Cash Holdings in European Firms

Transcription

1 Cash Holdings in European Firms Masterthesis Finance Tilburg University October 2011 Name: E.A.M. Mangnus ANR: Supervisor: prof. dr. V. Ioannidou Abstract The corporate cash holdings in the United States have increased dramatically over the past decades. This thesis studies if the rising trend in cash holdings is also existent in Europe by utilizing a dataset of 25 EU countries for the period from 1988 to The different motives for holding cash are tested. I research what firm characteristics are important determinants of cash holdings for European firms and how their relation changes over time. Special attention is paid to the crisis years. I also investigate if agency problems can explain the changes in cash holdings. The results show that in Europe corporate cash holdings have increased significantly in the period from 1988 to The increase in cash holdings can largely be explained by changes in firm characteristics, the most important ones being an increase in the R&D ratio, an increase in cash flow volatility and a decrease in the net working capital ratio. During the crisis years the cash ratio decreased slightly, which can be explained by a decrease in cash flow volatility, a decrease in the market to book ratio and an increase in the leverage ratio. The results also show that agency problems can partly explain the increase in corporate cash holdings in Europe. Keywords: Cash Holdings, Liquidity, Financial Crisis, Agency Problems

2 Table of Contents 1. Introduction 2 2. Theoretical and Empirical Literature Review Motives for holding cash The transaction motive The precautionary motive The tax motive The agency motive The pecking order theory Differences between Europe and the United States Liquidity constraints Cash flow risk Taxation Corporate governance The crisis years Data Description A First Overview of the Change in Cash Holdings over Time Empirical Analysis The determinants of corporate cash holdings Does the relation between firm characteristics and cash holdings change over time? What firm characteristics are most important in explaining cash holdings? The crisis years The agency motive for cash holdings Conclusion 55 References.57 Appendix

3 1. Introduction Over the past decades the cash holdings of U.S. firms have increased dramatically. The average cash to assets ratio more than doubled from 1980 to Bates, Kahle and Stulz (2009) show that this trend is not driven by the largest firms in the sample. Capital raising activities of IPO firms also cannot explain the growth in cash holdings. By using the model of Opler, Pinkowitz, Stulz and Williamson (1999), they investigate whether the increase in cash holdings results from changes in firm characteristics and whether the relation between firm characteristics and the cash ratio changes over time. The four most important changes in firm characteristics that led to larger cash holdings in the U.S. are decreasing working capital, increasing cash flow volatility, decreasing capital expenditures, and increasing R&D expenditures. Bates, Kahle and Stulz (2009) did not find evidence that the relation between firm characteristics and the cash ratio changed over time. They also did not find evidence for an increase in agency problems. Theory provides different motives for firms to hold cash. First, the transaction motive introduced by Keynes (1936) states that firms hold cash to avoid the transaction costs of selling illiquid assets and using capital markets to raise funds. In the precautionary motive cash holdings function as a buffer against possible adverse shocks in the future, especially when external financing is costly (Pinkowitz and Williamson, 2004). Firms might also hold cash to be able to take advantage of profitable investment opportunities in the future. Another motive for firms to hold cash stems from the tax incentives that multinationals face (Foley, Hartzell, Titman and Twite, 2007). The agency motive states that asymmetric information and conflicting interests between debt holders and equity holders raise a firm s costs of outside funds and can therefore cause firms to build up cash balances (Jensen and Meckling, 1976; Grinblatt and Titman, 2004). Also conflicting interests between equity holders and the management of the firm can lead to large cash holdings, since the management has an incentive to hold excess cash balances to pursue its own objectives at the expense of the firm s equity holders. Finally, according to the pecking order theory, firms have a preference to use internal finance over external finance for investments (Myers, 1984). There are several reasons why cash holdings may evolve differently in Europe compared to the United States. First, there are differences in liquidity constraints between European and U.S. firms. Nykvist (2008) argues that the access to capital markets is worse in Europe than in the United States. According to Aernoudt (1999) this is due to the European risk-averse culture, the lack of confidence of bankers in small and medium size enterprises (SMEs), the attitude towards failure and the absence of liquid secondary markets. The European Commission (1997) argues that banks charge excessive interest rates, have a high level of risk aversion, require too many guarantees and prefer short-term lending. Bottazzi and Da Rin (2003) find that the evolution of venture capital in Europe has lacked behind compared to venture capital in the United States, creating a large gap between venture capital markets in 2

