DIFFERENTIATED CORPORATE GOVERNANCE STRUCTURES AND FIRM INVESTMENTS: THE EVIDENCE FROM EMERGING MARKETS TANWEER HASAN

Transcription

1 Preliminary Draft DIFFERENTIATED CORPORATE GOVERNANCE STRUCTURES AND FIRM INVESTMENTS: THE EVIDENCE FROM EMERGING MARKETS TANWEER HASAN Walter E. Heller College of Business Administration Roosevelt University Schaumburg, IL PALANI-RAJAN KADAPAKKAM College of Business Administration University of Texas San Antonio, TX and P. C. KUMAR* Kogod School of Business American University Washington, DC * Corresponding author A preliminary draft submitted for presentation at the Eastern Finance Association Conference, 2002

2 2 DIFFERENTIATED CORPORATE GOVERNANCE STRUCTURES AND FIRM INVESTMENTS: THE EVIDENCE FROM EMERGING MARKETS Abstract The quality of corporate governance has been shown to have wide-ranging implications, e.g., on the performances of stock markets and on exchange rates. This study investigates whether the quality of corporate governance impacts investment decisions made at the micro level of the firm. We focus our study on emerging markets since it has been known that they have widely varying standards of corporate governance. We develop an index of corporate governance for the countries in our sample drawing on data from published sources. We find an association between our index of corporate governance and factors that affect firm investments. In countries with high standards of corporate governance, firm investments are driven by market factors. We find that firm investments are determined by internal resources in countries with relatively low standards of corporate governance. We attribute these variations to differential protection afforded to minority shareholders.

3 3 DIFFERENTIATED CORPORATE GOVERNANCE STRUCTURES AND FIRM INVESTMENTS: THE EVIDENCE FROM EMERGING MARKETS 1. INTRODUCTION Corporate governance deals with the ways in which suppliers of finance to corporations assure themselves of getting a return on their investment [Shleifer and Vishny (1997)]. Potential expropriation of investors is a consequence of the classic managerial agency problem that arises from the separation of ownership and control in the corporate form of organization. Another result is the misallocation of corporate resources thus diluting the benefits of value-enhancing investments in the firm. A set of rights that are embodied in their legal contracts protect suppliers of capital to the firm. For example, absolute priority rules ensure that debt holders have the first claims on the income stream and assets of the firm. Equity shareholders have the right to vote on the appointment of directors and other matters relating to the firm. However, an environment in which these rights may be enforced is essential to investors. Such an environment is characterized by high standards of corporate governance. The presence (or absence) of corporate governance has wide-ranging effects. Johnson, Breech, Boone and Friedman (2000) [JBBF] report that corporate governance variables explain a greater portion of the variation in exchange rates and stock market performance during the Asian crises than macroeconomic variables. In this study, we broaden the scope of corporate governance by examining its impact at the micro level on the firm s operations, specifically, on its capital investment decision. Two themes have evolved in the literature to explain corporate investments. The agency perspective posits that managerial ownership influences corporate investments, which, in turn, determine firm value. On the other hand, managerial compensation plans imply that it is market value of the firm that determines the ownership of the corporation in general, and managerial ownership, in particular. It is conceivable that firm ownership, its market value, and corporate investments are linked together in simultaneous relationships. Second, the literature on capital financing constraints

