Corporate cash holdings, cash holding
policies and the financial crisis
ABSTRACT
This study researches whether changes in firm characteristics explain the increase in cash holdings or the change in the demand function for cash explains the increase in cash holdings
during the financial crisis. Using a sample of Chinese and U.S. firms in the period between 2001-2013, I find that U.S. firms and Chinese firms share most of the determinants of corporate cash holdings. My results show that the financial crisis has changed both the firm characteristics and the cash holdings policies of Chinese and U.S. firms. It turns out that the changes in cash holdings cannot be considered as a pure unexplained shift in demand for cash.
Date: 16.01.2015 Student number : s1879723
Name: Rico van der Vaart Study Program: Msc Finance
1.
Introduction
Nowadays, corporations around the world have high amounts of corporate cash holdings. Studies like Bates, Kahle, and Stulz (2009) find a secular increase in cash holdings for U.S. firms over the period of 1980 to 2006. Understanding the phenomenon of high corporate cash holdings helps to find reasons for the slow recovery from the recession. This because hoarding of cash by large firms is a big problem for governments since it slows down economic growth. Academics argue that firms with large stockpiles of cash overinvest and make poor acquisitions. This affects shareholders value (see e.g., Dittmar and Mahrt-Smith, 2007; Harford, Mansi, and Maxwell, 2008) But why are firms from all around the world sitting on a mountain of cash?
Bates, Kahle, and Stulz (2009) find that most of the increase in cash holdings are explained by changes in firm characteristics. For instance, inventories and capital expenditures have fallen during the period of 1980 to 2006. The crisis can disturb the demand function for cash. So , I am wondering whether the firm characteristics can still explain changes in cash holdings. Subsequent, I research whether this U.S. phenomenon in cash holdings is systemically shared by China
External finance has become more expensive for companies during the financial crisis. As a consequence investments may slow down for firms which have no sufficient slack to fund all positive net present value (NPV) projects internally (see e.g., Campello, Graham, and Harvey, 2010; Duchin, Ozbas and Sensoy, 2010). The financial crisis in the second half of 2008 had dramatic effects on the financial sector globally. Banks have retreated from lending and Harford, Klasa, and Maxwell (2014) find for example that refinancing risk of debt has increased for firms. Campello, Graham and Harvery (2009) find evidence that firms cancel profitable investment opportunities during the financial crisis. This is a result of binding external financing constraints. Therefore, high cash holdings during the crisis and just after the crisis might be beneficial for shareholders when capital markets tightens and management use these cash holdings to invest in NPV projects according to the precautionary motive.
problems and currency volatility. Brazil, Russia, India, China and South Africa (BRICS) are the fastest growing emerging countries. They are called the BRICS countries together. Several studies hypothesizing that firms in countries with less developed credit markets should have more cash according to the precautionary motive of cash holdings ( see, e.g., Dittmar, Mahr-Smith, and Servaes, 2003; Kalcheva and Lins, 2007). The U.S. has the most developed credit and stock market compared to the BRICS countries, “see Appendix A”. Only South Africa has a more developed stock market of the BRICS countries. Firms from countries relying on credit from banks might have used up their excess cash to offset restrictions on bank credit during the financial crisis. U.S. firms rely more on the credit market than on the stock market compared to the BRICS countries, “see Appendix A”. Other characteristics that are common for emerging countries and uncommon for developed countries are high values on the GINI coefficient. This is a measure of income inequality and low values for gross domestic product (GDP) per capita. Firms in a country with higher growth opportunities have more investment opportunities. Companies from emerging markets have been able to seize more growth opportunities than companies from developed economies. Fig. 1 represents the difference in annual GDP growth between the U.S. and the BRICS countries over the period of 2001 to 2013. Especially in an economic environment with large growth opportunities, such as the BRICS, the potential impact of a breakdown in the financial sector could be enormous. Emerging economies such as the BRICS countries are increasingly gaining importance in global trade. There is a fall in the contribution to global trade by the U.S.. On contrary, the BRICS countries witness an increase in contribution to global trade over the years. China has been the major player in the BRICS economies for years now. Fig. 2 shows the trend in the contribution for the BRICS and the U.S. over the period of 2001 to 2012.
Fig. 1. Annual gross domestic product growth.
Figure represents the aggregate annual GDP growth of the BRICS countries together and the annual GDP growth of the U.S.. China has the greatest contribution to annual GDP growth for the BRICS ( mean of 9,80% over period 2001-2013) (source: World Bank’s World Development Indicators).
Fig. 2 Contribution to trade in the world.
100% contribution to trade is based on the total trade of the BRICS and the U.S. together. Trade is measured based on merchandize export in U.S. dollars. (source: World Bank’s World Development Indicators).
A macroeconomic shock might have changed this trend and disturbed the demand function for cash. Therefore, it is interesting to see whether the secular increase in cash holdings till 2006, found by Bates, Kahle, and Stulz (2009), is ongoing in the crisis and whether this increase in cash holdings is still caused by changes in firm characteristics. Another interesting focus in this paper is based on the believes of economists who argue that the economies of emerging countries will be those that will boost the world economy. The international monetary fund forecasts that the U.S. still has supremacy from the economic point of view in 2025, unless China comes closer rapidly. China will have the largest economy in 2050. China is chosen because it has the highest growth opportunities from the BRICS countries.
Nevertheless, there have been relatively few papers published about cash holdings based on non U.S. data, while the emerging countries’ financial systems, values and objectives sometimes conflict with the Western world.
The methods used for the analysis in this research draws on the methods used by Bates, Kahle, and Stulz (2009). I will test whether there is still a secular increase in cash holdings when the time interval includes the financial crisis. I will also compare the cash holdings of U.S. firms with the cash holdings of Chinese firms. Next to this I test with several dummy variables whether firm characteristics are still the determining factors for explaining the secular increase in cash holdings during the crisis or whether the demand function has changed for cash holdings. So, I will test whether cash holdings changed because firms moved along the demand curve for cash or because the demand curve shifted. Again I will do this for U.S. and Chinese firms. The term “cash holding policy” is interchangeably used with the term “demand function for cash” throughout this study.
The following questions are therefore leading in this study: Main research question:
• What is the effect of the financial crisis on cash holdings and cash holding policies for Chinese and U.S. firms?
Three sub question:
• To what extent do Chinese and U.S. firms share the same determinants of corporate cash holdings?
• To what extent is the change in cash holdings of Chinese and U.S firms during the financial crisis a result from changes in firm characteristics?
• To what extent is the change in cash holdings of U.S. and Chinese firms during the financial crisis a result of a shift in demand for cash holdings?
The dataset is comprised of firm-year observations from China and the U.S. from 2001-2013. I use 2008 as the start of the financial crisis because both China and the U.S. witnessed a significant decline in GDP growth that year. Moreover, the cash holdings of both countries are significant different from each other for the periods 2001-2007 and 2008-2013. I first examine the trend in U.S. and Chinese cash holdings before and after the financial crisis has started. The average cash holdings of Chinese firms are 17.2% before the beginning of the crisis and 21.7% after the beginning of the crisis. The average cash holdings of U.S. firms before the crisis are 18.5%. After the beginning of the crisis is the average cash holding increased to 20.2%. Bates et al. (2009) find that the secular increase in cash holdings is mostly explained by changing firm characteristics. They conclude that the cash holdings changed because firms move along the demand curve for cash. According to my research, the change in firm characteristics and change in the demand function for cash explain partially the change in cash holdings after the start of the financial crisis.
