Corporate cash holdings in Civil law
countries: Evidence from Western-Europe
Abstract
This paper investigates which determinants affect corporate cash holdings for civil law countries in Western-Europe. The results show that cash holdings differ significantly between three major families of civil law: Scandinavian, French and German civil law. Furthermore, cash holdings for countries operating within the French origin vary significantly from the average of this origin. The independent variables show significant different results for the law origins as well as for the corresponding countries. The significant different amount of cash holdings seems to be explained by the diverse independent variables and the varying impact they have in the different origins and countries.
Keywords: Cash holdings; Civil Law; Trade-off Theory; Financing Hierarchy Theory.
Author: Jeroen Meesterberends Student Number: 1386301
Supervisor: dr. N. Brunia
University of Groningen
1. Introduction
During the credit crunch many companies have been facing difficulties in raising money to invest. Therefore, companies are inclined to hold on more cash to be sure profitable investments can be made. Uncertainty about cash flows thus seems to influence the amount of corporate cash holdings. But which other variables influence corporate cash policy?
With the support of two leading theories, this study investigates which variables influence corporate cash holdings. Regarding the trade-off theory there is an optimal point where the costs of holding an extra dollar are balanced to the benefits of holding this extra dollar (Myers, 1984). Although there may be economies of scale in cash management, holding large amounts of cash may also induce managers to engage in wasteful spending (Dittmar et al., 2002). In contrast to the trade-off theory, no optimal amount of cash exists in the financing hierarchy theory. Following this theory, cash balances are the outcome of capital expenditures and investment decisions as is suggested by the pecking order theory. Whereas firms with low cash flows draw down their cash and issue debt when they need cash, firms with large cash flows pay dividend, pay off their debt and accumulate cash (Dittmar et al., 2002).
Several studies have investigated which firm-specific variables affect cash holdings and what the role of corporate governance is on cash holdings policy. Opler et al. (1999) have studied which determinants affect cash holdings policy in the US. With behalf of the financing hierarchy theory, the trade-off theory and the agency cost theory they conclude that firms that are large, pay dividend and have a high leverage ratio hold less cash than other firms do. A study by Harford et al. (2006) investigates publicly traded companies in the UK, thereby also including the influence of corporate governance variables. Their results for firm-specific variables are in compliance with the results of Opler et al. (1999). Also, they find a negative correlation between board size and cash holdings and between insider ownership and cash holdings. Institutional ownership has a positive influence on cash holdings.
obtained from La Porta et al. (1998). With the use of the tradeoff theory, cash flow theory and pecking order theory they find variables as leverage, size and liquid assets being negatively correlated with cash holdings. Furthermore, better investor protection is said to positively influence the amount of cash companies hold.
Former studies have been investigating the influence of both firm-specific variables and corporate governance variables. However, these variables differ per country and seem to be influenced by the law system they are operating in. Legal origin thus seems to matter for the amount of cash a company holds.
As is argued by Watson (1974), laws in different countries are transplanted from a few legal families or traditions. According to La Porta et al. (1998) these laws come from two broad traditions. One with English origin, referred to as common law, and one derived from Roman law called civil law. This latter civil law has three major families where modern laws stem from: Scandinavian, German and French civil law. Whereas the German and the French civil law are, just as common law, widespread all over the world, the Scandinavian law is bordered to Scandinavia.
La Porta et al. (1998) mention the difference between civil law and common law on several points. Overall, civil law provides investors with weaker legal rights than common law does. On the other hand, law enforcement quality is the highest in Scandinavian and German civil law countries, followed by common law countries and with French civil law providing the weakest quality of law enforcement. La Porta et al. (1998) show that legal systems matter for corporate governance and that firms have to adapt to the limitations of the legal systems they operate in. Furthermore, they show that for different measures of shareholders rights there exist significant differences between common and civil law countries. Also, civil law countries adopt more adaptive legal mechanisms to cope with weak investor protection.
shown by Dittmar et al. (2002). However, no distinction is made between the civil law systems.
This study will distinguish itself from other studies by investigating cash holdings in the German, Scandinavian and French civil law. Given the fact that the German and Scandinavian law system provide investors better legal protection and better law enforcement than the French system does, one would expect cash holdings to differ between these legal systems. Since countries operating in countries with weaker investor protection and law enforcement hold higher amounts of cash, as is argued by Dittmar et al. (2002), it is expected that countries within the French legal system hold more cash than countries within the German and Scandinavian system. For each law, countries from the Eurozone representing these laws are taken into account and determinants of corporate cash holdings are investigated. Due to the scope of this research and data unavailability, only one country is selected for the German origin as well as for the Scandinavian origin. The French origin consists of six countries.
The results of this study show that cash holdings within the Scandinavian origin are significant higher compared to cash holdings in the French and German origin. Cash holdings for countries within the French origin are quite diverse. Only the cash ratio in Italy is not significantly different from the French origin average. Furthermore, almost all the independent variables differ significantly between the origins and countries. Moreover, the signs and coefficients of the independent variables frequently show significant differences per law origin and country. This seems to explain the differences in corporate cash holdings between law origins and countries. The results found for the independent variables in this study are in compliance with earlier studies of Opler et al. (1999), Harford et al. (2006) and Ferreira and Vilela (2004). Size, leverage and net working capital express a significant negative relation with corporate cash holdings whereas cash flow and capital expenditures are found to be significantly positive with the amount of cash a firm holds.