4 Europe and the United States. A final reason for the underdeveloped capital markets in Europe is that the financial integration in European credit markets lacks behind (Baele, Ferrando, Hördahl, Krylova and Monnet, 2004). According to the theory, the high liquidity constraints compared to the U.S. should make the precautionary motive for cash holdings more important in Europe. There are also differences between Europe and the United States in the field of corporate governance. La Porta, Lopez-de-Silanes, Shleifer and Vishny (1998) argue that the legal protection of investors is for a large part determined by legal origin. The U.S. stems from a legal origin that results in better protection of investors than the legal origin of European countries. 1 The theory of La Porta, Lopez-de-Silanes, Shleifer and Vishny (1998) suggests that agency problems might be more severe in Europe than in the United States. There are several channels through which the financial crisis could affect corporate cash holdings. Schwert (1989) shows that stock volatility increases during recessions and financial crises. In times of a crisis the ability of firms to raise external finance is reduced, which makes firms more liquidity constrained (Arslan, Florackis and Ozkan, 2006). Duchin, Ozbas and Sensoy (2010) show that the supply of external financing declines during crises, making corporate investments more sensitive to internal resources. Therefore it is expected that the precautionary motive for cash holdings plays a dominant role during the financial crisis years. In this thesis, I investigate if the rising trend in cash holdings found in the United States is also existent in Europe. The different motives for holding cash are tested. I research what firm characteristics are important determinants of cash holdings for European firms and how their relation changes over time. Special attention is paid to the crisis years. I also investigate if agency problems can explain the changes in cash holdings. The research methods draw heavily on those used in Bates, Kahle and Stulz (2009). The results show that in Europe corporate cash holdings have increased significantly in the period from 1988 to The cash ratio increases with 0.25% per year on average, which is lower than the time trend of 0.46% found for the United States in the period from 1980 to The increase in cash holdings can largely be explained by changes in firm characteristics, the most important ones being an increase in the R&D ratio, an increase in cash flow volatility, and a decrease in the net working capital ratio. These findings confirm the precautionary and transaction motive for cash holdings. During the crisis years the cash ratio decreased slightly, which can be explained by a decrease in cash flow volatility, a decrease in the market to book ratio and an increase in the leverage ratio. Although Bates, Kahle and Stulz (2009) did not find evidence for increasing agency problems in the United States, the results in this thesis suggest 1 Only the United Kingdom stems from the same legal origin as the United States, which will be elaborated on in Section

5 that agency problems can partly explain the increase in corporate cash holdings in Europe, confirming the agency motive for cash holdings. The remainder of this thesis is structured as follows. Section 2 provides an overview of the theories on cash holdings and reviews the empirical literature on this subject. In Section 3 a description of the dataset is given. Section 4 provides a first overview of the change in cash holdings over time. In Section 5 the empirical analysis is conducted and the results are presented. Finally, Section 6 concludes. 4

6 2. Theoretical and Empirical Literature Review The literature on corporate cash holdings provides different theories on why firms hold cash balances. In Section 2.1 these theories are reviewed extensively. The theory is able to explain the dramatic increase in corporate cash holdings in the United States in the past decades. However, Europe differs from the United States in several fields. In Section 2.2 the empirical literature is reviewed to find differences between Europe and the U.S. that could have an influence on corporate cash holdings. Differences in corporate governance, taxation, liquidity constraints and cash flow risk are considered. Finally, in Section 2.3 the possible effects of the financial crisis on corporate cash holdings are considered Motives for holding cash There are different theories that explain the motives for firms to hold cash balances. First, the transaction motive states that firms hold cash to avoid the transaction costs of selling illiquid assets and using capital markets to raise funds. Firms can also hold cash as a buffer against possible adverse shocks in the future, or to take advantage of profitable investment opportunities in the future. Another motive for firms to hold cash stems from the tax incentives that multinationals face. The agency motive states that asymmetric information and conflicting interests between debt holders, equity holders and management can cause firms to build up cash balances. Finally, according to the pecking order theory, firms have a preference to use internal finance over external finance for investments. The following sections review the literature on the different motives for holding cash The transaction motive Firms hold cash balances to minimize transaction costs. This transaction motive was introduced by Keynes (1936) and is defined as the need of cash for the current transaction of personal and business exchanges. Keynes (1936) makes a distinction between the income motive and the business motive. The income motive states that a certain amount of cash is held in order to bridge the interval between receipt and disbursement of income. The business motive states that firms hold cash balances in order to bridge the interval between the time of incurring costs and the time of the receipt of sale proceeds. In a world with perfect capital markets the transaction motive would not exist, since holding liquid assets has no opportunity costs. But in reality the buying and selling of assets comes with costs. A firm that is short of liquid assets has several options. It can raise funds in the capital markets, liquidate existing assets, reduce dividends and investment, renegotiate existing financial contracts, or a combination of these actions (Opler, Pinkowitz, Stulz and Williamson, 1999). Selling assets or using the capital markets to raise funds is costly to the firm. Therefore the firm holds cash and liquid assets as a buffer to avoid 5

