The effect of repatriations on US corporate cash holding

(1)

The effect of repatriations on US corporate cash holding.

This paper investigates the extent to which repatriations has a

positive effect on US corporate cash holding. An empirical analyse is

used on panel data over the time period 2000-2006. The cash holding

is explained by the repatriation ratio and a number of other control

variables. Main results find a positive effect of repatriation ratio on

US corporate cash holding, even for firm fixed effect control.

Lotte de Wit

10642056

Economie en Bedrijfskunde

Financiering en Organisatie

Ieva Sakalauskaite

30-01-2017

(2)

2

Statement of Originality

This document is written by Student Lotte de Wit who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

3. Research Methodology………..12 3.1 Financial Data………12 3.1.1 Sample Selection………..12 3.1.2 Variable Construction………12 3.1.3 Dependent variable.………..13 3.1.4 Independent variable….………..13 3.2 Descriptive statistics……….……….15 4. Results………16 4.1 Table descriptive………16 4.2 OLS regression……….17

4.2.1 Regression model column (1)………17

4.3 Results of independent variables………19

5. Conclusion………21

(4)

4

1. Introduction

In the past recent years, researchers remark a growing trend of US corporations holding a significant amount of cash reserves on their balance sheets. Economic researchers, like Foley et al. (2009), find that the magnitude of corporate cash holding is, in part, a consequence of the tax incentive faced by US multinationals. However, not much attention is paid in existing academic literature on the potential impact of incentives for cash holding created by taxes. But nowadays, with the increasing interest by the media, it is one of the most discussed topics in the financial world.

At the end of fiscal year 2015, US non-financial multinationals have left roughly 1.1 trillion dollars of foreign earnings overseas in effort to skirt tax charges of moving profit back to the US1. Leaders in this are the five tech giants like Apple, who amassed together 504 billion dollars of cash. The reason for this large amount of cash holding, is the federal income taxes of 35% on repatriating foreign earnings. The US and many other countries tax foreign operations of domestic firms and grants tax credits for foreign income taxes paid abroad. For most US affiliates and holding companies, these taxes are equal to the

difference between foreign income tax paid abroad and the tax payments that would be due if foreign earnings were taxed at the US tax rate. Because US affiliates and holding

companies are unincorporated in the US, there aren’t immediate taxed by the US

government. These foreign profits earned are only exposed to the federal taxation when earnings are repatriated back. Therefore, US affiliates and holding companies can deferred taxes until they are repatriated. These tax burdens create incentives for US non-financial multinationals to hold the cash in low tax jurisdictions, when investment opportunities are unattractive.

Not much research has been done on the potential impact of repatriations on corporate cash holding. The earliest explanations offered by academic research was based on transaction costs. Karni (1973) developed the argument that firms hold cash to avoid the cost of being short liquid. Opler et al. (1999) provide evidence for precautionary reasoning to hold cash. Firms are more likely to hold cash when risk increase and asymmetric information

1_{Forbes, May 8 2015: U.S. Companies are spending like crazy and still have a record of 1.73 trillion of dollars in}

(5)

5 or agency costs make it difficult for firms to raise external capital. Building on that, Bates et al. (2006) argued that increases in cash holding is a consequence of the increases in the precautionary motive for corporate cash holding. In particular, when cash flows are high and firms are constrained, firms have a higher cash flow sensitivity of cash (Almeida, 2004).

In this paper we explore the possibility that the tax costs associated with repatriations has an effect on the magnitude of cash holding for US non-financial

corporation. The broad question that we posit is: How does repatriating costs influence the

cash holding of US firms after it is repatriated? To investigate the effect of repatriation on

US corporate cash holding, the cash ratio coefficient is regressed on the repatriation ratio and other explanatory variables (like leverage and Tobin’s Q). For the analysis, panel data is used from the WRDS Compustat Fundamental Annual database. The sample consists of 1,691 unique US non-financial corporations with 9,310 firm-year observations during the period of 2000-2006.

The empirical analysis shows that there is a significant effect of repatriation ratio on corporate cash holding. In the regression model the effect is positive and significant, even for firm fixed effect control. Other findings indicate that the explanatory variables leverage, profitability ratio and Tobin’s Q have a significant effect on cash holding. When controlling for fixed effects only leverage remains, next to repatriation ratio, significant.

The paper is organized as follows. The following section provides a literature review were the motives for cash holding will be discussed, a brief summary is given of the US international tax code and the introduction of the Homeland Investment Act in 2004 (which had consequences for US corporate cash holding). In the third section the methodology, data and hypothesis will be presented. The empirical results are explained in section four and finally the fifth section concludes.

