Descriptive statistics

In the present paper, we define the cash ratio of cash and marketable securities to net assets, which are computed as total assets less cash and marketable securities as the proxy for cash holdings. The main reason for netting out from assets is that cash holdings are mainly held for the transaction and precautionary motive, not for a firm’s profitability, which is mainly related to assets without cash and marketable securities. Although not reported in the present paper, we also use cash and marketable to sales ratio as cash ratio to measure, and find that this alternative measure does not affect our main conclusion. This result is also similar to those by Dittmar, Mahrt-Smith, and Servaes (2003) and Opler et al. (1999).

Next, shareholder rights (RIGHT) and rule of law (LAW), as proxies for the country-level governance, are the main explanatory variables in our samples. These data are from the IMD World Competitiveness Online. Shareholder rights index is scored from 0 to 10, which presents the best protection. Rule of law is the legal and regulatory framework index, which is scored from 0 to 10; the higher the score, the higher the legality.

corporate cash holdings. For example, Kim, Mauer, and Sherman (1998) report that firms facing higher costs of external financing with higher financial constraints have more volatile earnings; thus, they should hold more cash holdings. Opler et al. (1999) and Bates, Kahle, and Stulz (2009) find that small firms with strong investment opportunities and riskier cash flows hold larger liquid assets. Hence, investment opportunities, financial constraints, and firm specifics are also factors that affect corporate cash holdings.

How do financial constraints and firm specifics affect corporate cash holdings? First, the existence of asymmetric information between firms and investors makes external financial costly; thus, firms with severe asymmetric information should have more cash holdings to avoid passing up valuable investment opportunities that could earn more cash, which could enhance corporate governance (Myers and Majluf, 1984). Brennan and Hughes (1991) and Collins, Rozeff, and Dhaliwal (1981) suggest that large firms have less information asymmetry than small firms; thus, small firms have higher financial constraints. Thus, we use firm size (SIZE), which is computed as the natural logarithm of total assets, as the proxy for the financial constraints. In addition, we use Tobin’s q (Q), which is computed as the ratio of market value of equity plus book value of liability to book value of assets, as the proxy for the investment opportunities.

Second, firm specifics also affect cash holdings. Firms with higher idiosyncratic risk face higher cost of external financing; thus, they should hold more cash to avoid giving up valuable growth opportunities (Kim, Mauer, and Sherman, 1998; Bates, Kahle, and Stulz, 2009). Because DOL¹ is used to measure the idiosyncratic risk of a firm, we use it as the proxy for the idiosyncratic risk. Although not reported in the present paper, we also use standard deviation of cash flows divided by average total assets as idiosyncratic risk, and find

1Firms with higher idiosyncratic risk need to accumulate more cash to solve a higher frequency of cash flow shortfalls. This implies that firms with more volatile cash flows have higher idiosyncratic risk so that they would hold more cash (Bates, Kahle, and Stulz, 2009; Ferreira and Vilela, 2004; Harford, Mansi, and Maxwell, 2008;

Opler et al., 1999). In addition to the cash flow volatility, the degree of operating leverage (DOL) is also a method to measure the firm’s idiosyncratic risk. Hence, we use DOL, computed as the percentage change in EBIT to the percentage change in Sales, as the proxy for idiosyncratic risk.

that this alternative measure does not affect our main conclusion. Next, Baskin (1987) argues that the cost of funds used to invest in liquidity increases as the ratio of debt financing increases, which implies a reduction in cash holdings with increased debt in capital structure.

We therefore use leverage (LEVERAGE), computed as the ratio of total debt to total assets, as a substitute for cash. In addition to cash, another substitution effect is due to other liquid assets of firms because firms with sufficient liquid assets may not have to raise funds through the capital markets when they experience cash shortage. Hence, we use non-cash liquid assets (NWC), computed as the ratio of working capital minus cash and marketable securities to total assets minus cash and marketable securities, as substitutes for cash.

All the firm-level financial and country-level governance variables are defined in Table 2.

Table 2 Description of variables

This table shows the definition of variables. The full sample period is from 2002 to 2008. Firm-level accounting data are collected from Compustat Global Vantage. Country-level governance data, such as rule of law and shareholder rights, are collected from the IMD World Competitiveness Online. Net assets are computed as assets less cash and equivalents.

Measurement item Proxy

variable Definition Reference

Dependent variable

Cash holdings CASH The ratio of cash and marketable securities to net assets the IMD World Competitiveness Online.

The index is from 0 to 10 (best).

La Porta et al., 2002.

RIGHT The index of shareholder rights is collected from the IMD World Competitiveness Online. The index is from 0 to 10 (best). liability ) /book value of assets

Bates et at., 2009; Dittmar,

LEVERAGE The ratio of total debt to total assets

Other variables included in the main analysis but not reported in the table are (i) industry dummies and (ii) a common law dummy (Common). We compute the ratio of cash and

marketable securities to net assets within the Fama and French 48 industry categories (similar with Dittmar and Mahrt-Smith, 2007). There remain 42 industry categories in our paper for firms with missing values and for financial firms and utilities. The common law dummy is a dummy equal to 1 for the common-law countries and 0 for the civil-law countries.

Table 3 provides mean, median, and standard deviation of whole variables in our analysis for the period 2002–2008.

On average, firms of four countries hold 13.8% of CASH. This ratio is highly skewed, with a median of 6.9%. US firms have the largest standard deviation of CASH. The common-law countries (i.e., the US and the UK), on average, receive higher score than civil-law countries, such as Germany and France. This is similar with the findings of La Porta et al. (1998). Next, the common-law countries also have higher LAW and RIGHT than the civil-law countries.

Furthermore, firm specifics and financial constraints also affect cash holdings (Bates, Kahle, and Stulz, 2009; Dittmar, Mahrt-Smith, and Servaes, 2003; Dittmar and Mahrt-Smith, 2007; Faulkender and Wang, 2006; Opler et al., 1999; Ozkan and Ozkan, 2004; Pinkowitz, Stulz, and Williamson, 2006). According to previous empirical studies, we can expect that small firms with better investment opportunities, more idiosyncratic risk, and less liquidity will hold more cash holdings.

In this table, we find some conditions of control variables as follows. First, the common-law countries have higher Q than the civil-law countries; however, the civil-law countries have higher SIZE than the common-law countries. Second, NWC and LEVERAGE are substitutes for cash; firms can use them when they have cash shortfalls. Among these counties, we find that both LEVERAGE and NWC of the civil-law countries are higher than those of the common-law countries. Third, firms with higher idiosyncratic risk need to accumulate more cash to deal with a higher frequency of cash flow shortfalls. In Table 3, it reveals that the civil-law countries have higher idiosyncratic risk (DOL) than the

common-law countries.

Table 3 Summary statistics

This table provides mean, median, and standard deviation of whole variables in our analysis. The data are from 2002 to 2008. There are 735 samples for Germany, 812 for France, 6,027 for the US, and 1,456 for the UK. The US and the UK are common-law countries. Germany and France are civil-law countries. The total samples contain 9,030 observations. The detailed variable definitions are defined in Table 2.

Country Germany France United

Mean 0.121 0.039 0.096 0.027 0.062 0.080

Median 0.123 0.033 0.084 0.010 0.047 0.078

Standard 0.175 0.185 0.171 0.179 0.175 0.180

DOL

Mean 16.925 12.347 10.377 12.939 11.658 14.636

Median 3.431 2.398 2.190 2.338 2.264 2.915

Standard 58.699 41.116 32.183 51.590 41.887 49.908