7 these transaction costs. The optimal amount of liquid assets is the amount where the marginal costs of holding liquid assets are equal to the marginal costs of having a liquid assets shortage. The marginal costs of holding liquid assets are their lower expected return. Baumol (1952) creates a model in which a decision maker invests his wealth in an interest bearing asset and a non-interest bearing cash balance. Receipts and expenditures are not perfectly synchronized, which creates the need for transaction balances (Tobin, 1956). The transfer of funds between the two assets comes with transaction costs. So the decision maker faces a trade-off. On the one hand he earns interest from investing in the interest bearing asset, but on the other hand he incurs transaction costs from selling the illiquid interest bearing asset when he needs to make certain payments. The demand for money increases with the transaction costs of transferring funds from the interest bearing asset to the cash balance and decreases with the interest rate or the opportunity costs of the cash holdings. Miller and Orr (1966) adapt the money demand model of Baumol (1952) to include the variability of the cash flows. The model shows the same relation between transaction costs and the demand for money as the model of Baumol (1952). The model of Miller and Orr (1966) also shows that the demand for money is an increasing function of the variability of the cash flows. Mulligan (1997) shows there are economies of scale in the demand for money using a sample of 12,000 firms for the years The results indicate that large firms hold less cash as a percentage of sales than small firms The precautionary motive A second reason for firms to hold cash is to be able to cope with adverse shocks (Pinkowitz and Williamson, 2004). Firms want to be able to invest in positive NPV projects at all times, also when they are unable to generate enough internal cash to finance these projects. When external finance is costly, cash holdings accumulated in the past make sure that firms can still undertake positive NPV projects in the future. Especially when access to capital markets is limited this is an important reason for firms to accumulate cash holdings. This motive for holding cash originated from Keynes (1936) and is called the precautionary motive for cash holdings. Even when cash is not held as a buffer against future adverse shocks, firms might still hold cash balances in order to be able to take advantage of profitable investment opportunities in the future. This is called the speculative motive for cash holdings. Almeida, Campello and Weisbach (2004) model the relation between financial constraints and corporate liquidity demand for a large sample of manufacturing firms between 1971 and The precautionary motive predicts that firms with strong financing constraints maintain large cash balances. The need to hold cash is lower for financially unconstrained firms. Since there are opportunity costs to holding cash, firms face a trade-off between the profitability of current and future investments. The effect 6

8 of financial constraints is reflected by the firm s propensity to save cash out of cash flows, which is called the cash flow sensitivity of cash. According to the precautionary motive, the cash holdings of financially constrained firms should increase when cash flows are higher. In other words, financially constrained firms should have a positive cash flow sensitivity of cash. For financially unconstrained firms no such relation should be found. Almeida, Campello and Weisbach (2004) use different approaches to divide the firms into subsamples based on measures of financial constraints. As measures of financial constraints they use payout policy, asset size, bond ratings, commercial paper ratings, and an index that combines different measures. The results show that for firms that are financially constrained the cash holdings are positively related with cash flows. This cash flow sensitivity of cash holdings is not found for financially unconstrained firms. Han and Qiu (2007) extend the theoretical model of Almeida, Campello and Weisbach (2004) to include the uncertainty of future cash flows. In contrast to the model of Almeida, Campello and Weisbach (2004), it is not possible to fully diversify future cash flow risk in the model of Han and Qiu (2007). The focus of Han and Qiu (2007) is therefore not on the expected value of future cash flows but on the uncertainty of future cash flows. They estimate the relation between cash holdings and cash flow volatility for a dataset of U.S. firms in the period from 1997 to The results show that financially constrained firms increase their cash holdings in response to an increase in cash flow volatility, while the cash holdings of financially unconstrained firms are not sensitive to cash flow volatility. Both the results of Almeida, Campello and Weisbach (2004) and Han and Qiu (2007) support the precautionary motive of Keynes (1936). Pinkowitz and Williamson (2004) investigate the market value of cash held by firms in the United States. They estimate the value of a marginal dollar of cash to be around $1.20, which implies that the market values cash at a significant premium to its book value. This finding can be explained by the precautionary motive for cash holdings. Growth options and investment uncertainty increase the market value of cash holdings, while financial distress reduces the market value of cash. Remarkably, Pinkowitz and Williamson (2004) do not find evidence for a relation between capital market access and the value of cash holdings The tax motive A totally different explanation for holding cash stems from the tax incentives that multinational firms face (Foley, Hartzell, Titman and Twite, 2007). In the U.S. a firm s foreign income is taxed. 2 However, firms are allowed to defer their taxes if they do not repatriate their foreign income. So the U.S. tax system grants multinationals an incentive to retain earnings abroad, often in the form of cash holdings. Foley, 2 To avoid double taxation on foreign income, firms are granted a tax credit on the taxes they paid abroad. 7