(6)

6

2. Literature Review

In this section of the paper we will discuss the existing literature with regards to the different determinants for corporate cash holding. To get a better understanding why US corporations hold cash abroad we look at the international tax policy for US multinationals and what changed for them after the introduction of the Homeland Investment Act in 2004 (HIA). In the end we build towards our hypothesis related to US corporate cash holding.

2.1 Determinants of corporate cash levels

Nowadays, perhaps one of the most interesting subjects around corporate cash holding in the financial world are ‘why does US corporations hold excessively large cash on their balances sheets?’. In the existing literature published, large cash holding is tightly related to two main reasons: precautionary motive and repatriation taxes (Bates, 2009 and Foley, 2007). The first main reason is precautionary. Firms hold cash and equivalent liquid assets, because it provides firms flexibility that they need in their transactions. Two main factors here are: uncertainty and credit constraints. Firm who facing future uncertainty may find it beneficial to hold a significant larger amount of cash for security. For example, when a firm faces an acquisition, a firm may want to hold cash able to use for fast acquiring. Bates et al. (2009) showed that this increase in cash-to-assets ratio was related to precautionary existing motives. They constructed a measure of cash flow uncertainty and showed that firms with higher uncertainty in their future cash flows, had higher cash-to-assets ratios. The second main reason is related to taxes. Foley et al. (2009) showed that the increase in cash holding by US multinationals is caused by repatriation taxes. The US government tax their citizens based on their worldwide income. US corporation who operate abroad, are determined to the taxes already paid abroad and the taxes the US would imply. Such taxes only arise when profits, which are hold abroad, are repatriated back to the US. Therefore, firms have

incentives to rather hold earnings abroad than repatriate them, in case when foreign investments are limited and profitability is high.

Next to the precautionary and tax motive there are two other traditional theories for the determination of corporate cash holding. In the paper of Opler et al. (1999) they viewed the use of classic finance models to derive the optimal demand for cash needed by

(7)

7 corporations who incurs transaction costs associated with converting non-financial assets into cash and use the cash for payment. This is known as the transaction motive. Jensen (1986) argued the last motive for cash holding. When managers face poor investment opportunities, they rather retain cash than increase pay-out to shareholders. This agency problem which arise when managers hold discretionary cash hold with the transaction- and precautionary motive.

In this paper, the focus relies on the tax motive related to corporate cash holding and the contribution of repatriation on the magnitude of domestic corporate cash holding. For most US affiliates and holding companies abroad taxes are equal to the difference between foreign income taxes paid and tax payments that would be due if foreign earnings were taxed at the US rate. These US taxes can be deferred until earnings are repatriated. Such tax burdens can create incentives for US multinationals to hold the retained earnings as cash abroad when investment opportunities are unattractive. This, because multinationals can choose whether they invest in liquid securities domestically or abroad. To get a better understanding of this financial decision making, we will have a closer look at the US tax rules that apply to international activities for US multinationals.

2.2 U.S. International Tax policy

Nearly all countries in the world tax the income of corporation that operate within their borders. Additionally, the US and other countries tax the foreign income of their residents. However, to avoid double taxation of foreign income, U.S. law grants tax credits for foreign income taxes paid abroad (Foley, 2007). For this reason, the US law offers opportunities for US multinationals to defer US taxation on certain foreign profits until they are repatriated (Foley, 2007). However, this deferral is only possible for foreign affiliates and holding companies of a US parent that are incorporated in foreign countries. Branches, which are unincorporated foreign affiliates, their profits are immediately taxed by the US. The taxes due upon repatriation are the difference between foreign income taxes paid and the payment that would be due if earnings were taxes at the US rate. An example to illustrate this: suppose the US federal income tax rate is 35% and a US affiliate or holding company earns 100 dollars abroad and pays 20 dollars income taxes in the host country. When the US

(8)

8 affiliate or holding company wants to repatriate these earnings back to the US, they have to pay an additional tax payment of 15 dollars. If foreign income taxes paid exceed the amount that would be due if earnings were taxed at the US rate, then no additional taxes are due (Foley, 2007).

However, there are important caveats for firms to deferral US tax avoidance. Foley et al. (2007) explain them briefly. First, Under the Subpart F provisions of the US law income classified as ‘passive income’ are ‘deemed distributed’ and therefore immediately taxable by the US even if such passive income is not repatriated. The term ‘passive income’ consists of interest income and dividend received from investment in securities. Firm classify foreign cash holding as being necessary for their business operations and thus not subject to passive income tax treatment. These rules provide incentives for firms to engage in avoidance behaviour and suggest that certain kind of firms are more likely to benefit from a tax holiday (Dharmapala, 2011). Firms are often better off holding earnings in a low tax jurisdiction in liquid securities rather than repatriating them. Even though these earnings are deemed distributed because they are passive income and therefore taxable. Second, US tax obligations are determined by worldwide averaging. This allows firms that pay tax rates above US tax rate in a jurisdiction, to use this foreign tax credit to shield income repatriated from low tax location from US taxation. However, these foreign tax credits cannot be used to reduce tax obligations related to income earned within the US (Foley, 2007). Last, firms that face the alternative minimum tax (AMT) have lower repatriation costs than firms who faces regular taxes as a consequence of the lower statutory rate applied under the AMT.