When these costs and benefits are examined carefully together with the impacts of the main determinants of cash holdings, the CFO can create value by maximizing the amount of cash held in the firm.
This study is organized in the following way. Section 2 discusses the main theories and previous research. Section 3 will cover data and methodology, whereas in section 4 the results are analyzed. The conclusion and recommendations are presented in section 5.
2. Theoretical background
In this section, two theoretical models are discussed to identify which firm characteristics influence corporate cash holdings. Whereas the trade-off theory states there is an optimal amount of cash to be held, no such point exists in the financing hierarchy theory. This theory is built on the fundamentals of the pecking theory, which states that companies prefer internal above external financing.
2.1 Trade-off theory
capital expenditures are high firms want to be sure these investments can be made. Consequently, larger amounts of cash will be held by these firms to be sure of funding profitable investments (Opler et al., 1999).
The costs of holding cash are twofold. First of all, the opportunity costs of capital which arise from the lower return on liquid assets (Ferreira and Vilela, 2004). Another cost is the so called cost-of-carry: the difference between the return on cash and the interest that would have to be paid to finance an additional dollar of cash (Dittmar et al., 2002). Firms paying dividend have the opportunity to raise cash by cutting dividend (Ferreira and Vilela, 2004). By doing so they increase their cash holdings internally instead of raising cash from the capital market. As a consequence, the need to hold large amounts of cash diminishes. This induces a negative relation between dividend payment and corporate cash holding.
2.2 Financing hierarchy theory
In contrast to the trade-off theory, there is no optimal cash ratio in the financing hierarchy theory. The amount of cash held by a firm is simply the outcome of capital expenditures and investment decisions made by the firm (Dittmar et al., 2002). Because firms find equity expensive, among others due to information asymmetries, they issue debt when they need cash. When the firm has a sufficient amount of cash to invest in projects available, they repay outstanding debt. When there are no costs involved with holding cash, there is no reason to object if the firm has a large amount of cash (Opler et al., 1999). Firms with higher cash flows are therefore expected to hold higher amounts of cash. This relation is also found by Opler et al. (1999) and Ferreira and Vilela (2004). Developed capital markets improve the possibility for firms to obtain external financing. Better access to capital markets lowers transaction costs of raising new funds (Ferreira and Vilela, 2004). Firms operating in countries with better developed capital markets are therefore expected to hold less cash.
size is measured as the number of directors on the supervisory board. The board of directors has the responsibility to monitor the actions of top management. In this way the goal of monitoring is to safeguarden the interests of minority shareholders (Kusnadi, 2005). While large boards have more power to monitor managers, Jensen (1993) argues that they perform less effective. This is due to the fact that decision-making becomes slower because of the involvement of more people (Kusnadi, 2005). According to Lee and Lee (2009), firms with a strong board structure hold lower amounts of cash. Furthermore, strong board structure is among other things associated with small board sizes. Therefore, firms with larger boards of directors are expected to have lower corporate cash holdings.
2.3 Previous studies
Opler et al. (1999) focus in their study on the determinants and implications of cash holdings in US firms. Their results show that companies hold more cash than is predicted by the trade-off theory. Also, they find operating losses to be the reason for large changes in excess cash. Ozkan and Ozkan (2004) incorporate managerial ownership as determinant of corporate cash holdings. Their results show a significant negative relation between managerial ownership and cash holdings. Harford et al. (2006), who study firms in the United Kingdom, have a major focus on shareholder protection and make use of four measures of corporate governance: an index for anti-takeover provisions and governance, ownership concentration, executive compensation and board characteristics. They find cash holdings to be negatively related to the index for antitakeover provisions and also with insider ownership. Insider ownership is measured as the amount of shares held by the five largest shareholders to the total amount of shares outstanding.
countries with better investor protection hold less cash. Furthermore, they find liquid assets and leverage to have a negative influence on corporate cash holdings.
Whereas many studies have investigated the influence of firm characteristics, shareholder protection and law systems, less evidence is available about the influence in countries in specific. It is especially interesting how the influence of different determinants of cash holdings differ per country and especially what the role of legal law systems is. It is possible that firms operating in countries with weak legal rights for investors adopt stronger shareholder protection mechanisms to protect shareholders as is also mentioned by La Porta et al. (1998).
3. Data & Methodology
3.1 Data
In this research I make use of a sample of public companies from the Eurozone for the period of 2004 till 2008. The data is obtained from Amadeus. Countries in the Eurozone all have the Euro as their currency. In 1999 the Eurozone was introduced consisting of 11 countries1. These countries are used for this research and comprise of: Austria, Belgium, Finland, France, Germany, Italy, Ireland, Luxemburg, the Netherlands, Portugal and Spain. Austria is left out of the sample because of data unavailability. Ireland has a common law system which is outside the fit of this research. Therefore, Ireland is not taken into consideration in this study. Since Luxemburg does not have a clear classification for legal origin according to La Porta et al. (1998), this country is not taken into consideration. The selected countries are assigned to the three civil law origins. Finland is part of the Scandinavian law system while Germany belongs to the German civil law system. Belgium, France, Italy, the Netherlands, Portugal and Spain are part of the French origin. The firms are assigned to industries following the NACE Rev. 2 sections from Amadeus. All the variables used in this study, except for dividend policy, board size and market capitalization come from Amadeus. Dividend policy and the number of directors on the supervisory board are obtained from the annual reports of the individual companies. Market capitalization is obtained from the World Bank. Although I
only have information about dividend policy, board size and market capitalization for one year, these variables are used for the whole 2004-2008 period. According to La Porta et al. (1999), ownership patterns tend to be relatively stable. Furthermore, they argue that the fact that these variables do not come from different years is not a big problem. Firms with missing observations are excluded from the sample. After applying these criteria and, as a consequence, the exclusion of 589 firms the sample consists of 3690 firm-year observations. To remove outliers, the data is winsorized at their 1th and 99th percentile as is also done by Dittmar et al. (2002). Furthermore, this is in line with the reasoning of Armstrong (2001, p222) who states that “winsorizing can be used to equal the most extreme observations to a level you feel certain about”.
