment.
5
implies that the capital structure of a firm is determined by the trade-off between the costs and benefits of using debt. More specifically, the benefits are that the interest serves as a tax shield while the costs are the direct and indirect costs of expected bankruptcy (Frank and Goyal, 2009). Another perspective of trade-off theory by Jensen and Meckling (1976) focuses on the role of debt in mitigating agency problems and disciplining managers, debt would do so as it must be repaid in order to avoid bankruptcy. In terms of asset tangibility, tangible assets are easy to value and undergo only a small loss in valuation. As a result, bankruptcy is less likely when firms with high asset tangibility are in financial distress (Shleifer and Vishny, 1992; Bae and Goyal, 2009). Therefore, tangible assets serve as good collateral in debt financing. However, it is worth noting that because of the high liquidation value, creditors might opt for bankruptcy early to collect that value. Also, the agency costs of debt decrease for firms with high asset tangibility as it becomes harder for shareholders to replace low-risk with high-risk assets (Frank and Goyal, 2008). In other words, according to both perspectives of the trade-off theory tangible assets support debt financing. The pecking-order theory developed by Myers (1984) on the other hand argues that firms prefer internal over external funding, and debt over equity financing because of adverse selection. If there is low information asymmetry for tangible assets, it makes the issuing of equity less costly (Frank and Goyal, 2009). This implies that firms with high asset tangibility would hold less leverage. However, empirical findings primarily supported the positive relationship between asset tangibility and leverage (e.g., Rajan and Zingales, 1995; Lemmon, Roberts, and Zender, 2008)
6
agency concerns. Again, beware that creditors might opt for bankruptcy early when the liquidation value of collateral is high. The role of collateral is empirically tested on a dataset of Bolivian financial institutions by Berger, Frame, and Ioannidou (2011), they provide evidence that collateral is used to overcome the incentive conflict between borrower and lender. In short, providing collateral eases the constraints of external financing. The difference between tangibility and collateral lays in the inclusion of inventories in collateral. Even though inventories are tangible assets, they do not belong to the fixed assets of a firm. Furthermore, inventories would typically support short-term debt and the fraction of inventories of total assets has decreased overtime (Frank and Goyal, 2009). On that note, some intangible assets are also collateralizable. This will be discussed in more depth in Section 2.2. Nonetheless, fixed tangible assets remain a significant proportion of collateralizable assets of a firm.
To summarize, tangible assets can be collateralized which will increase the borrowing power and therefore leverage of a firm. Accordingly, it is hypothesized that:
H1: There is a positive association between asset tangibility and leverage.
2.2 The moderating effect of multinational firms
7
the last decades (Syverson, 2011), while asset tangibility has decreased and leverage has increased (Graham, Leary and Roberts., 2015). This might suggest a relationship between intangible assets and debt after all. Along with the rise of intangible assets, there is also an increase in multinational firms participating in global economic activity (OECD, 2018). Building on the internalization theory, these intangible assets cause firms to expand abroad through foreign direct investments rather than licensing or exporting (Morck and Yeung, 1991), implying that multinationals possess more intangible assets compared to domestic firms.
8
9
the privilege of posting their intangible assets as collateral, something that domestic firms cannot.
All in all, multinationals possess valuable intangible assets which have become more collateralizable. In addition, due to their international nature, multinationals have access to international capital markets where they can find lenders that do accept their intangible assets as collateral. As a result, tangible assets will be of less importance for multinational firms compared to domestic ones when it comes to their leverage policies. Accordingly, it is hypothesized that:
H2: The effect of asset tangibility on leverage is less pronounced (i.e., less positive) in
multinational firms compared to domestic firms.
H3: The effect of intangible on leverage is more pronounced (i.e., more positive) in
multinational firms compared to domestic firms.
3. Sample and methodology
3.1 Data sources and sample
10
regression analyses, therefore firms are required to have no missing data with the exception of the data retrieved from the Worldscope database. This ensures that the descriptive statistics in
Table 2 provide a fair overview of the variables used in the multivariate regression analyses.
Variables constructed based on the data retrieved from the Worldscope database are never in the same model at the same time, a no missing data requirement would unnecessarily exclude additional observations from the sample. In order to deal with extreme values all non-dummy variables are winsorized at the 1% and 99% percentiles. This also excludes extreme values that are a consequence of wrong data.
Table 1
Industry distribution.
3.2 Methodology
In order to test the hypotheses multivariate regression analyses are performed using the equation
𝑦𝑖,𝑡 = 𝛼 + 𝜷𝟏𝐴𝑖,𝑡+ 𝜷𝟐𝐴𝑖,𝑡× 𝑀𝑁𝐶𝑖,𝑡+ 𝜷𝟑𝑀𝑁𝐶𝑖,𝑡+ ∑ 𝜷𝒙𝐶𝑉𝑖,𝑡+ ƞ𝑗+ 𝜙𝑡+ 𝜀𝑖,𝑡 (1)
where i denotes firm, j denotes industry and t denotes year. The dependent variable 𝑦𝑖,𝑡 is both 𝐵𝑜𝑜𝑘 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 and 𝑀𝑎𝑟𝑘𝑒𝑡 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 in separate analyses. 𝐴𝑖,𝑡 captures the ratio of fixed tangible assets to total book value of assets when testing Hypothesis 1 and 2, while it captures the identifiable intangible assets to total book value of assets ratio for Hypothesis 3. Next, 𝑀𝑁𝐶𝑖,𝑡 is a dummy variable that identifies multinational firms based on either their
Industry code Industry name Firms Firm-Year 1 Consumer non-durables 338 2,813 2 Consumer durables 176 1,465
3 Manufacturing 668 5,968
4 Oil, gas, and coal extraction and products 329 2,535 5 Chemicals and allied products 163 1,543 6 Business equipment 1,525 9,439 7 Telephone and television transmission 273 2,113
8 Utilities 0 0
9 Wholesale, retail, and some services (laundries, repair shops) 662 5,286 10 Healthcare, medical equipment, and drugs 1,017 5,985
11 Finance 0 0
12 Other 1,016 7,793
11
percentage of foreign sales or foreign assets. The interaction term captures the moderating effect of being a multinational firm on the relationship between either fixed tangible assets and leverage or identifiable intangible assets and leverage. 𝐶𝑉𝑖,𝑡 contains a set of firm-level determinants of leverage. All variables are discussed in more depth in Sections 3.3 to 3.6 and an overview is provided in Appendix A. In order to capture systematic differences across industries an industry dummy is used, which is denoted as ƞ𝑗. The industry dummies are based on the Fama-French 12-industry classification; an overview of the industry distribution can be found in Table 1. In addition, a year dummy is used to capture common macroeconomic shocks, this is denoted as 𝜙𝑡.