Despite the tax treatment of passive income, firms which can choose between invest earnings from a low income tax jurisdiction in cash at home or in the low income tax

jurisdiction, often have the incentive to hold this amount of cash in the low tax income jurisdiction (Dharmapala, 2011). Another example to illustrate this: an incorporated US affiliate earns 100 dollars and pays 20 dollars of foreign income taxes. If the US affiliate immediately want to repatriate earnings, additional 15 dollars income taxes must be paid and the US parent can invest the remaining 65 dollars in liquid securities in the US. However, when multinationals don’t repatriate the earnings, it will be able to invest 80 dollars in liquid securities abroad. Despite the requirements of Subpart F for the firm to pay US taxes on the

(9)

9 earnings for investment, the firm will be better off holding cash abroad and deferral taxes on the original 100 dollars. (Dharmapala, 2011). Foley et al. (2007) argued that this evidence shows that such incentives are important for explaining large cash holdings of firms.

Furthermore, Foley et al. (2009) finds that affiliates of the same US parent in

different countries who are facing different repatriation costs follow different cash-holding patterns. In the analyse they find that affiliates in countries with low tax rates, which would face higher repatriation costs, are more reluctant to bring foreign profits back to the US. For an example: If the US parent have two affiliates, one in France and one in Switzerland, and Switzerland has a lower tax rate than France, than the affiliate would bring less foreign profit back to the US in contrast to the affiliate based in France.

These are all causes for business lobbyists to search for tax breaks to repatriate foreign profits. With the introduction of the one-time tax holiday in 2004, an unexpected amount of cash was repatriated back to the US. More than 300 billions of dollars were repatriated relative to the average of 60 billion dollars over the previous five years

(Dharmapala, 2011). The tax holiday was originally created by president George W. Bush and the Congress when the US economy showed signs of weakness. The main idea behind the temporary tax reduction was to ensure multinationals to repatriate foreign profits back to the US and invest domestically.

Homeland Investment Act of 2004:

On October 22 in 2004, an Act was launched by congressman were firms were given a one-time tax holiday on the repatriation of foreign earnings by incorporated U.S. multinationals. The Homeland Investment Act (HIA) was part of the American Jobs Creation Act (AJCA), which has to create more than 500.000 jobs over 2 years by raising investments back to the US. The earnings returned would be used for ‘permitted investments’ which included debt repayment, finance capital spending, funding for research and development, venture capital and acquisitions (Dharmapala, 2011). The HIA allowed firms to deduct 85% of their

repatriations from additional US taxes beginning in 2005, but were still due on the remaining 15%. Under the HIA firms continued to receive tax credits for foreign income taxes paid for these profits. Therefore, if U.S. multinationals earned 100 dollar of income abroad and paid

(10)

10 15 dollars to host income taxes, under the HIA 85% of foreign earnings would be exempt from U.S. repatriation taxes. Firms only had to pay 15% in US taxes on the remaining 15 dollars in earnings. The tax burden will be only 2.25 dollars (15% times 15 dollars) to repatriate instead of the original 15 dollars. If firms didn’t take the benefit to repatriate under the HIA by the end of the accounting year or year after the HIA was passed, the lower tax advantage was not available for future years.

Several restrictions have to be met for repatriations to be qualified for the HIA (Dharmapala, 2011). The first restriction was that repatriations had to be paid in cash. This restriction required affiliates who already invested their earnings in real assets and had low cash reserves to raise cash. The second restriction was that repatriations couldn’t exceed: 1) 500 million, 2) the earnings reported as permanently reinvested on the financial statement before June 30 in 2003, or 3) the amount of cash affiliates had historically repatriated to its US parent. The amount of qualifying repatriations was reduced with the total amount of debt outstanding form foreign subsidiary to related parties and by the amount of the increase in related-party debt between US multinationals and its foreign affiliate or holding company. This, to prevent companies from borrowing abroad of their U.S. parent in order to fund repatriations at the lower tax rate.