Cash ratio
Table 1. Cash ratio per origin for 2004-2008 period.
*** significant at 1%, ** significant at 5% * significant at 10%
Law origin Mean Median
Standard
deviation N
Scandinavian origin 0.15 0.08 0.21 330
German origin 0.13 0.07 0.19 1145
French origin 0.13 0.08 0.18 2185
Total Euro countries 0.14 0.08 0.19 3660
Wilcoxon-Mann-Whitney test German vs. Scand. Origin 2.67*** German vs. French Origin 2.94*** French vs. Scand. Origin 2.13**
Table 2 shows cash ratios for the individual countries. Most countries have cash ratios at around the average of the total sample of 14%. Only in Spain this ratio is somewhat lower and in Portugal the cash ratio is only one fourth of the average. A Wilcoxon-Mann-Whitney test is performed to examine how the different countries within the French origin relate to the average cash ratio of the French origin. Since the data used is not normally distributed, which is a requisite when performing a T-test, another test of equality is preferred. Following Lee and Lee (2009), the Wilcoxon-Mann-Whitney test is the most commonly used nonparametric method to compare two treatments when the underlying distribution is not normally distributed.
companies in the sample. Because of this low influence Portugal stays part of the French origin.
As can be seen in table A1 in the appendix, most cash ratios for the various industries are significant higher or lower than the average cash ratio of the sample. Remarkable are the significant higher cash ratios of administration activities, information & communication and public administration. These industries have higher cash flows than on average as can be seen in table A2 from the appendix. On the other hand, industries as accommodation, electricity and mining & quarrying have cash ratios which are significantly lower than on average. As can be seen in table A2, these industries are on average larger and pay more dividends.
Table 2. Cash ratio per country for 2004-2008 period.
*** significant at 1%, ** significant at 5% * significant at 10%
Country Mean Median
Standard deviation N Finland 0.15 0.08 0.21 330 Germany 0.13 0.06 0.19 1145 Belgium 0.15 0.10 0.17 140 France 0.15 0.11 0.16 805 Italy 0.12 0.07 0.18 485 Netherlands 0.15 0.07 0.22 345 Portugal 0.03 0.02 0.03 35 Spain 0.11 0.05 0.16 375
Total Euro countries 0.14 0.08 0.19 3660
Wilcoxon-Mann-Whitney test Belgium vs. French Origin 2.14**
France vs. French Origin 8.60*** Italy vs. French Origin 1.76* Netherlands vs French Origin 0.46 Portugal vs. French Origin 3.87***
Spain vs. French Origin 4.81***
Independent variables
Leverage is defined as the ratio between total debt and total net assets (Opler et al., 1999 and Ferreira and Vilela, 2004). Net Working Capital (NWC) is taken as a proxy for liquid assets and is taken as ratio between current assets minus current liabilities and net assets. The natural logarithm of total assets in 2008 is taken as a measure for size. To investigate the influence of dividend payment policy, a dividend dummy is created whereby the dummy takes a value of one when dividend is paid and a value of zero otherwise. Board size is defined as the total members of the supervisory board in 2008. Market capitalization is defined as the market capitalization of listed companies as percentage of GNP. This variable is taken to measure the influence of development of the capital market on cash holdings.
Table 3. Identification of variables
Variables Calculation Explanation variables
Predicted Sign
CAPEX (ΛA -ΛL)/NA
Capital Expenditures = change total assets with last year
minus change total debt with last year divided by net assets Positive
CFLOW P/NA Cash Flow = after tax profit divided by net assets Positive
LEV L/A Leverage = total liabilities divided by total assets Negative
NWC (WC-C)/NA
Net Working Capital = working capital - cash and cash
equivalents divided by net assets Negative
SIZE ln(A) Firm Size = logarithm of total assets Negative
DIV D Dividend = value of 1 when dividend is paid, otherwise zero Negative
BOARD B Board Size = number of board of outside directors Negative
MARKETCAP M
Market Capitalization = market capitalization of listed
companies (% of GDP) Negative
Symbols Description Source
A Total Assets Amadeus
L Total Liabilities Amadeus
NA Total Assets-Cash & Cash equivalents /
P After tax profit
Annual Reports
WC Working Capital Amadeus
C Cash & Cash equivalents Amadeus
D Dividend paid=1 if not=0
Annual Reports
B Number of directors on the board
Annual Reports
M Market Capitalization of listed companies (% of GNP) World Bank
larger than companies within the Scandinavian and German origin. NWC and leverage are significant lower in the French law system than in the two other law systems. Companies operating within the French law system have the largest boards of directors whereas the Scandinavian law has the lowest number of directors .