2.
Literature study
2.1. Imperfect markets and cash holdings
2.2. Cash holding trend
Bates, Kahle, and Stulz (2009) show a secular increase in the cash holdings of U.S. firms from 1980 to 2006. They report that the average cash-to-assets ratio has increased from 10.5% in 1980 to 23.2% in 2006 for U.S. firms. Another finding of their study is that the increase can mainly be explained by the change in firm characteristics over the years and less by changes in the relation between firm characteristics and cash holdings. Iskandar-Datta and Jia (2012) find a pronounced secular upward trend in cash holdings across seven industrialized countries over the period of 1991 to 2008. This trend is true for six out of seven industrialized countries in their sample. Only Japanese firms experience a decline in cash holdings in this period. Iskandar-Datta and Jia (2012) argue that in the Japanese case the cash holdings were quite high in the 1980s and that the cash holdings policy of Japanese firms have become more economically oriented towards the 2000s. Azar, Kagy, and Schmalz (2014) find other evidence for the increase in cash holdings of U.S. firms. They also studied the cash holding puzzle in the U.S. and find a negative effect of opportunity costs on cash holdings of U.S. firms in the period between 1945-2011. They argue that the current U.S. corporate cash holdings are not abnormal. Their finding is that the interest rates were close to zero in 1945, had a maximum of 15.0% in 1980, and are back close to zero in 2011. The trend in interest rates is opposed to the trend in the level of cash holdings. Therefore, the opportunity costs of holding cash were much higher around 1980 than they were around 1950 or 2011. Song and Lee (2012) find that East Asian firms increased median cash holdings from 6.7% in 1996 to 12.1% in 2006 and mean cash holdings from 11.0% in 1996 to 16.0% in 2006. The increase in cash holdings is after the Asian financial crisis of 1997-1998. Song and Lee (2012) did not find the secular increase in cash holdings before 1996, documented by Bates, Kahle, and Stulz (2009) for U.S. firms, in the sample of Asian firms. Most studies find an increasing trend in cash holdings. Therefore, I expect also higher cash holdings for U.S. and Chinese firms in the sample period of 2001-2013.
2.3. Adverse macroeconomic shocks, cash holdings and investments
has to convert cash substitutes into cash (see e.g., Miller and Orr, 1966). There are several possibilities for firms to raise cash. Firms can raise for example funds in the capital markets, liquidating existing assets, reducing dividends and investing or renegotiating existing financial contracts. This conversion costs money. In some situations these costs can be extremely high for firms. The precautionary motive arises then from the fact that firms hold cash to withstand future cash shortfalls when access to capital markets is costly (see e.g., Opler, Pinkowitz, Stulz and Williamson, 1999). This motive also suggests that firms with better investment opportunities hold more cash because adverse shocks and financial distress are more costly for them. Opler, Pinkowitz, Stulz and Williamson (1999) find evidence for this motive by using market-to-book ratios and R&D spending as proxies for investment opportunities. The growth opportunities for Chinese firms are higher than for U.S. firms. Therefore, I expect higher cash holdings for Chinese firms than for U.S. firms during the crisis.
In times of crisis external financing might become too expensive and difficult to obtain due to the tight financial market. Therefore, a firm tends to make efforts to increase cash holdings to avoid raising external capital both in the debt and stock market. Adverse macroeconomic shocks will lead to a growing wedge between the cost of internal and external funds. Denis and Sibilkov (2009) figure that cash reserves may be used to take advantage of unexpected investment opportunities. Firms use cash as a hedging tool in states of the world when cash flows are low and the investment opportunities high (see e.g., Acharya, Almeida and Campello, 2007). During the financial crisis of 2008 the cash holdings as a hedging tool for firms are not enough because Campello, Graham and Harvey (2010) find that firms are likely to postpone or cancel investment plans when the capital markets tights. Lins, Servaes and Tufano (2010) also find that firms use non-operational cash to hedge against future cash flow shocks in bad times, but use credit lines to increase the likelihood to not forego positive NPV projects in good times. There might be a precautionary savings role for excessive cash and this may benefit firms in times of dislocation in markets for external markets according to Duchin, Ozbas and Sensoy (2010). I expect that the cash holdings are higher during the crisis for both U.S. and Chinese firms and that the demand function of investments for cash will be more sensitive during the crisis than during the pre-crisis period.
2.4. Changing firm characteristics’ and changes in the demand function for cash
3.
Data and Methodology
3.1. Sample construction
3.2. General data description
The dependent variable in this study is the cash ratio. There are several measures for cash ratios but this study uses the definition of cash ratio used by Bates, Kahle, and Stulz (2009). I measure the cash ratio as cash and short term investments divided by total assets. Next to the variable ‘Cash and short term investments’ I will use the variable ‘Logarithm cash divided by net assets’ as the dependent variable as a robustness check. This proxy for cash holdings reduces the magnitude of the problem of extreme outliers in cash divided by the total book value of assets excluding cash.
Table 1 presents the cash holdings of U.S. and Chinese firms in the pre-crisis period and the period after the start of the crisis. Both cash holdings for U.S. firms and Chinese are bigger in the post-crisis period. Chinese firms hold 5.1% more in cash after the crisis started and U.S. firms hold 1.5% more in cash after the crisis started. The difference between the cash holdings of Chinese firms and U.S. firms in the pre-crisis and in the post-crisis period are statistically significant at the 1% level. The higher cash holdings in the post-crisis period for Chinese firms are according to the expectations. Firms with higher growth opportunities in times of an adverse macro-economic shock hold higher cash holdings than firms with less growth opportunities.
Table 1
Statistical differences between cash holdings U.S and Chinese firms. Table 1 summarizes the cash ratios of sample firms of China and U.S. for 2001-2013. It represents the pre-crisis (2001-2007) and the post-crisis (2008-2013) cash ratios. The cash ratio is measured as cash and short term investments to total assets. The number given in the row and column “Difference in cash holding” are p-values from t-tests to test whether the mean cash ratios are significant different from each other.