Firms responses to the HIA provide an opportunity to look at financial constraints, corporate governance and international tax policy (Dharmapala, 2011). The temporary tax holiday did effectively reduce the cost for US multinationals to repatriate foreign profits which was hold as reinvested earnings. The framers of the Act justified the tax holiday on the premise that US multinationals were financially constrained. If this premise was true, repatriated earnings could be used to invest in domestic projects that otherwise were not profitable at the higher cost of external capital (Bates, 2009). Looking at the existing literature, Hubbard (1998) reviews literature on financial constrained firms. Multinationals are in general though be less financially constrained than other firms. Therefore,

multinationals choose optimal levels of investment and employment before the tax-holiday, so when the tax holiday occurred they would not have to increase expenditures on capital and labour. The outcome of the act was therefore, for well-governed and not financially constrained multinationals, that any internal capital accessed under the HIA was returned to

(11)

11 shareholders. Furthermore, the response of multinationals to the HIA shows important implications. Altshuler and Grubert (2003) argued in their paper that US multinationals are able to use tax planning strategies that allows them, even in the absence of a tax holiday, to avoid repatriation taxes. In this paper, affiliates are limited to two alternatives: to repatriate dividends to the US parent or invest in its own real operations. If the affiliate achieves the equivalent of tax-free repatriation in the low tax subsidiary, they do not have to underinvest to obtain the benefits of tax deferral. Desai et al. (2004) found significance for repatriation taxes on repatriation decisions. Despite the availability to escape from the repatriation taxes, the repatriation taxes do impose burdens.

The US treasury Department issued explicit guidelines on how returned repatriated earnings should be spend. The retained earnings would be spent on research and

development, employment, visa versa. Certain uses, like executive compensation, stock redemptions and dividend would disqualify repatriations under the tax holiday (Dharmapala, 2011). Results from the HIA imply that almost all of returned repatriated cash went to the shareholders. A use that was disqualified for repatriations. The effects of the HIA were contradicting for which they were intended. In 2011 a permanent subcommittee on

investigations of the US senate, presented the actual expenditures from repatriations after the HIA by corporations. As a result, companies reduced their workforces, no further

investment in research and development was made, executives were compensated and cash balances radically increased2_.

This paper analyses the effect of the repatriation ratio on corporate cash holding. To be consistent with the explicit guidelines of the HIA, returned repatriated earnings had to be invested domestically. But results show, that after the tax holiday domestic cash balances sheets largely increased. Therefore, the effect of repatriation ratio should be positive related to corporate cash holding. To analyse this the following hypothesis is tested: the repatriation ratio has a positive effect on the corporate cash holding in the US.

2_{In 2011, US corporations show a record on domestic cash assets around 2 trillion dollars on their balance}

sheets. Indicating that the availability of cash is not constraint to hiring or domestic investment decisions. Allowing corporations to repatriate cash from abroad would be an ineffective way to create new jobs (Bean et al. 2011).

(12)

12

3. Methodology 3.1. Financial Data 3.1.1. Sample Selection

The financial data used for the analysis of cash holding is drawn from the Compustat Fundamental Annual database, which contains information about public companies in the US. For the panel data analyse, it covers the period of 2000-2006. For this time period we have to make some exclusion. First, only include U.S. public companies which have at least 100 million dollars in assets and are incorporated in the US. This, because international repatriation tax obligations do not apply for unincorporated companies (like branches). Second, eliminate financial firms (SIC 6000-6999) and utilities (SIC 4900-4949) from the data. These companies are subjected to regulatory supervision and constraints. They hold cash for other reasoning than economic reasons studied here. Third, we eliminate all firm year observations that show missing data for the financial information. The final sample consists of 1,691 unique firms with 9,310 firm-year observations over the period 2000-2006.

3.1.2. Variable construction

The collected financial data is used to fit in a multiple regression model that estimated the relationship between a company’s cash holding, repatriation ratio and firm characteristics to predict the expected level of a company’s cash holding. The data has been segregated for companies with foreign activities (multinationals) and non-foreign operating companies. Too distinct these different firms, the control variable Research and Development is included in the multiple regression. One of the firm characteristics of multinationals is the great

expenditure on research and development in contrast with non-foreign operating firms. Therefore, we use the variable Research and Development ratio as an interaction variable with Repatriation ratio to distinct multinationals from non-foreign operating firms

(RepatriationxR&D). For the analysis we use the cash ratio as the dependent variable, which is most widely and traditional measured as the cash scaled by total assets ratio3.