3.3 Dependent variables
To test the hypotheses a measure of leverage is required. In this paper leverage is measured in two ways: Book Leverage and Market Leverage. Where Book Leverage is measured as the ratio of book value of total debt to the book value of assets and Market
Leverage as the ratio of book value of total debt to the market value of assets (Lemmon, Roberts,
and Zender, 2008; Park, Suh and Yeung, 2013; Graham, Leary, and Roberts, 2015). Prior papers primarily consider market leverage (Burgman, 1996; Doukas and Pantzalis, 2003; Lee and Kwok, 1988; Mansi and Reeb, 2002) while, from the perspective of the lender, financial decision-making regarding credit focusses on book values (Chava and Roberts, 2008). In addition, Graham and Harvey (2001) conducted a survey from which the results suggest that managers use book leverage when setting their target leverage. Therefore, both measures are taken into account to ensure robustness of the results. Table 2 shows that the average Book
Leverage in the sample is 0.2718 with a standard deviation of 0.2202. The minimum value is
12
value (0.2346) is lower than the average, indicating that the majority of the observations have a lower Book Leverage ratio than the average. Next, the average Market Leverage in the sample is 0.1929 with a standard deviation of 0.1726. This is a lower average than Book Leverage which implies that on average the market value of assets exceeds the book value of assets. The minimum is 0.0002 for firms with almost no debt while the maximum value is 0.7400. Again, the median value (0.1479) is lower than the average. So, also the Market Leverage of the majority of the observations is lower than the average.
3.4 Independent variables
The main explanatory variable in Hypothesis 1 and 2 is Tangibility which is commonly measured as the ratio of net property, plant and equipment to the book value of assets (Desai, Foley, and Hines, 2004; de Jong, Kabir, and Nguyen, 2008; Lemmon, Roberts, and Zender, 2008; Mittoo and Zhang, 2008; Bae and Goyal, 2009; Park, Suh, and Yeung, 2013; Graham, Leary, and Roberts, 2015; Horsch, Longoni, and Oesch, 2020). Since property, plant and equipment are fixed tangible assets it serves as a reliable measure of the tangibility of a firm. As discussed in Section 2.1, a positive relationship with Book Leverage and Market Leverage can be expected. As shown in Table 2, the average value of Tangibility in the sample is 0.2750 with a standard deviation of 0.2386. The minimum value is 0.0059, which are firms that have almost no fixed assets or firms of which the fixed assets are almost fully depreciated, while the maximum is 0.9040. Also, since the median value (0.1935) is significantly lower than the average, the majority of the observations has a relatively lower amount of fixed tangible assets than the average.
For Hypothesis 3 the main explanatory variable is Intangible Assets.Section 2.2 of this
13
variable Intangible Assets is constructed as the ratio of identifiable intangible assets to the book value of assets (Graham, Leary, and Roberts, 2015), where the identifiable intangible assets are computed by subtracting the fixed and floating assets from the book value of assets. What remains is identifiable goodwill and other identifiable intangible assets (e.g., patents) of which the composition differs between firms. Identifiable goodwill is only present on the balance sheets of firms that are acquired in the past. It is therefore assumed that the majority of the variable Intangible Assets represents the identifiable intangible assets such as patents. Table 2 shows that the average Intangible Assets of the sample is 0.3071 with a standard deviation 0.2139. The minimum is 0.0117, which are firms with almost no identifiable intangible assets, while the maximum is 0.8421. In addition, since the median is 0.2615 the majority of observations will have relatively fewer identifiable intangible assets than the average.
Both variables are determinants of leverage and will therefore also serve as control variables. More specifically, when testing Hypothesis 1 and 2 Intangible Assets becomes a control variable and for Hypothesis 3 Tangibility becomes a control variable.
3.5 Moderating variables
14
Also, this measurement is in line with Doukas and Pantzalis (2003) and Mittoo and Zhang (2008). As a robustness test the distinction between multinational and domestic firms is also measured based on the foreign asset percentage of total assets. This would be a better identifier since it alleviates the problem of export sales being included in the foreign sales percentage (Kwok and Reeb, 2000; Reeb, Kwok and Baek., 1998). The dummy variable MNC Assets is constructed where multinationals are identified when foreign asset percentages are larger than 10% which is in line with the requirements of the U.S. Statement of Financial Accounting Standard No. 14 (FASB 1976) as mentioned earlier. As discussed in Section 2.2, the interaction term between MNC Sales, or MNC Assets, and Tangibility is expected to have a negative relationship with Book Leverage and Market Leverage. On the other hand, the interactions term between MNC Sales, or MNC Assets, and Intangible Assets is expected to have a positive relationship with Book Leverage and Market Leverage. As shown in Table 2, the average of
MNC Sales is 0.5549 indicating that approximately 55.5% of the firm-year observations are
identified as multinational. However, the average of MNC Assets is 0.3571 which implies that according to this measure only approximately 35.7% identifies as a multinational observation. The shortcoming of this identification technique is that firms can be identified as multinational in year t and as domestic in year t+1. Section 4.3 addresses this shortcoming in more detail.
Table 2
15
3.6 Control variables
In addition to the main explanatory variables, a set of firm-level determinants of leverage are selected from prior studies as control variables. Firstly, to measure the size of a firm Size is constructed as the natural logarithm of total sales (Mittoo and Zhang, 2008; Desai, Foley, and Hines, 2004; de Jong, Kabir, and Nguyen, 2008; Graham, Leary, and Roberts, 2015; Horsch, Longoni, and Oesch, 2020). Larger firms would face lower default risk, as they are more diversified (Frank and Goyal, 2009). As a result, larger firms would be able to attract more debt, implying a positive relationship with leverage. As shown in Table 2 the distribution of Size ranges from -0.1404 to 11.4639 with an average of 6.3959 and a median of 6.5084. Since the median is larger than the average the majority of the observations have higher sales than the average. Moreover, this will be more pronounced for the distribution of total sales, as Size is the natural logarithm of total sales. Therefore, the descriptive statistics of total sales will be discussed. In Table 2 it is provided that the average total sales are approximately $4.8075 billion with a standard deviation that is about three times as large ($13.4576 billion), implying that data points are spread out over a wider range of values. This can be observed as the minimum total sales are approximately $0.0009 billion, while the maximum total sales are approximately $95.2140 billion. The median is about $0.6707 billion, which is significantly lower than the average, indicating a negatively skewed distribution of total sales. This is primarily the reason why the natural logarithm is used in the multivariate regression analyses.
16
negative relationship exists between Market-to-Book and Market Leverage. Table 2 provides that the average of Market-to-Book is 1.8537 with a standard deviation of 1.2487. The maximum value is 8.0214 and the minimum is 0.5614. Firms with a Market-to-Book ratio below 1.0000 are firms with higher book value of assets than market value of assets and vice versa. Also, since the median value (1.4623) is lower than the average it can be concluded that the majority of the observations have lower Market-to-Book ratios than the average.
