The leverage of private SMEs
in high-income OECD countries
By C.E. Mauritzen
Supervisor: Dr. H. Gonenc
MSc International Financial Management
Faculty of Economics and Business
University of Groningen
June 2016
Abstract
This paper examines factors affecting the leverage of private SMEs in eight high-income OECD countries during the period 2006 to 2011. From testing firm- and country specific factors much of the previous research is proven to be accurate in the case of private SMEs as well. The results show that the size of private SMEs is negatively related to the use of
leverage and that in countries with a stable banking sector, or alternatively a stable stock market, the use of leverage among private SMEs is higher. The subprime crisis were not found to have any significant impact on the leverage of private SMEs, however, the leverage of private SMEs were positively affected by the systemic banking crisis that many countries experienced during, and after the subprime crisis.
I. INTRODUCTION ... 4
II. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT ... 6
2.1 MEASURING LEVERAGE ... 6
2.2 LEVERAGE AND PRIVATE SMES ... 7
2.3 FACTORS AFFECTING LEVERAGE ... 9
III. DATA AND METHODOLOGY ... 17
3.1 DATA ... 17
3.2 METHODOLOGY ... 26
IV. RESULTS ... 31
V. DISCUSSION ... 34
VI. CONCLUDING REMARKS ... 38
ACKNOWLEDGEMENTS ... 40
VII. REFERENCES ... 41
I. Introduction
The studies on leverage and capital structure of firms are numerous, many of them implement both firm- and country specific variables that try to explain variations in the leverage of firms, most commonly large firms are studied. However, this body of research only embodies the tip of the proverbial iceberg.
This thesis aims to study the invisible part of said iceberg, specifically what and how firm and country specific variables influence the leverage of unlisted, small and medium-sized enterprises in eight high-income OECD countries in the time period 2009 to 2011.
Private SMEs are important drivers of the economy in countries; they are responsible for creating jobs and overall production (Matteev, Poutziouris and Ivanov, 2013). However, their access to external finance is often constrained (IFC, 2009). Being private, these firms do not benefit from access to public capital markets, and therefore the use of leverage for financing is extensive (Brav, 2009). Previous research has found firm specific variables such as tangibility, size, profitability and growth opportunities to affect the capital structure of firms (Myers, 1984; Myers and Majluf, 1984; Rajan and Zingales, 1995; Booth, Aivazian, Demirguç-Kunt and Maksimovic, 2001; De Jong, Kabir and Nguyen, 2008). Along with these firm specific variables, also country specific variables relating to the development of bank sectors, stock and bond markets, and the legal enforcement are also proven to impact the leverage of firms by way of better access to external capital or having a legal framework that better facilitates for the use external finance (La Porta, López-de-Silanes, Shleifer and Vishny, 1998; Demirguç-Kunt and, 1999; Giannetti, 2003; De Jong, Kabir and Nguyen, 2008; Demirgüç-Kunt, Martinez-Peira and Tressel, 2015).
II. Literature Review and Hypotheses Development
2.1 Measuring leverage
In their study of public and private firm capital structure during the subprime crisis Demirgüç-Kunt et al. (2015) use three measures for leverage, where the first two measures captures how firms are financed with long-term debt and debt in general and the third measure captures the maturity of the debt. In this thesis I will adopt the first two of these measures; these measures will provide a picture of the extent to which firms finance themselves with debt in general, and more specifically long-term, given by total debt to total assets and long-term debt to total assets. Since this thesis is focused on private SMEs the measurements are calculated from book values. Giannetti (2003) also use book values when calculating measurements of leverage in her study of private firms because market values are notoriously hard to obtain for private firms.
In finance theory, a narrow measurement of leverage is usually adopted; Rajan and Zingales (1995) provide a good explanation of why broader definitions such as total liabilities over total assets, short- and long-term leverage over total assets and total debt to net assets will not capture the financing decisions of firms in a sufficient manner. Using total liabilities to total assets as a proxy for firm capital structure fail to provide any information of whether a risk of default is present. Total liabilities include items that are not relevant in assessing the financing of a firm; items such as these tend to be used for transaction purposes (Rajan and Zingales, 1995). Rajan and Zingales states that using the ratio of short-term and long-term leverage to total assets might provide a better picture of the leverage of firms, but this measure also has its shortcomings, among these are issues regarding trade credits. Another measure that is not affected by trade credits is the ratio of total debt to net assets; however, the authors argue that this measure might also include items that have nothing to do with the financing of firms1. Despite of the shortcomings identified by Rajan and Zingales (1995) broad definitions of leverage are adopted in this thesis to ease the data collection process and to ensure consistency in the measures, which are calculated using data from several different countries, which have different accounting standards and practices.
2.2 Leverage and private SMEs
As previously stated most of the current literature about country specific determinants of capital structure and the subprime crisis has been written in relation to large listed firms, however it is important to study how firm and country specific factors affects privately held SMEs. Although they do not wield the same kind of power as large multinational corporations, SMEs play an important part in their domestic economy by creating jobs and increasing overall production (Mateev et al. 2013), however, they do not benefit from the same accessibility of financing as larger firms do, which leads them to seek other ways to obtain financing and thus the capital structure of SMEs might be different than that of large firms (IFC, 2009). In order to properly study the leverage among SMEs a proper definition of SMEs are in order.
The definition of SMEs depends on how the countries in question choose to define small and medium-sized enterprises, for instance in the US, where numerous very large firms reside, the definition of SMEs is firms with less than 500 employees (United States International Trade Commission, 2010). The European Commission defines SMEs as firms that consist of less than 250 employees and have an annual turnover of less than 50 million euro or an annual balance sheet of less than 43 million euro (European Commission, 2006).