The independent variables are also significantly different between the Scandinavian and German law systems. This explains the differences found between the amount of cash holdings held in the various origins.
Table 4. Descriptive statistics of exogenous variables for 2004-2008 period: Capital expenditures; Cashflow; Leverage; NWC; Size; Dividend; Board Size.
Variable Mean Median
Standard deviation N CAPEX 0.07 0.06 0.24 3660 CFLOW 0.09 0.09 0.11 3660 LEV 0.16 0.13 0.14 3660 NWC 0.09 0.10 0.27 3660 SIZE 13.17 12.82 1.96 3660 DIV 0.63 1.00 0.48 3660 BOARD 9.65 8.00 5.23 3660
Table 5. Descriptive statistics of exogenous variables per origin for 2004-2008 period. *** significant at 1% ** significant at 5% * significant at 10%
CAPEX CFLOW LEV NWC SIZE DIV BOARD
Scand. Origin 0.07 0.13 0.19 0.12 12.47 0.86 6.45 German Origin 0.06 0.09 0.17 0.12 12.92 0.59 8.73 French Origin 0.08 0.09 0.15 0.08 13.41 0.62 10.64 Wilcoxon-Mann-Whitney test German vs. Scand. Origin 1.19 7.56*** 4.02*** 0.39 3.44*** 7.72*** 4.55*** German vs. French Origin 3.13*** 2.05** 1.35 5.05*** 8.41*** 1.66* 11.25*** French vs. Scand. Origin 0.16 9.81*** 5.24*** 3.81*** 8.87*** 7.14*** 15.44***
independent variables differ significantly between countries operating within the French origin and the average value of the French origin. The effect of the independent variables on the cash ratios seem to explain the differences in cash ratios.
Table 6. Descriptive statistics of exogenous variables per country for 2004-2008 period. *** significant at 1% ** significant at 5% * significant at 10%
CAPEX CFLOW LEV NWC SIZE DIV BOARD
Belgium 0.02 0.11 0.14 -0.08 12.79 0.82 9.61 Finland 0.07 0.13 0.19 0.12 12.47 0.86 6.45 France 0.09 0.08 0.15 0.06 13.63 0.70 9.37 Germany 0.06 0.09 0.17 0.12 12.92 0.59 8.73 Italy 0.05 0.06 0.13 0.09 12.87 0.40 14.12 Netherlands 0.08 0.11 0.14 0.17 13.30 0.58 7.22 Portugal 0.00 0.05 0.29 0.12 13.97 0.71 11.86 Spain 0.11 0.10 0.19 0.06 13.93 0.68 12.29 Wilcoxon-Mann-Whitney test Belgium vs. French Origin 1.66* 1.58 0.81 6.46*** 3.75*** 4.00*** 1.96* France vs. French Origin 1.19 0.78 1.21 3.04*** 5.58*** 3.43*** 4.98*** Italy vs. French Origin 3.25** 4.90*** 3.63*** 1.56 5.58*** 7.52*** 13.29*** Netherlands vs. French Origin 0.13 6.85*** 1.01 8.57*** 0.19 1.21 12.86*** Portugal vs. French Origin 2.36** 2.77*** 5.06*** 1.24 1.76* 0.96 1.56 Spain vs. French Origin 3.40*** 0.26 2.29** 1.75* 5.69*** 1.85* 5.42***
Table 7 shows the relation between four cash quartiles and the independent variables. Quartiles are constructed along the cash/assets range with the first quartile containing the lowest cash/assets range and the fourth quartile the highest range. Whereas capital expenditures and cash flow show a positive trend within the cash/assets range, net working capital and leverage follow a negative trend towards cash holdings. Firms with a lower cash ratio are larger which seems to hold since both Portugal and Spain have the lowest cash ratios and the largest companies. Also, the fact that firms with lower cash holdings have a larger board of directors is justified by the fact that Italy, Portugal and Spain have the largest board of directors and the lowest cash ratios.
cash/assets range. The results in table 7 show that only dividend does not significantly differ between the first and fourth quartile.
Table 7. Cash range for 2004-2008 period.
*** significant at 1% ** significant at 5% * significant at 10%
First Second Third Fourth
Wilcoxon-Mann Independent variable Quartile Quartile Quartile Quartile
Whitney Test Cash/assets range 0 to 0.03 0.03 to 0.07 0.07 to 0.16 0.16 to 1.27 CASH 0.02 0.05 0.11 0.37 40.74*** (0.02) (0.05) (0.11) (0.27) CAPEX 0.13 0.15 0.17 0.31 7.36*** (0.07) (0.07) (0.08) (0.15) CFLOW 0.07 0.08 0.09 0.11 9.84*** (0.08) (0.08) (0.09) (0.11) LEV 0.20 0.18 0.15 0.10 17.52*** (0.17) (0.16) (0.12) (0.06) NWC 0.21 0.19 0.12 -0.15 27.90*** (0.19) (0.18) (0.12) (-0.11) SIZE 13.32 13.42 13.19 12.60 9.03*** (13.02) (13.04) (12.82) (12.26) DIV 0.60 0.64 0.64 0.62 0.58 (1) (1) (1) (1) BOARD 10.22 10.20 9.61 8.79 6.13*** (9) (9) (9) (7) 3.2 Methodology
The aim of this study is to investigate the determinants of cash holdings in different law origins. To measure the influence of several exogenous variables, these variables are tested against the dependent cash ratio. First, the results for the total sample as well as for the different origins are presented. Hereafter, the results are controlled for country and industry effects. A Wilcoxon-Mann-Whitney test is performed to measure the differences between the law origins and between the different countries.
show a lower correlation than these critical values, each variable can be used in the regressions.
