1
Cash Holdings, Investor Protection and Multinationals: Evidence from the
European market
MSc. Thesis International Financial Management
By Thomas Swikker
Student number: s2015455
Student email: t.swikker@student.rug.nl
Supervisor: Mr. M. Reijenga
Field Key Words: Cash Holdings, Investor protection, Multinationals, Financial
crisis
Abstract:
2
Table of content
1. Introduction p.3
2. Theoretical background and hypotheses p.6
2.1 Why do firms hold cash? p.6
2.2 Investor protection and Corporate Cash Holdings p.8
2.3 Multinationals and Cash holdings p.9
2.4 The European crisis and Cash holdings p.10
3.Data and Methodology p.12
3.1 Sample p.12
3.2 Variables and Equation p.13
3.3 Descriptive Statistics p.16 4. Results p.18 4.1 Part A p.18 4.2 Part B p.21 4.3 Robustness check p.23 5. Conclusion p.24
6. Limitations and recommendations p.25
7. References p.26
3
1. Introduction
The amount of cash held by the non-financial firms all over the world accounted for $2.6 trillion of deposits at the end of 2014 (Forbes, 2014). This was a new record level in number of dollars. CNN Money (2015) states that the S&P 500 companies, financial companies excluded, accounted for $1.43 trillion in cash reserves. Although, cash is thought of as being unproductive, due to the low rate of return and the costs of holding cash, there is an upward trend in the amount of cash held by companies.
This upward trend is against the wishes of the shareholders. Shareholders strive for the highest returns on their investments and push for a higher pay out of dividends. An example of this is illustrated by Carl Icahn and Apple. As a minority shareholder, Icahn was not pleased with Apple’s cash position, saying that it was an ‘excessive liquidity’ position and pushed for a buyback of its own shares. After threatening Apple with filing for a shareholder proxy vote, Apple bought back $14 billion worth of its own shares. This illustration shows that the amount of cash held by a company could decrease by the influences the shareholders have on the company’s policy. As holding cash is against the interests of the shareholder and is seen as an unproductive asset, why do firms hold so much cash?
This paper tries to answer this question and investigates what impacts the amount of cash held by a company. There are several points in order to justify the amount of corporate cash
holdings, for instance using cash as a buffer giving the firm shelter in economic downturns. Some other plausible reasons for holding cash would be that the firm has higher investment and R&D expenditures or better investment opportunities (Dittmar, Mahrt-Smith and Servaes, 2003).
Another influence on corporate cash holdings is a country’s investor protection rate. Countries with a high rate of investor protection, will protect the shareholders by giving them more influences within the company. An investor could use this influence to lower the cash amount, to receive a higher pay out or a higher return on investment, as is illustrated by the example described above.
Additionally, this paper will investigate whether multinationals hold more or less cash than domestic firms do, this gives the paper an international aspect and sheds a light on what the effects of being a multinational company has on cash holdings. Theory states that
4 multinational companies increase their amounts of cash holdings, due to higher tax costs (Titman, Wei, and Xie, 2004). Furthermore, the media reports more frequently news on the large multinational companies increasing cash reserves. An example for this is illustration about Icahn and Apple described above, another example is the big cash reserves of CISCO, Google and Microsoft with an amount of cash held, of respectively, $60 billion, $70 billion and $96 billion (CNNMoney, 2015). However, some authors have found contradictory evidence regarding this topic. Example given, Fernandes and Gonenc (2015) found evidence that multinational companies hold less cash than domestic do, as they have economies of scale in cash management.
Furthermore, according to the precautionary motive (Bates, Kahle and Stulz, 2009), cash could be used as a buffer for adverse shocks. Therefore this paper will incorporate the
European economic crisis, starting at the end of 2007 until 2014, to see what the effect of this crisis is on the amount of cash held and whether this is in line with the precautionary motive. Taking all these assumptions into considerations the following research question is produced: What impacts the amount of cash held by a company?
To answer this question the following sub-questions are formed:
What impact does Investor protection have on the amount of cash held by a company? What impact does a multinational character of the firm have on the amount of cash held by a company?
5 The piling up of cash by the multinational companies is undergoing heavy media attention, this paper will add to this discussion. Additionally, by explaining the factors that influence
6
2. Theoretical Background and Hypotheses
2.1 Why do firms hold cash?
Cash is thought of as being unproductive, due to that the rate of return is low. This means that decreasing the cash holdings is more profitable due to the high marginal cost of holding cash assets in relation to more productive assets (Huang et al., 2013). Thus, one could argue that holding cash is more costly to hold, in respect to other assets and therefore firms should not hold cash. However, this would be in the perfect world where no motives exist for holding cash.
Previous research that has covered the topic of corporate cash holdings describe four motives for a company that determine the level of cash (e.g., Bates et al, 2009). The first motive is the transaction motive. This motive implies that a firm holds cash in order to deal with
transactions. Keynes (1936) argues that it is more beneficial for a firm to hold cash, instead of converting illiquid assets to cash in order to avoid the costs associated with transactions. Iskandar-Datta and Jia (2014) agree with this assumption and state that one of the primary believes to hold cash is, that it decreases transaction costs associated with financing through debt or issuing equity. They argue that the cash reserves held by the company should be adjusting the marginal costs of risking shortness of cash against the marginal costs of
investing in liquid assets. However, it is assumed that bigger firms hold less cash than smaller firms, due to that firms of greater size could establish economies of scale in order to deal with transaction costs. The bigger firms can acquire funds less costly, as they have superior access over small firms to financial markets and have economies of scale in cash management and raising funds (Pinkowitz, Stulz, & Williamson (2006) and Dittmar et al, 2003)). Several authors found evidence for these economies of scale in cash management (e.g. Mulligan, 1997, and Miller and Orr, 1966). Furthermore, Fernandes and Gonenc (2015) find evidence that multinational companies hold less cash than domestic firm do, and make the assumption that this is due to economies of scale in cash management.
7 The third motive looks at the international aspect of the firm and is known as the tax motive. This motive assumes that multinational companies are hording more cash than domestic companies do, as a result of taxes that multinational firms face when repatriating their foreign earnings (Foley, Hartzell, Titman and Twite, 2007). Titman et al. (2007) agree with this statement and found evidence for this motive while investigating the U.S. market.