9 Hartzell, Titman and Twite (2007) investigate whether higher tax costs associated with repatriation lead to larger cash holdings using a sample of U.S. firms for the period from 1982 to There are four main findings from their study. First, firms that face high repatriation tax costs hold larger amounts of cash than firms with low repatriation tax costs. Second, firms that face a high repatriation tax burden hold a larger share of their cash abroad. This increase in foreign cash holdings is not offset by lower domestic cash holdings. The third finding distinguishes between incorporated affiliates and branch affiliates. Unlike incorporated affiliates, the earnings of a foreign branch are taxed immediately. So for branches there is no incentive to retain large amounts of cash. In line with expectations, Foley, Hartzell, Titman and Twite (2007) find that incorporated affiliates hold higher levels of cash than branch affiliates. Finally, they find that financially constrained firms are less sensitive to repatriation tax costs. For firms that have a high level of leverage or a low debt rating there is no significant relation between their cash holdings and the repatriation tax costs. Foley, Hartzell, Titman and Twite (2007) show that the tax motive for holding cash is present in the United States. However, most European countries have a different tax system on foreign-source income than the system used in the United States. This will be elaborated on in Section The agency motive Agency costs can be another reason for firms to hold cash balances. Agency costs of debt arise out of differing interests between equity holders and debt holders and raise the costs of outside funds (Jensen and Meckling, 1976). The incentives of equity holders to maximize the value of their shares are not necessarily consistent with the incentive to maximize the total value of the firm s debt and equity (Grinblatt and Titman, 2004). On the contrary, equity holders often have an incentive to take actions that reduce the value to debt holders. Lenders anticipate the self-interested incentives of equity holders and therefore charge higher interest rates on their loans. So in the end, the equity holders bear the costs of their own future adverse incentives by paying higher interest rates on their borrowings. The agency costs of debt problem thus creates an incentive for equity holders to convince the debt holders that their goal is to maximize the total value of the firm instead of the value of equity. However, it may be difficult for firms to credibly commit to a policy of firm value maximization instead of equity value maximization. Grinblatt and Titman (2004) distinguish between four categories of problems that arise out of conflicts of interest between equity holders and debt holders. The first one is the debt overhang problem. When considering an investment opportunity, one can distinguish between the net present value to the firm and the net present value to the firm s shareholders. In some cases it can happen that the net present value to the firm is positive, while the net present value to the shareholders is negative. This can be the case when the firm has a high borrowing rate, so that the 8

10 created value accrues to the debt holders instead of the equity holders (Grinblatt and Titman, 2004). This debt overhang problem gives firms an incentive to pass up positive net present value projects when most of the project s benefits go to the debt holders. The debt overhang problem is also referred to as the underinvestment problem and can also be a consequence of senior debt obligations. Firms that have a large amount of senior debt outstanding may have difficulties in obtaining new financing, since this new debt will always be subordinated to the outstanding debt. This means that in times of default the new debt holders will be paid after the senior debt holders are paid. New debt holders might therefore be unwilling to lend to this firm, unless they can earn very high interest rates. The debt overhang problem is also relevant when firms use internally generated funds to finance an investment project. Equity holders might prefer the funds to be paid out as a dividend rather than investing the funds in positive net present value projects when most of the benefits go to the firm s debt holders. 3 A second problem is that the use of debt financing can lead to shortsightedness. A firm with a large amount of near-term debt obligations needs cash quickly. Therefore it might prefer investment projects that pay off quickly over long-term investment projects with a higher net present value. The shortsighted investment problem is related to the debt overhang problem. Firms with large amounts of senior debt may have great difficulties refinancing this debt with subordinated debt. Lenders will require high interest rates, making the refinancing costs for these firms very high. This generates an incentive for firms to select investment projects that pay off quickly. A third problem arising out of conflicts of interest between equity and debt holders is asset substitution. Firms with large amounts of debt have an incentive to invest in overly risky projects. The payoff to equity holders can be compared to the payoff from a call option on the firm s assets (Black and Scholes, 1973). The upside potential of the investment is unlimited, while the downside risk is limited to zero. Since equity holders have limited liability, they cannot lose more than the amount they invested in the firm. Option pricing theory states that the value of the option increases with the volatility of the underlying stock (Body, Kane and Marcus, 2009). Therefore it is beneficial for the equity holders to select risky investment projects. However, the benefit for the equity holders comes at the expense of the debt holders, since they are only subject to the downside risk and do not benefit from the upside potential of the risky projects. So by selecting riskier projects, firms can transfer wealth from their debt holders to their equity holders. Therefore equity holders may prefer a low net present value project with a high level of risk over a high net present value project with a low level of risk. If debt holders anticipate such behavior, they may be unwilling to provide debt financing to the firm. Equity holders must convince debt holders that they will commit to a policy of maximizing firm value instead of equity value. It may not be 3 In this case a possible solution for the debt overhang problem is to install covenants that restrict the amount of funds the firm can pay out in dividends. 9