This thesis will follow the definition of the European Commission. As stated, this thesis will include only private SMEs, while it is acknowledged that SMEs do not benefit from the same accessibility to external financing as large firms, there are differences of note in the leverage of private and public SMEs that are important to examine as well. As opposed to public firms, private firms rely almost entirely on financing through debt, Brav (2009) found that the leverage of private firms are approximately 50 per cent higher than that of public firms, the same goes for short-term debt, where private firms’ ratio of short-term debt were found to be twice as high than that of public firms. However, when it comes to obtaining external capital through issuing equity public firms tend to use this method about four times as often in comparison to private firms2. Brav (2009) attribute differences between private and public firms to what he calls level, and sensitivity effects. Level effects are the consequences of private firms having a higher cost of equity compared to that of public firms, thus making them more likely to opt for debt financing in comparison to public firms. The sensitivity
2 40 percent of the external capital raised in public firms comes from issuing equity, whereas only 10 percent of external finance
less sensitive to operating cash flows than that of private firms, Saunders and Steffen (2011) show that private firms pay higher borrowing costs than public firms, and Capmello, Graham and Campbell (2010) find that public firms enjoy less expensive lines of credit than private firms.
2.3 Factors affecting leverage
2.3.1 Firm specific factors
Several studies have focused on the firm specific determinants of capital structure and found that these determinants stay consistent in international settings as well, Rajan and Zingales (1995) found that in their sample of G7 countries that asset tangibility and size had a positive effect on leverage, while the market to book ratio and profitability of firms had a negative effect3, the findings of Booth et al. (2001) supports the validity of these firm specific factors in their study of the capital structure of firms in developing countries. Out of the four variables set forth by Rajan and Zingales this study will implement three of them, namely tangibility, size and profitability. In the case of asset tangibility finance theory stipulates that when firms have a large amount of tangible assets lenders might be more willing to supply the firm with debt because the tangibility of the firm will serve as a signal as to what extent the firm’s assets might serve as collateral (Rajan and Zingales, 1995). Tangibility has been shown to have a positive effect on leverage (Booth et al., 2001; De Jong et al., 2008; Rajan and Zingales, 1995). Hence, when obtaining finance, a private SME may offset the lenders’ perception of risk by having assets that can serve as collateral.
When mitigating risk diversification can serve as an efficient tool, Rajan and Zingles (1995) suggests that big firms tend to have a higher degree of diversification than that of smaller firms, thus the probability that a large firm will go bankrupt is less than that of a small firm (De Jong et al., 2008; Rajan and Zingales, 1995). This variable brings forth an interesting element to this research as the firms included in the sample are classified as SMEs, meaning that relative to the other firms in their industry and environment in general the SMEs are small firms and thus, according to theory, more exposed to the risk of bankruptcy. In the sample of this thesis small SMEs should have a higher risk of bankruptcy than that of larger
3 The variable size had a positive impact on leverage in all countries except Germany (negative). The variable profitability was
SMEs, but given that lenders supply whole populations of firms with loans, very small and very large firms, will they distinguish between small SMEs and large SMEs, or rather just classify them as small firms altogether? If the lenders do distinguish between small and large SMEs size is expected to have a positive relationship to leverage.
There have been discussions about whether profitability affects the leverage of firms positively or negatively; Myers and Majluf (1984) argue that the relationship between leverage and profitability is a negative one. The pecking order theory, which states that firms would rather finance their investments with internally generated funds than seek external financing, supports the arguments of Myers and Majluf (1984). In the case of private SMEs ownership concentration is relatively high compared to listed firms, therefore the concentration of power is high. When taking on additional leverage the power of the owners or managers of the SME, which often is the same person, will decrease. It stands to reason that in order to retain control of the firm the owners will want to finance their investment with internally generated funds, thus supporting the pecking order theorem. Jensen (1986) predicts a positive relationship, however this positive relationship is contingent on the efficiency of the market of corporate control, if efficiency is high pressure can be put on firms in order to make them pay out cash by acquiring more leverage. Following the pecking order theory and the expectations of Booth et al., (2001), De Jong et al., (2008) and Rajan and Zingales (1995) profitability is in this study expected to have a negative relationship to firm leverage.
H1a: Tangibility has a positive impact on the leverage of private SMEs. H1b: Firm size has a positive impact on the leverage of private SMEs. H1c: Profitability has a negative impact on the leverage of private SMEs.
H1d: Growth opportunities have a positive impact on the leverage of private SMEs.
Firm specific factors has been shown to explain some of the reasons behind firm capital structure, however, country specific factors such as development of financial sector, banking sector and legal institutions have also been found to play an important role in this question (Demirgüç-Kunt and Maksimovic, 1999; Graham and Harvey, 2001; Booth et al. 2001; Claessens, Djankov and Nenova, 2001; Fan, Titman and Twite, 2012; Bancel and Mittoo, 2004; Brounen, De Jong and Koedijk, 2006). It is reasonable to assume that the leverage ratios of private SMEs vary across countries because of differences in factors such as development of the financial and banking sector, and differences in the legal environment.
2.3.2 Country characteristics
When investigating how financial institutions impact corporate debt maturity Demirgüç-Kunt and Maksimovic (1999) state that the presence of agency costs will influence how firms choose to structure their financing in terms of debt versus equity. However, these agency costs can be controlled by financial contracts, and the authors argue that the impact of these financial contracts rely greatly on the legal environment of countries (Demirgüç-Kunt and Maksimovic, 1999). The better the enforcement of financial contracts and enforcement of laws in general, the risk of providing private SMEs with debt should decrease. If the legal system in a country is strong (weak) providers of debt, such as banks will have to rely less (more) on short-term debt when providing debt.
borrowers are under stricter contract enforcement on their debt when the degree of legal enforcement is higher (De Jong et al. 2008).
Another study, using the classifications of La Porta et al. (1998), find that where the legal system is more efficient companies rely more on long-term debt to finance their investments (Demirgüç-Kunt and Maksimovic, 1999). The same study finds that in countries where the legal framework facilitates for better creditor rights, firms has easier access to debt financing through the banking sector being more developed. It is arguable that better creditor protection rights may result in it being less risky for banks to issue long-term debt. Therefore, creditor rights protection could be positively associated with the use of leverage by offsetting the risks represented by lending to private SMEs. Alternatively, when creditor rights protection is high, lenders bear the most risk in the transaction, and thus the bankruptcy risks may be too great for SMEs and therefore they seek to obtain their financing in other ways, hence, creditor protection rights being negatively associated with the leverage of SMEs. In her study about how institutions affect the capital structure choices of unlisted firms, Giannetti (2003) find evidence supporting Demirgüç-Kunt and Maksimovic (1999), namely that creditor protection and the level of legal enforcement is positively associated with firm leverage and the use of long-term leverage for financing. Fan et al. (2012) present results along the same line of reason as Demirgüç-Kunt and Maksimovic (1999), and Giannetti (2003), that in countries of common-law origin debt ratios are generally lower and the debt maturity is predominantly long-term (Fan et al., 2012). Looking back at the classification of La Porta et al. (1998) it makes sense that a common-law country will have stronger shareholder protection rights and better legal enforcement. Thus, the total debt ratio of public firms will decrease because of high availability, and low cost, of equity in developed stock markets and shareholder protection being emphasised in the laws. However, the maturity of debt will increase; this might be due to the high degree of law enforcement in common-law countries. Risks tied to offering debt to private SMEs are expected to decrease, as the requirement for monitoring will be less with an efficient legal system.