31 _{The Compustat Fundamental Annual data which is used in calculating the cash ratio is defined as: Cash and}

(13)

13 The following model is estimated to analyse the empirical model:

(𝐶𝑎𝑠ℎ 𝑟𝑎𝑡𝑖𝑜)𝑖,𝑡

= 𝛽0+ 𝛽1∗ 𝑅𝑒𝑝𝑎𝑡𝑟𝑖𝑎𝑡𝑖𝑜𝑛𝑖,𝑡+ 𝛽2∗ 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡+ 𝛽3∗ 𝑇𝑜𝑏𝑖𝑛′𝑠𝑄𝑖,𝑡+ 𝛽4

∗ 𝑃𝑟𝑜𝑓𝑖𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦 + 𝛽5∗ 𝑅&𝐷 + 𝛽6∗ 𝑅𝑒𝑝𝑎𝑡𝑟𝑖𝑎𝑡𝑖𝑜𝑛𝑥𝑅&𝐷𝑖,𝑡+ 𝜀𝑖,𝑡

Subscript i denote the firms and the subscript t indicates for year. The included error term in the model is normally distributed and controls for the difference in equations during the empirical analysis.

3.1.3 Dependent variable.

The cash ratio coefficient is used as the dependent variable (Cash ratio) for the model. In the existing literature one of the main reasons why firms hold large amounts of cash was

because of the repatriation costs. Under the HIA repatriation costs decrease and US multinationals repatriated a large amount of cash back to the US. The requirements for firms, under the HIA, was to invest the cash into Research and Development, investment or hiring workers. For the empirical research the model predicts how much cash that has been repatriated, is hold in cash domestically.

3.1.4 Independent variables.

The independent variables used in the explanatory model for predicting the cash ratio are mainly disclosed in existing literature by Bates et al. (2009), Foley et al. (2007) and

Dharmapala (2009). The variables used and described below:

1. Leverage ratio – is affected by firm decisions. Opeler et al. (1999) and Kim et al. (1998) find for cash holding there are two effects. When firms can borrow easily from the external markets they are not concerned to hold a large amount of cash for precautionary reasons. But, on the other hand when leverage increases, the financial distress costs also increase which results in accumulating sufficient amounts of cash. Therefor leverage can have a positive or negative relation to cash holding. Bates et al. (2009) build on that debt is sufficiently constraining, so firms will use cash to reduce leverage which results in a negative relation according to cash holding. Archarya et

(14)

14 al. (2007) came with a hedging argument which …The leverage ratio is defined by Total Debt divided by Total Assets.

2. Tobin’s Q – Here used as a proxy for growth opportunities and asymmetric

information. Growth firms are expected to have a larger Tobin’s Q, because they are surrounded by relatively larger information asymmetries. Such companies have more value in intangible assets (like goodwill, patents and Research & Development) and therefore find it harder to raise external capital. As a result, they hold a larger amount of cash. Tobin’s Q is defined by the Market Value of Equity scaled by Total Assets.

3. Profitability ratio – The theoretical predictions regarding to cash holding are again doubles sided. On the one hand, in a world pecking order firms use profits to build up liquidity and tend to hold more cash (Opeler et al., 1999). On the other hand, profits provide an immediate source of liquidity. If cash and profits are substitutes (firms use cash to repay debt), there should be a negative relationship (Kim et al., 1998). The profitability ratio is defined by the Net Income divided by Total Assets.

4. Research and Development ratio – Is use as a proxy for growth opportunities and information asymmetries. Firms with higher R&D have larger profitability of default and faces a higher potential cost of financial distress. Therefore, companies with high R&D are expected to hold a larger amount of cash in reserve. Research and

Development is defined by the Research and development Expenses divided by Total Assets.

5. Repatriation ratio – The repatriation ratio here will be defined by the repatriation of foreign earnings from affiliates and holding companies to the US parent scaled by the total sum of firms Total Assets.

6. RepatriationxR&D – Is an interaction variable created between the control variables:

Repatriation ratio multiplied Research and Development ratio. This interaction

(15)

15 To control for fixed effects and potential endogeneity issues that are not captured by other variables, the variable SIC is included. In the panel data analysis, regression in column (3) control for firm fixed effects is done. This, to investigate the impact of the independent variable on the dependent variable over the industries, while controlling for time invariant effects.

3.2 Descriptive Statistics

All the variables discussed so far are summarized in Table 2. To deal with the distorting effects of outliers in the regression, large outliers are excluded from the data. The dependent variable, cash-to-assets ratio, is defined as cash and short-term investments scaled by total assets. The cash-to-assets ratio exhibits a mean of 0.21 and a standard deviation of 1.43. There is an extremely wide variation within the sample as indicated by the standard deviation of the cash ratio. Looking at the predictor variables, there is a wide dispersion as well. The Tobin’s Q has a mean of 1.22 and a standard deviation of 1.01 and

Research and Development ratio shows a mean of 0.1779 and a standard deviation of 1.495.

The independent variable repatriation ratio shows a mean of 0.14 and a standard deviation of 0.10.

The findings from our sample, respective to the cash-to-asset ratio, shows a relatively higher mean and standard deviation than documented in the existing literature. Dharmapala et al. (2011) who also use this sample, but only for the fiscal year 2004, shows a mean of 11.3% and a standard deviation of 13.98%. The difference in magnitude of the mean and standard deviations of the cash ratio might attribute that we only use data from Compustat instead of Compustat and the BEA database. Also we use a different time period.

