1
Title: Company’s financing choices in different developed financial
structures
Student: Joris van der Zee Student number: 10254250 University of Amsterdam
Faculty: Economics and Business Track: Finance and Organization Supervisor: Ieva Sakalauskaite Date: 02-02-2016
Abstract
This paper examines whether a country’s financial structure, as being market- based or bank-based, has any influence on the financing choices companies make. It also considers whether the recent debt crisis had any influence on the relationship between firms’
financing choices and a country’s financial structure. I will use data of industrial companies in highly developed financial systems. The financial choices will be characterized by the ratios: gearing, short-term debt, long-term and retained earnings. For the gearing ratio this paper looks at both private- and publicly listed firms, the rest of the data will consist of public companies. The results show significant differences for the first three ratios when comparing public companies in different financial structures. In bank-based economies companies prefer to have an higher debt-to equity, long-term debt and short-term debt ratio relative to market-based economies. The opposite of this holds for market-based economies. There is no significant difference to be found for the retained earnings ratio. The same holds for the debt-to-equity ratio of private companies. Finally, I find no sizable effect of the financial debt crisis on the relationship between company’s financing choices and the financial structure of countries.
2 Statement of Originality
This document is written by Student Joris van der Zee who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3 I. Introduction
There is a long standing debate by financial economists over the role and function of bank-based and market-based economies, and what the advantages and disadvantages of both systems are.1 The increasing integration of world capital markets and the recent financial debt crisis tilted the discussion towards the role of domestic financial structures to an even higher level.
On one hand, the existing empirical evidence suffers from several shortcomings: (i) it has analysed the effect a bank- or market-based system has on the economy as a whole, (ii) does not distinguish between the different ownership structures of firms in those economies by only using evidence from public firms, (iii) have not looked at the influence of an financial crisis on the relationship between behaviour of companies in different financial structures and (iv) focused on only the major economies. These factors are important in assessing the effects of the financial structure on firms’ financing choices and the objective of this paper is to give more insight in these areas of focus. 2
First, there is substantial empirical analysis about the functioning of the economy as a whole for different financial structures but little research which uses evidence from firm-level data. For instance, the paper of Levine (2002) is about the relative merits of a bank-based versus a market-bank-based financial system. Or the paper of Beck and Levine (2002) which focuses on market-based or bank-based financial systems and whether they are better at financing the expansion of industries. In particular industries that depend heavily on external finance, facilitating the formation of new establishments, and improving the efficiency of capital allocation across industries. Finally the paper of Demirgüç-Kunt and Levine (1999) uses collected data on a cross-section of roughly 150 countries to illustrate how financial systems differ around the world.
A paper that uses evidence from firm-level data is the one from Schmukler and Vesperoni (2000). In this paper, they use data from emerging markets to look at how firm-specific characteristics change when firms integrate with world markets.
Furthermore existing literature does not make any distinction regarding the ownership of companies. They only examine publicly listed firms. In most countries, like
1
Partly a citation of Demirgüç-Kunt and Levine (1999). 2
Note that the term ‘financing choices’ for companies is more commonly used as ‘financial structure’. But in this paper I use ‘financial structure’ for the different financial systems of countries (i.e. bank-based or market-based).
4 Germany and Italy, publicly listed firms are the exceptions rather than the rule. And quite a few private companies are much larger than average public companies. Since there are more privately owned companies than there are public companies in most economies, it would be significant to look at the private companies as well (Pagano, Panetta and Zingales, 1998). Especially because listed and unlisted companies in general finance themselves differently, which could have an effect on the financing choices companies make (Berk & DeMarzo, 2011).
Another important factor that other papers did not take into account is the recent financial debt crisis. The influence this financial debt crisis has on the relationship between choices companies make and different financial structures will be examined for the first time in this paper.
Finally, to evaluate the effect that economic systems have on the financing choices of companies, new data is needed. Historically, empirical research on the bank-based versus the market-based economy debate has centred on Japan, Germany, the United Kingdom and the United states. Although these major economies account for more than 50 percent of the worlds output, this work produced only insights in the operation of financial structures for these countries. Using more countries would help draw broader conclusions on differences between financial structures. This is in line with some recent papers of: Demirgüç-Kunt and Levine (1999); Beck and Levine (2002); Levine (2002); andDemirgüç-Kunt and Maksimovic (2002), which also use data of more than only these four countries to overcome this shortcoming.
Existing research shows no significant evidence for the relative merits and demerits of the financial structure of a country or industry. One could examine some alternatives like, for instance, the behaviour of companies in different financial structures. Average firm-level data about companies in different structures could say something about companies motives. It would be interesting to know how the financial structure of a country influences the choices companies make.
The fact that most of the literature is based on the major economies makes it difficult to draw broad conclusions. As mentioned, it would be good to use a broader national
information set, and look at the influence a financial debt crisis has on the relationship between financial structures and firms’ choices. This is why this paper looks at the financing choices companies make in a broad perspective of different developed financial structures,
5 including panel data of countries and companies before and during the financial debt crisis. It examines only industrial companies since they are production companies, which often need investments to start production. The financial choices will be characterized by the following ratios: gearing(%), short-term debt over total liabilities and debt, long-term debt over total liabilities and debt and retained earnings over total liabilities and debt.
The remainder of this paper will be organized as follows. Section II will present the literature review. Section III presents the methodology. Section IV describes the data with the different variables used. Section V will provide the empirical results and Section VI concludes.
II. Literature review
II.I. Differences in financial systems
In a market-based economy, most of the credit gets provided by capital markets such as bond and equity markets. In a bank-based economy most of the credit gets provided by banks (Schmukler & Vesperoni, 2001). Existing literature states some differences regarding a market- or bank-based economic system, and how these affect firm financing decisions.
The positive role of banks is related to (i) the ability to acquire information about firms and managers to improve the allocation of capital and corporate governance, (ii) the risk management of banks in bank-based economies and (iii) by mobilizing capital to exploit economies of scale. Besides that, the bank-based view states that powerful banks can monitor companies more easily and build up long-term relations which can prevent
asymmetric information problems and use their power against other creditors. Therefore, in banking-based systems, companies could be expected to take more loans. The improvement of the allocation of capital, improvement of corporate governance and the risk management can make banks more willing to lend. Also as banks can build up long-term relationships to prevent asymmetric information problems, companies in bank-based economies may likely have more long-term debt than in a market-based economy.
Whereas the bank-based view highlights the role of banks, the market-based view highlights the role of a well-functioning market in: (i) the greater incentives for research-firms because it is easier for these research-firms to profit from information by trading in big liquid markets (ii) enhancing corporate governance by easing takeovers and to tie managerial compensation to firm performance and (iii) facilitating risk management. Also the
market-6 based view states that powerful banks can protect established firms with close bank-firm ties from competition reducing the amount of loans to new firms and constraining economic growth. Besides, powerful banks can collude with firm managers against other creditors, which can hinder efficient corporate governance. A higher degree of competition in capital markets suggests a more effective transmitting of information and signals to investors, with beneficial implications for firm financing and economic performance (Levine, 2002).