By having access to a highly developed stock market firms may be inclined to substitute debt with equity, thus at first glance, developed stock markets seem to have a negative effect on firm leverage. However, Demirgüç-Kunt and Maksimovic (1999) suggests an alternative view on how the development of stock markets influences the capital structure of firms. Instead of having a negative impact on leverage by offering firms better access to equity financing, a highly developed stock market will offer creditors information about their borrowers that would not be available to them if the stock market were to be underdeveloped (Demirgüç-Kunt and Maksimovic, 1999). This information can be found in the price of securities. These quotes offer information that creditors such as banks might take in consideration when making decisions regarding lending to publicly traded firms, lending with less risk associated to the contract, and thus be able to offer more long-term loans (Grossman, 1974, and Grossman and Stiglitz, 1980). Banks would get information about the firms listed on said market, thus it stands to reason that the increased availability of debt financing would only be available to publicly listed firms. However banks are driven by profit as any company, and being able to provide long-term contracts of less risk might induce them to offer more debt financing to private SMEs because of them having a more diversified portfolio of borrowers. However, Demirgüç-Kunt and Maksimovic (1999) does not find any evidence of stock market size having any impact on the capital structure choices of small firms. De Jong et al. (2008) find that stock market development has an indirect impact on the use of debt amongst firms, more specifically a negative impact by way of equity being more accessible for firms. Giannetti (2003) find that in countries where stock markets are of low development the use of leverage for financing is higher. The literature reviewed is not in agreement regarding how the development of a stock market affects the capital structure of firms, it seems as it is dependent on the specifics of the firms in the sample included in the studies, Giannetti (2003) studies private firms, and find that the less developed a country’s stock market is the more leveraged the firms are. Demirgüç-Kunt and Maksimovic (1999) includes both public and private firms of all sizes in their study and find that the development of stock markets is not of significant importance for small firms when deciding on their capital structure. De Jong et al. (2008) also includes all types of firms in their study and report findings slightly similar to that of Giannetti (2003) where there is a negative relationship between stock market development and the use of leverage.
more leverage than that of firms without the same access, everything else being equal. Both stock market and bond market development shouldn’t have an impact on the leverage of private SMEs on accounts of these markets being inaccessible to these firms, however the development of these markets can serve as an indicator of overall financial activity and economic development in a country and thus influence the leverage of private SMEs indirectly, Beck, Demirguç-Kunt and Maksimovic (2008) support this statement in arguing that firms enjoy greater access to external finance in countries with a high degree of financial development, this especially relates to small firms.
H2a: Banking sector access has a positive impact on the leverage of private SMEs. H2b: Banking sector stability has a positive impact on the leverage of private SMEs. H2c: Bond market development has a positive impact on the leverage of private SMEs. H2d: Stock market access has a positive impact on the leverage of private SMEs. H2e: Stock market stability has a positive impact on the leverage of private SMEs. H2f: GDP growth has a positive impact on the leverage of private SMEs.
H2g: Legal enforcement has a positive impact on the leverage of private SMEs.
2.3.3 The subprime crisis, and its effect on the leverage of private SMEs
Although in itself the subprime crisis is not a determinant of the leverage of private SMEs the conditions it brought upon the economies around the world affected the leverage of firms. This subchapter will try to identify how the subprime crisis affected the leverage of private SMEs
as identified by Laeven and Valencia (2013). From their findings it’s possible to see that five of the countries included in this study experienced a systemic banking crisis that started in 2008. Knowing that five of the countries in this study plunged into a systemic banking crisis soon after the onset of the subprime crisis it’s expected that the subprime crisis and the subsequent banking crisis will have a negative effect on the leverage of private SMEs in all countries included in this study.
III. Data and methodology
The first section of this chapter describes the sample used in this thesis and also the dependent variables; the second section will cover the independent variables, both firm specific and country specific. Descriptive statistics can be found in table 2 while table 3 serves as a correlation matrix for the variables. Section 3.2 includes the specification of the models.
3.1 Data
3.1.1 Sample and dependent variables
Firm specific variables are collected from the ORBIS database from Bureau van Dijk. This database contains information of over 200 million companies worldwide.
3.1.2 Independent variables
Tangibility is calculated as the ratio of fixed assets to total assets, the means range from
0.35555 in Belgium to 0.5846 in the Slovak Republic. The highest standard deviation reported is 0.26959 from the Spanish sample. Size is given by the natural logarithm of operating revenue. The largest and smallest firm observed is located in Korea, and France, respectively.
Profitability is calculated by EBITDA divided by total assets, from table 2 we can deduct that
on average, the most profitable firms are located in Finland, and the firms with the lowest profitability are located in Italy. The last firm specific variable included is Growth; this variable measures the growth opportunities of the firm by dividing sales on total assets. The means tell us that on average, French firms have the highest growth opportunities, while Greek firms have the lowest growth opportunities.
Table 3 presents how the variables are correlated to each other. Tangibility is, rather surprisingly, negatively correlated with TDTA, but positively correlated with LTDTA. Size is positively correlated with TDTA; however, the variable is negatively correlated with LTDTA.
Profitability of SMEs is negatively correlated with TDTA, which suggests that more
profitable SMEs will finance their investments with internally generated funds; however
profitability is positively related to LTDTA. Lastly, the growth opportunities of SMEs are
positively correlated to the ratio of total debt to total assets, while it’s negatively correlated to the ratio of long-term debt to total assets. This correlation might suggest that SMEs have a hard time financing their growth opportunities with long-term debt, compared to short-term debt.