The fourth and last motive is known as the agency motive. Jensen and Meckling (1976) define the agency relationship as a commitment between the principal and the agent to carry out a duty on the principal’s (minority shareholders) behalf and thereby shifting a percentage of the decision making capabilities towards the agent (the controlling shareholders and managers). If both parties tend to increase the utilities to the highest level, one could conclude that the agent is not always performing in the interest of the shareholder, thus creating a conflict of interest. Huang et al. (2013) state that managers tend to pile up cash for personal gains or to enlarge their power within the company by increasing control of resources. These managers will invest in activities, which might not be the best for the company, but will enable the managers to build an empire. Stulz (1990) argues that shareholders are aiming for a higher payback on their investments in the style of dividends, rather than holding higher amount of cash. A decrease in cash holding would mean that it is rather difficult for managers to gain private benefits from the cash excess.
The consequences of the agency theory is that it leads to additional costs in order to keep the keep the agent’s interest aligned with the shareholder’s interests. Jensen and Meckling (1976) state that theses inflicted costs consist of three costs, namely, the additional costs received by the shareholder for monitoring the managers, the bonding costs of the agent (alignment of interest) and the residual costs. The residual costs are the costs associated with the divergence in interest of the shareholders and the managers of the firm.
8 The agency motive has an overlap with the transaction motive, as a higher agency problem lead to higher additional costs (higher transaction costs), therefore it is assumed that firms faced with higher agency costs hold more cash.
These four motives have different consequences in light of their relation towards the amount of cash held by a company. The next three sections will explain what the relation between these motives and the main research purposes of this paper are.
2.2 Investor protection and Corporate Cash Holdings
As described above the agency motive suggest that companies who experience more agency problems, will have a greater amount of cash holding compared to firms with less agency problems. However, there are some instruments used that help mitigate the agency problem effects.
One of these tools is a higher investor protection rate, forced upon by the authorities.
Companies that are required by law to comply with the higher investor protection regulations, will have an improved quality of corporate governance over companies that are active in a country accompanied with a lower investor protection rate. (Pagano, Roell and Zechner (2002)). Pagano et al. (2002) found evidence that companies that list or cross-list on
countries’ indexes where tight regulations exist, thus an existing of higher investor protection, companies will increase their quality of corporate governance in order to comply with these rules. Therefore, it can be said that a higher investor protection will be followed by a high corporate governance.
According to Chen et al. (2012) a higher quality of corporate governance will lead to a lower amount of cash held by the company. They state that, if the amount of cash held by the company is reflected by the needs of the insiders (managers, agents) and not by the desire of the shareholders to maximise the value of the company, the amount is greater than the amount needed for investments. The authors found evidence suggesting that a greater quality of corporate governance will diminish this amount of cash held. Dittmar et al. (2003) found similar results suggesting that in countries where shareholders have little protection, the amount of cash held is significantly higher. Porta et al. (2000) argues that corporate
9 These results are aligned with the agency motive for holding cash, whereas a higher investor protection leads to less agency costs, therefore, the company will hold less cash as the likelihood of piling up cash for a manager’s personal gain decreases.
However, several authors found contradictory evidence suggesting that an increase in shareholders protection will lead to an increase in cash holdings. Harford et al. (2008) state that firms with lower corporate governance have greater expenditures and are more frequent purchasers and therefore have less amount of cash held. Firms are able to hold more cash, as the agents are restricted to build an empire with the excess cash, and therefore cash will be used in an efficient manner. In addition, Iskandar-Datta et al. (2014) found similar evidence showing that a lower corporate governance will lead to a decrease in the amount of cash held by the company.
Although there are found mixed results in previous literature, the hypothesis is derived based on the agency motive and is stated as follows:
Hypothesis 1: Higher investor protection score will decrease the amount of cash held by the company
2.3 Multinationals and Cash holdings
The transaction motive and the tax motive give both different outcome on the relationship between a multinational firm and the amount of cash held by that firm. Based on the transaction motive, one could argue that multinational firms hold less cash than domestic firms do, assuming that multinationals are of greater size (economies of scale in cash
management). This is in line with the findings of Fernandes and Gonenc (2015), showing that there is a negative relationship between the level of multinationality and cash holdings. On contrast, the tax motive states that multinationals hold more cash as the tax consequences of repatriation is too high. This view is recognized and found evidence for by Foley et al., (2007) and Bates et al., (2009), after investigating the U.S. market and the cash holdings of the present multinationals. This view on multinationality and cash holdings is also consistent with the view of Titman, Wei and Xie (2004).
10 CISCO, Google and Microsoft alone accounted for respectively, $60 billion, $70 billion and $96 billion in cash held as a reserve.
Therefore, based on the mixed results found in previous literature and the contradictory claims of the transaction motive and the tax motive have on the relationship of multinationals and cash holdings the following hypotheses are derived:
Hypothesis 2a: A multinational company holds more cash than a domestic company. Hypothesis 2b: A multinational company holds less cash than a domestic company
2.4 The European crisis and Cash holdings
To show the relationship between the European crisis and the amount of cash held by a firm, a brief overview of the European crisis will be given. In this paper the European crisis is
defined as the period of 2008 to 2014 and, according to Berk (2016), occurred in four stages. The first stage was the liquidity crisis (summer 2007-autumn 2008), resulting in a collapse of several hundred non-banks mortgage lenders (Haan de, Oosterloo, Schoenmaker p.59, 2015). This lead to reported losses by banks as they were connected due to investments in these mortgages. The second stage was the solvency crisis (autumn 2008-spring 2010), already beginning in the autumn 2008, this crisis went into acceleration after the collapse of the Lehman Brothers in September 2008. In Europe, governments needed to rescue financial institutions to avoid systemic breakdowns (e.g. the bailout of Fortis) (Haan de, et al. 2015). A systemic breakdown occurs when both the non-financial and financial firms face a large default rate and cannot repay contracts on time, generating an exhaustion of the banking systems’ capital. The third stage was the European debt crisis (Spring 2010-2013). Several countries in Europe could not (re)finance their government debt. Leading to disbursements towards Greece, in order to prevent contagion from Greece to other European countries. The fourth stage started at July 2013 and is still in process. This stage is characterized by slow recovery, low growth and inflation.