Table 2 shows the descriptive statistics of each variable used in the tests. Leverage is the ratio of total debt to the book value of total assets. The mean leverage level for U.S. is 0.230 and the mean leverage level for Chinese firms is 0.233. Although the levels are close to each other, the difference in leverage levels between U.S. and Chinese firms is significantly different. NWC is the ratio of working capital minus cash and short term investments to the book value of assets minus cash. NWC is negative for Chinese firms and positive for U.S. firms. Proportionally, U.S. firms hold more liquid assets than cash in the firm on the contrary Chinese firms hold more cash than liquid assets in the firm. Investments are lower for Chinese firms than for U.S. firms. On the other hand the market to book value of U.S. firms is lower than for
China cash holding U.S. cash holding Difference in cash holding
Pre-crisis 0.171 0.189 <0.01
Post-crisis 0.222 0.204 <0.01
Chinese firms. Investment is the ratio of capital expenditures and acquisitions to the book value of assets. Market to book ratio comes directly from the data item “market to book value” of Worldscope. Firm size is the logarithm of the book value of total assets. The mean size of U.S. firms is 8.648 and the mean size of Chinese firms is 8.409. The sizes of the firms of U.S. and China are significantly different from each other. The dividend dummy equals one when a given firm pays common dividend in a given year otherwise the variable equals 0. In the descriptive statistics is the percentage dividend payers shown, this is the mean value of the annual proportion of firms which pay common dividend. For U.S. firms that pay dividend in a year is the mean 35.8% and the mean of Chinese firms that pays dividend in a year is 50.0%. Cash flow is measured by EBITDA to the total book value of assets. U.S. firms have lower mean cash flows than Chinese firms but the median cash flows is higher for U.S. firms than for Chinese firms. Industry sigma is measured as the average division (SIC code: 0100-0999 agriculture, 1000-1499 mining, 1500-1799 construction, 2000-3999 manufacturing, 4000-4999 transportation, communication, electric, 5000-5199 wholesale trade, 5200-5999 retail trade, 7000-8999 services) standard deviation of cash flows over the first five years of a given firm year observation. This variable shows the cash flow volatility of a firm over five years. U.S. firms have a mean industry sigma of 0.090 and Chinese firms have a mean industry sigma of 0.046. The cash flows are therefore more risky for U.S. firms. The difference between issuance and retirement of stock and debt is higher for Chinese firms than for U.S. firms, while Chinese firms hold also higher cash holdings than U.S. firms in this period. External finance measures the issuance and retirement of stock and debt for a given firm year observation.
Table 2
Descriptive statistics U.S. sample and Chinese sample.
Table 1 reports the number of firm-years, mean. median, maximum, minimum and standard deviation on each variable. It consists of panel data collected from World Scope over the period 2001-2013 on 32451 U.S. firm-years observations and on 23814 Chinese firm-years observations. Cash ratio is the ratio of cash and short term investments to book value of total assets. Leverage is the ratio of total debt to the book value of total assets. NWC is the ratio of working capital minus cash and short term investments to the book value of assets minus cash. Investment is the ratio of the of capital expenditures and acquisitions to the book value of assets. Market to book ratio comes directly from the data item “market to book value” of Worldscope. Firm size is the logarithm of the book value of total assets. Percentage dividend payers is the mean value of the annual proportion of firms with positive dividends. Cash flow is measured by EBITDA to the total book value of assets. Industry sigma is measured as the average division (SIC code: 0100-0999 agriculture, 1000-1499 mining, 1500-1799 construction, 2000-3999 manufacturing, 4000-4999 transportation, communication, electric, 5000-5199 wholesale trade, 5200-5999 retail trade, 7000-8999 services) standard deviation of cash flows over the first five years of a given firm year observation. External finance measures the issuance and retirement of stock and debt for a given firm year observation. The columns “statistical difference” give the p-values of the differences in mean values of the variables between the U.S. sample and the Chinese sample.
Variable Observations Mean Median Maximum Minimum STD Observations Mean Median Maximum Minimum STD Statistical difference (mean) Statistical difference (median) Cash ratio 32451 0.197 0.106 1.000 0000 0.227 23814 0.201 0.158 0.998 0.000 0.157 0.02 <0.01 Leverage 32331 0.230 0.188 1.000 0.000 0.226 23812 0.249 0.233 1.000 0.000 0.188 <0.01 <0.01 NWC 32096 0.053 0.039 0.617 -0.751 0.195 23715 -0.048 -0.043 0.515 -0.562 0.195 <0.01 <0.01 Investment 32266 0.076 0.047 0.563 0.000 0.086 23681 0.066 0.050 0.226 0.000 0.057 <0.01 0.92 Market to book ratio 28434 2.384 1.790 9.835 -0.365 2.004 19495 3.161 2.520 14.540 -13.819 2.697 <0.01 <0.01 Firm size 32451 8.648 8.634 11.902 6.702 0.925 23814 8.409 8.351 11.581 6.700 0.560 <0.01 <0.01 % Dividend payers 32077 0.358 0.000 1.000 0.000 0.479 14221 0.500 1.000 1.000 0.000 0.500 <0.01 <0.01 Cash flow 31323 0.066 0.107 0.557 -1.194 0.219 22838 0.092 0.086 0.433 -0.457 0.083 <0.01 <0.01 Industry sigma 32451 0.090 0.092 0.142 0.046 0.021 23814 0.046 0.048 0.070 0.015 0.007 <0.01 <0.01 External finance 29029 0.040 0.000 1.103 -0.427 0.179 23808 0.067 0.003 0.748 -10.753 0.186 <0.01 <0.01
The increase in cash holdings after the crisis started should also affect the policies of firms regarding to financial and investment policies. Therefore I report in Table 3 the trend in leverage and trend in investments of U.S. and Chinese firms. Leverage decreases in the first period of the sample period for U.S. firms. In 2008, the leverage level reaches a top of 24.4% from where it decreases again till 2010 and from this point in time it is increasing again till 24.5% in 2013. Investments of U.S. firms do not show any trend in time. Striking is the lowest level of investments in 2009. The credit crisis is not anymore a crisis only felt by the banking sector. The first part of the period of the Chinese sample shows the opposite trend in leverage for Chinese firms compared to U.S. firms. Leverage is increasing from 25.0% till 30.0% in 2005. After 2005, leverage is gradually decreasing to a leverage level around 22.0% in 2013. Just like for the U.S. sample, investments do not show any pattern for Chinese firms in the sample period of 2001-2013. Again, 2009 has the lowest level of investments between other years. Remarkable is the low level in investments in 2013 for Chinese firms.
Table 3
Trend in leverage and investments U.S. and Chinese firms Table 3 shows the annual mean value of leverage and investments of U.S. and Chinese firms over the period 2001-2013. Leverage is the ratio of total debt to the book value of total assets. Investment is the ratio of the of capital expenditures and acquisitions to the book value of assets. Panel A consists of U.S. firms and Panel B consists of Chinese firms.
Bates, Kahle, and Stulz (2009) find a statistically significant time trend in the cash ratio of the U.S. sample. They find that the coefficient on the time trend for the cash ratio corresponds to a yearly increase of 0.46% and has a p-value below 0.01 in the period between 1980 and 2006. This again corresponds with an increase of the cash ratio in 1980 of 10.50% to a cash ratio of 23.20% in 2006. They also state that this regression is only useful to characterize the evolution of the cash holdings during the sample period and that it will not make sense to
Panel A: U.S. Panel B: China Year Leverage Investment Leverage Investment
extrapolate the in-sample trend to future years. Therefore I will run the regression on the cash holdings again but now for the period between 2001 and 2013 and I will do this for U.S. firms and Chinese firms separately. To assess whether there is a statistically significant trend in the cash ratio, I estimate regressions of the cash ratio on a constant and time measured in years. The coefficient on the time trend for the cash ratio of U.S. firms corresponds to a yearly increase of 0.28% and a p-value below 0.01. This number implies a positive time trend in cash holdings over the period 2001-2013. The 0.28% time trend is lower than the time trend found by Bates, Kahle, and Stulz (2009) in the period 1980-2006. The coefficient on the time trend for the cash ratio of Chinese firms corresponds to a yearly increase of 0.58% and a p-value below 0.01. Therefore, the time trend for the cash ratio of Chinese firms is larger than for the time trend in cash ratio for U.S. firms in the period between 2001-2013.