11 easy for equity holders to credibly commit to such a policy. As a consequence, debt holders will adapt the terms of their loans to take into account the asset substitution problem. So in the end, it is again the equity holder that bears the costs of its own adverse incentives. Finally, the amount of debt in a firm can affect the firm s liquidation policy. When a firm s liquidation value is higher than its going concern value, liquidation is the best option for the firm. However, despite the lower going concern value, equity holders may still prefer to continue the firm s operations. The reason for this is that equity holders receive proceeds from the liquidation after all debt holders are paid off. Therefore, if the value of the outstanding debt exceeds the liquidation value of the firm, the equity holders will receive nothing from liquidation of the firm. When the firm continues to operate, there is a chance that the firm will become profitable again, providing value for the equity holders. Again this situation can be compared with option theory. Equity can be viewed as a call option on the value of the firm. When the value of the debt exceeds the liquidation value of the firm the option is out of the money and worth nothing if exercised. Continuing operations is analogous to higher risk and a longer maturity, which are factors that increase the value of a call option (Body, Kane and Marcus, 2009). Equity holders prefer continuing operations, since there is a chance that in the future the call option on the value of the firm will be in the money. These agency problems associated with holding large amounts of debt can be quite costly to firms. For a firm with high agency costs of debt it can be difficult to raise funds to invest in profitable projects. Therefore firms with high agency costs have an incentive to reduce their leverage and hold a large amount of liquid assets. For firms with valuable investment opportunities the costs of being short of funds is high. Therefore these firms are expected to hold larger cash balances than firms without valuable investment opportunities. Agency costs do not only arise out of conflicts of interest between equity holders and debt holders. Another form of agency costs stems from the conflicts of interest between the equity holders and the management of the firm. The management has an incentive to maintain excess cash balances to pursue its own objectives at the expense of the firm s equity holders. Opler, Pinkowitz, Stulz and Williamson (1997) discuss three channels through which management can use cash for its own objectives. First, entrenched management may hold excess cash out of risk aversion reasons. Second, large cash holdings give management more flexibility to pursue its own objectives. It allows the management to invest in projects it was not able to finance using the capital markets. Third, management may accumulate cash because it wants to keep funds in the firm instead of paying out excess cash to shareholders. Jensen 10

12 (1986) argues that entrenched management may use this excess cash to invest in negative net present value projects, when profitable projects are not available. 4 Dittmar, Mahrt-Smith and Servaes (2003) find empirical evidence for the agency cost motive of cash holdings. They find that firms in countries where shareholder rights are not well protected have significantly larger cash holdings than firms in countries with good shareholder protection The pecking order theory All previously described motives for firms to hold cash assume that firms optimize capital structures period by period by comparing the costs and benefits of leverage and cash holdings (Grinblatt and Titman, 2004). In other words, firms choose their capital structures in a static way. The pecking order theory states that there is no optimal amount of leverage and cash, but capital structures are determined in a dynamic way. According to Myers (1984) there are two central ideas of the pecking order theory. First, firms have a preference to use internal finance for investments. Second, when external financing is sought, firms have a preference for debt financing over equity financing. The costs of external finance arise out of information asymmetries. The firm s management has more information than outside investors. Therefore the firm may not be able to sell the securities for their actual value. As a consequence, the firm may choose to pass up a positive net present value investment to prevent issuing underpriced shares. The firm can avoid these information asymmetry costs by retaining enough internally generated cash to fund its future investment opportunities. If the firm requires external financing, the pecking order theory states that first the safest securities should be issued before more risky securities are used. Safe securities are securities that do not change much in value when the management s inside information becomes public. So according to the pecking order theory a firm issues the safest security first, which is straight debt. Next convertible bonds are used, and only as a last resort the firm issues equity. If the firm has sufficient funds to finance all positive net present value investments, it tends to pay off its debt and build up liquid assets Differences between Europe and the United States There are several reasons to believe that cash holdings evolve differently in Europe than in the United States. First, the literature provides several studies that find that liquidity constraints are more severe in Europe than in the United States. Second, there could be differences in cash flow risk. Another difference can be found in the tax systems, more specifically the way multinationals are taxed. Finally, there could 4 This third reason is less harmful for firms with good investment opportunities, since management can invest the excess cash in positive net present value projects, which is also beneficial to equity holders. 11