11 as external financing becomes too costly. The transaction costs motive could also explain a decrease in cash held by firms, as it becomes more expensive to finance an investment
externally. According to Puri, Rocholl and Steffen (2011), banks that are subjected to a crisis, reject more loans than they used to before a crisis. Ivashina and Scharfstein (2010) found similar results and state that loans to large borrowers decreased by 47% in the worst period of the crisis. These evidence show that it becomes more difficult for a firm to finance themselves through external channels, therefore internal funds are used and a drop in cash holdings is expected. However, as the transaction motive state, larger firms have more excess to external funds, therefore, a smaller change in cash holdings is expected for multinational firms. Thus, based on the discussion in this section on the relationship between the European crisis and corporate cash holdings, the following two hypotheses are derived
Hypothesis 3: The European companies experienced a decline in corporate cash holdings during the European crisis.
12
3. Data and Methodology
3.1 Sample
The data used in this paper is obtained from the following sources: Datastream, Worldscope, the Worldbank database, the article of Kaufmann, Kraay, and Mastruzzi (KKM, 2010) and the index of economic freedom of the Heritage Foundation & Wall Street Journal. The sample is constructed from the period 2003 to 2014 from firms of the following European countries that provided sufficient data: Finland, Iceland, Denmark, Sweden, Netherlands, Norway,
Switzerland, Luxembourg, United Kingdom, Austria, Germany, Belgium, Ireland, France, Spain, Portugal, Slovenia, Italy, Hungary, Slovakia, Greece, Poland, Serbia, and Romania. The selection of these countries allows for an analysis of the effect of the European financial crisis on cash holdings by dividing the sample in the periods 2003 to 2007 and the European crisis period in the years 2008 to 2014. Furthermore, following Huang et al. (2013), financial firms and utilities are excluded from the sample, as these firms might be subjected to liquidity constrains following government’s regulations. In addition, following Ramirez and Tadesse (2009), yearly observations with missing values or negative values on assets and sales are deleted, as well as observations with a foreign sales ratio being negative or above hundred per cent. To deal with the possibility of outliers, the same approach as Huang et al. (2013) is taken, by winsorizing the data at its first and 99th quantile. By winsorizing the data, the observations found outside the threshold will be replaced by the threshold quantile value, to mitigate the effect of outliers.
13
3.2 Variables and Equation
3.2.1 Dependent variable
The dependent variable, which is used in the equation, is the cash ratio. This ratio will be defined as the amount of cash, cash equivalent and short-term investments. This method is commonly used in prior papers covering the cash holding topic (e.g. Bates et al., 2009, and Ozkan & Ozkan (2004))
3.2.2 Independent variables
Investor protection. According to Aggarwal, Erel, Stulz, and Williamson (2010) countries’ traits are extensive powers that fabricate a countries’ corporate governance. Therefore, the investor protection variable will be defined by the index introduced by Kaufmann, Kraay and Mastruzzi (KKM hereafter, 2010). This index provides annual data covering six governance proxies, specifically, voice and accountability, political stability, government effectiveness, regulatory quality, rule of law and control of corruption from 2003 to 2014. Following Huang et al. (2013), the composition of this variable is the sum of the six governance indicators. The higher the sum of these indicators is, the more superior the protection of investor is.
Degree of mulitnationality. This variable will be defined as a dummy variable computed by the ratio of foreign sales over total sales (Pinkowitz, Stulz and Williamson, 2012, Fernandes and Gonenc, 2015, and Doukas and Pantzalis, 2003). Following Pinkowitz et al. (2012) and Fernandes and Gonenc (2015), a multinational company will be defined as a company with a foreign sales ratio greater than 25 per cent.
3.2.3 Control Variables Firm control variables
14 However, there are mixed results regarding what relationship size has with cash holdings. Several authors found evidence that bigger firms hold less cash than small firms do (Miller and Orr, 1966, Harford et al. 2008). Opler, Pinkowitz, Stulz and Williamson (1999) found contradictory evidence that bigger firms hold more cash. Therefore, no clear relationship between cash holdings and size is expected in the equation.
Research and Development expenses. This control variable is defined as the R&D expenses divide by the firms’ total sales. According to Dittmar et al. (2003) firms with higher R&D expenditures will possess more cash. This is in line with the precautionary motive, as holding a larger amount of cash will guarantee the firm with ability to pay for future
interesting investments. Therefore, it is expected that there will be a positive sign in this relationship.
Dividends paid. Following Opler et al. (1999) and Ozkan and Ozkan (2004) this variable is used in the equation. This is a dummy variable and is equal to one if the firm pays out common dividend in the observed year and otherwise equal to zero. It is expected that firms with dividends pay outs hold less cash. The firm could decrease the dividends pay out, in order to gain liquid assets, when faced by a shortage of cash.
Market to book ratio. Similar to the R&D expense, firms that face better investment opportunities will hold a greater amount of cash. Based on the precautionary motive, firms will hold more cash as they do not want to miss out on investment opportunities. Better investment opportunities are measured by the firm’s market to book ratio. This is the
calculated ratio between the firm’s market value and book value (Opler et al, 1999, and Ozkan & Ozkan 2004)
15 experiencing large amounts of debt need cash to service their debt. Therefore, it is expected that the leverage of a firm is negatively correlated with the amount of cash held.
Net working capital ratio (NWC). This ratio is used and calculated as, following Bates et al. (2009) and Ozkan & Ozkan (2004), the working capital minus cash, cash equivalents and short-term investment over the book value of the total assets. According to the
precautionary motive and the transaction motive, firms that hold more liquid assets will hold less cash, as liquid assets are easily transferable to cash. Therefore, there is a negative correlation expected between the net working capital ratio and cash holdings.
Country control variables
Inflation. According to Ramirez and Tadesse (2009), inflation has a negative relationship with cash holdings. Inflation is the cost to hold money, if inflation goes up, the level of cash held by the company is suspected to decline. The inflation data is collected from the World Bank database.
GDP. This is the proxy for a country’s economic development and is calculated as the Gross Domestic Product per capita. Ramirez and Tadesse (2009) argue that there is a positive relationship between GDP per capita and the cash holdings of a firm. The GDP data is
collected from the World Bank.