Next, the risk of default is lower for profitable firms (Bae and Goyal, 2009) which would result in higher leverage. However, according to the pecking order theory a firm prefers internal over external financing. Profitable firms generate more internal cash flows and will thus take on less debt, which suggests a negative relationship between Profitability and leverage (Frank and Goyal, 2009). In order to capture this effect, the variable Profitability is constructed and measured as the ratio of operating income before depreciation to the book value of assets (Desai, Foley, and Hines, 2004; de Jong, Kabir, and Nguyen, 2008; Lemmon, Roberts, and Zender, 2008; Bae and Goyal, 2009; Park, Suh, and Yeung, 2013; Graham, Leary, and Roberts, 2015; Horsch, Longoni, and Oesch, 2020). Table 2 shows that the average Profitability is 0.0659 with a standard deviation that is about three times as large (0.1964). The minimum is -0.9682 while the maximum is 0.3680. Firms with a value below 0.0000 have had negative operating income before depreciation. In addition, the median value (0.1061) is larger than the average indicating that the majority of the observations have higher operating income before depreciation to book assets ratios than the average.
17
Profitability over the trailing ten years (Lemmon, Roberts and Zender, 2008; Mittoo and Zhang,
2008; Fernandes, 2011; Park, Suh and Yeung, 2013; Graham, Leary, and Roberts, 2015) requiring at least four trailing years. For firms with less than ten, but at least four years with available data, the available trailing years are used. This measure is in line with Graham, Leary and Roberts (2015), and ensures that the standard deviations can be meaningfully interpreted.
Asset Volatility is measured as the variance of asset returns which is obtained by calculating the
yearly variances of equity returns and unleveraging them, assuming that other asset values are equal to their book values. (Frank and Goyal, 2009). The annualized variances of equity returns are based on daily stock returns, more information can be found in Appendix A. As shown in
Table 2 the average of Cash Flow Volatility is 0.1066 while the average Asset Volatility is
0.1895. The minimum and maximum value of Cash Flow Volatility are 0.0081 and 1.4240 respectively. For Asset Volatility on the other hand the minimum value is 0.0004 while the maximum is 7.6212. Both distributions have a lower median value (0.0508 and 0.0054, for
Cash Flow Volatility and Asset Volatility respectively) than the average. Therefore, the majority
of the observations will be less risky than the average.
Table 3
Correlation matrix. Full variable names: (1) Book Leverage, (2) Market Leverage (3) Tangibility, (4) Intangible Assets, (5) MNC Sales, (6) MNC Assets, (7) Size, (8) Market-to-Book, (9) Profitability, (10) Cash Flow Volatility, and (11) Asset Volatility.
18
4. Empirical results
Before performing the multivariate regression analyses to test the hypotheses, Pearson correlation coefficients are computed between the variables in order to secure the validity of the results. The correlation matrix can be found in Table 3. Based on the assumption that there is severe or hindering multicollinearity when correlation coefficients between independent variables go below -0.7 or -0.8 or exceed 7.0 or 0.8 there is no indication of severe multicollinearity problems.
4.1 Multinationals, asset tangibility and book leverage
Table 4 provides the results of testing Hypothesis 1 and 2 on Book Leverage based on
multivariate regression analyses. In Column 1 the relationship between Tangibility and Book
Leverage without controlling for other firm-level determinants of leverage is tested. What is
found is a significantly positive relationship (0.2026), which is as expected based on the positive single variate relation in Table 3. It implies that firms with relatively more tangible assets have higher book leverage ratios. Furthermore, when adding the firm-level control variables in
Column 2, the main relationship remains significantly positive and increases in terms of its
magnitude (0.3144). These results are in line with the empirical findings of Rajan and Zingales (1995), Lemmon, Roberts and Zender (2008), Hall (2012) and Park, Suh and Yeung (2013). In terms of economic significance, ceteris paribus, an increase of 0.1 in the tangibility ratio (Tangibility) with and without taken into account other determinants of leverage corresponds with an increase in the book leverage ratio of approximately 0.0314 and 0.0203 respectively. These results are consistent with Hypothesis 1. The relationship between MNC Sales and Book
Leverage is found to be significantly negative (-0.0369), indicating that on average
19
multinational and domestic firms. Furthermore, everything else being equal, on average the book leverage ratio of multinational firms will be approximately 0.0369 lower than that of domestic firms. In addition, in line with the expectations, a significantly positive relationship (0.2599) is found between Intangible Assets and Book Leverage. This is similar to the empirical findings of Lim, Macias and Moeller (2020). Keeping everything else equal, an increase of 0.1 in the identifiable intangible asset ratio (Intangible Assets) of a firm corresponds with an increase of approximately 0.0260 in the book leverage ratio. Also, Size is found to have a significantly positive relationship (0.0095) with Book Leverage. Empirical findings that suggest that larger firms tend to have higher book leverage ratios are consistent with Rajan and Zingales (1995), Desai, Foley and Hines (2004), Lemmon, Roberts and Zender (2008), Horsch, Longoni and Oesch (2020) and Lim, Macias and Moeller (2020). Note that since the variable Size is measured as the natural logarithm of total sales, its economic interpretation deviates from the rest of the control variables. Ceteris paribus, if the total sales of a firm increase by 10%, the book leverage ratio will increase by approximately 0.0009. Next, a significantly positive relationship with Book Leverage is found for Market-to-Book (0.0144). Although this contradicts expectations, similar findings are found by Frank and Goyal (2009). They argue that the market-to-book ratio might not be operating through the effect on debt but more through an effect on equity.1 Assuming that everything else remains equal, an increase of 0.1 in the market to book ratio (Market-to-Book) of a firm corresponds with an increase in the book leverage ratio of approximately 0.0014. As opposed to the abovementioned control variables, Profitability is found to have a significantly negative relationship (-0.2062) with Book Leverage. This is consistent with the empirical findings of Rajan and Zingales (1995), Desai, Foley and Hines (2004), Lemmon, Roberts and Zender (2008), Park, Suh and Yeung (2013), Graham, Leary and
1 When excluding firms with negative book values of equity the relationship between Market-to-Book and Book
Leverage becomes significantly negative while other results for other variables remain similar to the results in Table 4 except for the interaction term between Tangibility and MNC Sales, which becomes significant. This is
20
Roberts (2015), Horsch, Longoni and Oesch (2020) and Lim, Macias and Moeller (2020). The findings suggest that, everything else being equal, an increase of 0.1 in the profitability ratio (Profitability) corresponds with a decrease of approximately 0.0206 in the book leverage ratio. Lastly, both risk proxies are significantly negative with a magnitude of -0.0171 and -0.0450 for
Cash Flow Volatility and Asset Volatility respectively. For Cash Flow Volatility this is in line
with the empirical findings of Lemmon, Roberts and Zender (2008). It implies that, ceteris paribus, an increase in the standard deviation of the profitability ratio (Cash Flow Volatility) of 0.1 corresponds with a decrease in the book leverage ratio of approximately 0.0017. The findings for Asset Volatility are consistent the empirical findings of Frank and Goyal (2009). An increase of 0.1 in the variance of asset returns (Asset Volatility) corresponds with a decrease of approximately 0.0045 in the book leverage ratio. Note that this only holds when assuming everything else to remain equal. The signs and magnitudes of the control variables remain approximately the same across Columns 2 to 5.