3.3. Empirical strategy.
To analyse the impact of the financial crisis on corporate cash holdings, I compare cash holdings of firms before and after the beginning of the crisis as a function of their motives to hold cash. The impact of the crisis is researched by testing whether cash holdings changed because firms moved along the demand curve for cash or because the demand curve shifted. I am mostly interested in studying the role of firms’ investments opportunities in magnifying the impact of the crisis on cash holdings. Therefore I compare Chinese firms with U.S. firms in this study, because Chinese firms have higher growth opportunities than U.S. firms have. I use 2008 as the start of the crisis for non-financial firms because both China and the U.S. witnessed a significant decline in GDP growth in that year. Besides the crisis was in 2007 first only felt by the banking system. Later the crisis spread to non-financial firms. The cash holdings in the post-crisis period are significantly different from the cash holdings in the pre-crisis period “as noted in section 3.2”. The sample is therefore divided in the sub-period 2001-2007 and in the sub-period 2008-2013.
I use the model of Bates, Kahle, and Stulz (2009) as a basis for this study. I will estimate this model for China and for the U.S. separately. This model makes cash holdings depended on variables that proxy for the motives to hold cash.. The equation is as follows:
CashHolding(i,t) = ß(0) + ß(2)MKTtoBOOK(i,t-1) + ß(3)Investment(i,t-1) + ß(4)External finance(i,t) + ß(5)Cashflow(i,t) ß(6)Firmsize(i,t)+ + ß(7)NWC(i,t)+ ß(8)Leverage(i,t) + ß(9)
Industry sigma (i,t)+ ß(10)Dividend dummy(i,t)
(i) identifies the firm and (t) refers to the period in the sample.
Market to book ratio is a lagged variable because firms anticipate to the market to book value by increasing or decreasing cash holdings one year after the annual report. I use the variable ‘investment’ instead of the variables ‘acquisitions’ and ‘capital expenditures’ used in the model of Bates, Kahle, and Stulz (2009). I combine these two variables because there is not sufficient data available on the item acquisitions for firms in the sample. Song and Lee (2012) also use this variable instead of capital expenditure and acquisitions. I use the lagged value for investments to rule out reverse causality between cash holdings and investments. Reverse causality may exist because the firm’s investment and financing decisions cannot be separated in imperfect markets. The external finance variable is a contemporaneous variable, because firms should have more cash immediately after raising capital and cash should decrease as firms spend the capital raised (see e.g., Bates, Kahle, and Stulz, 2009). Leverage is also a contemporaneous variable because most of the firms hold a target leverage ratio which they do not change on a yearly basis.
I will check whether the firms’ demand function for corporate cash holdings has changed for U.S. and Chinese firms or whether these firms shifted along the demand curve for cash holdings during the sample period. Therefore I divide chapter 4 “Empirical results” in three subsections that each deal with one of the sub-questions posed “as noted in section 1”. In section 4.1 I check with Eq. (1) whether U.S. firms and Chinese firms share the same determinants for cash holdings. Subsequently in section 4.2. I check whether the firm characteristics have changed for U.S. and Chinese firms during the post-crisis period comparing to the pre-crisis period. The coefficients of the independent variables computed in section 4.1 of Eq. (1) are then an input to discuss whether the changing firm characteristics can explain the increase in cash holdings. In section 4.3. I test for the assumption of a changing demand function by checking whether the change in cash holdings is explained by an intercept change and coefficient change in Eq. (1) after the financial crisis comparing to the pre-crisis period. Therefore I use different indicator variables to the regression analysis of Eq. (1). The indicator variables are also used in the studies of Bates, Kahle, and Stulz (2009), Iskandar-Datta and Jia (2011) and Song and Lee (2012). I answer the research question based on the comparisons of the results found in section 4.2. and section 4.3.
Below I discuss the association between cash holdings and the independent variables used in Eq. (1). Next to this I give the signs which I expect for the coefficients of the independent variables from the regression analysis in relation to cash holdings in Eq. (1).
the firm’s cash than a firm with fewer, more stable growth opportunities. Bates, Kahle, and Stulz (2009) test the influence of the investment opportunity set on the cash holding measured by the market to book value ratio. They find empirical evidence that firms with better investment opportunities have higher cash holdings.
Investments - Proxies for investment opportunities of a firm are capital expenditures and acquisitions. Higher capital expenditures and acquisitions should lead to higher cash holdings according to the precautionary motive. But at the same time the higher the capital expenditures and acquisitions the higher is the debt capacity of a firm and the lower is the demand for cash (see e.g. Stulz, 2007). Bates, Kahle, and Stulz (2009) find that acquisitions negatively affect cash holdings and that the effect of capital expenditures is ambiguous.
External finance – Bates, Kahle, and Stulz (2009) argue that firms hold more cash after raising capital. They also conclude that the amount of cash decrease as firms spend raised capital. Bates, Kahle, and Stulz (2009) find a positive coefficient for net equity issuance and net debt issuance.
Cash flow - In general the pecking order theory of capital structure posits that firms with higher cash flows will have higher cash holdings. However, Bates, Kahle, and Stulz (2009) find ambiguous empirical evidence on the relationship between cash flow and cash holdings.
Firm size - Raising costs for external financing is higher for smaller firms than for larger firms because larger firms have scales of economies in raising external capital. The scale of economies arises because there is a significant fixed costs component in issuing external capital ( see e.g., Barclay and Smith, 1996). Bates, Kahle, and Stulz (2009) find in their empirical study that an increase in firm size would lead to a decrease in cash holding.
Net working capital - In literature it is assumed that the cost of converting cash substitutes into cash is much lower as compared to illiquid assets. Firms which have sufficient cash substitutes can use these substitutes instead of the capital markets for raising cash when the firm is short of liquid assets. Bates, Kahle, and Stulz (2009) find significant empirical evidence that NWC influence cash holdings negatively.
Leverage - A small increase in cash holdings serves the debtholders, because most of the value of a firm is in the hand of the debtholders. This in turn increases the debt value and not the equity value. However, the higher the debt ratio the higher the likelihood of financial distress. Bates, Kahle, and Stulz (2009) find a negative relationship between leverage and cash holdings.
use cash to hedge against changes in the expected value of future cash flow. Kim, Mauer, and Sherman (1998) show in their model that the optimal investment in liquidity is increasing in the variance of future cash flows. Bates, Kahle, and Stulz (2009) find a positive sign for industry sigma on cash holdings.
Dividend dummy - The precautionary motive for holding cash is weaker for firms that pay dividends. This is because they are seen as less risky and have greater access to capital markets than firms that do not pay dividends (see e.g., Bates, Kahle, and Stulz, 2009). Bates, Kahle, and Stulz (2009) find ambiguous evidence for the relationship between dividends paid and cash holdings.
coefficient estimates could be wrong. However, the sample is large enough to come up with accurate results. For the Jarque-Bera test “see Appendix D”.