A full list of the firm and country specific variables and their sources can be found in table A1 in the appendix.
Stability will provide a picture of the development of the sector, consisting of banks and
financial intermediaries that might provide private SMEs with leverage in a country.
On average, Spain has the most accessible, but least stable banking sector during the time period, the Slovak Republic has the least accessible one, and Finland has the most stable banking sector. From the correlation matrix in table 3 it’s possible to assess how the size and stability of the banking sector in countries are correlated with the dependent variables, Bank
Access and Bank Stability are both negatively correlated with all the dependent variables. In
order to measure the development of bond markets the variable Bond is constructed. This variable is given by the average of outstanding domestic private debt securities to GDP, and outstanding domestic public debt securities to GDP. Out of the countries in the sample Italy has the highest mean measure of Bond; the Slovak Republic has the lowest mean value of
Bond. The variables Stock Access and Stock Stability measure the accessibility and size of the
stock market, and the stability of the stock market, respectively. Stock Access is given by the average of stock market total value traded to GDP, and stock market capitalization to GDP.
Stock Stability is given by the average of stock market turnover ratio, and stock price
volatility. The calculation of the second element in Stock Stability is given in equation 1 below.
Equation 1)
1 − (𝑆𝑡𝑜𝑐𝑘 𝑝𝑟𝑖𝑐𝑒 𝑣𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦𝑖
100 )
Equation 1: Stock price volatility is the annual average volatility of the stock market index of country i.
Together the variables Stock Access and Stock Stability provide a picture of the overall development of the stock market in a country. The Republic of Korea (henceforth referred to as Korea) has both the most accessible and stable stock market out of all the countries in the sample.
All of the country specific independent variables discussed thus far in this section are collected from the Global Financial Indicators database of the World Bank.
Legal is a measure given by four variables provided by the World Bank’s World Governance
Indicators database. These variables are highly correlated to each other, a case of near multicollinearity might therefore be present, in order to adjust for this problem the variable
Legal is constructed as an average of the four measures to effectively measure the strength of
the legislative institutions, the strength of legal rules and the degree of legal enforcement in countries, De Jong et al. (2008) also used such a strategy to deal with the same problem. Table 3 shows how the variable is correlated with the dependent variables; Legal is negatively correlated to TDTA, and positively correlated to LTDTA. Looking back at section 2.3.2 it was expected that strong legal enforcement is positively related to the leverage of private SMEs, however table 3 shows a small negative correlation to total debt to total assets, and stronger positive correlation to long-term debt to total assets. Five sets of dummy variables are also included in this study in order to capture other unobserved effects that the main independent variables are not capable to capture. These dummies are year dummies, country dummies, industry dummies and two crisis dummies, one for the subprime crisis and another one for the banking crisis. The last dummy variable was developed by Laeven and Valencia (2013). This dummy is equal to one if a country experienced a systemic banking crisis during the period in question.
Table 1: Industry dummy codes
Dummy Code NACE Rev. 2 main section (Industry group)
A Agriculture, forestry and fishing B Mining and quarrying C Manufacturing
D Electricity, gas, steam and air conditioning supply
E Water supply; sewerage, waste management and remediation activities F Construction
G Wholesale and retail trade; repair of motor vehicles and motorcycles H Transportation and storage
I Accommodation and food service activities J Information and communication
L Real estate activities
M Professional, scientific and technical activities N Administrative and support service activities P Education
Q Human health and social work activities R Arts, entertainment and recreation S Other service activities
Table 2: Descriptive Statistics
TDTA LTDTA Tang Size Profit Growth Bank Access Bank Stab Bond Stock Access Stock Stab GDP Legal
Table 2: Descriptive Statistics, continued
TDTA LTDTA Tang Size Profit Growth Bank Access Bank Stab Bond Stock Access Stock Stab GDP Legal
Table 2: Descriptive Statistics, continued
TDTA LTDTA Tang Size Profit Growth Bank Access Bank Stab Bond Stock Access Stock Stab GDP Legal
Table 3: Correlation Matrix
Tang Size Profit Growth Bank Access Bank Stab Bond Stock Access Stock Stab GDP Legal TDTA -0.02359 0.01748 -0.15806 0.11883 -0.02405 0.06571 0.03018 -0.02797 0.04713 -0.06846 -0.01868 LTDTA 0.42837 -0.23506 0.01196 -0.16499 -0.02364 0.22620 -0.23505 0.14347 0.08571 -0.04517 0.22824 Tang 1.00000 Size -0.19148 1.00000 Profit 0.03188 -0.18961 1.00000 Growth -0.37638 0.10231 0.22058 1.00000 Bank Access -0.12269 0.14388 -0.20588 -0.00082 1.00000 Bank Stab 0.09116 -0.14844 0.20387 0.09588 -0.39383 1.00000 Bond -0.27927 0.16421 -0.20805 0.01327 0.63566 -0.49300 1.00000 Stock Access -0.00470 0.03753 0.05342 0.13384 0.31584 0.02902 0.03657 1.00000 Stock Stab -0.04086 0.21690 -0.09686 0.00139 0.43203 -0.10780 0.25399 0.72464 1.00000 GDP 0.09284 0.02099 -0.05599 -0.07325 -0.17139 -0.22290 -0.13132 -0.24562 -0.09240 1.00000 Legal 0.04321 -0.29115 0.28066 0.25517 -0.14446 0.61466 -0.37955 0.39791 0.01670 -0.14378 1.00000
Table 3: Correlation Matrix: The dependent variables TDTA and LTDTA represent the ratio of total debt to total assets, and the ratio of long-term debt to total assets. Tang is an abbreviation for tangibility, size represents the size of private SMEs, profit represents the profitability of said firms, and growth represents the growth opportunities of the firm. Bank Access and Bank Stab are variables that measure the accessibility and stability of the banking sector, Bond measures the development of bond markets. Stock Access and Stock Stab measures the accessibility and stability of stock markets. GDP measures the annual growth in gdp per capita and legal is a measure of the legal enforcement in a country. For a detailed
3.2 Methodology
This subchapter contains the reasoning behind the choice of methodology, specification of the models and a summary of the expected effect the independent variables will have on the dependent variables TDTA and LTDTA.