Corruption. Pinkowitz, Stulz and Williamson (2003) and Ramirez and Tadesse (2009) propose that there is a positive correlation between the level of corruption within a country and the cash holdings of a firm. They argue that firms active in countries with a high level of corruption, firms need to pay off people in order to smoothing the process and therefore, a higher level of cash is needed within the firm. This index is formed by Transparency International, who have ask businessmen and analysts their perception of corruption in the country. A higher score means a low perception of corruption
16 into 10 factors of economic freedom. The higher the sum of these factors, the greater the government interference and, therefore, the lower the quality of the banking system is.
3.2.4 The equation
The equation is a multiple regression model with cross-section and time effects used for an unbalanced pooled data and fabricated following the variables described in the previous sector. Therefore, the notation of the equation is specified as follows:
𝐶𝑎𝑠ℎ ℎ𝑜𝑙𝑑𝑖𝑛𝑔𝑠𝑖,𝑡= 𝛼 + 𝛽1∗ 𝐼𝑛𝑣𝑒𝑠𝑡𝑜𝑟 𝑃𝑟𝑜𝑡𝑒𝑐𝑡𝑖𝑜𝑛𝑖,𝑡+ 𝛽2∗ 𝑀𝑢𝑙𝑡𝑖𝑛𝑎𝑡𝑖𝑜𝑛𝑎𝑙𝑖𝑡𝑦𝑖,𝑡+ ∑ 𝐹𝑖𝑟𝑚𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖,𝑡
+ ∑ 𝐶𝑜𝑢𝑛𝑡𝑟𝑦𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖,𝑡+ 𝜀𝑖,𝑡
Where i stands for the firm observations at time t. This equation will be used to test the data
for the first and the second hypothesis. The model will be tested for a normal distribution, multicollinearity and autocorrelation. The Hausman test and the redundant fixed effect test will define whether the right model is used. The third hypothesis will be tested, by comparing the means of the company’s cash holding for the pre-crisis sample and the crisis sample, respectively 2003 to 2007 and 2008 to 2008. The fourth hypothesis is tested by using the above equation divided by the pre-crisis and crisis periods.
3.3 Descriptive Statistics
Table 1 shows the descriptive statistics and exhibit the main variables, with its means, standard deviation, median and number of observations. Panel A displays that the variable Cash holdings has a mean of 0,16, a standard deviation of 0,16, a median of 0,11 with 10.649 observations in total. Taking these numbers in consideration one could say that there are large fluctuations as the mean and standard deviation are similar, additionally, more than half of observations are below the mean. This can be concluded as the median of the cash holding variable is lower than the mean. This could be said for all the variables, except for the NWC ratio and firm size, displayed in Panel A. All these variables (Cash holdings, Capital expense, Market-to-book, R&D, and Leverage) have a standard deviation that is higher than or equal to the mean (respectively; 0.17 to 0,16; 0,04 to 0,04; 15,50 to 3,02; 0,12 to 0,05 and 0,21 to 0,21), and the median is below the mean (respectively; 0,11 to 0,16; 0,03 to 0,04; 1,86 to 3,02; 0,02 to 0,05 and 0,19 to 0,21).
17
Table 1. Descriptive statistics
Summary statistics of key variables for 2003-2014 split up in three panels. In Panel A the key firm statistics are given, describing the variables, Cash holdings (cash, cash equivalents and short-term investments over total assets), Capital expense (capital expense over total assets), Market-to-Book (ratio of market value to book value), R&D (R&D expense over total sales), Leverage (debt to total assets ratio), NWC ratio (NWC over total assets) and the yearly observations In Panel B shows the statistic of Panel A divided into a domestic firm sample (foreign sales/total sales<25%) and
multinational firms (foreign sales/total sales>25%). Panel C represent the descriptive statistics per country, it shows the cash holdings per country, a country’s KKMI score (2010, calculated by summing up the scores of the 6 governance indicators), the inflation rate (Worldbank), a country’s level of corruption (Transparency International) and a country’s GDP per capita (Worldbank) Panel C can be found in the appendix.
has 9600 yearly observation and the domestic firm subsample has 2630 yearly observations. The multinationals’ cash holdings have a mean of 0,147 to a mean of 0,184 by the domestic firms. This could already confirm the hypothesis that multinational companies hold lower cash than domestic companies, however statistical prove when controlled for the other
variables is needed. Additionally, Panel B shows that multinational firms have a lower capital expense (0,05 to 0,81), a lower market-to-book ratio (2,65 to 3,56), lower R&D ratio (0,047 to 0,059), a similar leverage ratio (0,218 to 0,217) and a higher NWC ratio (0,2 to 0,18).
Panel C shows the descriptive statistics per country with the variables (see appendix); Cash holdings, KKMI (investor protection index), inflation, corruption and GDP per capita. It shows that Serbia has the lowest cash holdings ratio of 0,04 and Norway the highest ratio of 0,24. The KKMI variable displays the lowest mean for Serbia, with a score of minus 1,55. The highest mean is found in Finland, with a score of 10,83. This indicates that the investor experiences the highest protection in Finland and the lowest protection in Serbia. The
Variable Mean Standard Deviation Median Observations
Panel A Summary statistics of key variables 2003-2015
Cash holdings 0,16 0,17 0,11 10649 Capital Expense 0,04 0,04 0,03 10649 Market-to- book 3.02 15,50 1,86 10649 R&D 0,05 0,12 0,02 10649 Leverage 0,21 0,21 0,19 10649 NWC ratio 0,002 0,001 0,0019 10649 Firm size 13.37 2.50 13.19 10649
Panel B: descriptive statistics per Domestic/MNCs firm
Cash holdings Capital expense
Market-to-Book R&D Leverage
18 inflation rate displays its highest rate in Romania and its lowest in Ireland. Regarding the corruption variable for the countries, the highest score and therefore the lowest corruption rate can be found in Denmark. The lowest score and therefore the highest corruption rate can be found in Serbia. The GDP per capita rate is found in Luxembourg, the lowest GDP per capita score is found in Serbia. And the last variable, the quality of the banking system, displays its lowest value in Serbia and its highest value in Ireland and Switzerland.