Regarding the financing choices of companies, the liquid markets in market-based economies can stimulate the use of term debt maturities. One reason to use short-term debt as a financial instrument is to take advantage of an upward sloping yield curve to reduce the firms’ interest expense. The yield is usually upward sloping, thus using short-term debt as a long-term source of debt capital financing could reduce the interest expense of a firm. The higher liquidity in market-based economies could indicate a shorter maturity because it would be easier to refinance the firm (Fosberge, 2012).
Besides, the enhancement of corporate governance makes it is easier to take over businesses and to ty managerial compensation to firm performance and risk management. This could make it easier to get equity financed in market-based economies, because a company (owner) can earn more trust.
II.II. Empirical Evidence
Schmukler and Vesperoni (2001) uses data from emerging markets to look at how firm-specific characteristics change when firms integrate with world markets. They find similar results for bank-based and market-based economies: Debt maturity (i.e. long- and short term debt-to-equity) tends to shorten when countries undertake financial
liberalization. Domestic firms that participate in international markets obtain better financing opportunities and extend their debt maturity. Another thing that empirical research suggested was that debt-equity ratios were consistently higher in market-based economies for public firms. This is surprising because one might expect equity values, relative to debt values, to be higher in market-based economies. A suggestion was that bank-based economies are liquidity constrained as banks, in general, do not issue enough credit to firms. This paper looked at the changes of financing choices for integrating firms of emerging economies. They used countries with an emerging underdeveloped market- and banking structure. One of the conclusions states that the financial sector of emerging markets, either bank- or market-based, needs further development and can potentially
7 benefit from integrating with international markets. The results suggested that the
difference between emerging and development markets is more important than the difference between a bank-based and market-based financial system.
The paper of Levine (2002) was about the relative merits of a bank-based versus a market-based financial system. This paper looks at the influence of the financial structure on economic growth. The result showed that although overall financial development is robustly linked with economic growth, there is no support for either the bank-based or market-based view.
Beck and Levine (2002) focused on market-based or bank-based financial systems and whether they are better at financing the expansion of industries that depend heavily on external finance, facilitating the formation of new establishments, and improving the efficiency of capital allocation across industries. They find evidence for neither the market-based nor the bank-market-based hypothesis.
At last, a paper of Demirgüç-Kunt and Levine (1999) used collected data on a cross-section of roughly 150 countries to illustrate how financial systems differ around the world. They found a clear pattern:
-Banks, other financial intermediaries, and stock markets all grow and become more active and efficient as countries become richer. As income grows, the financial sector
develops.
- In higher income countries, stock markets become more active and efficient than banks. Thus, financial systems tend to be more market based.
- Countries with a common law tradition, strong protection for shareholder rights, good accounting standards, low levels of corruption, and no explicit deposit insurance tend to be more market-based, even after controlling for income.
- Countries with a French civil law tradition, poor accounting standards, heavily restricted banking systems, and high inflation generally tend to have underdeveloped financial systems, even after controlling for income.
II.III. Financial structures and the effects of firms’ ownership
As mentioned, it would be significant to compare the ownership of companies across different financial structures. This could provide information about differences in financing choices of publicly- and privately listed companies in different financial structures, especially
8 since there are more privately owned companies than there are public companies in most economies.
From the firms’ perspective retained earnings would be a better source of funding than debt, while debt will be chosen above equity financing (Frank and Goyal, 2003).
According to the Pecking Order Theory of capital structure by Myers (1984), there are three sources of funding available for firms: retained earnings, debt and equity. Retained earnings have no adverse selection problems, debt only faces minimal adverse selection while equity is subject to serious adverse selection problems. For an outside investor, equity would be riskier than debt and would therefore need a bigger risk premium. The extent to which firms can choose the financing structure depends on their ownership structure and the financial environment in which they operate.
In general, the two main advantages of going public are greater liquidity and better access to capital (Berk & DeMarzo, 2011). Public companies can issue stock (equity) and bonds (public debt), while private firms can issue stock and only in some cases issue bonds. This is if they meet the requirements similar to public companies. To give an indication, Edwards et al. (2007) found out that 15,7% of the complete record of U.S. (OTC) secondary trades in corporate bonds between 2003 and 2005 were private.
This could have an impact on the financing choices the companies make in different financial structures, because there is a different pecking order in what companies would prefer to use as a financing instrument. One would expect that in a bank-based economy, both public and private firms would want to attract debt. We would observe more
differences in a market-based economic system as private firms have to rely more on equity financing, although they would prefer debt (Myers, 1984).
The existing empirical evidence on such effects is scarce because while data on publicly traded firms exist because they have to submit their balance sheet regularly to the market, data is hard to find for private firms. However I found data for the gearing ratio of private industrial companies, which I will use in this paper. This can help provide information about the trade-off choices companies make between debt and equity for different forms of ownership in different financial structures.
II.IV. The effects of financial structure during a financial crisis
Another important question to ask is the effect the recent financial debt crisis had on the relationship between the financial structure of a country and the firms’ financing
9 choices. In this paper we look at the most recent financial debt crisis which started in 2008. The increased default risk of companies and banks could have had an impact on the financial choices made during that period for two important reasons.
First, the turmoil in the financial markets during the financial debt crisis resulted in heightened security directed to the solvency of banks (De Nederlandse Bank [DNB], 2010). As Ivashina and Scharfsteinb (2010) showed the overall bank credit to US companies
declined during the financial debt crisis. New loans to large borrowers in the US fell by 79% from the peak of the credit boom (second quarter 2007) relative to the peak period of the financial crisis (fourth quarter 2008). Although in the short period from September to mid-October 2008 loans for commercial and industrial companies rose by about $100 billion, this was only driven by the drawdowns by corporate borrowers of existing credit lines, to ensure that they had access to funds in that insecure period, while overall the bank credit to
companies declined. Therefore, because a bank-based financial structure relies more on bank loans the fact that banks offer less loans in a debt crisis could result in less debt for companies in a bank-based economy, compared to a market-based economy.
Secondly in theory a higher share of inside debt should face a lower risk of default for the company. So an increase in the default risk should also lead to a decrease of debt that companies want to hold (Bennet et al., 2014). This is because interest and debt have to be paid through cash flows, that may be significant lower due to economic downturns. Therefore one could expect a lower leverage during a crisis period, which will result in a lower debt-to-equity ratio. As the increasing default risk applies to both public and private companies, both entities would prefer to deleverage during a crisis period.
Overall, I expect the financial debt crisis to have more impact on the relationship between the financing choices in a bank-based economy, than in a market-based economy.
III. Hypotheses and Methodology
To test my hypotheses about the relationship between financial structure and firm financing, I use the following measures of firm financing choices:
• gearing (%), a ratio which gives the level of debt-to-equity in percentages; • short-term debt ratio, such that short-term debt consists of loans and financial
10 • long-term debt ratio, where long-term debt consists of loans and financial obligations
lasting over one year;
• retained earnings ratio, were the retained earnings represent the percentage of the net earnings not paid out as dividends and retained as an investment to the core business of the company, or to pay debt (Berk & DeMarzo, 2011).
I develop seven hypotheses about the effect of financial structure on firm financial decisions based on the existing empirical literature and economic theory:
Hypothesis 1:
Public firms will have a higher gearing ratio in a market-based economy relative to a bank-based economy because of their better ability to raise capital.
Hypothesis 2:
Private firms will have a lower gearing ratio in a market-based economy relative to private firms in a bank-based economy. This because it will be harder for private companies to issue debt in a market based economy, even when they prefer debt.