3.2.1 Choice of model
The data sample used in this thesis is comprised of a combination of time series and cross-sectional components that provide us with information about specific firms (Brooks, 2014). In his book, Brooks list numerous ways to deal with data of such structure; among these are methods such as OLS (ordinary least squares), SUR (seemingly unrelated regression), and fixed or random effects models. Using a fixed effects model comes with considerable benefits because the intercept in said model will capture unobserved effects that affect the dependent variables This kind of model is used in studies by Giannetti (2003) and Demirguç-Kunt et al. (2015) to name a few. I will however use an OLS model with a rich set of dummy variables that may capture unobserved effects and thereby provide additional insights to the research objective of this thesis. Drawing inspiration from De Jong et al. (2008) where they first introduce a simple OLS model testing firm specific factors and in later models adding on country specific characteristics as well. It is worth noting that De Jong et al. also use SUR and WLS-models in their study. This thesis will only use the simple OLS model that De Jong et al. initially introduced, and develop it further by including dummies and other country specific factors.
3.2.2 Model specification
Model 1) 𝐿𝐸𝑉𝑖,𝑡= 𝛽0+ 𝛽1𝑇𝑎𝑛𝑔𝑖,𝑡+ 𝛽2𝑆𝑖𝑧𝑒𝑖,𝑡+ 𝛽3𝑃𝑟𝑜𝑓𝑖𝑡𝑖,𝑡+ 𝛽4𝐺𝑟𝑜𝑤𝑡ℎ𝑖,𝑡+ 𝛽5(𝑌𝑒𝑎𝑟𝐷𝑢𝑚𝑚𝑖𝑒𝑠) + 𝛽6(𝐶𝑜𝑢𝑛𝑡𝑟𝑦𝐷𝑢𝑚𝑚𝑖𝑒𝑠) + 𝛽7(𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝐷𝑢𝑚𝑚𝑖𝑒𝑠) + 𝜀𝑖,𝑡 Model 2) 𝐿𝐸𝑉𝑖,𝑡= 𝛽0+ 𝛽1𝑇𝑎𝑛𝑔𝑖,𝑡+ 𝛽2𝑆𝑖𝑧𝑒𝑖,𝑡+ 𝛽3𝑃𝑟𝑜𝑓𝑖𝑡𝑖,𝑡+ 𝛽4𝐺𝑟𝑜𝑤𝑡ℎ𝑖,𝑡+ 𝛽5(𝑌𝑒𝑎𝑟𝐷𝑢𝑚𝑚𝑖𝑒𝑠) + 𝛽6(𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝐷𝑢𝑚𝑚𝑖𝑒𝑠) + 𝛽7𝐵𝑎𝑛𝑘𝐴𝑐𝑐𝑒𝑠𝑖,𝑡+ 𝛽8𝐵𝑎𝑛𝑘𝑆𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖,𝑡+ 𝛽9𝐵𝑜𝑛𝑑𝑖,𝑡 + 𝛽10𝑆𝑡𝑜𝑐𝑘𝐴𝑐𝑐𝑒𝑠𝑠𝑖,𝑡+ 𝛽11𝑆𝑡𝑜𝑐𝑘𝑆𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖,𝑡+ 𝛽12𝐺𝐷𝑃𝑖,𝑡+ 𝛽13𝐿𝑒𝑔𝑎𝑙𝑖,𝑡+ 𝜀𝑖,𝑡 Model 3) 𝐿𝐸𝑉𝑖,𝑡= 𝛽0+ 𝛽1𝑇𝑎𝑛𝑔𝑖,𝑡+ 𝛽2𝑆𝑖𝑧𝑒𝑖,𝑡+ 𝛽3𝑃𝑟𝑜𝑓𝑖𝑡𝑖,𝑡+ 𝛽4𝐺𝑟𝑜𝑤𝑡ℎ𝑖,𝑡+ 𝛽5(𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝐷𝑢𝑚𝑚𝑖𝑒𝑠) + +𝛽6𝐵𝑎𝑛𝑘𝐴𝑐𝑐𝑒𝑠𝑖,𝑡+ 𝛽7𝐵𝑎𝑛𝑘𝑆𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖,𝑡+ 𝛽8𝐵𝑜𝑛𝑑𝑖,𝑡+ 𝛽9𝑆𝑡𝑜𝑐𝑘𝐴𝑐𝑐𝑒𝑠𝑠𝑖,𝑡 + 𝛽10𝑆𝑡𝑜𝑐𝑘𝑆𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖,𝑡+ 𝛽11𝐺𝐷𝑃𝑖,𝑡+ 𝛽12𝐿𝑒𝑔𝑎𝑙𝑖,𝑡+ 𝛽13(𝑆𝑢𝑏𝑝𝑟𝑖𝑚𝑒) + 𝛽14(𝐵𝑎𝑛𝑘𝑖𝑛𝑔𝐶𝑟𝑖𝑠𝑖𝑠) + 𝜀𝑖,𝑡
In all models the dependent variable LEVi,t is a measure of the ratio of total debt to total
assets, and long-term debt to total assets for firm i at time t.
The firm specific variables 𝑇𝑎𝑛𝑔𝑖,𝑡 , 𝑆𝑖𝑧𝑒𝑖,𝑡 , 𝑃𝑟𝑜𝑓𝑖𝑡𝑖,𝑡 and 𝐺𝑟𝑜𝑤𝑡ℎ𝑖,𝑡 represents the tangibility, size, profitability and growth opportunities of private SMEs, respectively. The country specific variables 𝐵𝑎𝑛𝑘𝐴𝑐𝑐𝑒𝑠𝑠𝑖,𝑡 , 𝐵𝑎𝑛𝑘𝑆𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖,𝑡 , 𝐵𝑜𝑛𝑑𝑖,𝑡 , 𝑆𝑡𝑜𝑐𝑘𝐴𝑐𝑐𝑒𝑠𝑠𝑖,𝑡, 𝑆𝑡𝑜𝑐𝑘𝑆𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖,𝑡, 𝐺𝐷𝑃𝑖,𝑡 and 𝐿𝑒𝑔𝑎𝑙𝑖,𝑡 represents the accessibility and size of the banking sector, stability of the banking sector, development of bond markets, the accessibility and size of the stock market, the stability of the stock market, the GDP growth per capita and the degree and strength of legal enforcement, respectively. The calculation methods of these variables are given in section 3.1.2 and are also presented in table A1 in the appendix. Lastly, 𝜀𝑖,𝑡 is the error term. All of the explanatory variables are for the ith firm at time t.
observations included in each model, for reference the total sample consists of 25771 observations. The results are presented in table 4.