4. Results
In this section the data will be examined regarding the four hypotheses developed in the theoretical background section. Firstly, the first two hypothesis have been tested, this was done by using the equation provided in the Data and Methodology section. This will be a time and cross sectional multiple regression, so factors can be controlled across entities and over time. Furthermore, several test statistics are provided in the appendix, (namely; a redundant fixed effects test, the Hausman test, a test for normality, a test for autocorrelation, and a correlation matrix) to examine whether the right model and whether this model is biased. Secondly, the third hypothesis will be tested by comparing the means of the two subsamples (pre-crisis vs. crisis period). A t-test will provide a significance level for both periods’ means. The fourth hypothesis will be tested in same manner as the first and the second hypotheses, however, the equation is divided in the periods of the pre-crisis and crisis and in the
subsample multinational and domestic firms.
4.1 Part A
This section will analyse the first and second hypotheses, namely:
Hypothesis 1: a higher investor protection score will decrease the amount of cash held by the company
19
Table 2 Regression equation on Cash Holdings
This table presents the effect of investor protection and mulinationality on cash holdings in the European market. The sample period is 2003-2014 and include non-financial and non-utilities firms. The regression sample has 1118 unique firms with a total of 8224 yearly observations. The dependent variable is cash holding (cash, cash equivalents and short-term investments) and the main independent variables are KKMI (sum of six governance indices, 2010) and Multinationality (a dummy variable which equals 1 if the foreign sales/total sales is greater than 25%, otherwise 0). The equation is controlled by several firm and country control variables, described in previous sections. The results show the coefficient next to the relevant variables and the t-statistics are shown in brackets underneath the coefficient. The cash holding variable in winsorized at 1% at both ends, in order to mitigate the effect of possible outliers. The equation is extended (not shown in this table, see appendix), by a lagged variable of cash holding, in order to eliminate autocorrelation. After performing the Hausman test and the redundant fixed effect test, conclusions were made that regression is most suitable with cross-section fixed effects and period fixed effects (see appendix for related tests). Additionally, the regression was tested for non-normality, however, the sample size is big enough to be biased by non-normality (see appendix).
*** statistically significant at the 0,01 level ** statistically significant at the 0,05 level * statistically significant at the 0,10 level
Variable
Intercept 0,316 ***
(5,91)
Investor protection (KKMI) -0,002
20 The effects of multinationality and investor protection on the amount of cash held are shown in table 2. This table shows the regression after correcting the model for autocorrelation effects, by adding a lagged cash holding variable to the equation, and mitigating the effect of possible outliers, by winsorizing at the 1% level. Additionally, after taking the Hausman test and the redundant fixed effect test, it is concluded that a regression model with both cross-section and period fixed effects is the most appropriate model.
The table shows that the coefficient for the investor protection index (KKMI) shows that investor protection has a negative effect on the amount of cash holdings, which is in line with both the agency motive and transaction cost motive for holding cash. However, the variable is not statistically significant at any significance level, therefore hypothesis one ‘a higher
investor protection score will decrease the amount of cash held by the company’ cannot be accepted.
The coefficient of -0,01 of the multinational dummy variable in the table indicates that multinational companies hold less cash than domestic companies do. The coefficient is statistically significant at the 0,01 level and, therefore, we reject hypothesis 2a and accept hypothesis 2b stating that ‘A multinational company holds less cash than domestic company does’. This found evidence is contrary to the tax motive examined by Foley et al. (2007), stating that multinational companies hold more cash than domestic firms do due to tax reasons, and is consistent with the transaction cost motive and Fernandes and Gonenc et al findings, stating that multinational companies hold less cash as they have economies of scale in cash management.
Although the statistically significance of the multinational dummy variable, it only explains a small portions of the cash held by a company. More importantly, the table shows that both the NWC ratio and R&D costs explain a much larger portion of cash holdings, especially the NWC variable. This variable shows a coefficient of -1,016, which is statistically significant at the 0,01 level, and is therefore the variable which influence the cash holdings the most. For every increase by one, all other variables stay equal, the amount of cash decreases by 1,016. This negative relationship is line with the precautionary and transaction motive, stating that companies that hold more liquid assets, hold less cash, as liquid assets are easily transferable to cash.
21 This evidence is contradictory to the precautionary motive and the findings of Dittmar et al. (2003), arguing that there is a negative relationship between R&D expense and cash holdings, as holding a larger amount of cash will guarantee the company the ability of paying for future valuable investments.
The other variables that are statistically significant up to a 0.10 significance level are firm size, capital expenditure and the quality of banking systems. The variable firm size shows a coefficient of -0,016, significance level of 0,01, showing a consistent view with the
transaction motive, arguing that bigger firms have economies of scale in cash management. The coefficient of capital expenditure is, as stated in the table, -0,003 at a significance level of 0,01. This shows that an increase of capital expenditures have a negative effect on cash
holdings. This relationship is consistent with the findings of Bates et al. (2009). The last variable that is significant at a significance level of 0,1 is the Quality of banking. Contrarily, to the assumptions of Ramirez and Tadesse (2009), the relationship found in the equation is positive, with a coefficient of 0,001. Arguing that for every decrease in the quality of banking, the firm’s cash holdings will increase. Although these variables are all significant, they only explain a small portion of the amount of cash held within a company.
4.2 Part B
In this section, hypothesis three and four will be analysed and answered.
Hypothesis 3: The European companies experienced a decline in corporate cash holdings during the European crisis.
Hypothesis 4: The European multinational companies experienced a lower decline in corporate cash holdings than the European domestic companies.
The third hypothesis is tested through a comparison of the cash holding means for the pre-crisis period and pre-crisis period. The fourth hypothesis is answered by using the same equation used for hypothesis one and two, but divided in the pre-crisis period and the crisis period and a sub-sample is made for domestic and multinational firms.
22
Table 3.Means per period/sample
This table presents the overall means and the mean of domestic and multinational firms, divided into the periods pre-crisis (2003-2007) and crisis (2008-2014). A test for equality of means has been done in order to fabricate the presented p-values of the corresponding t-statistics
mean. This could indicate that multinational are better protected in times of trouble, as they could collect liquid assets easier through external financing, relating to their superior access to financial markets.