In order to examine the potential relationship between firm characteristics and the probability of holding cash it is important to create a model that is capable of predicting cash holding after repatriation for non-foreign operating firms and multinational firms. Reliable models are delivered by papers as Opler et al. (1999) and Kim et al. (1998) about research on corporate cash holding. To investigate the firm characteristics and repatriation we use a simple OLS regression. The following specification will be employed and fitted for the whole 1,691 firms (i) sample and 9,310 Firm year observations (t) for the period of 2000-2006.

(16)

16 Before analysing the effect of the dependent variable on the independent variables, the correlation between the explanatory variables are evaluated. These correlations for the regression are shown in Table 1.

Table 1: Correlation Matrix

Cash Repatriation Leverage Tobin’s Q Profitability R&D RepatriationxR&D

Cash 1 Repatriation 0.0314 1 Leverage -0.0384 -0.0167 1 Tobin’s Q 0.0664 0.0173 -0.4589 1 Profitability -0.0137 0.0057 -0.1374 0.1386 1 R&D -0.002 -0.026 0.0069 -0.0243 0.0692 1 RepatriationxR&D -0.0016 0.0195 0.0078 -0.0253 0.0565 0.9578 1

Multicollinearity problems may arise when there is high correlation between explanatory variables. The correlations shown in Table 1 are all correlated around the value 0.14 or lower, which indicates that from these correlations no multicollinearity should be expected. Only the interaction variable RepatriationxR&D ratio shows a high correlation ratio, but this is logical. Interactions terms and its components are usually correlated and so are correct. Because of the low correlation between the other explanatory variables, it can be expected that no multicollinearity will arise in the analysis and the explanatory variables are

independent.

4. Results

In this section, the results of the empirical analyse are provided. First, only the results of the independent variable on the dependent variable are provided. Second, the results of the explanatory variables on the dependent variable are showed. Last, the results of the explanatory variables on the dependent variable are provided controlling for fixed effects.

4.1. Table descriptive

The results of the panel data analysis are provided in Table 2. The model contains the dependent variable cash-to-assets ratio regressed against the explanatory variable. In column (1) only the independent variable repatriation ratio is included and regressed against the dependent variable. In column (2) the dependent variable is regressed against all

(17)

17 explanatory variables. Last, in column (3) the dependent variable will be regressed against the explanatory variable controlling for fixed effects. For each estimated coefficient, the standard deviation is provided below in parentheses. The table also shows the R-squared, which explains how much the variance in the dependent variable can be explained by the variance in the independent variable. Below the R-squared the values are provided. The F-value indicates the significance of the model. These F-F-values are based on clustered

standard errors. The reason for clustering is that there might be correlation between standard errors of industries, nut not across industries.

Table 2: results empirical analyze Dependent variable Cash ratio

Independent variables (1) (2) (3) Repatriation ratio 0.486*** (0.15) 0.457** (0.16) 0.416** (0.16) Leverage -0.191* (0.11) -0.494** (0.21) Tobin’s Q 0.041*** (0.01) (0.006) -0.001 R&D ratio 0.001 (0.01) 0.008 (0.01) Profitability ratio -0.165** (0.07) 0.036 (0.04) RepatriationxR&D -0.014 (0.11) -0.069 (0.11) Constant 0.164*** (0.02) 0.178*** (0.05) (0.09) 0.348

Firm Fixed Effects No No Yes

N 9,310 9,310 9,310

R-squared 0.0011 0.0058 0.0037

F-value 10.23** 70.26*** 5.60***

df 1, 1690 6, 1690 6, 1690

Robust standard errors are parentheses below the estimated coefficients. Significance is indicated as follows: * if p<0.10, ** if p<0.05 and *** if p<0.01. Firm fixed affects only included in column (3). N is the total

observations. R-squared indicates the variance in the dependent variable by the variance of the independent variable(s). F-values indicates the significance of the models, with below the degrees of freedom used. 4.2. OLS Regression

4.2.1 Regression model column (1).

In the first column of Table 2, only the explanatory variable repatriation ratio is included. The estimated coefficient is positive related to the cash ratio. This indicates when

(18)

18 repatriations increase firm cash holding will increase. Furthermore, this positive effect is significant, as is the model as a whole. Also the Wald-value for the model is low en shows significance for a level of 0.0014. To find more information about the different effect for the effects of repatriation on cash holding between multinationals and non-foreign operating firms, we include more variables in column (2).