Table 4: Regression output.
(1) (2) (3)
TDTA LTDTA TDTA LTDTA TDTA LTDTA
Table 4: Regression output (continued)
TDTA LTDTA TDTA LTDTA TDTA LTDTA
Table 4: Regression output (continued)
TDTA LTDTA TDTA LTDTA TDTA LTDTA
IV. Results
This section will report the results presented in table 4, in section V these results are discussed and section VI will provide a conclusion of this thesis as a whole.
Firm specific factors
Tangibility has over all the three models positive coefficients related to both TDTA and LTDTA, especially for the ratio of long-term debt to total assets, therefore hypothesis H1a is supported. The results on the firm specific variable Size are mixed, in the first model the coefficient related to TDTA is negative, small and significant at the 1% level, and the coefficient related to LTDTA is also negative at the same significance level. In model number two, where country specific factors are added and country dummies are dropped the first coefficient for size is small and negative, however not significant, while the second coefficient for size is still negative and significant at the 1% level. The third model also has negative coefficients for Size, thus no support is found for hypothesis H1b.
Profitability has large, negative and highly significant coefficients in all three models, for both TDTA and LTDTA, which lends support to hypothesis H1c. Growth opportunities was hypothesised to have a positive impact on the leverage of private SMEs and the results seem to support this hypothesis by presenting positive coefficients in all of the three models towards TDTA. On the other hand, growth opportunities have negative coefficients towards LTDTA, however, since growth opportunities have a positive impact on the ratio of total debt to total assets it is concluded that H1d is supported.
Year, country and industry dummies
the only ones towards TDTA that are significant, at the 5%, 5% and 10% level, respectively. In the second set of coefficients the only one that is significant is the one for 2007, at the 10% level. In this second set negative coefficients for some of the year dummies appear, this tells us that for this model, when controlling for country specific characteristics, the year 2011 was not the year with the lowest coefficients.
Country dummies are only included in the first model where country specific characteristics are not controlled for. The coefficients for Belgium and France have no significant impact on TDTA, while Spain and Korea have negative coefficients that are significant at the 1% level. The coefficients for Italy and Finland are the only ones that seem to have a highly significant (1% level), and positive impact on the ratio of total debt to total assets, Italy has by far the largest coefficient. For LTDTA all of the country dummy coefficients are positive and highly significant. Finland has the largest coefficient, followed by Spain and Belgium.
are relatively large and highly significant for all the three models. The next industry dummy of note is Q, which represents human health and social work activities. The coefficients are negative in all of the models and significant at the 1% level for TDTA. The last dummy included is the one for arts, entertainment and recreation (R), this dummy has coefficients, which are negative and significant at the 1% level towards TDTA, and positive and significant at the 5% level towards LTDTA, with the exception of model one where the significance level is at 10%.
Country specific factors
The accessibility of the banking sector seems to have a positive impact on the leverage of private SMEs in model two and three; the results are significant at the 1% level in model two and at the 10% level in model three which lends support to hypothesis H2a. The stability of the banking sector was hypothesised to have a positive effect on the leverage of private SMEs; this hypothesis is supported by large and significant coefficients in both model two and three. Bond market development has a negative, but non-significant coefficients towards TDTA in both model two and three, however the coefficients are negative and highly significant towards LTDTA in both of the models, therefore hypothesis H2c is not supported. The access to stock markets seems to negatively impact the leverage of firms in model two and three, H2d is not supported. The stability of stock markets have a positive and highly significant impact on all of the dependent variables in models two and three, this lends support to hypothesis H2e. The coefficients for GDP are negative, but not statistically significant, therefore there is no support for hypothesis H2f. Legal enforcement is found to have a positive impact on LTDTA in models two and three at the 1% level. No significant impact is found to exist between legal enforcement and TDTA. It can be argued that legal enforcement affects impacts the leverage of private SMEs positively by way of being positively related to the ratio of long-term debt to total assets, thus, hypothesis H2g is supported.
Crisis dummies
V. Discussion
Based on the findings of previous studies such as those of Rajan and Zingales (1995), Booth et al. (2001) and De Jong et al. (2008), the firm specific variables were expected to impact the leverage of private SMEs in a specific way. Tangibility and size were in the aforementioned studies found to have a positive impact on the leverage of firms, while profitability was found to have a negative impact. Myers (1984), following the pecking-order theory, argued that growth opportunities would have a negative impact on the leverage of firms.
In the case of tangibility it seems as though the lenders (banks and other financial intermediaries) are willing to provide private SMEs with more long-term debt, and leverage in general, if the assets of the SMEs are tangible. Tangible assets seems to be an efficient mitigator of risk when it comes to lending to small and opaque borrowers, there is less need for monitoring, and the banks can afford to let more time go by in-between the renegotiation of debt contracts. This fact also holds true when controlling for crises.
for the ratio of long-term debt to total assets. When financing investments with short-term debt private SMEs are giving up more control to the ones providing this debt by agreeing to stricter monitoring and more frequent renegotiations of debt contracts. Following the logic of Myers (1984), and Myers and Majluf (1984) private SMEs and firms in general seek to retain as much control as possible. More profitable SMEs avoid the use of external finance, especially short-term debt.