3.4. Direction of data to the research questions
According to the current studies to the precautionary motive, cash holdings of financially constrained firms should increase more than for financially unconstrained firms after macroeconomic shocks ( see e.g., Almeide, Campello, and Weisback, 2004). Therefore I test the cash holdings trends of financially constrained and financially unconstrained firms. The results are based on three measurements for financially constrained firms. Namely, a firm is classified as a financially constrained firm in a year when it does not pay dividends, is small in firm size and belongs to the top 30% in leverage ratio (see e.g., Song and Lee, 2012).
3.4.1. Firm size
Fig. 3 shows the cash holdings for each firm size quintile of Chinese firms for the period 2001-2013. The smallest firms show an increase in cash holdings over the sample period with a very steep increase during the financial crisis. The cash holdings of the largest firms do not show a pattern in cash holdings. This result corresponds with the results found by Bates, Kahle, and Stulz (2009), the smallest firms seem to exhibit the largest increase in cash holdings. Smaller firms are also seen to be more financially constrained than large firms. This corresponds with the steep increase in cash holdings for the smallest firms during the financial crisis.
Fig. 3. Evolution of cash holdings by firm size quintile from 2001 to 2013 for Chinese firms. Firm size is measured
Fig. 4 shows the cash holdings for each firm size quintile of U.S. firms for the period 2001-2013. All the firm size quintiles show an increase in cash holdings over the sample period. The cash holdings increase more for the smallest firms than for the largest firms. This result corresponds with the results found by Bates, Kahle, and Stulz (2009), in that the smallest firms seem to exhibit the largest increase in cash holdings
Fig. 4. Evolution of cash holdings by firm size quintile from 2001 to 2013 for U.S. firms. Firm size is measured
as the logarithm of the book value of total assets.
3.4.2. Dividend paying firms
Fig. 5. Evolution of cash holdings for (non)dividend paying firms from 2001 to 2013 for U.S. and Chinese firms.
A dividend paying firm has a positive common dividend payout in a firm-year observation. A non-dividend paying firms pays not common dividend in a given firm-year observation.
3.4.3. Leverage
Song and lee (2012) use highly leveraged firms as financially constrained firms. Where highly leveraged means that a firms belong to the top 30% of the leverage ratios of the sample firms. According to the precautionary motive of cash holdings financially constrained firms tend to have more cash holdings due to costly external financing. Therefore, the highly leveraged firms should have higher cash holdings than the lower leveraged firms. Fig. 6 shows the cash holdings of high and low leveraged firms in the U.S. and China in the period 2001-2013. Here you can see that not the high leveraged firms but the low leveraged firms have the highest cash holdings. An argument for this is that cash can be seen as negative debt. Another measurement used for debt is net debt, which is all the outstanding debt minus cash. Thus firms with more cash hold lower debt levels.
Fig. 6. Evolution of cash holdings for high/low leveraged firms from 2001 to 2013 for U.S. and Chinese firms.
4.
Empirical study
4.1. Determinants of corporate cash holdings
In this section I conduct test whether the China sample and the U.S. sample share the same determinants of corporate cash holdings. The results are presented in Table 4. The regression is based on Eq. (1), “see section 3.3”. The discussion is based on Panel A. Panel B and C are used as a robustness check.
Table 4
Regressions for determinants of corporate cash holdings of U.S and Chinese firms. Table 4 reports regression estimates for the determinants of cash holdings for Chinese and U.S. firms over the period 2001-2013. Cash flow is measured by EBITDA to the total book value of assets. Dividend dummy equals 1 when a firm-year observation has a positive common dividend payout and 0 otherwise. External finance measures the issuance and retirement of stock and debt for a given firm year observation. Firm size is the logarithm of the book value of total assets. Industry sigma is measured as the average division standard deviation of cash flows over the first five years of a given firm year observation. Investment is the ratio of the of capital expenditures and acquisitions to the book value of assets. Leverage is the ratio of total debt to the book value of total assets. Market to book ratio comes directly from the data item “market to book value” of Worldscope. NWC is the ratio of working capital minus cash and short term investments to the book value of assets minus cash. Panel A shows results using an OLS regression with year and firm fixed effects, Panel B shows results using an OLS regression and Panel C shows results using an OLS regression on the logarithm of cash to net assets with year and frim fixed effects. ***,**, or * indicates that the coefficient estimates is significant at the 1%, 5%, or 10% level, respectively. Standard errors (in brackets) are heteroskediasticity-consistent.
Panel A Panel B Panel C
Cash ratio China Cash ratio U.S. Cash ratio China Cash ratio U.S. Log. cash/net assets China
Log. cash/net assets U.S. Intercept 0.056 [0.057] 0.493*** [0.045] 0.163*** [0.020] 0.423*** [0.013] -1.904*** [0.251] 0.436** [0.172] Cash-Flow 0.111*** [0.016] 0.039*** [0.007] 0.103*** [0.015] -0.102*** [0.009] 0.724*** [0.085] 0.141*** [0.026] Dividend 0.016*** [0.004] 0.004 [0.003] 0.037*** [0.003] -0.049*** [0.002] 0.058*** [0.014] 0.005 [0.013] External finance 0.024** [0.011] 0.116*** [0.008] 0.022** [0.009] 0.226*** [0.011] 0.075 [0.050] 0.337*** [0.028] Firm Size 0.017*** [0.006] -0.037*** [0.005] 0.000 [0.002] -0.023*** [0.001] 0.138*** [0.029] -0.177*** [0.019] Industry sigma -0.384 [0.370] 0.620*** [0.088] 0.787*** [0.167] 0.879*** [0.050] -1.974 [1.503] 3.127*** [0.357] Investment -0.135*** [0.021] -0.169*** [0.169] -0.219*** [0.019] -0.473*** [0.012] -0.302*** [0.096] -0.594*** [0.047] Leverage -0.114*** [0.011] -0.153*** [0.008] -0.208*** [0.008] -0.339*** [0.007] -0.428*** [0.052] -0.644*** [0.033] Market to book ratio 0.001**
By the results it follows that U.S. and Chinese firms share the same signs on the coefficients of cash flow, dividend, external finance, investment, leverage, market to book ratio and NWC. Dividend is not statistically significant for U.S. firms, and external finance, market to book ratio and NWC are statistically significant at the 5% level for Chinese firms. The rest of the coefficients with the same signs for Chinese and U.S. firms are statistically significant at the 1% level. The positive sign on cash flow indicates that both U.S. and Chinese firms use the pecking order theory regarding to cash flows. U.S. firms are more sensitive to cash flows because the coefficients’ magnitude on cash flow is more of less three times larger for U.S. firms than for Chinese firms. The coefficient on external finance is significantly positive meaning that the firms have more cash directly at hand when raising external capital and do not spend it immediately. Cash should decrease as firms spend the capital raised. The negative sign on investment means that the firms use cash to spend on investments and that the capital expenditures of the investments can be used as collateral. The coefficient on leverage is significantly negative supporting the theory that the cost of investing in liquid assets increases with debt financing. Since most of the value of a firm is in the hands of the debtholder when leverage is high. The negative sign on NWC is according to the substitution of inventory, accounts receivables and credit lines into cash. A firm needs less cash when it has highly liquid assets in place. The Chinese firms are more sensitive to NWC than U.S. firms. Chinese firms have lower cash holdings per one unit of NWC compared to U.S. firms.
other signs for the coefficients on cash flow and dividends compared to the OLS regression with fixed effects. Bates, Kahle, and Stulz (2009) also find ambiguous results of the signs on the coefficients on dividends and cash flow in different models in their study. The negative sign on cash flow in the U.S. sample is closer connected to the argument of Kim, Mauer, and Sherman (1998) who argue that the higher the cash flow, the greater the ability of the firm to generate cash and less likely to be subjected to financial constraints.