Hypothesis 3:
I expect bank-based economies to have a higher long-term debt ratio compared to market-based economies, as banks prefer to build up long-term relationships with
companies. Hypothesis 4:
I expect market-based economies to have more short-term debt compared to bank-based economies because of the liquid markets.
Hypothesis 5:
I expect the retained earnings to be the same in a market-based and in a bank-based economy. Following the pecking order theory, companies in both financial structures will prefer retained earnings above debt and equity.
Hypothesis 6:
I expect the financial debt crisis to have an effect on the relationship between the financial structure of a country and the firms’ financing choices, but there to be no difference between public and private companies as both face increasing default risk.
11 Hypothesis 7:
I expect the financial debt crisis to have more impact on the relationship between the financing choices in a bank-based economy, than in a market-based economy. This because of the solvency problems banks faced during the crisis.
To test the hypotheses, I will first use panel data of 107 countries between 2001 and 2011 to classify the financial structure and whether countries have a highly developed financial structure. Once the financially developed countries are identified and assigned as market- or bank-based, I use firm-specific data of the chosen financial developed countries. The data has to contain the following ratios: gearing, short-term debt over total debt, long term debt over total debt and retained earnings. Unfortunately there was not enough data to compare both public and private companies for all of the ratios, the gearing ratio was the only one with sufficient data.
I used the following econometrical regression models to assess the impact of
financial structure on the financing choices of companies in financially developed countries:
𝐺𝑒𝑎𝑟𝑖𝑛𝑔 𝑟𝑎𝑡𝑖𝑜 𝑜𝑓 𝑖𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙 𝑓𝑖𝑟𝑚 = 𝛼 + 𝛼1 ∗ 𝑃𝑢𝑏𝑙𝑖𝑐 + 𝛼2 ∗ 𝐷𝑒𝑏𝑡𝐶𝑟𝑖𝑠𝑖𝑠 + 1𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒 − 𝐴𝑔𝑔𝑟𝑒𝑔𝑎𝑡𝑒 +2 𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒 − 𝐴𝑔𝑔𝑟𝑒𝑔𝑎𝑡𝑒 ∗ 𝑃𝑢𝑏𝑙𝑖𝑐 +3𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒 − 𝐴𝑔𝑔𝑒𝑔𝑟𝑎𝑡𝑒 ∗ 𝐷𝑒𝑏𝑡𝐶𝑟𝑖𝑠𝑖𝑠 + 𝑆ℎ𝑜𝑟𝑡 − 𝑡𝑒𝑟𝑚 𝑑𝑒𝑏𝑡 𝑡𝑜𝑡 𝑡𝑜𝑡𝑎𝑙 𝑑𝑒𝑏𝑡 𝑟𝑎𝑡𝑖𝑜 𝑜𝑓 𝑝𝑢𝑏𝑙𝑖𝑐 𝑖𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙 𝑓𝑖𝑟𝑚 = 𝛼 + 𝛼1 ∗ 𝐷𝑒𝑏𝑡𝐶𝑟𝑖𝑠𝑖𝑠 +1𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒 − 𝐴𝑔𝑔𝑟𝑒𝑔𝑎𝑡𝑒 +2 𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒 − 𝐴𝑔𝑔𝑟𝑒𝑔𝑎𝑡𝑒 ∗ 𝐷𝑒𝑏𝑡𝐶𝑟𝑖𝑠𝑖𝑠 + 𝐿𝑜𝑛𝑔 − 𝑡𝑒𝑟𝑚 𝑑𝑒𝑏𝑡 𝑡𝑜𝑡 𝑡𝑜𝑡𝑎𝑙 𝑑𝑒𝑏𝑡 𝑟𝑎𝑡𝑖𝑜 𝑜𝑓 𝑝𝑢𝑏𝑙𝑖𝑐 𝑖𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙 𝑓𝑖𝑟𝑚 = 𝛼 + 𝛼1 ∗ 𝐷𝑒𝑏𝑡𝐶𝑟𝑖𝑠𝑖𝑠 +1𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒 − 𝐴𝑔𝑔𝑟𝑒𝑔𝑎𝑡𝑒 +2 𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒 − 𝐴𝑔𝑔𝑟𝑒𝑔𝑎𝑡𝑒 ∗ 𝐷𝑒𝑏𝑡𝐶𝑟𝑖𝑠𝑖𝑠 + 𝑅𝑒𝑡𝑎𝑖𝑛𝑒𝑑 𝑒𝑎𝑟𝑛𝑖𝑛𝑔𝑠 𝑜𝑓 𝑝𝑢𝑏𝑙𝑖𝑐 𝑖𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙 𝑓𝑖𝑟𝑚 = 𝛼 + 𝛼1 ∗ 𝐷𝑒𝑏𝑡𝐶𝑟𝑖𝑠𝑖𝑠 + 1𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒 − 𝐴𝑔𝑔𝑟𝑒𝑔𝑎𝑡𝑒 +2 𝑆𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑒 − 𝐴𝑔𝑔𝑟𝑒𝑔𝑎𝑡𝑒 ∗ 𝐷𝑒𝑏𝑡𝐶𝑟𝑖𝑠𝑖𝑠 +
12 The main explanatory variable will be Structure-Aggregate, which will be a constant variable
that determines the financial structure of a country. Public will be a dummy variable that equals 1 if a company is publicly listed, and 0 if it is private. DebtCrisis will be a dummy variable for whether the industrial firm is finding itself within the financial crisis or not. When the value is 1 the companies are in the years of the financial debt crisis (2008-2011), when the value is 0 they are in the years before the financial debt crisis. Within these regressions, all of the countries used will have been first classified as financially developed.3
IV. Data
This section first describes the indicators of the financial structure in section (A.I.) and discusses them in section (A.II.). Next, Section (B) will explain how to classify countries in financial development and Section (C) will say something about the dependent variables. Section (D) will eventually discuss the other variables.
A.I. Indicators of financial structure
To examine the relation between the financial choices of industrial companies and whether the economy is bank- or market-based we first need the right measurement of financial structure. While there is no uniformly accepted indicator for this, I follow Beck and Levine (2002); Levine (2002); and Demirgüç-Kunt et al.(2009) who use two variables:
Structure-Size and Structure-Activity. Structure-Size measures the size in total shares outstanding of stock markets relative to the size of bank. Structure-Activity indicates stock market liquidity and measures the activity of the market compared to the size of bank. In both cases, higher values indicate a more market-based financial system and together these two variables will give a good indication of the country’s financial structure.
Structure-Size equals the logarithm of Stock Market Capitalization to GDP divided by Bank Credit to GDP:
Structure- Size = 𝐿𝑁 (𝑆𝑡𝑜𝑐𝑘 𝑀𝑎𝑟𝑘𝑒𝑡 𝐶𝑎𝑝𝑖𝑡𝑎𝑙𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑡𝑜 𝐺𝐷𝑃𝐵𝑎𝑛𝑘 𝐶𝑟𝑒𝑑𝑖𝑡 𝑡𝑜 𝐺𝐷𝑃 )
3
Note that the financial structure has been shown to be affected by countries’ legal systems, for example, creditors’ rights, civil vs. case law, labour protection. If you include variables for both legal system and financial structure, the structure variable (Structure-Aggregate) will show the effects on firm financing decisions excluding the legal system, which is not relevant for the research question. This paper examines the effects of financial structures on firms’ financing choices, irrespective of what causes a country to be more bank- or market-based (Demirgüç-Kunt and Maksimovic, 1995).