The results report that the growth opportunities of private SMEs have a positive impact on the ratio of total debt to total assets and a negative effect on the ratio of long-term debt to total assets. These results suggest that when private SMEs have positive growth opportunities their access to debt, particularly short-term debt increases, and their access to long-term debt decreases. This dynamic could be explained by the fact that for banks and other providers of leverage to be willing to finance the growth opportunities of private SMEs a certain degree of control is demanded. This control comes in the form of better monitoring and less time between debt contract renegotiations by way of supplying the SMEs with short-term debt. Looking at the literature reviewed it becomes clear that financing the growth opportunities of private SMEs is seen as a risky endeavour because of the small degree of transparency private SMEs offers (Rajan and Winton, 1995; Pettit and Singer, 1985), therefore private SMEs looking to act on their growth opportunities will be supplied mainly with short-term debt.
dummies may be explained by the worsening economic conditions, given by the onset of the subprime crisis in 2008 and the subsequent systemic banking crisis that a number of the countries included in the sample experienced.
Controlling for industry effects by using a set of industry dummies it is possible to see how private SME’s ratios of total debt to total assets and long-term debt to total assets are affected by their respective industries. A reasonable assumption can be made of that, in general, firms that operate in industries that require a high amount of fixed assets, such as manufacturing, production, construction and the like, will have more leverage than firms that operate in, for example, retail and service industries. This assumption is made based on that private SMEs operating in industries such as those mentioned first are required to own a large amount of fixed assets to total assets, therefore having a higher degree of tangibility compared to the SMEs operating in the latter industries. However, no such results can be derived from table 4. A possible explanation for the lack of such evidence might be that the SMEs are indirectly associated with these industries by way of providing support services to larger firms, thus, suggesting that the high degree of fixed assets that might characterize certain industries is not applicable in the case of private SMEs. The most obvious industry effects are those connected to private SMEs engaged in real estate activities or human health and social work activities. SMEs in these two industries have noticeably smaller ratios of total debt to total assets. This might be due to less need for debt financing among SMEs in these industries, through higher profitability, or less need for investments. Another observation is that the coefficients for any of the industry dummies hardly change in model three when controlling for the subprime and banking crisis.
stability remains large, positive and significant. These results suggest that having a stable banking sector will mitigate the adverse effects that occurrences such as the subprime crisis and the banking crisis might cause.
the banking sector, stock market stability and accessibility, bond market development, and the effectiveness and strength of the legal enforcement will affect the leverage of private SMEs to a larger extent than that of proxies of overall economic development such as GDP growth per capita. Rather surprisingly the effects of both the subprime crisis and the banking crisis are positive, however, it is only the effect of the banking crisis on the ratio of long-term debt that is statistically significant. The relationships were expected to be negative, however I am not controlling for specific factors that can reduce the adverse effects the crises were expected to have. Among such factors are government interventions in the form of credit guarantee schemes or other schemes designed by monetary authorities to keep said adverse effects at a minimum. Looking at the country dummies included in model one, it is possible to compare the size and significance of the coefficients to the data in table 2. Finland has the largest positive coefficient towards the ratio of long-term debt of all the countries included, on the other end of the spectrum, the Slovak Republic can be found. Looking at the descriptive statistics of the two countries in table 2 it is clear that Finland “score” quite high on the factors that are proven to have a positive impact on leverage. Finland has the highest mean banking sector stability, a moderate degree of stock market stability and strong legal enforcement. The Slovak Republic has the lowest accessibility to the banking sector and stock market, the lowest stock market stability in addition to low values relating to legal enforcement. However, more in-depth analyses on country specific factors has to be done to fully map out how the different characteristics of countries affect the leverage of not only private SMEs, but also firms in general.
VI. Concluding remarks
Booth et al. (2001), and De Jong et al. (2008). One exception might be the fact that a private SME is by nature less transparent and therefore seen as more risky. Therefore, by virtue of being a private, small- or medium sized enterprise, size does not seem to matter for the ratios of leverage included in this study. Rather surprisingly, the ratios of large SMEs seem to be smaller than that of other private SMEs. Measures such as bank and stock market development are broken down to two different variables, one giving a picture of the size and accessibility of each sector/market and the other measuring the stability of each sector/market. This approach is to the best of the author’s knowledge novel. The results are significant and mostly positive with the exception of the accessibility of stock markets impacting the leverage of private SMEs negatively. The results suggest that even though a sector is small and inaccessible the stability of such sector is more important for the leverage of private SMEs as these measures may serve as proxies for overall confidence in the economic development in a country and thereby mitigate some of the adverse effects that occurrences like the subprime crisis and banking crisis brings.
This thesis provides some interesting directions for further research, the first of which relates to using more detailed measures on the development of markets and strength of institutions in order to break them down more and identify the functions that is of the biggest importance. The second direction relates to studying how government schemes, initiated to mitigate the effects of an economic crisis, affect firms such as private SMEs.
Acknowledgements
I would like to extend my gratitude towards my supervisor Dr. Gonenc for his useful inputs and swift replies on all of my inquiries. I am also very grateful for the support I have received
VII. References
Alves, P., Francisco, P., 2015. The impact of institutional environment on the capital structure of firms during recent financial crises. The Quarterly Review of Economics and Finance 57, 129-146.
Bancel, F., Mittoo, U., 2004. Cross-country determinants of capital structure choice: A survey of European firms. Financial Management 33, 103-132.
Beck, T., Demirguç-Kunt, A., Maksimovic, V., 2008. Financing patterns around the world: Are small firms different? Journal of Financial Economics 89, 467-487.
Booth, L., Aivazian, V., Demirgüç-Kunt, A., Maksimovic, V., 2001. Capital structures in developing countries. Journal of Finance 56, 87-130.
Brav, O., 2009. Access to capital, capital structure, and the funding of the firm. Journal of Finance 64, 263-308.
Brooks, C. 2014. Introductory Econometrics for Finance, third edition. Cambridge University Press, New York.
Brounen, D., De Jong, A., Koedijk, K., 2006. Capital structure policies in Europe: Survey evidence. Journal of Banking and Finance 30, 1409-1442.
Campello, M., Graham, J.R., Campbell R.H., 2010. The real effects of financial constraints: Evidence from a financial crisis. Journal of Financial Economics 97, 470-487.
Claessens, S., Djankov, S., Nenova, T., 2001. Corporate risk around the world. Cambridge University Press, 305-338
Custódio, C., Ferreira, M., Laureano, L., 2013. Why are US firms using more short-term debt? Journal of Financial Economics 108, 182-212.