4.2. Changes in firm characteristics
The average cash ratio for Chinese firms before the crisis is 17.1% and the average cash ratio after the impact of the financial crisis on cash holdings is 22.2%. The average cash ratio for U.S. firms before the crisis is 18.9% and the average cash ratio after the impact of the financial crisis on cash holdings is 20.4%. The difference between those two averages is statistically significant at the 1% level, “as noted in section 3.2”. A reason for the increase in cash holdings can be that firms shifted along the demand curve for cash holdings because the firm characteristics changed during the crisis. Firm characteristics might have been changed because for example more smaller firms than bigger firms did not survive the crisis. Another example is that demand during the crisis has decreased, which may affect growth opportunities of firms and indirectly the market to book ratios of firms. Fama and French (2004) find that the composition of firms has recently changed due to an influx of newly listed firms. This trend may be ongoing after their study. The new firms are smaller, have more growth opportunities and have higher cash flow risk. Table 5 presents the comparison of firm characteristics of Chinese and U.S. firms between the pre-crisis period and the post-crisis period.
Table 5
Comparison between firm characteristics in pre-crisis period and post-crisis period. Table 5 reports the mean and median firm characteristics for the pre-crisis period (2001-2007) and the post-crisis period (2008-2013) for Chinese and U.S. firms. Cash ratio is the ratio of cash and short term investments to book value of total assets. Cash flow is measured by EBITDA to the total book value of assets. Dividend dummy equals 1 when a firm-year observation has a positive common dividend payout and 0 otherwise. External finance measures the issuance and retirement of stock and debt for a given firm year observation. Firm size is the logarithm of the book value of total assets. Industry sigma is measured as the average division standard deviation of cash flows over the first five years of a given firm year observation. Investment is the ratio of the of capital expenditures and acquisitions to the book value of assets. Leverage is the ratio of total debt to the book value of total assets. Market to book ratio comes directly from the data item “market to book value” of Worldscope. NWC is the ratio of working capital minus cash and short term investments to the book value of assets minus cash. Panel A presents the Chinese sample and panel B presents the U.S. sample. The values for the mean differences come from t-tests and the p-values for the median differences come from the Wilcoxon test. ***,**, or * indicates that the coefficient estimates is significant at the 1%, 5%, or 10% level, respectively.
finance, because the firms hold higher levels of external capital after the beginning of the crisis. Firms in both countries increases in size after the beginning of the financial. The firms in the Chinese sample increased by 0.281 to 8.526. The firms in the U.S. sample increased by 0.142 to 8.718. Some smaller firms do not tend to survive the financial crisis. Industry risk also increases for both countries. This corresponds with the instability of the crisis at the market. This variable is somehow biased because it uses lagged values of cash flows in calculating the cash flow risk. Therefore the post-crisis period includes cash flow volatility of the pre-crisis period. Investments after the start of the financial crisis are lower than investments before the crisis, this corresponds to a negative shock in the supply of external finance and a decrease in demand after the financial crisis. This also corresponds with the findings of Campello, Graham, and Harvey (2010) who find evidence that firms forego profitable investment opportunities
Panel A: sample China Panel A: sample U.S.
Pre- crisis Post- crisis Difference Pre- crisis Post crisis Difference Pre-crisis Post-crisis Difference Pre-crisis Post-crisis Difference Cash-Flow 0.083 0.098 0.015*** 0.083 0.088 0.005*** 0.079 0.053 -0.026*** 0.113 0.101 -0.012*** % Dividend payers 0.445 0.535 0.090*** 0.000 1.000 1.000*** 0.350 0.365 0.015*** 0.000 0.000 0.000** External finance 0.055 0.075 0.020*** 0.002 0.004 0.002*** 0.039 0.042 0.003 0.000 0.000 0.000*** Firm Size 8.245 8.526 0.281*** 8.205 8.475 0.270*** 8.576 8.718 0.142*** 8.566 8.705 0.139*** Industry sigma 0.044 0.048 0.00*** 0.045 0.051 0.006*** 0.088 0.093 0.005*** 0.087 0.101 0.014*** Investment 0.068 0.065 -0.003*** 0.049 0.050 0.001*** 0.079 0.073 -0.006*** 0.049 0.044 -0.005*** Leverage 0.280 0.227 -0.053*** 0.271 0.201 -0.070*** 0.229 0.232 0.003 0.190 0.186 -0.004
Market to book ratio 2.938 3.322 0.282*** 2.300 2.700 0.400*** 2.488 2.285 -0.203*** 1.900 1.690 -0.210***
NWC -0.101 -0.011 0.090*** -0.090 -0.001 0.089*** 0.063 0.043 -0.020*** 0.046 0.034 -0.012***
during the financial crisis. Leverage decreases significantly for Chinese firms and not for U.S. firms. The leverage ratio increases insignificantly for U.S. firms. The market to book ratio for Chinese firms increases insignificantly after the beginning of the financial crisis, this does not corresponds with a decrease in demand and a shock to growth opportunities during the financial crisis. The market to book ratio for U.S. firms decreased significantly after the financial crisis started. NWC increases for Chinese firms and decreases for U.S. firms. According to the substitution effect of NWC for cash you might expect that firms with an increase in NWC hold lower cash holdings. But at the same time when you use up NWC, there can be a timely effect in the increase in cash holdings due to a decrease in NWC.
The increase in % dividend payers and external finance for U.S. and Chinese firms after the beginning of the financial crisis corresponds with an increase in cash holdings after the pre-crisis period because all these variables have a positive coefficient in the OLS regression with fixed effects “as noted in section 4.1”. The decrease in investments for U.S. firms and Chinese firms also corresponds with an increase in cash holdings after the pre-crisis period, because the coefficient of investment is negative in the OLS regression with fixed effects “as noted in section 4.1”. The increase in firm size of U.S. firms does not correspond with an increase in cash holdings regarding to the OLS regression with fixed effects. Cash flow has a positive coefficient in the OLS regression and therefore the decrease in cash flow for U.S. firms after the pre-crisis period is not according to higher cash holdings in the post-crisis period. The increase in cash flow volatility for Chinese firms after the pre-crisis period does not corresponds with an increase in cash holdings because this coefficient has a positive sign in the OLS regression with fixed effects. Leverage and NWC have both a negative coefficient in the OLS regression and therefore the increase in NWC for Chinese firms and the increase in leverage for U.S. firms does no correspond with an increase in cash holding after the pre-crisis period. The results presented here are ambiguous. Some of the firm characteristics have positive effects on the increase of cash holdings after the pre-crisis period and some variables have offsetting effects on the level of cash holdings. Therefore, I test in section 4.3 whether the demand function for cash holdings can explain the changes in cash holdings of U.S. and Chinese firms after the beginning of the financial crisis.