In Column 3 the moderating effect of being a multinational firm is examined on the relationship between Tangibility and Book Leverage. Here multinationals are identified based on their foreign sales percentage of total sales. In this model the main relationship remains significantly positive (0.3094) implying that for domestic firms relatively more tangible assets are associated with relatively more debt. More specifically, ceteris paribus, an increase of 0.1 in the tangibility ratio (Tangibility) corresponds with an approximately 0.0309 increase in the book leverage ratios of domestic firms. MNC Sales remains significantly negative (-0.0404). The interaction term between Tangibility and MNC Sales is found to be positive but insignificant. Based on these results Hypothesis 2 cannot be rejected or confirmed.
21
implies that, ceteris paribus, an increase of 0.1 in the tangibility ratio (Tangibility) corresponds with an increase of approximately 0.0329 in the book leverage ratio of firms. MNC Assets is also significantly negative (-0.0247) ensuring that on average multinationals have lower book leverage ratios than domestic firms. More specifically, everything else being equal, the book leverage ratio of multinationals lays approximately 0.0247 lower than domestic firms. In
Column 5 the main relationship remains significant and increases to 0.3401 in terms of its
magnitude, this represents the relationship between the fixed tangible assets to book value of assets ratio and the book leverage ratio for domestic firms. Ceteris paribus, an increase of 0.1 in the former corresponds with an increase of approximately 0.0340 in the latter.MNC Assets
also remains significant and negative (-0.0144). Furthermore, the interaction term between
Tangibility and MNC Assets is found to be significantly negative (-0.0362) indicating that the
main relationship is less pronounced for multinational firms than for domestic firms. When assuming that everything else remains equal, an increase of 0.1 in the tangibility ratio (Tangibility) corresponds with an increase of approximately 0.0304 in the book leverage ratio for multinational firms. As this is lower than what is found for domestic firms (0.0340), the main relationship is less pronounced in multinational firms. These results are consistent with
Hypothesis 2.
In summary, the empirical results confirm Hypothesis 1 and provide partial evidence for
Hypothesis 2. The findings reveal that tangible assets are, with and without controlling for other
determinants of leverage, positively related to the book leverage ratio. In terms of the moderating effect of being a multinational firm, when identifying multinationals based on the foreign sales percentage of total sales insignificant findings are found. This implies that
Hypothesis 2 cannot be rejected or confirmed. On the other hand, when multinationals and
22
leverage ratios. This suggests that only a physical operational presence abroad reduces the importance of tangible assets on book leverage ratios, the identification of multinationals based on foreign sales can also include large exporting firms that only have a physical presence in the U.S. It also suggests that multinationals with a physical operational presence abroad enjoy reduced frictions in international debt markets (Jang, 2017) and might be able to collateralize their other assets (i.e., intangible assets) in foreign countries, as these markets differ in their collateral requirements (Hall, 2012).
Table 4
Multinationals, asset tangibility and book leverage. Book Leverage (1) (2) (3) (4) (5) Tangibility 0.2026*** (0.0042) 0.3144*** (0.0053) 0.3094*** (0.0063) 0.3293*** (0.0054) 0.3401*** (0.0061) MNC Sales -0.0369*** (0.0023) -0.0404*** (0.0032) Tangibility * MNC Sales 0.0126 (0.0087) MNC Assets -0.0247*** (0.0023) -0.0144*** (0.0036) Tangibility * MNC Assets -0.0362*** (0.0096) Intangible Assets 0.2599*** (0.0057) 0.2604*** (0.0057) 0.2675*** (0.0060) 0.2670*** (0.0060) Size 0.0095*** (0.0058) 0.0094*** (0.0006) 0.0084*** (0.0006) 0.0085*** (0.0006) Market-to-Book 0.0144*** (0.0008) 0.0144*** (0.0008) 0.0161*** (0.0009) 0.0161*** (0.0009) Profitability -0.2062*** (0.0067) -0.2056*** (0.0067) -0.2064*** (0.0070) -0.2077*** (0.0071) Cash Flow Volatility -0.0171***
(0.0065) -0.0176*** (0.0065) -0.0202*** (0.0068) -0.0191*** (0.0068) Asset Volatility -0.0450*** (0.0011) -0.0450*** (0.0011) -0.0464*** (0.0012) -0.0464*** (0.0012) Constant 0.1462*** (0.0030) 0.0236*** (0.0052) 0.0255*** (0.0054) 0.0073 (0.0055) 0.0038 (0.0056) Adj. R2 0.0675 0.1804 0.1804 0.1732 0.1735 Number of observations 44,940 40,088 40,088 36,740 36,740 Industry dummy Yes Yes Yes Yes Yes Year dummy Yes Yes Yes Yes Yes *** Denotes statistical significance at 1% level.