De Jong, A., Kabir, R., Nguyen, T. T., 2008. Capital structure around the world: The roles of firm- and country-specific determinants. Journal of Banking and Finance 32, 1954-1969.
Demirgüç-Kunt, A., Maksimovic, V., 1999, Institutions, financial markets and firm debt maturity. Journal of Financial Economics 54, 295-336.
Demirgüç-Kunt, A., Martinez-Peira, M.S., Tressel, T., 2015, The Impact of the Global Financial Crisis on Firms’ Capital Structure. World Bank Group (Working paper)
Diáz-Diáz, N.L., García-Teruel, P.J., Martínez-Solano, P., 2016. Debt maturity structure in private firms: Does the family control matter? Journal of Corporate Finance 47, 393-411.
European Commission, 2016. User guide, the SME Definition. European Commission. February 2016.
Fan, J., Titman, S., Twite, G., 2012. An international comparison of capital structure and debt maturity choices. Journal of Financial and Quantitative Analysis 47, 23-56.
Faulkender, M., Petersen, M., 2006. Does the source of capital affect capital structure? Review of Financial Studies 19, 45-79.
Gianetti, M., 2003. Do better institutions mitigate agency problems? Evidence from corporate finance choices. The Journal of Financial Quantitative Analysis 38, 185-212.
González, M., 2014. The financial crisis and corporate debt maturity.
http://dx.doi.org/10.2139/ssrn.2502798. Available at SSRN: http://ssrn.com/abstract=2502798
Graham, J., Harvey, C., 2001. The theory and practice of corporate finance: Evidence from the field. Journal of Financial Economics 60, 187-243.
Grossman, S.J., 1976. On the efficiency of competitive stock markets where trades have diverse information. Journal of Finance 31, 573-585.
Grossman, S.J., Stiglitz, J.E., 1980. On the impossibility if informationally efficient markets. American Economic Review 70, 393-408.
International Finance Corporation, 2009. The SME banking knowledge guide, International Finance Corporation (IFC), The World Bank Group, Washington, DC.
Ivashina, V., Scharfstein, D., 2010. Bank lending during the financial crisis of 2008. Journal of Financial Economics 97, 319-338.
Jensen, M., 1986. Agency costs of free cash flow, corporate finance and takeovers. American Economic Review 76, 323-339.
Jensen, M., Meckling, W., 1976. Theory of the firm: Managerial behaviour, agency costs and ownership structure. Journal of Financial Economics 3, 305-360.
La Porta, R., López-de-Silanes, F., Shleifer, A., Vishny, R., 1998. Law and Finance. Journal of Political Economy 106, 1113-1155.
Laeven, L., Valencia, F., 2013.Systemic Banking Crises Database. IFM Economic Review 61, 225-270.
Mateev, M., Poutziouris, P., Ivanov, K., 2013. On the determinants of SME capital structure in Central and Eastern Europe: A dynamic panel analysis. Research in International Business and Finance 27, pp. 28-51.
Myers, S.C., 1977. Determinants of corporate borrowing. Journal of Financial Economics 5, 147-175.
Myers, S.C., 1984. The capital structure puzzle. Journal of Finance 39, 575-592.
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.
Pettit, R.R., Singer, R.F., 1985. Small business finance: A research agenda. Financial Management 14, 47-60.
Rajan, R., Winton, A., 1995. Covenants and collateral as incentives to monitor. Journal of Finance 50, 1113-1146.
Rajan, R., Zingales, L., 1995. What do we know about capital structure? Some evidence form international data. Journal of Finance 50, 1421-1460.
Santos, J., 2011. Bank corporate loan pricing following the subprime crisis. The Review of Financial Studies 24, 279-312.
Saunders, A., Steffen, S., 2011. The cost of being private: Evidence from the loan market. The Review of Financial Studies 24, 4091-4122.
United States International Trade Commission, 2010. Small and medium-sized enterprises: Overview of Participation in U.S. exports. United States International Trade Commission. January 2010.
VIII. Appendix
Table A1: Variables, definition and sources.
Variable Short definition Source
TDTA The ratio of total debt to total assets Bureau van Dijk: Orbis
LTDTA The ratio of long-term debt to total assets Bureau van Dijk: Orbis
TANG The ratio of fixed assets to total assets Bureau van Dijk: Orbis
SIZE The natural logarithm of operating revenue Bureau van Dijk: Orbis
PROFIT EBITDA divided by total assets Bureau van Dijk: Orbis
GROWT Sales divided by total assets Bureau van Dijk: Orbis
BANK ACCESS
The average of deposit money banks' assets to GDP, central banks assets to GDP, and domestic credit to the private sector.
World Bank: Global Financial Indicators
BANK STABILITY
The average of liquid assets to deposits and short-term funding, and bank capital to assets.
World Bank: Global Financial Indicators
BOND
The average of outstanding domestic private debt securities to GDP, and outstanding domestic public debt securities to GDP.
World Bank: Global Financial Indicators
STOCK ACCESS
The average of stock market total value traded to GDP, and stock market capitalization to GDP.
World Bank: Global Financial Indicators
STOCK STABILITY The average of stock market turnover ratio, and stock price volatility (See Eq. 1 for calculation of stock price volatility)
World Bank: Global Financial Indicators
GDP Annual GDP growth per capita. World Bank: World Development Indicators
LEGAL The average of the measures of corruption, rule of law, government strength, and regulatory environment.
World Bank: Worldwide Governance Indicators
INDUSTRY DUMMIES
Equal to 1 if the industry of a firm corresponds with one of the industries listed in the “NACE Rev. 2 Main Section" list, and 0 otherwise.
Constructed
COUNTRY DUMMIES
Equal to 1 if the firm is located in a country that corresponds with the dummy and 0 otherwise.
Constructed
YEAR DUMMIES
Equal to 1 if the observation is in a year that corresponds with that of the dummy, and 0 otherwise.
SUBPRIME Equal to 1 in the years 2008 and 2009, and zero otherwise. Constructed BANKING CRISIS Equal to 1 if the country of the firm experienced a systemic banking crisis in the time period, and 0 otherwise.