4 4 indicates that, while they may be imposed for a variety of institutional reasons, their important consequence may be to limit corporate investments. This study attempts to integrate these themes. It specifically examines whether the presence (absence) of high quality corporate governance standards serves to mitigate (tighten) financing constraints on corporate investments. We analyze a sample of corporations drawn separately from each of seven East Asian emerging markets, namely, Hong Kong, Indonesia, Korea, Malaysia, Singapore, Taiwan, and Thailand. 1 We develop an index of corporate governance (ICG) for each country drawing information from published sources. We estimate a common model of corporate investments for each emerging market. We further break down each country s data into sub-samples to examine whether the investment relationship is affected by variations in firm size and age. We attempt a rank ordering of the emerging markets based on the degree of dependence of corporate investments on the independent variables. We relate this rank ordering with a rank ordering of each emerging market s ICG value. Our research suggests that in environments with high standards of corporate governance, firm investments are related more to market factors than to the firm s internal resources. In environments with low standards of corporate governance, firm investments relate more to their internal resources than market factors. Specifically, we find that corporate investments in Singapore, which has the highest ICG score, are determined largely Tobin s q to the exclusion of other independent variables. On the other hand, corporate investments in Indonesia, which has the lowest ICG score, are determined by the firm s internal resources, such as cash flow. The other countries in our sample fall in levels between these extremes. The rest of this paper is organized as follows. The next section discusses the factors affecting firm investments, specifically, the impacts of ownership structure and corporate value, financing constraints which result in the firms relying on their internal resources to finance their investments, and corporate governance. Section 3 develops the perspective that in environments with high corporate governance standards, firm investments are easily financed by market-based capital; whereas, in environments with 1 We recognize that Hong Kong and Singapore do not conform strictly to IFC s definition of emerging markets.

5 5 low corporate governance standards, firm investments are likely to be financed with limited internally generated capital. Section 4 develops an index of corporate governance for the countries in our sample drawing on previous work by LaPorta, Lopez-de-Silanes, Shleifer, and Vishny [(1997), (1998)] [LLSV]. Section 5 describes the model of firm investments and presents the empirical results. It also investigates whether firm size and age influence the basic relationships. Section 6 organizes the countries in our sample into four levels. Singapore is placed in the first level and its firm investments are seen to be determined by market variables; Singapore is also associated with the highest ICG score. Indonesia is placed in the last level and its firm investments are determined by internal resources; Indonesia is also associated with the lowest ICG score. The other emerging markets find their places in levels between these extremes. Finally, section 7 concludes with some policy implications. 2. FACTORS INFLUENCING CORPORATE INVESTMENTS Several diversified themes characterize the literature on corporate investments. Ownership structure influences firm investment, which in turn affects corporate value. However, managerial compensation plans may dictate that firm market value determines corporate ownership. Thus, a circular relationship is posited among ownership structure, corporate value, and firm investment. A second approach suggests that external financing constraints, such as lack of access to capital markets, may limit firm investments. Finally, corporate governance (or the lack thereof), which encompasses the rights of investors and the legal environment in which these rights are enforced, may be a determinant of varying market performances across countries [JBBF]. We posit that corporate governance may be an exogenous factor influencing the ownership-firm investment-corporate value nexus. The following sub-sections expand these themes. 2.1 Ownership Structure, Corporate Value, and Firm Investment The distribution of corporate ownership between insiders and external investors has been posited to affect firm value. The impact of ownership structure on the firm s value is described as a two-stage process in which corporate investment is the linking variable: in the first stage, ownership structure influences investment, which in turn affects firm value in the second stage [Jensen and Meckling (1976)]. Empirical analyses

6 6 reveal that the relation between ownership structure and corporate value is nonmonotonic [Morck, Shleifer and Vishny (1988) and McConnell and Servaes (1990)]. Low levels of managerial ownership suggest close alignment of managerial and shareholder interests and any marginal increase of managerial ownership will result in enhanced corporate value. However, high levels of managerial ownership signify management entrenchment and the consequent insulation of managers from the discipline of market forces; thus a marginal increase in managerial ownership will result in decreased corporate value. Conventional wisdom has accepted the exogeneity of ownership structure to investment and valuation. However, Demsetz and Lehn (1985) argue that ownership structure may be determined endogenously in an equilibrium relation. Managers, for example, are observed to prefer equity compensation when their firm performance is expected to improve with consequent increases in corporate values. Furthermore, managerial compensation is positively correlated with corporate performance implying that ownership structure is an endogenous outcome of the compensation contract process. A reversal in the causal relation results in ownership structure being determined by corporate value instead of corporate value being determined by ownership structure. Thus, a circular relation exists among ownership, corporate investment, and firm value [Cho (1998)]. 2.2 Liquid Assets and Firm Investment Earlier work has concluded from the observed positive relationship between cash flow and firm investment that there are external financial constraints limiting these investments. These studies estimate the sensitivity of a firm s investment to measures of internal funds after controlling for the availability of investment opportunities. In other words, if a model for investments includes both Tobin s q and internal cash flow, in a perfect world, q should be the only influencing variable and the impact of cash flow should be insignificant. A positive, significant coefficient for the internal cash flow variable has been interpreted as indicating the presence of external financing constraints. The challenge faced in the empirical analyses lies in the identification of the appropriate variables that reflect the financial and institutional structures responsible for the unobservable financial constraint.