Although panel data sets are commonly used in finance, researchers treat possible biases in standard errors in many cases incorrect (Petersen, 2008). Petersen argues that methods like OLS and Fama-MacBeth produce biased standard errors when applied to panel data sets. The results then under- or overestimate the true value. Thompson (2005) proposes a method where standard errors from firm clustering and time clustering are combined. This method can be followed by using Generalised Least Squares (GLS). With GLS the data is clustered on time and on firm level.
When performing GLS there are two options to deal with standard errors. Fixed effects and random effects can be used to correct for standard errors. With fixed effects, cross-section or period-specific means are removed from the dependent variable and the exogenous regressors (Baltagi, 2001). Fixed effects assume time independence and therefore controls for year effects. The random effects approach proposes different intercept terms for each entity and these intercepts are constant over time. The relationships between the explanatory and explained variables are assumed to be both cross-sectional and temporarily (Brooks, 2002). Since a central assumption in random effects estimation is that the random effects are uncorrelated with the explanatory variables this has to be checked (Brooks, 2002). This is possible with a Hausman test. When testing the regressions in this study, the Hausman test provides evidence that the p-value for every single test is below 1% as can be seen in table A4 from the appendix. This means that there is correlation between the random effects and the explanatory variables. According to Brooks (2002), the random effects model is not appropriate and the fixed effects model is preferred.
The regression is:
CASH
ίt =α
ίt+ β1CAPEX
ίt+ β2CFLOW
ίt+ β3LEV
ίt+ β4NWC
ίt+
β5SIZE
ίt+ β6DIV
ίt+ β7BOARD
ίt+ µ
ίt+ ν
ίtwhere CASH is the cash ratio, CAPEX is stated as capital expenditures divided by net assets, CFLOW is cash flow divided by net assets, LEV is total liabilities divided by total assets, NWC is working capital minus cash and cash equivalents divided by net assets, SIZE is the natural logarithm of total assets, DIV takes a value of 1 if dividend is paid and zero otherwise and BOARD is the total number of directors on the board.
µ
ίt is the individual specific effect of the disturbance term.ν
ίt is the other disturbance part which varies over time and entities (Brooks, 2002).4. Results
To investigate the impact of the independent variables on corporate cash holdings, several regressions are performed. First, the independent variables are regressed against the cash ratio for the total sample over the whole period. The regression is performed with Generalized Least Squares (GLS) since GLS is better able to cope with standard errors and gives more precise predictions than Ordinary Least Squares (Petersen, 2008).
Hereafter, the same regressions are performed for the selected origins and countries to examine the impact of the different independent variables on cash holdings. The values found for the different origins and the corresponding countries are tested for equality with a Wilcoxon-Mann-Whitney test.
Portugal hold the lowest amount of cash. Ferreira and Vilela (2004) also find Portugal to hold the lowest amount of cash. However, they find Italy having the highest cash holdings.
Table 8. Regression on cash holdings for 2004-2008 period.
CASHίt = αίt + β1CAPEXίt + β2CFLOWίt + β3LEVίt + β4NWCίt + β5SIZEίt + β6DIVίt + β7BOARDίt + μίt + νίt
*** significant at 1% ** significant at 5% * significant at 10% Independent variable GLS Country dummies
Country and Industry
dummies
Constant 0.41*** N.A. N.A.
Capital expenditures shows a positive relation with cash holdings which is consistent with the view that higher capital expenditures raise the need for large cash reserves. The positive sign of the cash flow ratio is consistent with both the trade-off and the financing hierarchy theory. The negative relation between cash holdings and the independent variables leverage, net working capital and size are in compliance with the theoretical predictions. Whereas the financing hierarchy theory predicts a negative relation between cash holdings and dividend payment, the results show that firms paying dividend experience higher cash levels. Board size has a significant negative impact on cash holdings which is in compliance with the theoretical prediction and the results found by Harford et al. (2006).
In order to measure how cash holdings differ between countries, the same regression is performed without including one country dummy. By excluding for example the country dummy for Belgium the values of the other country dummies show how the cash holdings of these countries differ from cash holdings in Belgium. The same regression is performed for every single country.
As can be seen in table A6 from the appendix significant differences exist between cash holdings in various countries. Although Belgium, France, Italy, the Netherlands, Portugal and Spain are all part of the French law system, their cash ratios differ significantly from each other. The cash ratio of the French origin is thus subject to differences between countries which are part of this origin.
Also the effect of board size is varying between the origins. Larger boards result in significant lower cash holdings in the French and German systems while the effect in Scandinavia is insignificant positive. When controlled for industry effects, the relation becomes negative. Furthermore, supervisory boards in Scandinavia are smaller then in the other two law origins. The results for the different origins are to a large extent consistent with the findings for the total sample presented in table 8. However, the signs of capital expenditures, dividend and board size in the results for the total sample are influenced by differences between the law origins. Values of the industry dummies are presented in table A7. from the appendix.