13 The Stock market capitalization ratio is the value of listed shares to GDP. This equals the value of domestic equities listed on domestic exchanges to GDP. The bank credit ratio to GDP is used to measure the size of bank. According to Levine (2002) it is also possible to use the total banking system assets divided by GDP as an indicator to measure the size of bank, which will yield similar results.
Structure-Activity equals the logarithm of Stock Market Value Traded to GDP divided by Bank Credit to GDP
Structure- Activity= LN (𝑆𝑡𝑜𝑐𝑘 𝑀𝑎𝑟𝑘𝑒𝑡 (𝑡𝑜𝑡𝑎𝑙)𝑣𝑎𝑙𝑢𝑒 𝑡𝑟𝑎𝑑𝑒𝑑 𝑡𝑜 𝐺𝐷𝑃𝐵𝑎𝑛𝑘 𝐶𝑟𝑒𝑑𝑖𝑡 𝑡𝑜 𝐺𝐷𝑃 )
For the Activity, I use the ratio of total shares traded on the stock market exchange to GDP. This indicator is frequently used to gauge market liquidity. The total value traded to GDP will be divided by the same factor as the Structure-Size to get the Structure-Activity indicator.
Both of the variables are strongly positively related, with a correlation coefficient of 0.95 significant at the 1%-level (table 8 in the appendix). This is why I take the first principal component of Structure-Size and Structure-Activity to explain these variables as one. This variable will be called Structure-Aggregate.
A.II. Discussion of financial structure measure
According to Levine (2002) there are some important aspects to keep in mind in the collection of data for the Structure-Size and Structure-Activity indicator. Before determining the Structure-Size and Structure-Activity ratios, one first has to determine whether countries have developed or underdeveloped financial systems, as otherwise the Structure-Activity and Structure-Size can give biased outcomes. Having an underdeveloped financial system can wrongly indicate whether countries are market- or bank-based as small financial structure indicators (i.e. bank-based) can result either because a country’s banks are comparatively well-developed or because its markets are relatively underdeveloped. Similarly, the activity and size financial structure measures can be large either because the country has well-developed markets, or because it has poorly developed banks.
For instance, Levine’s research (2002) suggests that Structure-Activity makes the attractive classification of marking countries as Turkey, Mexico and Brazil as being very market-based, even though their sum total value traded to GDP ratio is about one-sixth of that of the United States. This low total value traded to GDP indicates that the countries
14 need to have extremely low bank development to get to the high Structure-Activity ratio. The same holds for the Structure-Size variable. Levine identified Ghana, Jamaica, and Zimbabwe as having highly market-based financial systems because these countries have very small and underdeveloped banking systems, rather than because of particularly well developed stock markets.
That is why this paper only focuses on developed financial systems. Another argument to do this is that middle-lower income economies will have less reliable data than more
developed systems.
B. Assessment of financial development
To avoid the bias related to using data from financially underdeveloped countries, I selected highly developed financial systems from the 107 countries in my sample. In the classification process I used the same indicators as Demirgüç-Kunt and Levine (1999). A country will be defined as an underdeveloped financial system as the following hold: (1) Claims of Deposit Money Banks on the Private Sector / GDP is less than the sample mean and (2) Total Value Traded as a Share of GDP is less than the sample mean. The Claims of Deposit Money Banks on Private Sector / GDP is a general indicator of bank activity in the private sector and measures deposit money bank credits to (and other claims on) the private sector as a share of GDP where the credits to the public sector is excluded. Total Value Traded as a Share of GDP equals the value of the trades of domestic equities on domestic exchanges divided by GDP and measures the value of stock transactions relative to the size of the economy. When countries have a low value of the first indicator, it shows that they have a poorly developed banking system. Similarly, if countries have a below-average value for the second indicator it shows they have poorly developed markets. While this
classification system also has problems, it helps in comparing financial structures across a broad cross-section of countries.
Within my research, I will only consider countries that have an average value above indicator (1) and indicator (2), so that neither the banking nor the market system will be poorly developed. Out of 107 countries the average value for indicator (1) is 61.49% and for indicator (2) 35.69% (see Table 10 in the appendix). From the 107 countries of my panel data, I computed that 25 countries were financially developed, (see Table 11 in the
15 appendix).4 and 19 of them had useable data. The resulting final sample of the financial structure of the countries are presented in Table 1.
4 The database used from the World Bank consist of indicators of financial development and structure across countries and over time includes a range of indicators (31 indicators in total), starting from 1960 until 2011, that measure the size, activity, and efficiency of financial intermediaries and markets. The authors are: Thorsten Beck , Asli Demirguc-Kunt , Ross Eric Levine , Martin Cihak and Erik H.B. Feyen. I used the latest version which was November 2013.
16 Table 1: Ranked Structure Indices
COUNTRY STRUCTURE-SIZE COUNTRY
STRUCTURE-ACTIVITY COUNTRY
STRUCTURE-AGGREGATE
South Africa 1,03 United States 1,54 United States 2,17
Hong Kong SAR 1,01 Hong Kong SAR 0,86 Hong Kong SAR 1,76
United States 0,81 Finland 0,49 South Africa 1,29
Singapore 0,59 Korea, Rep. 0,47 Finland 0,80
Switzerland 0,37 Switzerland 0,34 Singapore 0,77
Finland 0,30 South Africa 0,30 Switzerland 0,73
Malaysia 0,21 Singapore 0,16 Korea, Rep. 0,29
Australia 0,08 Spain -0,15 Australia 0,04
Israel -0,10 Australia -0,15 France -0,25
France -0,20 Japan -0,16 Japan -0,30
Korea, Rep. -0,25 France -0,20 Malaysia -0,56
Japan -0,29 Netherlands -0,22 Israel -0,57
Thailand -0,48 China -0,30 Spain -0,58
Iceland -0,51 Germany -0,54 Netherlands -0,63
China -0,56 Italy -0,62 China -0,66
Netherlands -0,60 Israel -0,67 Thailand -0,92
Spain -0,61 Thailand -0,68 Iceland -0,97
Germany -0,87 Iceland -0,70 Germany -1,15
Italy -0,91 Malaysia -0,97 Italy -1,26
*Notes:
Panel data World Bank between 2001 - 2011
Structure-Aggregate= the first principal component of Structure-Activity [log(Total value traded / commercial banks claims on the private sector)] and Structure-Size [log(Market capitalization / commercial bank claims on the private sector)]
Structure-Size = LN(stock market capitalization to GDP / bank credit to GDP)
Structure-Activity= LN(stock market total value traded to GDP / bank credit to GDP)
17 There are relatively more bank-based economies in my dataset than there are market-based economies. Italy is relatively the most banking oriented country with a minimum value of -1.26, this in contrast to the United States which is the most market oriented country of my sample with a maximum value of 2.17. The mean value of the Structure-Aggregate is 1.58𝑒−06with a standard deviation of 0.98. The mean value is close to zero because of the standardization of the components.