23
4.2 Multinationals, asset tangibility and market leverage
In Table 5 the robustness of the results in Section 4.1 is tested by replacing the dependent variable Book Leverage with Market Leverage. In Column 1 the relationship between
Tangibility and Market Leverage without controlling for other firm-level determinants of
leverage is tested. A significantly positive relationship is found (0.2381), implying that firms with relatively more tangible assets have higher market leverage ratios. When the firm-level control variables are added in Column 2, the main relationship remains significantly positive and its magnitude increases to 0.2587. This is consistent with the empirical findings of Rajan and Zingales (1995), de Jong, Kabir and Nguyen (2008), Lemmon, Roberts and Zender (2008), Frank and Goyal (2009), Fernandes (2011) and Park, Suh and Yeung (2013). With and without taking into account other determinants of leverage, an increase of 0.1 in the tangibility ratio (Tangibility) corresponds with an increase in the market leverage ratio of approximately 0.0259 and 0.0238 respectively. Note that this is the case when assuming everything else to remain equal. These results are consistent with Hypothesis 1. MNC Sales is found to be significant and negative (-0.0287), suggesting that on average multinational firms have lower market leverage ratios compared to domestic firms.Prior research supports these findings (Lee and Kwok, 1988; Burgman, 1996; Mittoo and Zhang, 2008). More specifically, keeping everything else equal, the market leverage ratio of multinational firms is approximately 0.0287 lower than that of domestic firms. In terms of the effect of the control variables on Market Leverage the significance and signs do not change compared to the effect on Book Leverage except for
Market-to-Book, which turns negative. This can be explained by the fact that Market-to-Book
24
(Rajan and Zingales, 1995; Chen et al., 1997; de Jong, Kabir and Nguyen, 2008; Lemmon, Roberts and Zender, 2008; Fernandes, 2011; Park, Suh, and Yeung, 2013), Profitability (Rajan and Zingales, 1995; de Jong, Kabir and Nguyen, 2008; Lemmon, Roberts and Zender, 2008; Frank and Goyal, 2009; Park, Suh and Yeung, 2013; Graham, Leary and Roberts, 2015; Horsch, Longoni and Oesch, 2020; Lim, Macias and Moeller, 2020), Cash Flow Volatility (Doukas and Pantzalis, 2003; de Jong, Kabir and Nguyen, 2008; Lemmon, Roberts and Zender, 2008) and
Asset Volatility (Frank and Goyal, 2009). Since the magnitudes remain approximately the same
25
In Column 3 the moderating effect of being a multinational firm is examined on the relationship between Tangibility and Market Leverage. Here multinationals are identified based on their foreign sales percentage of total sales and the main relationship remains significantly positive (0.2444). This represents the relationship between Tangibility and Market Leverage for domestic firms. It implies that, ceteris paribus, an increase of 0.1 in the tangibility ratio (Tangibility) corresponds with an increase of approximately 0.0244 in the market leverage ratio of domestic firms. MNC Sales remains significantly negative (-0.0386). The interaction term between Tangibility and MNC Sales is significantly positive (0.0365). More specifically, assuming that everything else remains equal, an increase of 0.1 in the tangibility ratio (Tangibility) corresponds to an increase in the market leverage ratio of approximately 0.0281 for multinational firms. Since this is larger than what is found for domestic firms (0.0244), the main relationship is more pronounced in multinationals. This is inconsistent with Hypothesis 2. In order to test the robustness of the abovementioned results MNC Sales is replaced with
MNC Assets in Column 4 and 5. The main relationship remains significant and positive in Column 4 and increases to 0.2665 in terms of its magnitude. This implies that, keeping
everything else equal, an increase of 0.1 in the tangibility ratio (Tangibility) corresponds with an increase of approximately 0.0267 in the market leverage ratio. MNC Assets is found to be significantly negative (-0.0251), suggesting that the on average market leverage ratio in multinationals is approximately 0.0251 lower compared to domestic firms. This is however, when assuming everything else to remain equal. In Column 5 the relationship between
Tangibility and Market Leverage for domestic firms is significantly positive (0.2783). This
suggests that, ceteris paribus, an increase of 0.1 in the tangibility ratio (Tangibility) corresponds with an increase in the market leverage ratio of domestic firms of approximately 0.0278. MNC
26
tangibility ratio (Tangibility) corresponds with an increase of approximately 0.0239 in the market leverage ratio of multinationals. As this is lower than what is found for domestic firms (0.0278), it indicates that the main relationship is less pronounced in multinational firms. This is consistent with Hypothesis 2.
Table 5
Multinationals, asset tangibility and market leverage. Market Leverage (1) (2) (3) (4) (5) Tangibility 0.2381*** (0.0032) 0.2587*** (0.0038) 0.2444*** (0.0046) 0.2665*** (0.0039) 0.2783*** (0.0044) MNC Sales -0.0287*** (0.0016) -0.0386*** (0.0023) Tangibility * MNC Sales 0.0365*** (0.0063) MNC Assets -0.0251*** (0.1660) -0.0138*** (0.0026) Tangibility * MNC Assets -0.0398*** (0.0069) Intangible Assets 0.1621*** (0.0041) 0.1634*** (0.0041) 0.1660*** (0.0043) 0.1654*** (0.0043) Size 0.0045*** (0.0004) 0.0043*** (0.0004) 0.0043*** (0.0004) 0.0044*** (0.0004) Market-to-Book -0.0414*** (0.0006) -0.0413*** (0.0006) -0.0414*** (0.0006) -0.0414*** (0.0006) Profitability -0.1444*** (0.0049) -0.1429*** (0.0049) -0.1442*** (0.0051) -0.1456*** (0.0051) Cash Flow Volatility -0.0300***
(0.0047) -0.0316*** (0.0047) -0.0299*** (0.0049) -0.0287*** (0.0049) Asset Volatility -0.0201*** (0.0008) -0.0201*** (0.0008) -0.0204*** (0.0009) -0.0203*** (0.0009) Constant 0.0936*** (0.0023) 0.1225*** (0.0038) 0.1280*** (0.0039) 0.1124*** (0.0040) 0.1085*** (0.0040) Adj. R2 0.1166 0.2916 0.2922 0.2835 0.2841 Number of observations 44,940 40,088 40,088 36,740 36,740 Industry dummy Yes Yes Yes Yes Yes Year dummy Yes Yes Yes Yes Yes *** Denotes statistical significance at 1% level.
** Denotes statistical significance at 5% level. * Denotes statistical significance at 10% level.
To summarize, the empirical results confirm Hypothesis 1 and partly confirm
Hypothesis 2. It is revealed that tangible assets are positively related to the market leverage
27
with Hypothesis 2 are found, while multinationals identified based on foreign assets provide results consistent with Hypothesis 2. Again, suggesting that a physical operational presence abroad reduces the importance of tangible assets, this time for market leverage ratios. Identifying multinationals based on the foreign sales percentage can also include large exporting firms of which their physical presence is restricted to the U.S. Furthermore, as international debt markets differ in their collateral requirements (Hall, 2012) and the physical operational presence reduces frictions in these markets (Jang, 2017), the results suggest that multinationals with foreign assets might be able to collateralize their other assets (i.e., intangible assets) in foreign countries.
4.3 Robustness tests
In order to ensure the robustness of the results in Table 4 and 5, robustness tests are performed by increasing the threshold in identifying multinational firms on foreign sales and assets from 10% to 25%. The shortcoming of this initial identification technique is that firms can be identified as multinational in year t and as domestic in year t+1. Therefore, another identification technique is introduced to further investigate the robustness of the results. Here one observation above the 10% or 25% threshold will identify the firm as a multinational for all years of available data2.
When increasing the initial threshold from 10% to 25%, approximately 43.4% of the firm-year observations are identified as multinational based on MNC Sales. However, based on
MNC Assets only approximately 19.5% firm-year observations are identified as multinational.
The robustness tests using this identification technique can be found in Columns 1 and 2 of both
2 Multivariate regression analyses with MNC Sales and MNC Assets as the foreign sales and asset percentage
28
Panel A and B in Table 6. In Panel A the dependent variable is Book Leverage. The interaction
between Tangibility and MNC Sales remains insignificant and all other findings remain the same. Also, when Market Leverage is the dependent variable in Panel B, the results do not differ from the initial results. This suggests that when multinationals are identified based on their foreign sales percentage and Book Leverage is the dependent variable, Hypothesis 2 cannot be rejected or confirmed. When replacing the dependent variable with Market Leverage the results suggests that Hypothesis 2 should be rejected. On the other hand, when identifying multinationals based on their foreign asset percentage Hypothesis 2 can be confirmed for both dependent variables.