4.2.2 Regression model column (2)

In the second column (2), more explanatory variables are included in the model. The model is significant for a level of 0.01. The variables repatriation ratio and Tobin’s Q are positive related to the cash ratio and significant at a level of 0.01. Profitability ratio is negatively related and shows significance at the level of 0.05. Further, is Leverage negatively related and significant for a level of 0.10. Even if the R-squared value of the model is low, variables show significance at certain levels. It indicates that the effect size is small, but isn’t bad or useless.

4.2.3 Regression model column (3)

In column 3, the dependent variable and the explanatory variables are tests controlling for fixed effects. With the use of fixed effects, the impact of the independent variables on the dependent variable can be controlled. The results in column (3) present the effect of repatriation ratio and the firm characteristics on the cash-to-asset ratio. The estimated repatriation ratio coefficient shows a positive and significant effect at a level of 0.01 on the cash ratio. Also, the estimated coefficient of leverage indicates a negative but significant effect at a level of 0.05. The other estimated coefficients of the explanatory variables show no significance at a level of 0.10 when the model is controlled for fixed effects. The fixed effects have removed the time-invariant characteristics and assess the net effect of the explanatory variables on the dependent variable. The R-squared of the model is lower relative to the previous model. The reason behind this is when controlling for fixed effects, the explanatory effects of the intercepts are removed. In the previous regression in column (2) the explanatory effects aren’t removed which result in a higher R-squared.

(19)

19

4.3 Results of independent variables

Repatriation ratio:

The estimated coefficients for the independent variable Repatriation ratio in columns (1), (2) and (3) are respectively 0.49, 0.46 and 0.42 positive related to the cash ratio. For simplicity, the coefficients result all around 0.45. Which indicates that an increase of 1 in the

repatriation ratio lead to an increase of 0.46 for the cash-to-asset ratio. According to the existing literature, Foley et al (2007) find that if lower repatriation tax burdens decrease cash holding, firms should hold less cash abroad. Since cash held abroad is a substitute for cash held domestically, then lower repatriation tax burdens should increase domestic cash holding. The coefficient of the variable repatriation ratio is there for consistent with the literature.

Leverage:

The estimated coefficient of leverage in column (2) and (3) are respectively -0.19 and -0.49. These outcomes are negatively related to the dependent variable. It implies that if leverage increase by 1, cash-to-asset ratio decrease by -0.19 and -0.49. For the model in column (2), leverage is significant at a level of 0.10 and in column (3) for the level of 0.05. Opeler et al (1999) and Kim et al. (1999) report the possibility of a negative relationship between leverage and cash holding. The reason for this is that debt and cash can act as substitutes and firms with lower leverage are tend to hold higher cash and vice versa.

Tobin’s Q

The control variables Tobin’s Q provided in column (2) and (3) are respectively 0.041 and -0.001. The estimated coefficient in column (2), shows a positive relationship related to the cash ratio. Which indicate an increase in Tobin’s Q lead to an increase in cash holding. Furthermore, the coefficient shows significance at a level of 0.10. The positive relationship between the cash ratio and the Tobin’s Q ratio, can be interpreted that firms with high growth prefer to use own internal funding sources to finance their growth projects and therefore they hold more cash. This falls in line with the pecking order theory (financing hierarchy theory) were external financiers mitigate the risk of growth projects by imposing a

(20)

20 higher required return, and therefore a higher cost of capital for the firm. In column (3) the estimated coefficient of Tobin’s Q is negatively and significant related to the cash ratio when fixed effects are included. This result indicates that there is a decrease in the cash ratio coefficient over time, which is the effect of time-varying factors that are not captured by the included explanatory variables.

Research and Development:

The estimated coefficients for Research and Development ratio are respectively positive related to the cash-to-asset ratio provided in column (2) and (3). Which indicates an increase in Research and Development ratio leads to an increase in cash ratio. This falls in line with existing literature, where firms with high R&D expenses are expected to hold a larger amount of cash on their balance sheet, because of the larger profitability of default and therefore they face a higher potential cost of financial distress (Bates, 2009). Opeler et al. (1999) and Bates et al. (2006) also document in their literature for the positive relationship between cash holding and Tobin’s Q and Research and Development ratio. The positive relationship between the Research and Development expenses and the acquisition expenses are partly evidence in favour of the precautionary and transaction motive relative to

accumulating cash. Firms try to avoid the instance for being short liquid when they need to fund investments, especially when future cash flows are uncertainly high.

Profitability ratio:

The control variable profitability ratio has a negative relationship and significance related the dependent variable cash-to-assets ratio in column (2). If the profitability ratio increases with 1, the cash-to-asset ratio decreases with -0.19. According to the literature, Kim et al (1999) give an explanation for the substitute effect of cash and profits. The profits are used to repay debt and therefore reduce cash holding. In column (3) there is a positive and insignificance effect related to the cash ratio when controlling for fixed effects. This results indicates that there is an increase in the cash holding ratio over time. According to the literature, the profitability ratio can be positive when firms use profits to build up liquidity and therefore hold more cash (Opeler et al., 1999).