C. Dependent Variables
In my final sample there is a total of 151.267 companies used for the dependent
variables. The gearing ratio is the only variable for where there are also private companies. The companies used for the sample of firms used is provided in Table 2 below:
Table 2: Total amount of companies used
COUNTRIES Gearing ratio (Public companies) Gearing ratio (Private companies) Long-term to total debt Short-term to total debt Retained earnings Total Amount Australia 676 4 748 300 1061 2.789 Switzerland 134 267 135 137 151 824 China 1.520 1359 1626 1388 1710 7.603 Germany 368 12277 389 375 465 13.874 Spain 77 12788 77 95 87 13.124 Finland 94 1317 96 90 100 1.697 France 418 20991 414 397 423 22.643 Hong Kong 104 700 105 121 181 1.211 Israel 224 5 255 286 398 1.168 Iceland 7 96 7 7 7 124 Italy 135 20581 142 150 133 21.141 Japan 2.660 23079 2217 2285 2306 32.547 Korea, Republic 780 9998 804 655 842 13.079 Malaysia 579 4867 563 485 667 7.161 Netherlands 68 879 63 56 70 1.136 Singapore 393 57 427 313 468 1.658 Thailand 388 3538 421 360 482 5.189 United States 749 103 1047 450 1482 3.831 South Africa 119 1 128 102 118 468 Total: 9.493 112907 9664 8052 11151 151.267 Note:
The companies are selected (i) because they had sufficient data between 2006-2011 (ii) they are public or large/ very large private firms (iii) they are industrial companies. The gearing ratio has privately and publicly listed firms. The long-term to debt, short-term to debt and retained earnings to debt ratios are all publicly listed firms.
18 These companies are selected because: (i) they have sufficient data for every year between 2006 and 2011 (ii) they are public or large/very large private firms (iii) they are industrial companies.5 The dataset only provides data starting in 2006 and shows the most results until 2011, so the panel is between 2006 and 2011. This will also include the financial debt crisis (2008-2011). I only focus on large and very large private companies because public companies are mostly considered as large/very large, which will be better when comparing both forms of ownership (Schmuckler and Vesperoni, 2001). Also large/very large companies mostly have a good internal reporting system of results and so are more likely to have more useable data for the independent variables.
Companies are considered to be very large if they match at least one of the following conditions:
Operating Revenue >= 100 million EUR (130 million USD) Total assets >= 200 million EUR (260 million USD)
Employees >= 1,000 Listed
Notes:
Companies with ratios Operating Revenue per Employee or Total Assets per Employee below 100 EUR (130 USD) are excluded from this category.
Companies for which Operating Revenue, Total Assets and Employees are unknown but have a level of Capital over 5 million EUR (6.5 million USD) are also included in the category.
Companies are considered to be large when they match at least one of the following conditions:
Operating Revenue >= 10 million EUR (13 million USD) Total assets >= 20 million EUR (26 million USD)
Employees >= 150 Not Very Large
5
These ratios are found at ORBIS company database. ORBIS is a comprehensive dataset with global company data available, both for listed and unlisted companies.
19 Notes:
Companies with ratios Operating Revenue per Employee or Total Assets per Employee below 100 EUR (130 USD) are excluded from this category.
Companies for whose Operating Revenue, Total Assets and Employees are unknown but have a level of Capital comprised between 500 thousand EUR (650 thousand USD) and 5 million EUR (6.5 million USD) are also included in the category.
For each year, there is an average taken per country which will be used as an indicator for the dependent variable for every country. This is separately done for the public and private firms used for the gearing(%) ratio. The weight of every country regarding their dependent variables is in this case same. The financing choices of industrial companies will be represented using 4 different variables: gearing, short-term debt, long term debt and retained earnings. These were all divided by the total liabilities and debt which includes: total current liabilities, total long-term interest bearing debt, minority interest, deferred taxes, provisions and other long-term liabilities. The total current liabilities are all short-term liabilities, namely: accounts payable, short-term debt, current portion of long term debt, and other current liabilities. The variables gearing (%), long-term debt and retained earnings were to be found in the used database. The variable short-term debt had to be defined in a different way. Following the paper of Degryse et al. (2012) and Muijs (2015) which used the same database, the short-term debts measure is the same as loans and other short-term debt. So for the dependent variables we use:
Table 3: Proxies of dependent variables
Proxy Factor Definition
Gearing ratio (%) debt-to-equity ratio
Long-term debt ratio long-term debt (USD) / total liabilities and debt (USD)
Short-term debt ratio loans and other short-term debt (USD) / total liabilities and debt (USD) Retained earnings ratio retained earnings (USD) / total liabilities and debt (USD)
I. First, as explained in the methodology, the gearing ratio represents the debt-to-equity ratio of capital in percentages. In graph 1 presented below, the gearing ratio is shown for publicly listed firms. This is in order of market oriented countries to bank oriented countries. The average gearing ratio is 81.68% with a standard deviation of
20 0,00 50,00 100,00 150,00 200,00 250,00
Graph 1: gearing (%) ratio publicly listed firms
Gearing % 2006 Gearing % 2007 Gearing % 2008 Gearing % 2009 Gearing % 2010 Gearing % 2011
33.24%. Looking at this graph one could make the suggestion that the more bank-based a country gets, the higher the gearing ratio for publicly listed firms. For this sample there are 9.493 companies used.
II. Graph 2 shows the average gearing ratio for private firms. In this sample there are 112.907 companies used. As explained the private companies are only large and very large firms. The average gearing ratio of the private companies per country is
101,70% with a standard deviation of 66,39%. Graph 2 also goes from relatively market-based to bank-based countries. At first sight, there is no clearly visible pattern to be seen in graph 2. Also the standard deviation is very high, which indicates a lot of fluctuation between the average gearing ratios of privately listed firms in different years for every country.
21 0 10 20 30 40 50 60
Graph 3: average long-term to total liabilities
and debt (%) public companies
Average long-term debt ratio 2006 Average long-term debt ratio 2007 Average long-term debt ratio 2008 Average long-term debt ratio 2009 Average long-term debt ratio 2010 Average long-term debt ratio 2011 0 50 100 150 200 250 300 350 400
Graph 2: gearing (%) ratio private firms
Gearing % 2006 Gearing % 2007 Gearing % 2008 Gearing % 2009 Gearing % 2010 Gearing % 2011
III. The long-term debt to total debt is presented in graph 3. The mean of the long-term to total debt is 21.22% and the standard deviation 8.17%. Compared to the other dependent variables, the long-term debt ratio has a low standard deviation. This means that there is not much fluctuation between the different years and the countries. As one can see, the long-term to total debt gets slightly higher when a country’s financial structure is becoming more bank-based, this could suggest a significant difference of the long-term debt to total debt for industrial companies in different financial structures.
22 0 10 20 30 40 50
Graph 4: average short-term debt to total
liabilities and debt (%) public companies
Average short-term debt ratio 2006 Average short-term debt ratio 2007 Average short-term debt ratio 2008 Average short-term debt ratio 2009 Average short-term debt ratio 2010 Average short-term debt ratio 2011
IV. Following the paper of Degryse et al. (2012) and de Jong (2002) the short-term debt variable is the same as loans and other short-term debt. Graph 4 shows the results of the average short-term debt ratio for every country from relatively market- to bank-based economies. The mean of the short-term debt ratio for different years and countries is 21.56% with an standard deviation of 8.01%. Looking at the graph, one can see that 3 out of 8 of the market-based economies and 5 out of 11 of the bank-based economies have an above average ratio. This could suggest there to be a difference in short-term to total debt ratio in different financial structures.