To compare the means of cash holdings during the pre-crisis and crisis period, a test for equality of means is provided to analyse the mean differences between the period and for the domestic and multinational firms. Table 3 present the related numbers on the next page. As presented in the column of overall means, the means show little differences between the pre-crisis and crisis period. Therefore, we could not accept hypothesis 3, stating that the European companies’ experiences a decrease, when faced by economic crisis. Reasons for this could be, that the companies spend less and therefore keep the same level of cash held. This decrease in spending is shown in figure 1., which indicates a small decline of
expenditures from 2008 to 2015.
Although, there is no differences between the overall means of the separate periods, there is a difference between the means of the domestic companies’ cash holdings between the pre-crisis and the pre-crisis period. It states that the mean of domestic cash holdings decreases from 0,1939 to 0,1786, whereas the mean of multinational’s cash holding hardly changed. This could indicate that the multinationals are still able to finance investments through external channels, whereas domestic firms need to finance investment internally through cash. Table 4 presents the coefficients of multinational and domestic firms calculated by using all the control variables described earlier. The coefficients of domestic and multinational firms during the pre-crisis period is, respectively, 0,005 at a significance level of 0,01 and -0,026 at a significance level 0,01. This is in line with the evidence found for the second hypothesis, stating that multinational companies hold less cash than domestic companies do. The
Means per period/sample
Overall
Mean MNCs Domestic P-value
Pre-Crisis(2003-2007) 0,1554 0,1445 0,1939 0,000
Observations 4379 3415 964
Crisis (2008-2014) 0,1546 0,1480 0,1786 0,000
23
Table 4. The effect of the European crisis on Domestic and Multinational firms cash holdings This table presents the coefficient of domestic and multinational firms divided for the pre-crisis and crisis period, derived by using the same equation with the same control variables as hypothesis 1 & 2. The statistic is shown in the brackets. The full table with all control variables can be found in the appendix (table D). coefficients of the domestic and multinational companies during the crisis period are,
respectively, 0,026 and -0,05. This shows that both companies hold more cash during a crisis period, however these two coefficients are both statistically not significant at any significance level. Therefore we could not accept hypothesis 4, stating that European multinational
companies experienced a smaller change in corporate cash holdings than European domestic companies do.
4.3 Robustness check
In order to make the evidence sounder, a robustness check is provided at table E in the appendix. It shows a regression model with multiple thresholds determining whether a company is multinational or domestic. There are four threshold used for a multinational, namely, a 15 per cent, a 20 per cent, a 30 per cent and a 35 per cent foreign sales over total sales ratio. All these threshold had a negative coefficient of, respectively, -0,02, -0,019, -0,009 and 0,06, and are at least statistically significant at a significance level of 0,05. Therefore, they are all consisted with the earlier findings within this paper and, thus, making this evidence sounder.
The effect of the European crisis on Domestic & Multinational firms cash holdings
Pre-crisis (2003-2007) Crisis (2008-2014)
Domestic firms 0,005*** 0,026 (1,422) (3,069) Multinational firms -0,026*** -0,005
(-2,99) (-1,14)
24
5. Conclusions
Cash holdings are an important tool as they are used in different ways. They could be used as a tool to eliminate transaction costs related to external financing, it could be used as buffer for adverse-risks or to deal with agency costs related to the different interest of shareholders and managers. There has been several studies done regarding this topic, however none of them used solely European firms as a sample and used the European crisis as an additional effect that may explain the level of cash held by the European firm. Due to the sample and the effect of the European crisis, this paper makes a contribution to the current literature base.
Therefore, this research adds to the established debates regarding the cash holding topics, finding statistical evidence for factors that influence the amount of cash hold by a company. The research question and sub-questions are answered, by using data from 2003 to 2014 from European non-financial companies, enabling this paper to extend the current literature in two ways. Firstly, evidence is found that multinational companies hold less cash than domestic companies do. This is contrary the studies of the tax motive (Titman et al. 2007, Foley et al. 2009), stating that multinational companies hold less cash as the tax consequences of repatriation is too high. However, the findings are consistent with the research of Fernandez and Gonenc (2015), stating that multinationals could create economies of scale in cash
management. And secondly, although there is no evidence supporting the other hypotheses on the relationship between investor protection, the European crisis and cash holdings, some other interesting statistics are found. The regression model shows that several other variables explain the level of cash holdings, especially the NWC ratio and R&D costs. The results show that the NWC ratio has the highest influence on cash holdings, stating that there is a negative relationship between them. This is in line with the transaction and precautionary motive, arguing that firms with an increase amount of liquid assets, hold less cash, as these liquid assets are easily transferable to cash assets. The regression model shows that there is a negative relationship between R&D expenditures and cash holding. This evidence is
25
6. Limitations and Recommendations
Limitations of this research might be that this paper only used one proxy for investor
protection, whereas other researchers, for instance Huang et al. (2013) used three proxies for investor protection. However, due to a lack of data for certain countries this was not an option. In addition to the restrictions of available data, some determinants of cash holdings were not used as a control variable as the data could not be derived from the available systems used for gathering the data. An example of this is the variable corporate governance. And lastly, due to lack of knowledge on other statistical software programs, the Eviews software is used to derive results. However, with this software it is impossible to perform a
heteroskedasticity test for panel data, therefore, the found results should be interpreted with precaution.
A recommendation for future research is finding out the rationale behind the found result of the negative R&D expenses and the amount of cash holdings. As this result is contradictory to previous found evidence, it might be interesting to determine the rationale behind this
relationship. In addition to this, it would interesting to investigate other continents, to find out whether the same results apply to those continents
26
7. References
Aggarwal, R., Erel, I., Stulz, R., & Williamson, R. (2010). Differences in governance
practices between US and foreign firms: Measurement, causes, and consequences. Review of Financial Studies, 23(3), 3131-3169.
Bates, T. W., Kahle, K. M., & Stulz, R. M. (2009). Why do US firms hold so much more cash than they used to. The Journal of Finance, 64(5), 1985-2021.
Berk, J, M,. (2016). Monetary Policy and Financial Regulation: Part 1- Monetary Policy.
Retrieved from
https://nestor.rug.nl/webapps/blackboard/execute/content/file?cmd=view&content_id=_76734 62_1&course_id=_147930_1&framesetWrapped=true
Campello, M., Graham, J. and Harvey, C. (2010), ‘The real effects of financial constraints: evidence from a financial crisis’, Journal of Financial Economics, 97.