13 be differences in corporate governance. The following sections review the literature on these possible differences between Europe and the United States and elaborate on the implications for corporate cash holdings Liquidity constraints Nykvist (2008) argues that the access to capital markets is worse in Europe than in the United States. This implies that European firms face more severe liquidity constraints than U.S. firms. According to Aernoudt (1999) the major reasons for the underdeveloped capital markets are the European risk-averse culture, the lack of confidence of bankers in small and medium size enterprises (SMEs), the attitude towards failure and the absence of liquid secondary markets. In 1997 the European Commission organized a round table of bankers and SMEs in order to improve the relation between bankers and SMEs (European Commission, 1997). The main conclusions of the European Commission were that banks charge excessive interest rates, have a high level of risk aversion, require too many guarantees and prefer short-term lending. Bottazzi and Da Rin (2003) research the financing of European entrepreneurial firms and make a comparison with entrepreneurial firms in the United States. They find that European firms face larger credit constraints than U.S. firms. The main reason for this is that the evolution of venture capital in Europe has lacked behind compared to venture capital in the United States, creating a large gap between venture capital markets in Europe and the United States. Bottazzi and Da Rin (2003) show that since the late nineties credit constraints for European entrepreneurial firms were decreasing. The main changes responsible for this finding were the introduction of the euro, the increasing supply of venture capital, the creation of new equity markets, and the implementation of policies targeted at financing entrepreneurial firms. Although these developments led to a fast growing venture capital market in Europe, they are still far from closing the gap with the U.S. venture capital market. Another reason for the underdeveloped capital markets in Europe is the lacking financial integration in European credit markets. Baele, Ferrando, Hördahl, Krylova and Monnet (2004) measure the financial integration in Europe. They consider the money market, the corporate bond market, the government bond market, the equity market and the credit market. They find that integration is high in all markets except for the credit market. The main reasons for the slow integration in the European credit market are the low international competition pressures, asymmetric information problems and switching costs. There are several studies that research credit constraints by investigating the relation between finance and investment. When credit constraints play a role, the investment of a firm that has difficulties in obtaining external finance should be more sensitive to internal funds and cash flows (Whited, 1992). Whited indeed finds evidence for a positive relation between internal funds and investment for U.S. firms. 12

14 The heterogeneity of the European countries makes it difficult to come to a general conclusion on this matter for entire Europe. Several studies have been conducted for European countries. Devereux and Schiantarelli (1989) find that in the United Kingdom cash flow is positively related to investment. However, they do not find evidence for a relation between internal funds and investment. Van Ees and Garretsen (1994) study the relation between internal funds and investment in The Netherlands. Their results show a positive relation that is much stronger than for the United States, implying that Dutch firms suffer more from credit constraints. Another strand of literature researches liquidity constraints for potential entrepreneurs by looking at the relation between household wealth and the propensity to start a business (Hurst and Lusardi, 2004). If there are indeed liquidity constraints, this would mean that low wealth households experience difficulties in receiving financing to start up a business. Therefore the presence of liquidity constraints implies that wealthy households are more likely to become entrepreneurs than households with low wealth. Hurst and Lusardi (2004) do not find evidence for liquidity constraints in the United States. Nykvist (2008) claims that the private equity and venture capital markets in Europe are poorly developed. She uses the methodology of Hurst and Lusardi (2004) to test if entrepreneurs are subject to liquidity constraints in Sweden. She finds a positive relation between wealth and entrepreneurship, confirming the presence of liquidity constraints in Sweden. Johansson (2000) conducts a similar research in Finland. The results also show the existence of liquidity constraints. Holtz-Eakin and Rosen (2005) find in their study that Germany faces more severe liquidity constraints than the United States. Again, the heterogeneity of European countries makes it difficult to say anything about Europe in general. Within Europe large differences with respect to the development of capital markets exist. Firms in Central and Eastern Europe face severe financial constraints from their underdeveloped capital markets (Pissarides, 1999). The economic uncertainty makes banks unable or unwilling to provide long-term financing to firms. The high inflation pushes up interest rates. Also the development of bond and stock markets in this part of Europe is slow. Firms in Central and Eastern Europe face more severe liquidity constraints than in the rest of Europe. There are different studies on liquidity constraints in Central and Eastern European countries confirming the claims of Pissarides (1999). Budina, Garretsen and De Jong (2000) investigate the impact of liquidity constraints on investment performance in Bulgaria. They find that firms in Bulgaria face more liquidity constraints than Western European firms. Similar results are found for Czech Republic (Anderson and Kegels, 1997) and Estonia (Mickiewicz, Bishop and Varblane, 2004). The findings above indicate that European firms might face more severe liquidity constraints than U.S. firms. This makes it harder for European firms to attract capital when profitable investment opportunities arise. Based on these findings one would expect the precautionary motive to be more 13