V. Graph 5 shows the retained earnings ratio. For most of the countries the retained earnings ratio is around 0, for some it is negative. This comes from the negative retained earnings (USD). Negative retained earnings can appear if the amount of the loss exceeds the amount of profit previously recorded in the retained earnings account as beginning retained earnings. The mean is 4.18 with a standard deviation of 10.85. The retained earnings are computed as a ratio, and is therefore not presented in percentages.
23 -70 -60 -50 -40 -30 -20 -10 0 10
Graph 5: average retained earnings to total
liabilities and debt ratio
Average retained earnings ratio (USD)
2006
Average retained earnings ratio (USD)
2007
Average retained earnings ratio (USD)
2008 Average retained earnings ratio
(USD) 2009
Average retained earnings ratio (USD)
2010
Average retained earnings ratio (USD)
2011
D. Other Variables
As discussed in the introduction and the methodology there are two important other factors which has to be considered: the financial debt crisis and whether a company is public or not. These will be tested in the model with the use of a dummy variable. In this research 2006 and 2007 are before the financial debt crisis, and the years after will be seen as the financial debt crisis. This is in line with the research of Bennet et al. (2014) which showed a peak in the credit boom the second quarter of 2007 but a sharp decline in 2008.
V. Results
In this chapter we look at the outcome of regression on the dependent variables. First, we look at the gearing ratio, afterwards at the long-term to total debt ratio, then the short-term to total debt ratio and eventually to the retained earnings.
In table 4 below, the regression results are shown for the gearing ratio. The coefficient and robust standard error of the Structure-Aggregate in regression 1 are
respectively -9.102 and 3.825. An increase of the Structure-Aggregate variable by 1% means a decrease of the gearing ratio by 9% with a standard deviation of 3.825%. For the first regression the p-value is 0.018. In this case one can say that Structure-Aggregate has a
24 significant effect on the gearing ratio of companies (private and public) with an alpha of 5%. This means one can assume that the more market-based a country is, the lower the gearing (%) ratio for industrial companies. Similarly, the more bank-based a country is, the higher the gearing ratio. In the second regression I add the variables Public and Structure-Aggregate * Public.
The coefficient of the Structure-Aggregate in the second regression becomes a lot less significant than in the first regression. Instead, the variable Public becomes (very) significant at a 1% level and the variable Structure-Aggregate * Public at a 5% level. This suggest that the indicated difference from the first regression can almost totally be explained by the public companies. Which shows that public companies have a significant lower average gearing ratio in market-based economies.
One can also add the year- and the country effects. The fixed country effects are added because this paper does not assume that all differences across countries are driven by the variable Structure-Aggregate. Countries in general should have some individual effects, and an additional effect of the Aggregate. The variables for Public and Structure-Aggregate * Public are still significant when adding the year- and the country effects in the third and fourth regression, which reduces the chance that something else is driving the results. The coefficients stayed the same, but the robust standard deviations reduced and the R^2 went up by adding these two variables. In this case, the model fits the data even better if the year and especially the country effects are included.
A reason for the differences in significance between private- and public firms can be the ability to gain debt. According to the pecking order theory companies prefer business investments from retained earnings above debt, and debt above equity. In general it is more difficult for private companies to obtain debt because they are not always able to issue bonds, this is why they are often more reliant on alternative forms of capital. It could be that one country gives better alternatives for this than other countries, which could be the
reason for the gearing (%) ratio to fluctuate more among private companies and why there is no significant pattern to be found.
The results show a negative effect between the gearing (%) ratio and the variables Public and Structure-Aggregate * Public. This is in line with the suggestion made in the data section for graph 1: the more market-based a country is, the lower the gearing (%) ratio for public industrial companies. Similarly, the more bank-based a country is, the higher the
25 gearing (%) ratio. Hypothesis 1 stated the opposite of this result. This hypothesis was mainly based on the empirical results found by Schmukler and Vesperoni (2000). One of the obvious reasons for the different result can be that Schmukler and Vesperoni only focussed on emerging economies with underdeveloped market- and banking-systems, as they were looking at the effect of financial market integration of companies. This can result in a bias which can cause wrong indicators of financial structure. The results as shown in Table 4 are more in line with general statements of theory. As one should expect that equity values for companies in a market-based economy would be higher than in a bank-based economy. A reason for this can be the ease to get equity financed in a market-based economy. This because of the higher liquidity and because a company owner can earn more trust. As mentioned in the literature review, this is because it is easier to take over businesses and to ty managerial compensation to firm performance and risk management. Another reason can be the trade-off between debt and equity.
When we look at the variables Debtcrisis and Structure-Aggregate * Debtcrisis in regression 5 to 9 we identify no significant effect. Also not when we include the country and year effects. This indicates that the financial debt crisis does not have a significant effect on the relationship between the debt-to equity ratio of industrial companies and the financial structure of a country.
According to theory companies lower their debt-to-equity ratio when facing a debt crisis, because of the increasing default risk. But the results show something different. This contradicts with hypotheses 6 and 7.
26 Table 4: Regression results gearing (%) ratio
This table reports coefficient estimates of 9 robust regressions for industrial companies with gearing ratio (%) as the main dependent dependent variable. The sample consist of 9.493 public companies and 112.907 private companies from 19 different financially developed countries. All of these companies have sufficient data between the chosen panel (2006-2011). Robust standard errors reported in parentheses.
*,**, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Gearing ratio Regression {1} {2} {3} {4} {5) {6} {7} {8} {9} Structure-Aggregate −9.102** (3.825) −0.859 (6.545) −0.859 (6.594) 3.186 (4.602) −8.202 (6.582) −8.202 (6.623) −4.157 (7.530) 0.041 (8.129) −0.945 (11.377) Public −17.889*** (6.807) −17.889*** (6.859) −17.889*** (4.750) −17.889*** (6.830) −17.889*** (6.845) Structure-Aggregate * Public −16.486** (6.940) −16.486** (6.989) −16.486*** (4.995) −16.486** (6.963) −14.514 (11.807) Debtcrisis −5.076 (7.737) −11.525 (12.821) −11.525 (9.317) −5.076 (7.531) −5.076 (7.547) Structure-Aggregate * Debtcrisis −1.350 (8.102) −1.350 (8.158) −1.350 (6.527) −1.350 (7.285) 0.129 (13.938) Structure-Aggregate * Public * Debtcrisis −2.958 (14.635)
Year effects No No Yes Yes No Yes Yes No No
Country effects No No No Yes No No Yes No No
R^2 0.0279 0.0791 0.0858 0.5956 0.0300 0.5398 0.5445 0.0813 0.0814
27 Table 5 shows the results with the long-term debt ratio as the main dependent
variable. In regression 1 only the Structure-Aggregate variable is added to the model. This gives a coefficient of -1.887 with a robust standard deviation of 0.704. The long-term debt ratio is measured in percentages which means that an 1% increase of the
Structure-Aggregate means a -1.887% decrease of the long-term debt ratio. The p-value of the Structure-Aggregate 0.008, this is (very) significant at a 1% level.Hence, the same holds for the long-term debt ratio as for the gearing (%) ratio: the more market-based a country is, the lower the long-term debt ratio for public industrial companies. Similarly, the more bank-based a country is, the higher the long-term debt ratio.