7 7 Several segmenting variables have been used to impound external financing constraints. In the first study of this issue, Fazzari, Hubbard and Petersen (1988) (FHP) use dividend payout ratio as the proxy variable. The underlying intuition is that firms treat dividend payments as the outcome of a passive residual policy, i.e., the payments of dividends are subordinated to investment opportunities; thus, the firm s dividend payout ratio is a good indicator of the availability of surplus internal funds. Firms with low dividend payouts are identified as being financially constrained, whereas high dividend paying firms do not fall into this category. FHP report that the investment-cash flow sensitivity is higher for low dividend payout firms compared to high dividend payout firms. Kaplan and Zingales (1997) (KZ) reexamine the low dividend payout firms in the FHP sample to fine-tune the classification of firms into constrained and unconstrained categories by objective criteria drawn from 10-K reports. They find that firms with true financial constraints (as per their definition) do not necessarily have low dividend payout levels and vice-versa as assumed by FHP. More importantly, contrary to the FHP results, KZ find that financially constrained firms exhibit lower cash flow-investment sensitivity. They conclude by issuing a general caution about the use of the investment-cash flow coefficient as an indicator of financial constraint. Most of the earlier studies had focused their attention on a particular country. For example, FHP and Vogt (1994) examine the issue in the context of firms in USA; Schaller (1993) considers a group of Canadian firms; Athey and Laumas (1994) direct their study towards Indian firms; Hoshi, Kashyap and Scharfstein (1991) study Japanese firms; and Devereux and Chiantarelli (1990) investigate the issue for UK corporations. In a departure from previous studies, Kadapakkam, Kumar and Riddick (1998) (KKR) examine the issue separately for six OECD countries Canada, France, Germany, Japan, UK and USA. Furthermore, they investigate the direct influence of firm size, which very few of these earlier studies had considered. KKR first examine the entire sample of firms in each country, and report that corporate investment is dependent on the availability of internal resources in all countries. They then repeat the analysis segmenting each country s sample using three measures of firm size. Contrary to their expectations, KKR find that investment-cash flow sensitivity is generally highest in the large firm size group and smallest in the small firm size group. The authors deduce that

8 8 the explanations for these findings are grounded in managerial agency considerations, and in the greater flexibility enjoyed by large firms in timing their investments. They conclude that the degree of sensitivity of a firm s investments to its cash flows cannot be interpreted as an accurate measure of its access to capital markets. Firm size may be a preferred measure since small firms are known to have less access to external markets. 2.3 Corporate Governance and Firm Investment The separation of ownership and control is at the root of the classic agency problem in the corporate form of organization which may be resolved by better investor protection. Managerial opportunism resulting in expropriation of investors or misallocation of corporate resources reduces value-enhancing investments in the firm. In exchange for their resources, investors accept a set of rights embodied in legal contracts. Debt holders are offered a stream of interest payments with a prior claim on the income stream of the firm over other capital suppliers. But in the event of default on these payments, debt holders accept the right to seize possession of any collateral offered. Equity shareholders are residual claimants on the income stream and assets of the firm, and in exchange for this status, they accept the entitlement to vote on the appointment of the directors of the firm. These contracts confer on investors the rights, and thus the means, to protect their investments in the corporation. LLSV (1998) imply that these rights form a set of necessary, but not sufficient, conditions. What is needed additionally is the legal environment in which these rights may be enforced. Differences in legal protections offered to investors may determine variations in corporate investments among countries. Thus, corporate governance is an exogenous factor in the relationship linking ownership, firm investment, and corporate value. 3. CORPORATE GOVERNANCE AND THE LIQUID ASSETS-FIRM INVESTMENT NEXUS This section suggests that the extent to which the availability of liquid assets influences firm investment is conditioned by the corporate governance environment. Furthermore, while large shareholders have the ability to deal with the managerial agency problem, minority shareholders rely on an effective corporate governance climate to protect their interests. Thus the presence (or absence) of corporate governance has wide-