Table 9. Regression on cash holdings per origin for 2004-2008 period
CASHίt = αίt + β1CAPEXίt + β2CFLOWίt + β3LEVίt + β4NWCίt + β5SIZEίt + β6DIVίt + β7BOARDίt + μίt + νίt
*** significant at 1% ** significant at 5% * significant at 10%
Independent variable Scandinavian Origin French Origin German Origin GLS Industry GLS Industry GLS Industry
dummies dummies dummies
Constant 0.45*** N.A. 0.40*** N.A. 0.32*** N.A.
(9.91) (21.31) (12.23) CAPEX -0.02 0.01 0.05*** 0.03*** -0.02 -0.01 (-0.77) (0.33) (4.08) (3.08) (-1.23) (-0.98) CFLOW 0.30*** 0.28*** 0.06** 0.08*** 0.22*** 0.19*** (4.43) (4.77) (2.28) (3.26) (6.30) (5.55) LEV -0.45*** -0.23*** -0.21*** -0.17*** -0.24*** -0.20*** (-8.92) (-4.33) (-10.41) (-8.35) (-10.44) (-8.53) NWC -0.54*** -0.64*** -0.48*** -0.50*** -0.49*** -0.52*** (-21.79) (-26.05) (-49.94) (-51.12) (-38.21) (-36.84) SIZE -0.01*** -0.02*** -0.01*** -0.02*** -0.01** -0.01*** (-3.38) (-4.11) (-8.54) (-10.04) (-2.32) (-4.07) DIV -0.03 -0.03 0.00 0.00 0.04*** 0.04*** (-1.54) (-1.54) (0.76) (0.48) (4.93) (5.40) BOARD 0.00 0.00 0.00*** 0.00*** -0.01*** -0.01*** (0.60) (-0.57) (-3.69) (-2.81) (-6.75) (-6.43) 4.1 Robustness Tests
The results in table 10 show that, after the inclusion of marketcap, the independent variables still show significant results. Marketcap is positive and insignificant under the GLS regression, but significantly negative after controlling for country and industry effects. This negative relation is in compliance with the results found by Ferreira and Vilela (2004). Companies are inclined to hold less cash when they know they have better access to capital markets. This is in support of the precautionary motive for holding cash.
Table 10. Regression on cash holdings including marketcap for 2004-2008 period.
CASHίt = αίt + β1CAPEXίt + β2CFLOWίt + β3LEVίt + β4NWCίt + β5SIZEίt + β6DIVίt + β7BOARDίt + β8MARKETCAPίt + μίt + νίt
*** significant at 1% ** significant at 5% * significant at 10%
Independent GLS Country Country and Industry
Variable dummies dummies
Constant 0.41*** N.A. N.A.
(27.64) CAPEX 0.02** 0.02* 0.01 (2.53) (1.84) (1.50) CFLOW 0.14*** 0.14*** 0.15*** (6.98) (6.96) (7.66) LEV -0.23*** -0.22*** -0.19*** (-16.01) (-15.47) (-12.96) NWC -0.49*** -0.50*** -0.53*** (-66.22) (-68.03) (-69.77) SIZE -0.01*** -0.02*** -0.02*** (-10.85) (-12.75) (-13.19) DIV 0.01*** 0.02*** 0.02*** (2.94) (3.74) (3.47) BOARD 0.00*** 0.00** 0.00** (-5.52) (-2.05) (-2.16) MARKETCAP 0.00 0.00*** 0.00*** (0.13) (-3.54) (-3.39)
effects. The simultaneous determination of cash holdings, leverage and dividends seems to have no influence on the results except for board size.
Table 11. Regression on cash holdings without leverage and dividend.
CASHίt = αίt + β1CAPEXίt + β2CFLOWίt + β3NWCίt + β5SIZEίt + β6BOARDίt + μίt + νίt
*** significant at 1% ** significant at 5% * significant at 10%
Independent GLS Country Country and Industry
Variable dummies dummies
Constant 0.43*** N.A. N.A.
5. Conclusion & Discussion
In this study I have investigated which variables explain corporate cash holdings in civil law countries. With a dataset containing 3660 firm-years, cash holdings for eight countries from the Eurozone are analyzed for the period 2004-2008. The dependent cash-to-assets ratio is regressed against company specific variables. Firms operating within the Scandinavian law origin hold significant higher amounts of cash compared to companies operating in the German and French law origin. Furthermore, almost all the independent variables differ significantly between the origins.
Whereas capital expenditures has a significant positive influence on cash holdings in the French origin, this relation is negative in the German origin. Dividend has a significant negative influence on cash holdings in the Scandinavian origin, while the influence is positive in the German and French origin. The opposite relation holds for board size. Although the impact of board size is positive in the Scandinavian law system, the influence of board size on cash holdings is significantly negative within the German and French law systems. The significant different amount of cash holdings thus seem to be explained by the diverse independent variables and the varying impact they have in the different origins.
France, Spain, Portugal, Belgium and Italy show significant different amounts of cash holdings within the French origin compared to the average of this origin. The values of the independent variables for the different countries within the French origin differ significantly from each other which is also found for the three law origins. Corporate cash holdings within the French civil law system thus seem to be subject to differences between the countries which are part of this origin. Size, leverage and net working capital show a negative relation with cash holdings comparable to the findings of Harford et al. (2006), Opler et al. (1999), Ozkan and Ozkan (2004) and Ferreira and Vilela (2004). The significant positive influence of cash flow and capital expenditures on cash holdings are in compliance with the financing hierarchy theory.
comprehensive as possible, there are some limitations to keep in mind when analyzing the results.