Chen, Q., Chen, X., Schipper, K., Xu, Y., & Xue, J. (2012). The sensitivity of corporate cash holdings to corporate governance. Review of Financial Studies, hhs099.
Conerly, B. (2015). Business Cash Holdings: Too Much?. Available:
http://www.forbes.com/sites/billconerly/2015/03/25/business-cash-holdings-too-much/#2715e4857a0b2c91ae083d37. Last accessed 20th Jan 2016.
Dittmar, A., Mahrt-Smith, J., & Servaes, H. (2003). International corporate governance and corporate cash holdings. Journal of Financial and Quantitative analysis, 38(01), 111-133. Doukas, J. A., & Pantzalis, C. (2003). Geographic diversification and agency costs of debt of multinational firms. Journal of Corporate Finance, 9(1), 59-92.
Fernandes, N., Gonenc, H., (2015). Multinationals and Cash Holdings. Unpublished working paper. IMD International, University of Groningen
Foley, C. F., Hartzell, J. C., Titman, S., & Twite, G. (2007). Why do firms hold so much cash? A tax-based explanation. Journal of Financial Economics, 86(3), 579-607.
27 Haan de, J, Oosterloo, S, Schoenmaker, D. (2015). Financial markets and institutions: A European perspective. Cambridge University Press, Cambridge.
Harford, J., Mansi, S.A., Maxwell, W.F., (2008). Corporate governance and firm cash holdings in the US. Journal of Financial Economics, (3).
Huang, Y., Elkinawy, S., & Jain, P. K. (2013). Investor protection and cash holdings: Evidence from US cross-listing. Journal of Banking & Finance, 37(3), 937-951.
Iskandar-Datta, M. E., & Jia, Y. (2014). Investor protection and corporate cash holdings around the world: new evidence. Review of Quantitative Finance and Accounting, 43(2), 245-273.
Ivashina, V., & Scharfstein, D. (2010). Bank lending during the financial crisis of 2008. Journal of Financial economics, 97(3), 319-338.
Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of financial economics,3(4), 305-360.
Jensen, M. C. (1986). Agency cost of free cash flow, corporate finance, and takeovers. Corporate Finance, and Takeovers. American Economic Review,76(2).
Kaufmann, D., Kraay, A., & Mastruzzi, M. (2010). The worldwide governance indicators: A summary of methodology. Data and Analytical Issues, World Bank Policy Research Working Paper, (5430).
Keynes, J. M. (1936), ‘The general theory of employment, interest and money’, London: Macmillan and Co. Limited.
Kogut, B., & Zander, U. (1993). Knowledge of the firm and the evolutionary theory of the multinational corporation. Journal of International Business Studies, 24(4), 625–645.
La Porta, R., Lopez‐de‐Silanes, F., Shleifer, A., & Vishny, R. W. (2000). Agency problems and dividend policies around the world. The Journal of Finance,55(1), 1-33.
28
Mulligan, C. B. (1997). Scale economies, the value of time, and the demand for money: Longitudinal evidence from firms. Journal of Political Economy, 105(5), 1061-1079. Opler, T., Pinkowitz, L., Stulz, R., & Williamson, R. (1999). The determinants and implications of corporate cash holdings. Journal of financial economics,52(1), 3-46.
Ozkan, A., & Ozkan, N. (2004). Corporate cash holdings: An empirical investigation of UK companies. Journal of Banking & Finance, 28(9), 2103-2134.
Pagano, M., Röell, A. A., & Zechner, J. (2002). The geography of equity listing: why do companies list abroad?. Journal of finance, 2651-2694.
Pinkowitz, L., Stulz, R. M., & Williamson, R. (2003). Do firms in countries with poor
protection of investor rights hold more cash? (No. w10188). National Bureau of Economic
Research.
Pinkowitz, L., Stulz, R., & Williamson, R. (2006). Does the contribution of corporate cash holdings and dividends to firm value depend on governance? A cross‐country analysis. The
Journal of Finance, 61(6), 2725-2751.
Pinkowitz, L., Stulz, R. M., & Williamson, R. (2012). Multinationals and the high cash
holdings puzzle (No. w18120). National Bureau of Economic Research.
Puri, M., Rocholl, J., & Steffen, S. (2011). Global retail lending in the aftermath of the US financial crisis: Distinguishing between supply and demand effects. Journal of Financial
Economics, 100(3), 556-578.
Ramirez, A., & Tadesse, S. (2009). Corporate cash holdings, uncertainty avoidance, and the multinationality of firms. International Business Review, 18(4), 387-403.
Stulz, R. (1990). Managerial discretion and optimal financing policies. Journal of financial
Economics, 26(1), 3-27.
29
Table A. Descriptive statistics cont.
This table is the continuation of the descriptive statistic found in table 1 within the methodology section. This table show the means of the variables cash holdings, Investor protection (KKMI), Inflation, Corruption, GDP per Capita and the banking quality as well as the observations per country
8. Appendix
Country Observations
Cash
holdings KKMI Inflation Corruption
GDP per Capita
Banking quality
Panel C . Descriptive summary per country
Austria 358 0,09 9,29 1,8 7,87 45073 70,9 Belgium 271 0,14 7,72 1,77 7,39 42552 70,2 Denmark 323 0,16 10,66 2,04 9,3 55458 76,4 Finland 780 0,13 10,83 1,71 9,25 45256 73,5 France 1766 0,15 7,13 1,43 7,07 39529 62,6 Germany 2167 0,17 8,58 1,22 7,95 40835 70,9 Greece 391 0,09 3,04 1,57 4,14 24742 61,1 Hungary 39 0,08 4,56 3,75 5,08 12591 65,6 Iceland 26 0,06 9,34 4,72 8,75 49483 73,5 Ireland 101 0,12 8,71 0,5 7,5 52044 80,0 Italy 503 0,13 3,40 1,81 4,56 34924 62,1 Luxembourg 13 0,11 9,87 3,14 8,34 98251 75,8 Netherlands 432 0,14 9,78 1,47 8,7 48184 74,8 Norway 625 0,24 10,00 4,48 8,66 83376 68,5 Poland 625 0,12 4,00 2,47 4,84 11137 61,9 Portugal 103 0,09 5,98 1,78 6,28 21094 63,7 Romania 22 0,07 0,56 9,56 3,69 7422 60,2 Serbia 20 0,04 -1,55 8,67 3,39 5150 55,9 Slovakia 44 0,10 4,38 1,92 4,52 15306 67,5 Slovenia 53 0,04 5,52 2,23 6,17 22004 61,5 Spain 444 0,10 5,31 1,84 6,43 29361 68,8 Sweden 1312 0,21 10,22 1,6 9,12 51000 71,1 Switzerland 1235 0,20 10,11 0,67 8,84 69512 80,0 United Kingdom 2519 0,18 8,21 2,49 8,05 41390 77.3 0 0,05 0,1 0,15 0,2 0,25 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Fig. A, Means per year/segment
30
Fig. B Normality test
This graph shows that there is a non-normal distribution of the errors. However, the sample size is ought to be big enough to ignore this issue.