29 Table 6
Robustness checks. In Columns 1 and 2, MNC Sales and MNC Assets are measured on the observational level with a threshold of 25%. In Columns 3 and 4, MNC Sales and MNC Assets are measured based on the firm-level with a threshold of 10%. In Columns 5 and 6, MNC Sales and MNC Assets are measured based on the firm-level with a threshold of 25%. Panel A Book Leverage (1) (2) (3) (4) (5) (6) Tangibility 0.3233*** (0.0059) 0.3395*** (0.0058) 0.3108*** (0.0064) 0.3379*** (0.0061) 0.3165*** (0.0060) 0.3408*** (0.0058) MNC Sales -0.0290*** (0.0032) -0.0333*** (0.0032) -0.0296*** (0.0030) Tangibility * MNC Sales -0.0118 (0.0091) -0.0014 (0.0083) -0.0079 (0.0083) MNC Assets -0.0094** (0.0043) -0.0076** (0.0031) -0.0026 (0.0032) Tangibility * MNC Assets -0.0470*** (0.0113) -0.0362*** (0.0083) -0.0521*** (0.0088) Constant 0.0145*** (0.0053) 0.0005 (0.0055) 0.0287*** (0.0053) 0.0118** (0.0051) 0.0227*** (0.0051) 0.0080 (0.0051) Adj. R2 0.1791 0.1726 0.1763 0.1701 0.1763 0.1703 Number of observations 40,088 36,740 43,844 42,363 43,844 42,363
Industry dummy Yes Yes Yes Yes Yes Yes
Year dummy Yes Yes Yes Yes Yes Yes
Control variables Yes Yes Yes Yes Yes Yes
Panel B Market Leverage (1) (2) (3) (4) (5) (6) Tangibility 0.2562*** (0.0043) 0.2753*** (0.0042) 0.2437*** (0.0047) 0.2733*** (0.0044) 0.2490*** (0.0044) 0.2770*** (0.0042) MNC Sales -0.0298*** (0.0023) -0.0337*** (0.0023) -0.0300*** (0.0022) Tangibility * MNC Sales 0.0225*** (0.0066) 0.0207*** (0.0060) 0.0190*** (0.0060) MNC Assets -0.0109*** (0.0031) -0.0117*** (0.0022) -0.0061*** (0.0023) Tangibility * MNC Assets -0.0400*** (0.0082) -0.0273*** (0.0060) -0.0423*** (0.0064) Constant 0.1185*** (0.0038) 0.1059*** (0.0040) 0.1331*** (0.0038) 0.1167*** (0.0037) 0.1273*** (0.0037) 0.1121*** (0.0037) Adj. R2 0.2903 0.2821 0.2843 0.2781 0.2837 0.2780 Number of observations 40,088 36,740 43,844 42,363 43,844 42,363
Industry dummy Yes Yes Yes Yes Yes Yes
Year dummy Yes Yes Yes Yes Yes Yes
Control variables Yes Yes Yes Yes Yes Yes
30
Lastly, in Columns 5 and 6 of both Panel A and B in Table 6 the threshold of 10% is increased to 25% for the firm-level identification technique. Based on MNC Sales there are 2,721 multinational and 3,104 domestic firms in the sample while according to the MNC Assets measure 1,581 and 3,976 firms are identified as multinational and domestic respectively. As explained earlier, more observations than the initial analyses are used with this measure. Here the interaction between Tangibility and MNC Sales on the dependent variable Book Leverage in Panel A remains insignificant. Again, the results in Panel B on the dependent variable Market
Leverage do not differ from the initial results. Therefore, the same conclusions regarding Hypothesis 2 should be made.
In summary, the interaction term between Tangibility and MNC Sales remains insignificant in its relationship with Book Leverage for all different identification techniques. Also, the same results are found when Market Leverage is the dependent variable. Therefore, the initial results are robust. As mentioned before, Hypothesis 2 only holds for multinationals firm with a physical operational presence abroad. This might be because identified multinationals based on MNC Sales can also include large exporting firms with no foreign assets.
4.4 Multinationals, intangible assets and leverage
Table 7 provides the results of the multivariate regression analyses performed to test Hypothesis 3. Note that Tangibility is included in the control variables for these analyses as it
remains an important determinant of leverage. Book Leverage is the dependent variable in Panel
A. In Column 1 it is found that the relationship between Intangible Assets and Book Leverage
31
in the book leverage ratio of firms. MNC Sales is significantly negative (-0.0369), meaning that on average multinationals have lower book leverage ratios than domestic firms. More specifically, everything else being equal, the book leverage ratio of multinationals is approximately 0.0369 lower than that of domestic firms. In Column 2 the moderating effect of being a multinational firm is examined on this relationship. Here multinationals are identified based on their foreign sales percentage of total sales. In this model the main relationship remains significantly positive (0.2752) implying that for domestic firms relatively more intangible assets are associated with relatively more debt. More specifically, ceteris paribus, an increase of 0.1 in the intangible ratio (Intangible Assets) corresponds with an approximately 0.0275 increase in the book leverage ratios of domestic firms. MNC Sales remains significantly negative (-0.0277). The interaction term between Intangible Assets and MNC Sales is found to be negative and significant (-0.0303). Assuming that everything else remains equal, an increase of 0.1 in the tangibility ratio (Tangibility) corresponds with an increase in the book leverage ratio of approximately 0.0245 for multinational firms. The results suggest that the main relationship is less pronounced in multinational firms, which is inconsistent with Hypothesis 3.
As a robustness test the identification of multinational firms is also done by using the foreign asset percentage of total assets in Column 3 and 4. What is found in Column 3 is that the main relationship remains approximately the same (0.2675). This implies that, ceteris paribus, an increase of 0.1 in the intangible asset ratio (Intangible Assets) corresponds with an increase of approximately 0.0268 in the book leverage ratio of firms. MNC Assets is also significantly negative (-0.0247) ensuring that on average multinationals have lower book leverage ratios than domestic firms. More specifically, everything else being equal, the book leverage ratio of multinationals is approximately 0.0247 lower than that of domestic firms. In
Column 4 the main relationship remains significant and increases to 0.2707 in terms of its
32
value of assets ratio and the book leverage ratio for domestic firms. Ceteris paribus, an increase of 0.1 in the former corresponds with an increase of approximately 0.0271 in the latter. MNC
Assets also remains significant and negative (-0.0213). Furthermore, the interaction term
between Intangible Assets and MNC Assets is found to be negative but insignificant. Based on these results Hypothesis 3 cannot be rejected or confirmed.