9 9 ranging effects; the recent Asian crisis has been attributed to ineffective corporate governance systems in some of these countries [JBBF]. This section elaborates on these issues and develops the objectives of this study. 3.1 Corporate Governance and Concentration of Shareholdings Concerted action by a small number of investors who collectively have claims to a large cash flow stream is more effective than actions taken by a large number of small investors with vested control rights. Shleifer and Vishny (1998) observe in this regard that concentration of ownership leverages up legal protection. Large shareholders are more effective in dealing with the managerial agency issue as they are likely to be more unanimous in their objectives and to have adequate control for enforcement. However, large shareholders can be effective only if their control is reinforced by the extent of legal protection afforded to their votes. Thus an efficient legal environment protects shareholder rights against interference by management. Effectual corporate governance by minority shareholders is possible only in countries with relatively sophisticated legal systems. As a result, there is likely to be more concentration of shareholdings in countries with poor investor protection. 3.2 Corporate governance and the Asian crisis JBBF (2000) examine whether weaknesses of legal institutions in enforcing corporate governance principles were responsible for the exchange rate depreciation and stock market declines in emerging market countries during the recent Asian crisis. They argue that a decline in the expected returns on investments (say, as a result of a combination of some exogenous factors) together with increases in managerial expropriation would result in a reduction of local and foreign investors confidence. This development would result in increases in capital outflows and declines in capital inflows leading to exchange rate depreciation and stock market declines. The authors find from their cross-country regressions that corporate governance variables explain a greater proportion of the variation in exchange rates and stock market performance during the Asian crisis than do macroeconomic variables. They conclude that countries with weak

10 10 corporate governance traditions (particularly, inadequately enforced minority shareholder rights) are vulnerable to these variations The Impact of Corporate Governance on the Liquid Assets-Firm Investments Relationship The objective of this study is to integrate two separate themes in the literature on the determinants of the firm s real investments quality of corporate governance in the economic environment in which the firm operates and the availability of liquid assets to the firm. The degree of observance of corporate governance principles varies across countries. We expect that countries with strong corporate governance traditions, firms would find adequate financing for their investments from market sources. Such firms should exhibit less firm investment-liquid assets dependency than countries with weaker corporate governance traditions. In the latter countries (with weaker corporate governance traditions), expropriation of investors or misallocation of corporate resources (over/under-investment) would result in sub-optimal value-enhancing investments in the firm. Markets would not be forthcoming in financing firm investments. In such instances, firms would be dependent on their internal resources to finance their investments, which would be constrained by the extent of availability of liquid assets. 3. CORPORATE GOVERNANCE STRUCTURES IN EMERGING MARKETS LLSV [(1997), (1998)] infer that a necessary condition for the widely-held corporation (a la Berle and Means), characterized by dichotomy of ownership and control, is likely to be more prevalent in countries with good legal protection of minority shareholders. In countries with poor traditions of protecting the interests of minority shareholders, losing control of the corporation involuntarily and being denigrated to the status of a minority shareholder is a costly proposition for controlling shareholders. Hence, they would take all necessary action to ensure preservation of their control, such as accumulating voting rights. Controlling shareholders would have less interest in 2 Karmin (2000) also attributes the poor performances of emerging markets to inadequate standards of coporate governance the equitable treatment of minority shareholders and the timely and accurate disclosure of information. There is substantial resistance to reforming the corporate governance climate in Russia, Korea, Brazil and Poland. It is expected that reforms