15 important in Europe than in the United States. However, the heterogeneity within Europe makes it difficult to make claims about Europe in general. The literature showed that the capital markets in Central and Eastern European countries are underdeveloped compared to Western European countries. Therefore any differences with the U.S. are likely to be largest for these countries Cash flow risk One of the main reasons for the increase in cash ratios in the United States is the fact that over the past decades firms cash flows have become riskier (Bates, Kahle and Stulz, 2009). Kearney and Poti (2008) study the development of the idiosyncratic volatility of European stocks in the period from 1974 to They find that European stocks have become more volatile over time and that idiosyncratic risk is the main component of this increase. So in the past decades cash flow volatility has increased in both the U.S. and Europe. This is not surprising considering the increasing financial integration and globalization. According to Baele (2005) the U.S. volatility spillover effects in European equity markets increased significantly over the 1980s and 1990s. Also the U.S. continues to be the dominating influence in European equity markets. All in all, the literature does not give reasons to expect cash flow volatility to develop differently in Europe than in the United States. Brandt, Brav, Graham and Kumar (2009) find evidence that in the U.S. idiosyncratic volatility decreased from 2003 onwards to levels that are lower than the period before They claim that the rise in idiosyncratic volatility during the 1990s was an episodic phenomenon rather than a time trend. Because of the large influence of the U.S. on European financial markets, the decrease in cash flow risk in the late 2000s can also be expected in Europe Taxation The tax system on foreign-source income may also cause corporate cash holdings in Europe to be different than in the United States. The U.S. implements a credit system, which implies that tax liabilities in the host country are credited against taxes in the home country (De Mooij and Ederveen, 2003). The credit system is applied in combination with a deferral system. Profits of foreign affiliates that are reinvested are not taxed until they are repatriated to the parent company. Foley, Hartzell, Titman and Twite (2007) show that such a system gives firms an incentive to maintain earnings abroad, often in the form of cash holdings. They also show that this increase in foreign cash holdings is not offset with lower domestic cash holdings. 14

16 Most European countries have an exemption system to tax the foreign-source income of multinationals. 5 In an exemption system the foreign-source income of multinationals is exempt from home-country taxation. The firm only pays taxes to the host country and faces no taxes from the home country when it repatriates its foreign income. Therefore, unlike the deferral system, the exemption system does not create an incentive for firms to maintain cash balances abroad. According to the tax motive, U.S. multinationals have an incentive to maintain cash balances abroad. For European multinationals this incentive does not exist. Therefore one would expect taxation to be of minor importance in explaining corporate cash holdings in Europe Corporate governance Another difference between Europe and the U.S. can be found in the field of corporate governance. The legal protection of investors differs a lot across countries. La Porta, Lopez-de-Silanes, Shleifer and Vishny (1998) examine legal rules covering protection of corporate shareholders and creditors, the origin of these rules and the quality of their enforcement. Law and the quality of its enforcement are found to be important determinants of security holders rights and how well these rights are protected. La Porta, Lopez-de-Silanes, Shleifer and Vishny (1998) recognize that laws in different countries stem from a few broad legal families. First, there is the common law, which is English in origin. Second, there is the civil law, which derives from Roman law. Within civil law a distinction can be made between French, German and Scandinavian law. These four legal families have spread around the world, be it through imperialism or imitation. La Porta, Lopez-de-Silanes, Shleifer and Vishny (1998) classify countries to these four legal families in order to study differences in corporate governance between them. They find that legal origin can partly explain differences in legal protection of investors. Civil-law countries give investors weaker legal rights than common-law countries. This result is found both for shareholder rights and creditor rights. The French civil-law countries have the weakest protection of investors. Also the quality of law enforcement is higher in common-law countries than in French civil-law countries. 6 Again, the heterogeneity within Europe makes it difficult to draw a general conclusion on corporate governance in Europe. Wójcik (2006) shows that corporate governance differs a lot across European countries. For example, Scandinavian countries score well on measures of corporate governance, while for Southern European countries the scores are generally low. Nevertheless, it is not impossible to compare corporate governance in Europe with the United States for the following reason. The United States has a common law origin, while all European countries, apart from the United Kingdom, have a 5 Greece, Ireland, Spain and the United Kingdom have a credit system, while all other European countries have an exemption system (De Mooij and Ederveen, 2003). 6 La Porta, Lopez-de-Silanes, Shleifer and Vishny (1998) also show that the weak legal protection in civil-law countries results in higher ownership concentration in these countries compared to common-law countries. 15

17 civil law origin. Since La Porta, Lopez-de-Silanes, Shleifer and Vishny (1998) show that investor rights are better protected in common-law countries than in civil-law countries, it can be expected that corporate governance is worse in Europe than in the United States. 7 In a study on the relation between investor protection and corporate valuation they again find evidence that investor protection is stronger in countries with a common law origin than in countries with a civil law origin (La Porta, Lopez-de-Silanes, Shleifer and Vishny, 2002). According to the agency cost motive for holding cash, bad investor protection gives management the opportunity to use excess cash for its own benefits. Bates, Kahle and Stulz (2009) did not find evidence for agency problems in the United States. Following the theory of La Porta, Lopezde-Silanes, Shleifer and Vishny (1998), agency problems should be more severe in Europe than in the United States The crisis years The financial crisis erupted in August 2007 with the bursting of the housing bubble in the United States. Contaminated balance sheets of financial institutions and the excessive use of complex and opaque financial products were causing the effects to spread rapidly around the world (European Commission, 2009). By 2008 Europe suffered severe recessions, which continued in By 2010 most member states experienced at least a minimal degree of economic growth again (Cameron, 2010). It is hard to say in what year the financial crisis ended in Europe, because this is different for each country. Also for many countries the crisis is still not over. The economic contraction in the EU as a whole ended in the third quarter of 2009, but only in the second quarter of 2010 the growth rate in the EU returned to the level experienced before the crisis (Cameron, 2010). There are different channels through which the financial crisis can affect corporate cash holdings. First, the volatility of the cash flows increases in times of a crisis. Schwert (1989) shows that stock volatility increases during recessions and financial crises in the U.S. from 1834 to He states that volatility of stock returns reflects uncertainty about future cash flows and discount rates and is therefore an important business cycle indicator. Second, firms face more severe credit constraints during a crisis. Arslan, Florackis and Ozkan (2006) study the effects of a severe financial crisis in Turkey. 8 A financial crisis not only hampers the central function of financial markets, but also increases adverse selection and moral hazard problems. In times of a crisis the ability of firms to raise external finance is reduced, which makes firms more liquidity 7 Although a distinction can be made between, French, German and Scandinavian civil law, they all have weaker investor protection than common law (La Porta, Lopez-de-Silanes, Shleifer and Vishny, 1998). 8 Turkey suffered severe financial crises in November, 2000 and February, 2001 (Aslan, Florackis and Ozkan, 2006). 16