4.3. Demand function for cash holdings after the pre-crisis period
Table 6
Estimation of regressions on cash ratios with post crisis influence.
Table 6 reports the results of regressions on cash ratio for U.S. and Chinese firms over the period 2001-2013. Cash ratio is used as the dependent variable in model 1, 3, 5, and 6. The logarithm of cash/net assets is used as the dependent variable in the models 3 and 4. Cash flow is measured by EBITDA to the total book value of assets. Dividend dummy equals 1 when a firm-year observation has a positive common dividend payout and 0 otherwise. External finance measures the issuance and retirement of stock and debt for a given firm year observation. Firm size is the logarithm of the book value of total assets. Industry sigma is measured as the average division standard deviation of cash flows over the first five years of a given firm year observation. Investment is the ratio of the of capital expenditures and acquisitions to the book value of assets. Leverage is the ratio of total debt to the book value of total assets. Market to book ratio comes directly from the data item “market to book value” of Worldscope. NWC is the ratio of working capital minus cash and short term investments to the book value of assets minus cash. Post-crisis dummy equals 1 after the crisis (2008-2013) and 0 otherwise. Model 1 and 2 are OLS regressions with firm fixed effects and model 3 and 4 are OLS regressions with firm fixed effects. ***,**, or * indicates that the coefficient estimates is significant at the 1%, 5%, or 10% level, respectively. Standard errors (in brackets) are heteroskediasticity-consistent.
Model 1 and 2 uses cash holdings as the dependent variable in the OLS regression. Model 1 estimates the regression of the U.S. sample and model 2 estimates the regression of the Chinese sample. The post-crisis dummy has a significantly positive coefficient at the 1% level in model 1, meaning that the cash holdings increases after the beginning of the financial crisis for exogenous reasons that are unrelated to firm characteristics. So, there is a significant upward shift in demand for cash after the beginning of the financial crisis for U.S. firms. The coefficient of the post-crisis dummy in model 2 is statistically significant at the 2% level. Therefore Chinese firms also witness an increase in cash holdings after the beginning of the financial crisis for exogenous reasons that are unrelated to firm characteristics. The coefficient of the post-crisis dummy of the U.S. sample is the same as the coefficient on the post-crisis dummy for the Chinese sample. Hence, the shift in demand for corporate cash holdings after the pre-crisis period is the same for U.S. and Chinese firms.
more sensitive to investment. The coefficient on investments decreases from -0.166 to -0.236 ( -0.166 – 0.070). Therefore Chinese firms hold more cash after the beginning of the crisis regarding to investments. The sensitivity of cash holdings to NWC changes in sign and in magnitude. Namely, it decreases from 0.036 to -0.051 ( 0.036 – 0.087). Chinese firms increase cash holdings by an increase in one unit of NWC before the crisis and they decrease cash holdings by an increase in one unit of NWC after the beginning of the crisis. There is a large increase in the magnitude of the coefficient of leverage after the beginning of the crisis, namely the coefficient decreases from -0.051 to -0.140 (-0.051-0.089). The intercept decreases with 1.6%, at the same time this decrease is insignificant.
The change in cash holdings for U.S. and Chinese firms can be partially explained by a change in cash holding policy after the beginning of the crisis. On contrary, cash holdings reacts less sensitively to some determinants of cash holdings after the beginning of the crisis. These determinants cannot explain the increase in cash holdings.
5.
Conclusion
This study examines to what extent the change in cash holdings of Chinese and U.S. firms after the beginning of the financial crisis of 2007 is a result of changes in firm characteristics and a result in the shift in demand for cash. It tests whether cash holdings changed because firms moved along the demand curve for cash or because the demand curve shifted after the financial crisis. I test this by researching the cash holdings and firm characteristics before the beginning of the crisis in the period 2001-2007 and by researching the cash holdings and firm characteristics after the beginning of the crisis in the period of 2008-2013. Next to this I compare whether an emerging country reacts differently to the financial crisis comparing to a developed country regarding to the demand for cash (cash holding policy). Using a sample of Chinese and U.S. firms in the period between 2001-2013, I find that U.S. firms and Chinese firms share most of the determinants of corporate cash holdings. The cash holdings policies based on cash flow risk and firm size are different for both countries. Chinese firms increase cash holdings by an increase in firm size, while U.S. firms decrease cash holdings by an increase in firm size. Decreasing the cash holdings by an increase in firm size is according to the transaction motive of cash.
U.S. firms. The increase in dividend payers and external finance for U.S. and Chinese firms after the beginning of the financial crisis corresponds with an increase in cash holdings after the pre-crisis period. Investments are declining after the beginning of the financial crisis and this corresponds with an increase in cash holdings. Other changes in firm characteristics like the change in cash flow and firm size of U.S. firms do not correspond with an increase in cash holdings after the beginning of the financial crisis. The changes in cash flow risk and NWC for Chinese firms do not correspond with an increase in Chinese cash holdings after the beginning of the crisis. Therefore, I have also researched whether the demand curve for cash has changed after the financial crisis. It turns out that the changes in cash holdings cannot be considered as a pure unexplained shift in precautionary demand for cash. From this study it follows that the change in sensitivity of cash holdings to the variables “investment” and “leverage” leads to higher cash holdings for Chinese firms. The change in sensitivity of cash holdings to the variable “dividend” leads to higher cash holdings for U.S. firms.
My results show that the financial crisis has changed both the firm characteristics and the cash holdings policies of Chinese and U.S. firms. The firms moved along the demand curve for cash as well the demand curve for cash holdings shifted after the beginning of the financial crisis. At the same time other factors then firm characteristics do not explain the increase in cash holdings for both U.S. and Chinese firms. Next to this, Chinese and U.S. firms exhibit commonality in the determinants of cash holdings. The correlation between firm attributes and cash holdings documented by Bates, Kahle, and Stulz (2009) is therefore not spurious. Furthermore, the results confirming the results found by Song and Lee (2012) in that there is a shift in demand to cash after the beginning of a crisis. At the same time, the results confirm the findings of Bates, Kahle, and Stulz (2009) in that firm characteristics explains the increase in
cash holdings.
APPENDIX A
Fig. A.1. Stock market / Credit market.
Stock market capitalization to GDP divided by domestic credit available to private sector to GDP. (source: World Bank’s World Development Indicators).
Fig. A.2. Stock market. Stock market capitalization to
GDP. (source: World Bank’s World Development Indicators).
Fig. A.3. Credit market. Domestic credit available to
APPENDIX A
Fig. A.4. GINI coefficient. The GINI coefficient represents the income distribution of a nation’s residents. 0% expresses perfect equality and 100% expresses maximal equality. It is measured as a mean over the period 2001-2013. (source: World Bank’s World Development Indicators).
Appendix B
Redundant Fixed Effects Tests Equation: Untitled
Test cross-section and period fixed effects
Effects Test Statistic d.f. Prob.