Adding the dummies for different years and countries and the Debtcrisis variable does not make too much difference for the results of the Structure-Aggregate variable. One can assume that the financial structure of a country has a significant effect on the long-term debt ratio of industrial companies.
In the third regression the variable Debtcrisis gives a significant effect when adding the country effects. This paper assumes that not all the differences between the countries are driven by the Structure-Aggregate effect, countries can have some individual effects as well. The significance of the Debtcrisis can show that the variable had some influence on the long-term-debt ratio, but not on the relationship between the financial structure and the long-term debt ratio.
The results of Table 5 are in line with hypothesis 3. The theory behind this is that banks in general would prefer to build op long-term relationships with companies, so bank-based economies (with relatively more bank credit) would likely have more long-term debt to total debt.
28 Table 5: Regression results long-term debt (%) ratio
This table reports coefficient estimates of 4 robust regressions for industrial companies with long-term debt to total liabilities and debt as the main dependent variable. The sample consist of 9.664 public companies from 19 different financially developed countries. All of these companies have sufficient data within the chosen panel (2006-2011). Robust standard errors reported in parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1%
levels, respectively.
Long-term debt ratio
Regression {1} {2} {3} {4} Structure-Aggregate −1.887*** (0.704) −2.134* (1.242) −1.429*** (0.383) −1.429*** (0.387) Debtcrisis 0.608 (1.599) 0.608** (0.297) −0.055 (0.483) Structure-Aggregate * Debtcrisis 0.372 (1.513) 0.372 (0.280) 0.372 (0.276)
Year effects No No No Yes
Country effects No No Yes Yes
R^2 0.0505 0.0520 0.9719 0.9749
Number of observations 9.664 9.664 9.664 9.664
Table 6 below shows the results of the regressions of the short-term debt ratio. These results are quite surprisingly, as they show a significant negative effect of the short-term ratio on the financial structure of countries, as well as for the long-short-term ratio (Table 5). The first regression shows the (very) significant effect of the Structure-Aggregate variable on the short-term debt ratio, without the Debtcrisis dummy and the country and year effects. It shows that even with a more specified model, by adding the Debtciris and both the year- and the country fixed effects, the outcome is still significant for public companies. This means that the effect is not explained by something different rather than the Structure-Aggregate variable.
For the short-term ratio one can assume that there is a difference between bank-based and market-bank-based economies. Namely, that public companies in bank-bank-based economies prefer to have more short-term debt relatively to companies in market-based economies. The financial debt crisis of 2008-2011 does not have a significant influence on the relationship between the short-term debt ratio and the financial structure.
29 Both the short- and long-term ratios could give a significant result because of the indicator total debt and liabilities. This indicator contains more than only the short- and the long-term debt. The ratios could therefore be significant higher for both the short- as the long-term debt in a market-based economy, as they do not have to add up to one. Table 4 shows that the debt-to-equity ratio of companies in bank-based economies are assumingly higher than in a market-based economy. A possible explanation for the higher short- and long-term ratio could be that bank-based economies in general have more debt than market-based economies.
Besides, greater liquidity does not necessary imply short-term financial instruments. Market-based economies should also be capable to offer liquid long-term instruments.
Another explanation could be stage financing of banks. This type of financing enables banks to monitor companies in different stages of their investment projects. In this way they might replace long-term loans for a series of short-term contracts in bank-based systems (Schmuckler and Vesperoni, 2001).
Table 6: Regression results short-term debt (%) ratio
This table reports coefficient estimates of 4 robust regressions for industrial companies with short-term tot total liabilities and debt as the main dependent variable. The sample consist of 8.052 public companies from 19 different financially developed countries. All of these companies have sufficient data within the chosen panel (2006-2011). Robust standard errors reported in
parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1%
levels, respectively.
Short-term debt ratio
Regression {1} {2} {3} {4} Structure-Aggregate −2.369*** (0.651) −2.451** (1.128) −2.621*** (0.540) −2.621*** (0.532) Debtcrisis 0.666 (1.562) 0.666 (0.403) 0.263 (0.877) Structure-Aggregate * Debtcrisis 0.123 (1.388) 0.123 (0.368) 0.123 (0.372)
Year effects No No No Yes
Country effects No No Yes Yes
R^2 0.0829 0.0845 0.9375 0.9425
30 The retained earnings ratio shows a significant result when only Structure-Aggregate is added as a variable to the model. After including some control variables in regression 2, 3 and 4, one can find no significant effect for public firms. This suggests that something else rather than Structure-Aggregate is driving the results. The output of Table 7 shows no significant effect between the retained earnings ratio and the financial structure of a country. This is in line with Hypothesis 5. The theory behind this is, following the pecking order theory, that companies in both financial structures will prefer retained earnings above debt and equity. The expected retained earnings will therefore be the same in a market- and bank-based economy.
In the third regression the variable Debtcrisis becomes significant when adding the country effects. This result could indicate that the debt crisis had an influence on the retained earnings ratio, when looking at the individual country effects. Only not on the relationship between the financial structure and the retained earnings ratio.
Table 7: Regression results retained earnings ratio
This table reports coefficient estimates of 4 robust regressions for industrial companies with retained earnings to total liabilities and debt as the main dependent variable. The sample consist of 11.151 public companies from 19 different financially developed countries. All of these companies have sufficient data within the chosen panel (2006-2011). Robust standard errors reported in parentheses. *, **, and *** indicate statistical significance at the 10%, 5%, and 1%
levels, respectively.
Retained earnings ratio
Regression {1} {2} {3} {4} Structure-Aggregate −3.102** (1.344) −1.586 (1.045) 0.765 (1.077) 0.765 (1.144) Debtcrisis −1.637 (1.865) −1.637** (0.760) −1.226 (0.858) Structure-Aggregate * Debtcrisis −2.274 (2.172) 0.127 (−1.54) −2.274 (1.463)
Year effects No No No Yes
Country effects No No Yes Yes
R^2 0.0775 0.0918 0.8632 0.8699
31 VI. Conclusions
This paper examines the financing choices of industrial companies in different developed financial structures. The financing choices are defined by the following ratios: the gearing ratio, long-term debt ratio, short-term debt ratio and the retained earnings ratio. I used data of public and private industrial firms in developed financial economies and focused on the average levels for each country.
First, I had to define the indicator of financial structure. Second, I had to determine whether a country is financially developed. And third, I had to examine the different ratios for the selected countries. For the gearing ratio, this paper looks at both public- and private firms. For the remaining ratios, it looks at publicly listed firms, since there was no data available for the privately listed firms. Besides the variable for financial structure and a dummy variable for the ownership, this paper looks at the influence the financial debt crisis has on the relationship between the choices companies make and the financial structure. It also includes a dummy variable for the year- and the individual country effects.
The results show that there are differences in the financing choices companies make in different financial structures. In bank-based economies, companies prefer to have higher total debt- to equity, long-term debt and short-term debt ratios. The opposite of this holds for companies in a market-based economy. The results show no significant differences of the retained earnings ratio for companies in different financial structures.