18 constrained. Arslan, Florackis and Ozkan (2006) find that during a financial crisis investments become more sensitive to internal funds. The precautionary motive for holding cash is therefore more prevalent in times of a crisis. Duchin, Ozbas and Sensoy (2010) find similar results for U.S. firms during the financial crisis. They show that the supply of external financing contracted during the crisis and as a result corporate investments were more sensitive to internal resources. Also Ivashina and Scharfstein (2008) find evidence for larger credit constraints during the crisis. They study bank lending during the financial crisis in 2008 and find that during the crisis lending fell across all types of loans. This decline in the supply of funding imposes larger credit constraints on firms. Finally, Campello, Graham and Harvey (2010) survey CFO s in the United States, Europe and Asia to assess credit constraints during the financial crisis. In the U.S. almost 90% of constrained CFO s stated that they had to forego profitable investment opportunities because they were unable to attract external finance. Similar results were found for Europe and Asia. The literature provides ample evidence that firms faced larger credit constraints during the financial crisis. The precautionary motive for holding cash predicts that firms maintain cash balances in order to be able to invest in profitable projects when credit is constrained. Also the fact that cash flow volatility increases during crises is a reason for firms to hold more precautionary cash. The expectation therefore is that during the financial crisis firms hold larger amounts of cash. 17

19 3. Data Description For the analysis I utilize a panel dataset that contains annual fundamentals of European firms for the years 1988 to A panel dataset has both a cross-sectional and a time series dimension, which allows for an analysis of different firms over time. Every firm-year observation contains information on cash, assets and all independent variables which are described later in this section. The data is gathered from the WRDS Compustat Global database. The dataset of Bates, Kahle and Stulz (2009) ranges from 1980 to However, they use the Compustat North America database. For the Compustat Global database there is only data available from 1987 onwards. Therefore I use a sample for the period from 1988 to The benefit from using this sample is that it allows me to analyze the effects of the financial crisis in the years from 2008 to The sample includes all surviving and nonsurviving firms for the sample period. The selection of European firms is based on their ISO country codes. 10 I select the firms that are incorporated in one of the 27 countries that are presently members of the European Union, which are Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden and the United Kingdom. 11 As conditional statements I require firms to have positive assets and positive sales for a given year to be included in the sample. 12 Following most empirical studies on cash holdings, I exclude financial firms (SIC-codes ), since there could be capital requirements that oblige these firms to maintain cash balances. 13 Also utilities (SIC-codes ) are excluded, since their cash holdings might be under regulation. This leaves a sample of 70,224 observations for 6,950 unique firms. Table A1 in the Appendix shows the number of observations and unique firms per country. The sample consists for the largest part of firms from Western European countries, which account for 50,973 observations and 5,006 firms. 14 Table A2 in the Appendix shows the data items that were gathered from Compustat Global. The problem with the Compustat Global database is that it does not contain share price data, which is an important variable in my research. Therefore I use Datastream to find the shareprices and number of common shares outstanding for the firms in the sample. I also use Datastream to gather yearly data about T-bill yields, inflation and exchange rates. Since the Compustat Global data and the share prices from 9 There are only 21 firm-year observations for the year 1987, which is therefore excluded from the sample. 10 The ISO (International Standards Organization) code identifies the country where the company headquarters is established. 11 Bulgaria and Romania are missing from the Compustat Global database, so the sample consists of firms from 25 EU countries. 12 Bates, Kahle and Stulz (2009) also use these requirements in the construction of their dataset. 13 SIC-code stands for Standard Industrial Classification Code. 14 Austria, Belgium, France, Germany, Ireland, Luxembourg, the Netherlands, and the United Kingdom are classified as Western European countries. 18