11 11 selling their shares in the open market place. On the other hand, in countries offering good legal protection of minority shareholders, controlling shareholders are less apprehensive of being expropriated, or taken advantage of, if they were to lose control through a takeover. In such countries, controlling shareholders may not be confronted with the tradeoff between reducing their ownership rights by selling a part of their shares to enhance liquidity or to diversify their portfolios, and outright expropriation. Thus corporations are likely to have highly concentrated share ownership in countries which do not afford adequate minority investor protection. The magnitude of the potential consequences of inadequate minority shareholder protection may be influenced by firm size. Controlling shareholders in small firms have an overwhelming degree of control which offers them more expropriation opportunities. With a view to preserving their dominance, these controlling shareholders are unlikely to expand the small firm s investments by a new share issue with its potential for loss of control. Even if they were to approve the issuance of new shares, given the potential for expropriation, market response to the share issue is likely to be unenthusiastic. The net result is that investments in small firms are likely to be more dependent on internal resources. The degree of control by controlling shareholders in large firms is less and hence their potential for expropriation is relatively less severe. Hence, if the controlling shareholders were to expand firm investments with a new share issue, market response is likely to more encouraging. Thus, investments in large firms are less likely to be dependent on internal resources. The foregoing discussion leads to empirically testable hypotheses. Comparative country measures of corporate governance, which are central to such an empirical investigation, are discussed in the following section. 4.1 An Index of Corporate Governance We create an index of corporate governance (ICG) by aggregating the values assigned to seven variables from published data. This index is described for a set of seven emerging market countries, Hong Kong, Indonesia, Korea, Malaysia, Singapore, Taiwan and Thailand in Table 1. will be adopted in these countries when it is recognized that companies with good corporate governance practices provide better insulation against market downturns.

12 12 Table 1: Index of Corporate Governance Variable Hong Indonesia Korea Malaysia Singapore Thailand Taiwan Kong Judicial Efficiency Corruption Rule of Law Enforceable Minority Shr. Holder Rights Anti Director Rights Creditor Rights Accounting 69 69* Standards ICG Source: JBBF (2000) *Value of the acccounting standards variable is not available for Indonesia. It is estimated as the average value for the other six countries. Following definitions of variables are from LLSV (1998), except for Minority Shareholder Rights which has been extracted from JBBF (2000): Judicial Efficiency: Source: Business International Corporation (BIC - country risk rating agency). Efficiency and integrity of the legal environment as it affects business, particularly foreign firms. It may be taken to represent investors assessments of conditions in the country in question. Average of values obtained in the period Values are scaled between 0 and 10, with lower scores implying lower efficiency levels. Corruption: Source: International Country Risk (ICR - country risk rating agency). ICR s assessment of corruption in the government. Lower scores indicate that high government officials are likely to demand special payments and illegal payments are generally expected throughout lower levels of government in the form of bribes connected with import and export licenses, exchange controls, tax assessment, policy protection and loans. Average of values obtained in the period Values are scaled between 0 and 10, with lower scores implying higher levels of corruption.

13 13 Rule of law: Source: ICR. Assessment of law and order tradition in the country. Average of values obtained in the months of April and October of the monthly index obtained in the period Values are scaled between 0 and 10, with lower scores implying less tradition for law and order. Minority Shareholder Rights: Source: Flemings Research. An alternative measure of corporate governance which tries to capture the extent of shareholder rights in practice. It considers the disclosure of information, transparency of ownership structures, management and special interest groups, adequacy of the legal system, whether the standards that are set are actually enforced, and if the boards of companies are independent and the rights of minority shareholders are upheld. Values are scaled between 1 and 5, with lower scores implying less shareholder rights. Anti-director Rights: LLSV (1998) create this index by aggregating shareholder rights. The authors form this index by adding 1 when (i) the country allows shareholders to mail their proxy form to the firm, (ii) shareholders are not required to deposit their shares prior to the general shareholders meeting, (iii) cumulative voting or proportional representation in the board of directors is allowed, (iv) an oppressed minorities mechanism is in place, (v) the minimum percentage of share capital that entitles a shareholder to call for an extraordinary shareholders meeting is less than or equal to 10 percent (the sample median), or (vi) shareholders have preemptive rights that can be waived only by a shareholder s vote. Values of the index range between 1 and 6. Creditor Rights: LLSV (1998) create this index by aggregating different creditor rights. The authors form this index by adding 1 when (i) the country imposes restrictions, such as creditors consent or minimum dividends to file for reorganization; (ii) secured creditors are able to gain possession of their security once the reorganization petition has been approved (no automatic stay); (iii) secured creditors are ranked first in the distribution of the proceeds that result from the disposition of the assets of a bankrupt firm; and (iv) the debtor does not retain the administration of its property pending the resolution of the reorganization. Values of the index range between 0 and 4. Accounting Standards: LLSV (1998) create this index by examining companies 1990 annual reports for their inclusion or exclusion of 90 items and rating them accordingly. The authors note that these items fall into seven categories (general information, income statements, balance sheets, funds flow statements, accounting standards, stock data, and special items). A minimum of three companies in each country were studied. The companies represent a cross-section of various industry groups; industrial companies represented 70 percent, and financial companies represented the remaining 30 percent.