Table 7
Multinationals, intangible assets and leverage. Panel A Book Leverage (1) (2) (3) (4) Intangible Assets 0.2599*** (0.0057) 0.2752*** (0.0074) 0.2675*** (0.0060) 0.2707*** (0.0068) MNC Sales -0.0369*** (0.0022) -0.0277*** (0.0036) Intangible Assets * MNC Sales -0.0303***
(0.0093)
MNC Assets -0.0247***
(0.0023)
-0.0213*** (0.0041) Intangible Assets * MNC Assets -0.0109 (0.0107) Constant 0.0236*** (0.0052) 0.0185*** (0.0055) 0.0073 (0.0055) 0.0064 (0.0055) Adj. R2 0.1804 0.1806 0.1732 0.1732 Number of observations 40,088 40,088 36,740 36,740 Industry dummy Yes Yes Yes Yes Year dummy Yes Yes Yes Yes Control variables Yes Yes Yes Yes Panel B Market Leverage (1) (2) (3) (4) Intangible Assets 0.1621*** (0.0041) 0.1852*** (0.0053) 0.1660*** (0.0043) 0.1668*** (0.0049) MNC Sales -0.0287*** (0.0016) -0.0147*** (0.0026) Intangible Assets * MNC Sales -0.0461***
(0.0068)
MNC Assets -0.0251***
(0.0017)
-0.0243*** (0.0029) Intangible Assets * MNC Assets -0.0028 (0.0078) Constant 0.1225*** (0.0038) 0.1147*** (0.0040) 0.1124*** (0.0040) 0.1121*** (0.0040) Adj. R2 0.2916 0.2924 0.2835 0.2835 Number of observations 40,088 40,088 36,740 36,740 Industry dummy Yes Yes Yes Yes Year dummy Yes Yes Yes Yes Control variables Yes Yes Yes Yes *** Denotes statistical significance at 1% level.
33
In Panel B the dependent variable is replaced by Market Leverage. The main relationship is found to be significantly positive (0.1621) in Column 1. This implies that, ceteris paribus, an increase of 0.1 in the intangible asset ratio (Intangible Assets) corresponds with an increase of approximately 0.0162 in the market leverage ratio of firms. MNC Sales is significantly negative (-0.0287), meaning that on average multinationals have lower market leverage ratios than domestic firms. More specifically, everything else being equal, the market leverage ratio of multinationals is approximately 0.0287 lower than that of domestic firms. In
Column 2 the moderating effect of being a multinational firm is examined on the relationship
between Intangible Assets and Market Leverage. Here multinationals are identified based on their foreign sales percentage of total sales, the main relationship remains significantly positive (0.1852). This represents the relationship between Intangible Assets and Market Leverage for domestic firms. It implies that, ceteris paribus, an increase of 0.1 in the intangible asset ratio (Intangible Assets) corresponds with an increase of approximately 0.0185 in the market leverage ratio of domestic firms. MNC Sales remains significantly negative (-0.0147). The interaction term between Intangible Assets and MNC Sales is significantly negative (-0.0461). More specifically, assuming that everything else remains equal, an increase of 0.1 in the tangibility ratio (Tangibility) corresponds to an increase in the market leverage ratio of approximately 0.0139 for multinational firms. Since this is smaller than what is found for domestic firms (0.0185), the main relationship is less pronounced in multinationals. This is inconsistent with Hypothesis 3.
In order to test the robustness of the abovementioned results MNC Sales is replaced with
MNC Assets in Column 3 and 4. The main relationship remains significant and positive in Column 3 (0.1660). This implies that, keeping everything else equal, an increase of 0.1 in the
34
suggesting that on average the market leverage ratio in multinationals is approximately 0.0251 lower compared to domestic firms. This is however, when assuming everything else to remain equal. In Column 4 the relationship between Intangible Assets and Market Leverage for domestic firms is significantly positive (0.1668). This suggests that, ceteris paribus, an increase of 0.1 in the intangible asset ratio (Intangible Assets) corresponds with an increase in the market leverage ratio of domestic firms of approximately 0.0167. MNC Assets remains significantly negative (-0.0243). The interaction term between Intangible Assets and MNC Assets is negative but insignificant. Based on these results Hypothesis 3 cannot be rejected or confirmed.
To summarize, the empirical results do not confirm Hypothesis 33. If multinational and domestic firms are distinguished based on foreign sales inconsistent results with Hypothesis 3 are found, while multinationals identified based on foreign assets provide results on which
Hypothesis 3 cannot be rejected or confirmed. In other words, while the results for Hypothesis 2 suggest that the international physical operational presence of multinationals might make
posting intangible assets as collateral more likely it cannot be confirmed.
5. Conclusion
This paper investigates the relationship between tangible assets and leverage ratios. Tangible assets would serve as good collateral and should therefore increase the borrowing power of firms, and in turn increase their leverage ratios. In addition, the moderating effect of being a multinational firm is tested on this relationship. The asset tangibility of firms has decreased while leverage ratios have increased, if you add that intangible assets have become more important and collateralizable for firms over the last decades one might suggest that there
3 Robustness checks using the same multinational identification techniques used in Table 6 are also performed.
35
is a relationship between intangible assets and leverage ratios. Since multinational firms are the firms that possess relatively more valuable intangible assets compared to domestic firms, they would depend less on tangible assets in debt financing. Therefore, the positive relationship between tangible assets and leverage ratios would be less pronounced (i.e., less positive) in multinational firms compared to domestic firms. In turn this suggest that the relationship between intangible assets and leverage is more pronounced (i.e., more positive) in multinationals firms compared to domestic firms. These insights are tested using a sample of 6,167 unique firms with 44,940 firm-year observations in the 2002-2019 period.
36
being a multinational on the relationship between intangible assets and leverage opposite results are found when identifying multinationals based on foreign sales. When multinationals are identified based on foreign assets insignificant results are found.
These findings have significant and potentially broad implications for practicing managers of multinational firms with a physical presence abroad. This physical international presence creates value in alleviating financial constraints of collateral requirements in international debt markets, this provides an advantage against otherwise similar domestic or large exporting firms. Managers should exploit this in their debt financing policies by building relationships with foreign lenders if these firms only have loans from U.S. lenders to shifting loans from domestic to foreign lenders. By doing so managers will be able to attract more debt with the same assets. The advantage of attracting more debt than their otherwise similar domestic and large exporting firms lays primarily in the fact that interest payments can be deducted from the tax payable. This results in a margin for which they, for example, could lower the prices to gain customers that otherwise would have purchased their goods or services at competitors.
37
countries. Third, if these multinationals are more flexible in debt financing, they should be less affected by disruptions in the U.S. capital market than otherwise similar domestic and large exporting firms, further research should seek to confirm this.
In addition, the findings contribute to the literature regarding capital structure and more specifically the leverage policies of multinational firms. Where prior literature primarily focussed on the determinants of a firm’s capital structure, this paper provides insight in how the relationship between two of these determinants and leverage differs between multinational and domestic firms. Also, where prior literature explored the differences in leverage ratios between these two types of firms, this paper explores the role of (in)tangible assets in order to explain these differences.
38
Appendix A. Variable definitions
This appendix provides the construction of the variables used in the multivariate regression analyses. Foreign sales and foreign asset percentages are retrieved from Worldscope. The yearly variances of equity returns are based on daily stock returns retrieved from CRSP. All other data is retrieved from Compustat North America and are in millions of USD. All data items represent the values at the end of a firm’s fiscal year. Worldscope and Compustat codes are provided between brackets.
Variable Definition and measurement
Debt in Current Liabilities (DLC) This item represents the total amount of short-term notes and the current portion of long-term debt (debt due in one year).