Cross-section F 19,903,331 -2741.21 0.000 Cross-section Chi-square 30,402,896,763 2741 0.000 Period F 13,002,714 -12.2095 0.000 Period Chi-square 175,953,717 12 0.000 Cross-Section/Period F 19,987,552 -2753.21 0.000 Cross-Section/Period Chi-square 30,550,277,577 2753 0.000
Redundant fixed effects tests U.S. firms
Correlated Random Effects - Hausman Test Equation:
Untitled
Test period random effects
Test Summary Chi-Sq. Statistic Chi-Sq. d.f.
Prob.
Period random 72,793,579 9 0.000
Hausman test period random effects U.S. firms
Correlated Random Effects - Hausman Test Equation: Untitled
Test cross-section random effects
Test Summary Chi-Sq. Statistic Chi-Sq. d.f. Prob.
Cross-section random 1,435,302,849 9 0.000
APPENDIX B
Redundant Fixed Effects Tests Equation: Untitled
Test cross-section and period fixed effects
Effects Test Statistic d.f. Prob.
Cross-section F 5,929.051 -2233.892 0.000 Cross-section Chi-square 10,169,600,166 2233 0.000 Period F 16,126,699 -12.892 0.000 Period Chi-square 239,829,830 12 0.000 Cross-Section/Period F 6,007,143 -2245.892 0.000 Cross-Section/Period Chi-square 10,293,227,536 2245 0.000
Redundant fixed effects tests Chinese firms
Correlated Random Effects - Hausman Test Equation: Untitled
Test cross-section random effects
Test Summary Chi-Sq. Statistic Chi-Sq. d.f. Prob.
Cross-section random 352,488,508 9 0.000
Hausman test cross-section random effects Chinese firms
Correlated Random Effects - Hausman Test Equation:
Untitled
Test period random effects
Test Summary Chi-Sq. Statistic Chi-Sq. d.f. Prob.
Period random 33,859,497 9 0.000
APPENDIX C
Correlation matrix variables U.S. firms
Correlation matrix variables Chinese firms
Cash flow Dividend External finance Firm size Industry sigma Investment MKTB NWC Leverage Cash_flow 1.000 0.269 -0.421 0.300 -0.102 0.124 0.005 0.119 -0.021 Dividend 0.269 1.000 -0.120 0.423 -0.170 0.019 0.022 0.042 0.023 External finance -0.420 -0.120 1.000 -0.138 0.047 0.049 0.068 -0.134 0.083 Firm size 0.300 0.423 -0.138 1.000 -0.154 0.141 0.094 -0.096 0.211 Industry sigma -0.102 -0.170 0.047 -0.154 1.000 0.010 0.089 -0.081 -0.113 Investment 0.124 0.019 0.049 0.141 0.010 1.000 0.071 -0.114 0.127 MKTB 0.005 0.022 0.068 0.094 0.089 0.071 1.000 -0.171 -0.065 NWC 0.119 0.042 -0.134 -0.096 -0.081 -0.115 -0.171 1.000 -0.204 Leverage -0.021 0.023 0.083 0.211 -0.113 0.127 -0.065 -0.203 1.000
Cash flow Dividend External finance
APPENDIX D
Normality test residuals for U.S. firms (Jarque-Bera test)
References
Acharya, V., Almeida, H., Campello, M., 2007. Is cash negative debt? A hedging perspective on corporate financial policies. Journal of Financial Intermediation 16, 515-554.
Almeida, H., Campello, M., and Weisback M., 2004, The Cash Flow Sensitivity of Cash, Journal of Finance 59, 1777-1804.
Azar, J., Kagy, J., Schmalz, M., 2014. Can changes in the cost of cash resolve the corporate cash puzzle? Unpublished working paper. University of Michigan, Michigan.
Barclay, M., Smith, C., 1995. The maturity structure of corporate debt. Journal of Finance 50, 609-631.
Bates, T., Kahle, K., Stulz, R., 2009. Why do U.S. firms hold so much more cash than they used to? Journal of Finance 5, 1985-2021.
Brooks, C., 2012. Introductory econometrics for finance. Cambridge University Press, Cambridge.
Campello, M., Graham, J., Harvey, C., 2010. The real effects of financial constraints: evidence from a financial crisis. Journal of Financial Economics 97, 470–487.
Denis, D., Sibilkov, V., 2009. Financial constraints, investment, and the value of cash holdings. Review of Financial Studies 23, 247-269.
Dittmar, A., Mahrt-Smith, J., 2007, Corporate governance and the value of cash holdings. Journal of Financial Economics 83, 599–634.
Duchin, R., Ozbas, O., Sensoy, B., 2010. Costly external finance, corporate investment, and the subprime mortgage credit crisis. Journal of Financial Economics 97, 418-435.
Eije von, H., 2012. What causes differences in cash holdings between listed and unlisted firms around the world? Unpublished working paper. University of Groningen, Groningen
Fama, F., French, K., 2004. New lists: Fundamentals and survival rates. Journal of Financial Economics 73, 607-636.
Faulkender, M., Wang, R., 2006. Corporate financial policy and the value of cash. Journal of Finance 61, 1957-1990.
Harford, J., 1999. Corporate cash reserves and acquisitions. Journal of Finance 54, 1969-1997.
Harford, J., Mansi, S.A., Maxwell, W.F., 2008. Corporate governance and firm cash holdings in the US. Journal of Financial Economics 87, 535-555.
Iskandar-Datta, M., Jia, Y., 2012. Cross-country analysis of secular cash trends. Journal of Banking & Finance 36, 898–912.
Kahle, K., Stulz, R., 2012. Access to capital, investment, and the financial crisis. Unpublished working paper. University of Arizona, Arizona.
Kalcheva, I., Lins, k., 2007. International evidence on cash holdings and expected agency problems. Review of Financial Studies 20, 1087-1112.
Keynes, J., 1936. The general theory of employment, interest and money. Harcourt Brace, London.
Kim, C., Mauer, D., Sherman, A., 1998. The determinants of corporate liquidity theory and evidence. Journal of Financial and Quantitative Analysis 33, 335-359.
Lins, K., Servaes, H., Tufano, P., 2010. What drives corporate liquidity? An international survey of cash holdings and lines of credit. Journal of Financial Economics 98, 160-176.
Miller, M.H., Orr, D., 1966. A model of the demand for money by firms. Quarterly Journal of Economics 80, 413-435.
Modigliani, F., Miller, M., 1958, The cost of capital, corporation finance and the theory of investment. The American Economic Review 48, 261-297.
Myers, S., Majluf, N., 1984, Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics 13, 187-221.
Opler, T., Pinkowitz, L., Stulz, R., Williamson, R., 1999. The determinants and implications of corporate cash holdings. Journal of Financial Economics 52, 3-46.
Pinkowitz, L., Williamson, R., 2004. What is a dollar worth? The market value of cash holdings. Unpublished working paper. Georgetown University, Washington DC.
Pinkowitz, L., Stulz, R., Williamson, R., 2012. Multinationals and the High Cash Holdings Puzzle. Unpublished working paper. The Ohio state University, Ohio
Song, K.R., Lee,Y., 2012. Long-term effects of a financial crisis: evidence from cash holdings of East Asian firms. Journal of Financial and Quantitative Analysis 3, 617–641.