14 14 5. EMPIRICAL DESIGN, RESULTS, AND ANALYSES We first determine the extent to which the availability of liquid assets determines firm investment in the entire sample in each of the seven emerging market countries. 3 We then segment the entire sample into three groups (low, medium and high) by firm size and repeat the analysis for each size group. Two different measures of firm size are used, i.e., book value of total assets and market value of firm equity. 5.1 Empirical Procedure We estimate the following relationship: I it = b 1 CF it + b 2 CS it-1 + b 3 Q it-1 +b 4 S it-1 + e it, where I it is the change in the level of net fixed assets (investment) in the ith firm during year t. Investment is normalized by the level of net fixed assets to account for differences across firms. CF it is the ratio of cash flow generated by the firm during the year, measured as the sum of net income plus depreciation, amortized intangibles, and deferred taxes less dividends, to the level of net fixed assets at the beginning of the year. CS it-1 is the ratio of cash stock, calculated as the sum of cash and marketable securities held at the beginning of the year, to the level of net fixed assets at the beginning of the year. Q it-1 (a proxy for Tobin s q) is measured as the ratio of firm market value to firm book value. Firm market value is calculated as the sum of market value of outstanding common equity and book value of long-term debt and preferred stock at the beginning of the year. Market values of equity have been calculated on the basis of the average of the high and low values of the stock price for the year. Firm book value is measured as the book value of common equity plus the book value of long term debt and preferred stock. A positive relationship between investment levels and Tobin s q is expected, since a higher value for the latter indicates better financial prospects, either due to increased profit levels and/or due to a decrease in the cost of capital, both of which signify enhanced investment opportunities. S it-1 is the ratio of lagged sales to net fixed assets and is included in the model to reflect the sales accelerator theory of investment; e it is the error term. 3 Liquid assets in our discussion broadly include cash flow from operations and cash stock in current assets. However, the two variables are treated as separate independent variables in the regressions.

15 15 Under the null hypothesis that firms have ready access to external capital, variables measuring internal funds should not influence investment levels. However, under the alternate hypothesis that lack of availability of internal funds constrains firm investment, positive coefficients are expected for both cash variables. These hypotheses are stated as follows: H 0 : b 1 = b 2 = 0 H A : b 1 > 0 and b 2 > Description of the Data Appendix 1 provides the descriptive statistics for the seven countries. 4 Mean investment as a proportion of net fixed assets (I it ) ranges from 0.32 (Indonesia) to (Korea). The dispersion of the investment variable within each country is relatively uniform (range of standard deviation (sd) = ). Mean cash flow values (CF it ) vary between 0.60 (Hong Kong) and 0.11 (Malaysia and Taiwan) with maximum variability in Hong Kong (sd = 1.00) and minimum in Malaysia (sd = 0.13). 5 The mean value of the relative cash stock variable falls between 0.83 (Korea) and 0.09 (Thailand). The dispersion of relative cash stock ranges between 1.03 (Hong Kong) and 0.18 (Malaysia, Taiwan). Thailand registers the highest mean value of Tobin s q (2.36), while Korea has the lowest value (0.96). Standard deviation for Tobin s q is the highest for Malaysian firms (1.61) and the lowest in Korea (0.53). 5.3 Results for Aggregate Country Samples We present the regression results for the aggregate sample in each of the seven countries in Appendix 2. The use of panel data calls for transformation of the variables for each firm in the regressions to adjust for fixed effects and random effects. The variables have been transformed into deviations from the time series mean value. 6 If access to external capital markets is unconstrained, then our measures of internal funds should not influence investment levels. The results summarized in Appendix 2 4 The extreme values defined as the top and bottom one percent of the sample observations have been deleted for all variables. The elimination of outliers leads to more robust results. 5 Depreciation data for Hong Kong were not reported in the data base. Hence, the annual depreciation values used in the computation of net fixed assets and cash flow were imputed as follows: Buildings: 1/30; property: 1/10; vehicles: 1/15; other assets: 1/7. Information obtained from other sources confirmed that straight line depreciation is the common method adopted. 6 See Hsiao (1998) for a discussion of this procedure.