(21)

21 RepatriationxR&D:

The interaction explanatory variable RepatriationxR&D ratio results in negative and

insignificance effects related to cash holding in column (2) and (3). This interaction variable is included in the regression too distinct the multinationals from the non-foreign operating firms. The negative coefficient indicates that an increase leads to a decrease in cash holding. According to Bates et al. (2006) and Dharmapala et al. (2009), when tax cost burdens

decreases, US multinationals who have operating activities abroad, have an incentive to repatriate earnings. When under the HIA the federal income tax decreases, multinationals repatriate for more than 300 billion dollars in cash back to the US. When earnings are repatriated back, multinationals couldn’t hold it in cash on their balance sheet because this was disqualified under the HIA.

5. Conclusion

This paper investigates the relationship between repatriation ratio and the corporate cash holding for non-financial firms in the US. To investigate this effect, an empirical analysis is used for the panel data. The main findings from the analysis on the annual data are, that the coefficient of repatriation is positive and significant in all the estimated models. In the first model, only the independent variable repatriation ratio is included and showed a positive coefficient and significance at a level of 0.01. In the following two models, in which one model is controlled for firm fixed effects, repatriation showed also a positive and significant effect on cash holding for a level of 0.01. These results connect with existing literature, were repatriation costs influence financial decision making according to corporate cash holding (Foley, 2007 and Bates, 2009). Other findings in the empirical analyse, showed significance in model two for leverage ratio, profitability ratio and Tobin’s Q. In model three, when

variables are controlled for firm fixed effects, only leverage remained significant. Empirical results for multinationals showed for both models a negative coefficient related to corporate cash holding and insignificance. This consist with the literature of Darmapala (2011), were multinationals were not allowed to hold cash when they repatriated earnings back under the HIA but had to invest it. Later results find that US corporations on the contrary, hold the cash in large cash reserves on their balance sheet. From the empirical analysis it can be concluded

(22)

22 that the hypothesized positive effect of repatriation on corporate cash holding is supported in this sample.

For further research on the impact of repatriation on corporate cash holding, more databases could be used. In this empirical analyse only Compustat is used, but a lot of research can be done in combination with other databases. For example, the BEA index (Bureau of Economic Analyse). In the BEA database you can find which US parent has affiliates and holding companies and where they are located. Therefore, research can be done on specific industries and their foreign cash holding and repatriations. To improve the empirical model more explanatory variables could be included and the time period could be expanded. The question if there will ever be another tax holiday in the future is uncertain.

(23)

23

Bibliography

Almeida, H., Campello, M., Weisbach, M.S. (2004). The cash flow sensitivity of cash. Journal

of Finance 59, 1777-1804

Altshuler. R., Grubert, H. (2003) Repatriation taxes, repatriation strategies and multinationals financial policy, Journal of Public Economics 87, 73-107

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, 19845-2021

Bean, E.J., Roach, R.L., Murphy, A.F., Johns, M.F., Martineau, M.J., Robertson, M.D., (2011) Repatriating offshore funds: 2004 tax windfall for select multinationals. Washington, D.C.: US Government Printing Office

Desai, M.A., Foley, F.C., Hines, J.R., (2004) A multinational perspective on capital structure choice and internal capital markets, Journal of Finance 6, 2451-2487

Dhammika Dharmapala, C. F. (2011) Watch What I Do, Not What I Say: The Unintended Consequences of the Homeland Investment Act, Journal of Finance 66, 753-787 Foley, F. C., Kahle, K. M., Stulz, R. M., (2007) Why do firms hold so much cash? A tax-based

explanation, Journal of Financial Economics 86, 579-607

Hubbard, R. Glenn. (1998) Capital market imperfections and investment, Journal of Economic

Literature 36,193-225

Karni, E. (1973):The transaction demand for cash: incorporation of the value of time into the inventory approach. Journal of Political Economy 81, 1216-1225

Kim, C., Mauer, D.C., Sherman, A.E. (1998) The determinants of corporate liquidity: Theory and evidence. Journal of Financial and Quantitative Analysis 33, 335-358

Opler, T., Pinkowitz, L., Stulz, R., Williamson, R. (1999) The determinants and implications of corporate cash holding. Journal of Finance economics 52, 3-46

Pinkowitz, L., Stulz, R.M., Williamsom, R. (2013) Multinationals and the High Cash Holding

Puzzle (NBER Working Paper No. 18120). Retrieved from National Bureau of

The effect of repatriations on US corporate cash holding