Long-Term Debt (DLTT) This item represents debt obligations due more than one year from the firm’s balance sheet date.
Total Debt DLC + DLTT
Book Value Assets (AT) This item represents the total assets/liabilities of a firm at a point in time.
Book Leverage Total Debt / AT
Stockholders’ Equity (SEQ) This item represents the common and preferred shareholders’ interest in the firm.
Price End Fiscal Year (PRCC_F) This item represents the stock price and the end of the fiscal year. Common Shares Used to Calculate Earnings Per Share (CSHPRI) This item represents the average or actual number of common shares
outstanding, excluding dilution (conversion of convertible preferred stock, convertible debentures, options and warrants).
Market Value Assets AT – SEQ + (PRCC_F * CSHPRI)
Market Leverage Total Debt / Market Value Assets
Net Property, Plant and Equipment (PPENT) This item represents the cost, less accumulated depreciation, of tangible fixed property used in the production of revenue.
Tangibility PPENT / AT
Foreign Sales % of Total Sales (WC08731) This item represents the percentage sales generated from operations in foreign countries of total sales.
MNC Sales 1 if WC08731 > 10%; 0 if otherwise
Foreign Asset % of Total Assets (WC08736) This item represents the percentage assets from foreign operations before adjustments and eliminations of total assets.
MNC Assets 1 if WC08736 > 10%; 0 if otherwise
Cash and Short-Term Investments (CHE) This item represents cash and all securities readily transferable to cash as listed in the current asset section.
Receivables (RECT) This item represents amounts on open account owed by customers for goods and services sold in the ordinary course of business and claims against others collectible in cash.
Inventories (INVT) This item represents merchandise bought for resale and materials and supplies purchased for use in production of revenue.
Intangible Assets (AT – PPENT + CHE + RECT + INVT) / AT
Net Sales (SALE) This item represents gross sales (the amount in actual billings to customers for regular sales completed during the period) reduced by cash discounts, trade discounts, and returned sales and allowances for which credit is given to customers, for each operating segment.
Size The natural logarithm of SALE
Market-to-Book Market Value Assets / AT
Operating Income Before Depreciation (OIBDP) This item represents Net Sales, adjusted for cost of goods sold and selling, general, and administrative expenses.
Profitability OIBDP / AT
Cash Flow Volatility The standard deviation of Profitability calculated over the trailing ten years, requiring at least four trailing years. For firms with less than ten, but at least four year with available data, the available trailing years are used.
Equity Returns This item represents the change in the total value of an investment in a common stock over some period of time per dollar of initial investment. It is based on a purchase on the most recent day when the security had a valid price.
39
References
Adam, T., Goyal, V., 2008. The investment opportunity set and its proxy variables. Journal of Financial Research 31, 41-63.
Bae, K., Goyal, V., 2009. Creditor rights, enforcement, and bank loans. Journal of Finance 64, 823-860.
Berger, A., Frame, W., Ioannidou, V., 2011. Tests of ex ante versus ex post theories of collateral using private and public information. Journal of Financial Economics 100, 85–97.
Burgman, T., 1996. An empirical examination of multinational corporate capital structure. Journal of International Business Studies 27, 553-570.
Carey, M., Post, M., Sharpe, S., 1998. Does corporate lending by banks and finance companies differ? Evidence on specialization in private debt contracting. Journal of Finance 53, 845-878.
Chava, S., Roberts, M., 2008. How does financing impact investment? The role of debt covenants. Journal of Finance 63, 2085-2121
Chen, C., Cheng, C., He, J., Kim, J., 1997. An investigation of the relationship between international activities and capital structure. Journal of International Business Studies 28, 563-577.
Conyon, M., Haß, L., Vergauwe, S., Zhang, Z., 2019. Foreign experience and CEO compensation. Journal of Corporate Finance 57, 102-121.
De Jong, A., Kabir, R., Nguyen, T., 2008. Capital structure around the world: the roles of firm- and country-specific determinants. Journal of Banking & Finance 34, 1954-1969.
Desai, M., Foley, C., Hines, J., 2004. A multinational perspective on capital structure choice and internal capital markets. Journal of Finance 59, 2451–2487.
Doukas, J., Pantzalis, C., 2003. Geographical diversification and agency costs of debt of multinational firms. Journal of Corporate Finance 9, 59-92.
Edwards, D., 2001. Patent backed securitization: blueprint for a new asset class. Gerling NCM Credit Insurance Publication.
Fatemi, A., 1988. The effect of international diversification on corporate financing policy. Journal of Business Research 16, 17-30.
Fernandes, N., 2011. Global convergence of financing policies: evidence for emerging-market firms. Journal of International Business Studies 42. 1043-1059.
Frank, M., Goyal, V. 2008. Trade-off and pecking order theories of debt. In: Eckbo, B. (Ed.), Handbook of Corporate Finance: Empirical corporate Finance Volume 2 editor. Elsevier/North-Holland.
Frank, M., Goyal, V., 2009. Capital structure decisions: which factors are reliably important? Financial Management 38, 1-37.
Graham, J., Leary, M., Roberts, M., 2015. A century of capital structure: the leveraging of corporate America. Journal of Financial Economics 118, 658-683.
Graham, J., Harvey, C., 2001. The theory and practice of corporate finance: evidence from the field. Journal of Financial Economics 60, 187-243.
Gu, F., Lev, B., 2004. The information content of royalty income. Accounting Horizons 18, 1-12.
Hall, T., 2012. The collateral channel: evidence on leverage and asset tangibility. Journal of Corporate Finance 18, 570-583.
Hart, O., Moore, J., 1994. A theory based on the inalienability of human capital. Quarterly Journal of Economics 109, 814-879.
Horsch, P., Longoni, P., Oesch, D., 2020. Intangible capital and leverage. Journal of Financial and Quantitative Analysis, 1-24.
IRS (1994-2013). Internal Revenue Service (IRS) Reports on Corporation Income Tax Returns, 1994-2013. <https://www.irs.gov/statistics/soi-tax-stats-corporation-complete-report-1994-to-2013>.
Jang, Y., 2017. International corporate diversification and financial flexibility. Review of Financial Studies 30, 4133-4178.
Jensen, M., Meckling, W., 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economic 3, 305-360.
Kaplan, R., Norton D., 2004. Strategy maps: converting intangible assets into tangible outcomes. Harvard Business School Press.
Kwok, C., Reeb, D., 2000. Internationalization and firm risk: an upstream-downstream hypothesis. Journal of International Business Studies 31, 611-629.
Larkin, Y., 2013. Brand perception, cash flow stability, and financial policy. Journal of Financial Economics 110, 232-253.
Lee, K., Kwok, C., 1988. Multinational corporations vs. domestic corporations: international environmental factors and determinants of capital structure. Journal of International Business Studies 19, 195-217. Lemmon, M., Roberts, M., Zender, J., 2008. Back to the beginning: persistence and the cross-section of