Reference List:
Armstrong, J.S., 2001, “Principles of forecasting: a handbook for researchers and practitioners”. Springer Science and Business Media, Inc.
Baltagi, B.H., 2001, “Econometric Analysis of Panel data”. Wiley Chichester.
Baskin, J. 1987, “Corporate liquidity in games of monopoly power”. Review of
Economics and Statistics vol 69, pp 312-319
Baum, C.F. and Stephan, A., 2007, “The Effects of Industry-Level Uncertainty on Cash Holdings: The Case of Germany”. Economic Inquiry vol 47. no.2, pp 216-225.
Brooks, C., 2008, “Introductory Econometrics for Finance”.Second edition, New York: Cambridge University Press.
Denis, D.J. and Sibilkov, V. (2009), “Financial constraints, investments and the value of cash holdings”. Review of Financial Studies . Oxford University Press
Dittmar, A., Mahrt-Smith J. and Servaes, H., 2003, “International Corporate Governance and Corporate Cash Holdings”. Journal of Financial and Quantitative Analysis vol 38, pp 111-133.
D’Mello, R.., Krishnaswami, S and Larkin, P. J. (2004), “Determinants of Corporate Cash Holdings: Evidence from Spin-offs”, Working paper.
Fama, E. and French K., 2002, “Testing Tradeoff and Pecking Order Predictions about Dividends and Debt”. Review of Financial Studies 15, pp 1-33.
Frank, M.Z. and Goyal, V.K, 2000, “Testing the pecking order theory of capital structure”. Journal of Financial Economics vol.67, pp 217-248.
Graham, A. and Coyle, B, 2000, “Cash Flow forecasting and Liquidity”. Glenlake
Dearborn.
Guney, Y., Ozkan, A. and Ozkan, N., 2006, “International Evidence on the Non-Linear Impact of Leverage on Corporate Cash Holdings”. Journal of Multinational Financial
Management vol. 17. no.1, pp 45-60.
Harford, J., Mansi, S.A. and Maxwell, W.F., 2006, “Corporate Governance and Firm Cash Holdings”. University of Washington
Jensen, M.C. and Meckling, W.H., 1976, “Theory of the firm: Managerial Behavior, agency costs and ownership structure”. Journal of Financial Economics, Vol. 3, No. 4, pp 305-360
John, T.A., 1993, “Accounting measures of corporate liquidity, leverage and costs of financial distress”. Financial Management vol.22. no.3, pp 91-100.
Kusnadi, Y., 2005, “Corporate governance mechanisms and corporate cash holdings”. Working Paper, SSRN.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A. and Vishny, R., 1998, “Law and Finance”. Journal of Political Economy 106, pp 1113-1155.
Lee, K.W. and Lee, C.F., 2009, “Cash holdings, corporate governance structure and firm valuation”. Review of Pacific Basin Financial Markets and Policies vol.12, no.3, pp 475-508.
Levine, R. and Zervos, S., 1998, “Stock Markets, Banks and Economic Growth”.
American Economic Review.
Malmendier, U. and Tate G., 2005, “Superstar CEOs,”. Stanford University,
Myers, S., 1984, “The Capital Structure Puzzle”. Journal of Finance 39, pp 572-592.
Niskanen, M. and Niskanen, J, 2007, “Cash Holdings in Finnish SMEs”. University of Kuopio, Finland.
Opler, T., L. Pinkowitz, R. Stulz, and R. Williamson, 1999, “The Determinants and Implications of Corporate Cash Holdings”. Journal of Financial Economics 52, pp 33- 46.
Ozkan, A. and Ozkan, N., 2004, “Corporate Cash Holdings: An Empirical Investigation of UK Companies”. Journal of Banking and Finance vol.28, pp2103-2134.
Petersen, M., 2008, “Estimating standard errors in finance panel data sets: Comparing approaches”. Northwestern University.
Petit, J, 2007, “Strategic Corporate Finance, Applications in Valuation and Capital Structure”. Wiley Finance.
Thompson, S. (2005), “A Simple Formula for Standard Errors that cluster by both firm and time”. Harvard Working Paper.
Tirole, J. 2006, “The Theory of Corporate Finance”. Princeton University Press.