Table B multicollinearity test
This table shows that there are possible cases of multicollinearity between investor protection, GDP per capita and the quality of banking, as they exceed the threshold of 0,5. However, this will be ignored as they are used as a control variable and are not correlated with the independent variables
Cash Holdings MNCs KKMI CAPEX Dividend Leverage NWC R&D Firm size Inflation Banking GDP Corruption Cash Holdings 1 -0,099 0,087 -0,1667 -0,269 -0,284 0,112 0,325 -0,269 -0,013 0,064 0,121 0,085 MNCs -0,099 1 0,169 -0,025 0,136 -0,011 -0,122 -0,023 0,245 -0,047 0,048 0,138 0,184 KKMI 0,087 0,169 1 -0,012 0,026 -0,083 -0,017 0,088 0,078 0,399 0,612 0,592 0,184 Capex -0,1667 -0,025 -0,012 1 0,118 0,062 -0,039 -0,095 0,095 0,103 -0,021 -0,027 -0,034 Dividend -0,269 0,136 0,026 0,118 1 -0,16 -0,197 0,41 -0,004 0,018 0,014 0,022 0,022 Leverage -0,284 -0,011 -0,083 0,062 -0,02 1 -0,066 -0,098 0,138 -0,012 -0,081 -0,035 -0,089 NWC 0,112 -0,122 -0,017 -0,039 -0,16 -0,066 1 0,138 -0,339 0,073 0,105 -0,06 -0,009 R&D 0,325 -0,023 0,088 -0,095 -0,197 -0,098 0,138 1 -0,202 -0,021 0,028 0,067 0,101 Firm size -0,269 0,245 0,078 0,095 0,41 0,138 -0,339 -0,202 1 -0,004 -0,107 0,143 0,079 Inflation -0,013 -0,047 0,399 0,103 -0,004 -0,012 0,073 -0,021 -0,004 1 -0,007 0,001 0,014 Banking 0,064 0,048 0,612 -0,021 0,018 -0,081 0,105 0,028 -0,107 -0,007 1 0,492 0,338 GDP 0,121 0,138 0,592 -0,027 0,014 -0,035 -0,06 0,067 0,143 0,001 0,492 1 0,425 Corruption 0,085 0,184 0,184 -0,034 0,022 -0,089 -0,009 0,101 0,079 0,014 0,338 0,425 1
31 *** statistically significant at the 0,01 level
** statistically significant at the 0,05 level * statistically significant at the 0,10 level
Pre-crisis Crisis
Domestic firms 0,026 *** 0,005
(-1,273) (1,142)
Multinational firms -0,026 *** -0,005
(-2,995) (-1,142)
Investor protection (KKMI) -0,007 -0,001
(-1,268) (-0,304) Dividend -0,023 *** 0,007 (-2,981) (1,972) NWC -0,699 ** -1,74 *** (-2,046) (-11,001) Market to Book 0,000 -0,0002 1,012 (-2,488) Firm size -0,009 -0,018 *** (-1,324) (-4,296) Leverage -0,001 0 (-5,951) (0,716) Capital expenditure -0,001 * -0,002 *** (-1,929) (-7,059) R&D -0,205 *** -0,105 *** (-3,495) (-10,156) Inflation -0,003 0,0003 (-1,359) (0,425) GDP per capita 0 0 (-0,978) (-0,439) Corruption 0,024 -0,007 (1,968) (-1,522) Quality of banks 0,001 0,001 (0,837) (0,929)
The effect of the European crisis on domestic and MNCs' cash holding
Table D. The effect of the European crisis on Domestic and Multinational firms cash holdings
32 *** statistically significant at the 0,01 level
** statistically significant at the 0,05 level * statistically significant at the 0,10 level
15% 20% 30% 35%
Multinationality -0,020 -0,013 *** -0,009 *** -0,006 **
(-4,937) (-3,387) (-2,754) (-2,169)
Investor protection (KKMI) -0,002 -0,002 -0,002 -0,002
(-0,626) (-0,687) (-0,737) (-0,795) Dividend 0,002 0,002 0,002 0,002 (0,709) (0,713) (0,774) (0,783) NWC -1,031 -1,024 *** -1,013 *** -1,019 *** (-8,183) (-8,115) (-8,025) (-8,081) Market to Book -0,000 -0,000 0,000 0,000 (-1,584) (-1,536) (-1,567) (-1,581) Firm size -0,016 -0,016 *** -0,016 *** -0,016 *** (-1,816) (-6,058) (-6,007) (-5,985) Leverage 0,000 0,000 0,000 0,000 (-1,816) (-1,717) (-1,598) (-1,585) Capital expenditures -0,003 -0,003 *** -0,003 *** -0,003 *** (-10,077) (-10,058) (-10,062) (-10,059) R&D -0,128 -0,129 *** -0,129 *** -0,129 *** (-12,259) (-12,306) (-12,349) (-12,378) Inflation 0,000 0,000 0,000 0,000 (-0,026) (-0,038) (-0,055) (-0,040) Gdp per capita 0,000 0,000 0,000 0,000 (0,117) (0,221) (0,307) (0,378) Corruption -0,004 -0,004 -0,004 -0,004 (-1,265) (1,869) (-1,265) (-1,269) Quality of banks 0,001 0,001 0,001 0,001 * (1,851) (1,869) (1,883) (1,921)
Robustness Check of the effect of Multinationality on Cash holdings Table D. Robustness Check of the effect of Multinationality on Cash holdings