Regarding the crisis variable, the financial debt crisis had no sizable effects on the relationship between firms’ financing choices and the financial structure of countries. When adding the country effects in the regression, we see that the debt crisis variable gives a significant result for the long-term debt and retained earnings ratio. Presumably, it did have an effect on the values for the individual countries, just not on the relationship between the dependent variables and the financial structure.
The regression results of the debt-to-equity ratio for privately listed firms in different financial structures show no significant effect. It would be very interesting to further investigate the behavior of private firms in different financial structures, if one could find useful data for this. Also instead of total liabilities and debt, on could use a ratio that only consist of the long-term and short-term debt to say more about the differences of these ratios in bank-based or market-based economies.
32 References
Beck, T., Demirgüç-Kunt, A., and R. Levine, (2009). ‘’ Financial Institutions and Markets across Countries and over Time’’, Policy Research Working Paper 4943, pp. 1-14. Beck, T., Demirgüç-Kunt, A. Levine, R., Cihak, M., and E. Feyen, (2013). ‘’ Financial
Development and Structure Dataset’’, Subset of Global Financial Development Database. The World Bank.
Beck, T., and R. Levine, (2002). “Industry Growth and Capital Allocation: Does Having a Market- or Bank-Based System Matter?”, Journal of Financial Economics, 64, pp. 147-180.
Bennett, R., Guntay, L., and H. Unal, (2015). ‘’Inside Debt, Bank Default Risk, and
Performance during the Crisis’’, Journal of Financial Intermediation, 24, pp. 487-513. Berk, J. and P. DeMarzo, (2011). Corporate Finance. Harlow, England: Pearson Education
Unlimited, pp. 778-779.
Cleland, J. (1996). ‘’Demographic Data Collection in Less Developed Countries 1946-1996’’, Journal of Demography, 50(3), pp. 433-450.
Degryse, H., Goeij, de, P., and P. Kappert, (2012). ‘’The impact of firm and industry
characteristics on small firms’ capital structure’’, Small Business Economics, 38(4), pp. 431-447.
Demirgüç-Kunt, A., and R. Levine, (1999). “Bank-Based and Market-Based Financial Systems: Cross-Country Comparisons”, Manuscript. The World Bank.
Demirgüç-Kunt, A. , and V. Maksimovic, (2002).’’ Funding growth in bank-based and market- based financial systems: evidence from firm-level data’’, Journal of Financial
Economics, 65(3), pp. 337-363.
Demirgüç-Kunt, A., and V. Maksimovic, ( 1995). “Stock Market Development and Firm Financing Choices”, Policy Research Working Paper 1461. The World Bank. De Nederlandse Bank, (2010). ‘’ De solvabiliteitsratio van het Nederlandse bankwezen’’,
Statistische Bulletin, pp. 1-13.
Edwards, A., Harris, L., and M. Piwowar, (2007). ‘’Corporate Bond Market Transaction Cost and Transparency’’, Journal of Finance, 62(3), pp. 1421-1451.
Fosberg, R. (2012). ‘’ Determinants of short-term debt financing’’, Research in Business and Economics Journal’’, William Paterson University, pp. 1-11.
Ivashina, V., and D. Scharfstein, (2010). ‘’Bank lending during the financial crisis of 2008’’, Journal of Financial Economics, 97(3), pp. 319-338.
33 Levine, R. (2002). “Bank-Based or Market-Based Financial Systems: Which is Better?”,
Journal of Financial Intermediation, 11, pp. 1-30.
Muijs, W. (2015).’’The Impact of the Financial Crisis on the Determinants of Capital Structure: Evidence from Dutch Listed Firms’’ University of Twente, pp. 1-10.
Murray, F., and G. Vidhan, (2003). ‘’Testing the pecking order of capital structure’’, Journal of Financial Economics, 67(2), pp. 217-248.
Pagano, M., Panetta, F., and L. Zingales, (1998). ’’Why Do Companies Go Public? An Empirical Analysis’’, Journal of Finance, 53, pp. 27-64.
Schmukler, S., and E. Vesperoni, (2001). ‘’Firms’ Financing Choices in Bank-Based and Market-Based Economies’’, MIT Press, pp. 347-375.
Sturzenegger, F., and J. Zettelmeyer, (2006). ‘’Debt Defaults and Lessons from a Decade of Crisis’’, MIT Press, pp. 1-48.
34
Appendix
Table 8: Regression Statistics
r=0.95 P-Value=1,98966741602519E-10 Alpha=0.01 0.00<0.01 DF=N-2 SIGNIFICANT
There is a significant positive relationship between the Structure-Size and Structure-Activity, r(18), p<0.01
Table 9: principal component analysis
KMO and Bartlett's Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .500
Bartlett's Test of Sphericity Approx. Chi-Square 10.418
Df 1
Sig. .001
Table 10: Average values of Deposit Money Banks on the Private Sector / GDP & Total Value Traded as a Share of GDP Across Countries
Private Credit by Deposit Money Banks to GDP (%)
Stock Market Total Value Traded to GDP (%) Total number of countries 107 107 Total number of observations 1063.00 1063.00 Total value 65359.94 37936.54 Average value 61.49 35.69
35 Table 10: Developed Financial Systems
Country name
WB country
code WB Region WB Income Group
Average Private Credit by Deposito to GDP (%) (2001-2011)
Average Stock Market Value to GDP (%) (2001-2011)
Australia AUS High-income OECD members High-income OECD members 143,64 123,37 Canada CAN High-income OECD members High-income OECD members 120,27 77,89 Switzerland CHE High-income OECD members High-income OECD members 157,03 218,91 China CHN East Asia and Pacific Upper-middle-income economies 111,02 81,86 Germany DEU High-income OECD members High-income OECD members 110,91 63,62 Denmark DNK High-income OECD members High-income OECD members 187,89 57,19 Spain ESP High-income OECD members High-income OECD members 153,76 132,98 Finland FIN High-income OECD members High-income OECD members 74,67 122,60 France FRA High-income OECD members High-income OECD members 97,43 79,30 United Kingdom GBR High-income OECD members High-income OECD members 166,07 170,87 Hong Kong SAR, China HKG
High-income nonOECD
members High-income nonOECD members 148,20 349,15
Iceland ISL High-income OECD members High-income OECD members 154,98 73,57 Israel ISR High-income OECD members High-income OECD members 88,64 45,35
Italy ITA High-income OECD members High-income OECD members 93,94 50,35
Jordan JOR Middle East and North Africa Upper-middle-income economies 74,36 64,01 Japan JPN High-income OECD members High-income OECD members 106,64 87,61 Korea, Rep. KOR High-income OECD members High-income OECD members 90,38 145,41 Malaysia MYS East Asia and Pacific Upper-middle-income economies 107,55 44,33 Netherlands NLD High-income OECD members High-income OECD members 169,32 136,28 Norway NOR High-income OECD members High-income OECD members 73,73 44,14 Singapore SGP
High-income nonOECD
members High-income nonOECD members 96,87 112,68
Sweden SWE High-income OECD members High-income OECD members 101,25 130,86 Thailand THA East Asia and Pacific Upper-middle-income economies 96,16 48,96 United States* USA High-income OECD members High-income OECD members 55,39 254,52 South Africa ZAF Sub-Saharan Africa Upper-middle-income economies 68,81 91,66 Notes:
*USA bank development just below average level, but used because of existing literature / extreme market development Total of 25 countries compared to sample mean