Appendix
Table A1. Cash ratio per industry for 2004-2008 period. *** significant at 1%, ** significant at 5% * significant at 10%
Industry Mean Median
Standard deviation N Accommodation 0.05 0.02 0.11 70 Administration 0.20 0.12 0.23 230 Agriculture 0.12 0.08 0.11 72 Construction 0.17 0.12 0.21 200 Electricity 0.05 0.03 0.07 85 Financial services 0.16 0.13 0.13 110
Information and Communication 0.21 0.09 0.29 205
Manufacturing 0.13 0.06 0.18 1240
Mining and quarrying 0.07 0.03 0.12 45
Public administration 0.24 0.13 0.27 205 Real estate 0.12 0.08 0.13 770 Retail 0.13 0.06 0.20 150 Transportation 0.11 0.06 0.15 145 Wholesale 0.12 0.07 0.15 175 Average 0.14 0.08 0.19 3660
Test of means (t-statistics) Accommodation vs. Average 7.26***
Administration vs. Average 6.84***
Agriculture vs. Average 0.72
Construction vs, Average 4.72***
Electricity vs. Average 7.00***
Fin. Services vs. Average 4.79***
Information vs. Average 3.33***
Manufacturing vs. Average 2.88***
Mining vs. Average 4.59***
Public adm. vs. Average 6.25***
Real Estate vs. Average 2.11**
Retail vs. Average 0.51
Transportation vs. Average 3.56***
Table A2. Descriptive statistics of exogenous variables per industry for 2004-2008 period. *** significant at 1% ** significant at 5% * significant at 10%
Industry CAPEX CFLOW LEV NWC SIZE DIV BOARD
Accommodation 0.07 0.06 0.33 0.03 14.22 0.85 10.97 Administration activities 0.08 0.11 0.14 -0.03 12.01 0.43 7.81 Agriculture 0.03 0.08 0.10 0.06 11.95 0.47 6.61 Construction 0.11 0.10 0.16 0.07 14.11 0.77 11.36 Electricity 0.06 0.09 0.25 0.04 15.01 0.88 14.50 Financial services 0.14 0.06 0.11 0.11 11.95 0.35 7.77 Information & Communication 0.08 0.10 0.15 -0.08 13.54 0.51 11.02 Manufacturing 0.06 0.09 0.13 0.18 13.04 0.64 9.70 Mining & Quarrying 0.09 0.10 0.15 0.16 13.83 0.66 7.68 Public administration 0.12 0.11 0.16 -0.08 12.41 0.58 9.44 Real estate 0.06 0.08 0.20 0.09 13.74 0.64 9.83
Retail 0.05 0.10 0.16 0.11 12.74 0.53 8.97
Transportation & Storage 0.05 0.13 0.20 -0.07 13.35 0.72 10.41
Wholesale 0.07 0.10 0.12 0.17 12.73 0.71 8.57
Table A3. Correlationmatrix
CAPEX CFLOW LEV NWC SIZE DIV BOARD MARKETCAP CAPEX 1.00 0.26 -0.17 -0.09 0.03 0.05 -0.02 0.09 CASHFLOW 0.26 1.00 -0.14 -0.09 0.04 0.23 -0.07 0.10 LEVERAGE -0.17 -0.14 1.00 0.03 0.27 0.01 0.10 0.02 NWC -0.09 -0.09 0.03 1.00 -0.07 -0.03 -0.10 -0.02 SIZE 0.03 0.04 0.27 -0.07 1.00 0.36 0.56 0.12 DIVIDEND 0.05 0.23 0.01 -0.03 0.36 1.00 0.21 0.15 BOARDSIZE -0.02 -0.07 0.10 -0.10 0.56 0.21 1.00 -0.12 MARKETCAP 0.09 0.10 0.02 -0.02 0.12 0.15 -0.12 1.00
Table A4. Hausman test
Correlated Random Effects - Hausman Test Equation: Untitled
Test cross-section random effects
Test Summary Chi-Sq. Statistic Chi-Sq. d.f. Prob. Cross-section
Table A5. Redundant fixed effects test Redundant fixed effects
Redundant Fixed Effects Tests Equation: Untitled
Test cross-section and period fixed effects
Effects Test Statistic d.f. Prob.
Cross-section F 11.75169 -7,412,939 0.0000 Cross-section Chi-square 5081.053 741 0.0000 Period F 15.88022 -42,939 0.0000 Period Chi-square 78.90268 4 0.0000 Cross-Section/Period F 11.79519 -7,452,939 0.0000 Cross-Section/Period Chi-square 5106.129 745 0.0000
Table A6. Regression on cash ratio differences per country for 2004-2008
CASHίt = αίt + β1CAPEXίt + β2CFLOWίt + β3LEVίt + β4NWCίt + β5SIZEίt + β6DIVίt + β7BOARDίt + μίt + νίt
*** significant at 1% ** significant at 5% * significant at 10%
Table A7. Regression on cash holdings per origin for 2004-2008 period.
CASHίt = αίt + β1CAPEXίt + β2CFLOWίt + β3LEVίt + β4NWCίt + β5SIZEίt + β6DIVίt + β7BOARDίt + μίt + νίt
*** significant at 1%, ** significant at 5% * significant at 10%
Independent variable Scandinavian Origin French Origin German Origin GLS Industry GLS Industry GLS Industry
Dummies dummies dummies
Dummy Accommodation 0.34*** -0.03* 0.54*** (6.54) (-1.89) (9.25) Dummy Administration 0.46*** -0.02* 0.36*** (11.02) (-1.86) (12.49) Dummy Agriculture 0.43*** -0.03 0.30*** (8.74) (-1.17) (8.29) Dummy Construction 0.61*** 0.07*** 0.42*** (12.68) (6.15) (11.17) Dummy Electricity 0.42*** -0.01 0.34*** (6.26) (-0.81) (9.66)
Dummy Fin. Services 0.42*** 0.05*** 0.34***
(8.79) (3.68) (9.54) Dummy Information 0.36*** 0.04*** 0.35*** (6.93) (3.13) (11.43) Dummy Manufacturing 0.51*** 0.06*** 0.41*** (11.94) (6.94) (14.57) Dummy Mining 0.44*** 0.01 0.32*** (7.65) (0.43) (8.78) Dummy Publishing 0.56*** 0.02* 0.37*** (12.05) (1.65) (13.77)
Dummy Real Estate 0.31*** 0.04*** 0.36***