16 16 indicate that the coefficients for cash flow are positive and significant in all countries except Singapore. The coefficient for the cash stock variable is significantly positive in Hong Kong, Indonesia and, Korea and Taiwan, while Tobin s q has significantly positive coefficients in Hong Kong, Korea, Singapore, and Thailand. The coefficient for lagged sales is significantly positive in Malaysia, Singapore, Taiwan, and Thailand. It can be argued that cash flows may be related to Tobin s q, since they imply potentially better investment opportunities and thus feed back in higher Tobin s q values. We re-estimate the model discarding cash flow to assess the incremental contribution of this variable. The results of these truncated regressions are similar to the results described in the preceding paragraph. We observe that the explanatory power of the regression has declined in Indonesia, Korea and Malaysia. We therefore conclude that cash flow variable has independent explanatory power and reflects effects not captured by Tobin s q in at least these three countries. 7 Charts 1 and 2 display the coefficients of cash flow and Tobin s q respectively from the regressions in relation to the ICG scores for the seven countries. There appears to be a declining relationship between the cash flow coefficient and ICG score. Corporate investments are driven by the availability of cash flow in countries with low ICG scores. Hong Kong and Singapore, which exhibit the highest ICG scores, have the smallest cash flow coefficients. The relationship between the coefficient for Tobin s q and ICG score exhibits a different pattern. Hong Kong and Singapore have high ICG scores and higher values for the Tobin s q coefficient, whereas the other countries with lower ICG scores are clustered towards the bottom of the vertical axis. 8 These charts offer broad support for our hypothesis that firms operating in environments that offer high standards of corporate governance have access to capital market to fund their investments. On the other hand, firms operating in poor corporate governance environments rely on internal sources for financing their investments. 9 7 In the case of Malaysia, dropping the cash flow variable results in a positive coefficient for the cash stock variable, although it is still insignificant. 8 The exception to the patterns observed in the charts is Malaysia. Even with a high ICG score (109.2), it has the highest cash flow coefficient and the lowest coefficient for Tobin s q. 9 We also considered an alternate index of corporate governance. The Economist (4.7.01) published evaluations of three elements of corporate governance environment in the countries in our sample, namely, quality of the legal system, quality of corporate governance, and

17 17 (Insert Charts 1and 2 here) 5.4 Does Firm Size Matter? Impediments to corporations accessing capital markets may explain the strong positive relationship between investments and internal cash flow (except for Singapore). A better test of this hypothesis can be obtained by segmenting each country s sample. The lack of internal funds should constrain investments more severely for firms with greater costs of accessing external capital markets. Such costs may be manifest in higher floatation costs and/or may be associated with informational asymmetries between corporate insiders and external investors. However, firms having greater costs of access to external capital markets may be identified by their unique characteristics. For example, there may be less information asymmetry regarding the prospects of large, mature firms, whereas outsiders may have less information than insiders about the future of small, young firms. Claessens, Djankov and Lang (2000) [CDL] in their examination of corporate ownership structure in E. Asian countries provide another justification for segmentation of the sample. These authors conclude that size and age of the firm are important factors affecting differences in the extent of separation of ownership and control across and within these countries. 10 Recognizing that the need for effective corporate governance stems from the separation of ownership and control, we use firm age and two measures of firm size to segment our sample. The two definitions of firm size, namely, book value of total assets and market value of equity, validate our transparency. These evaluations are on a 10-point scale with a score of 10 indicating poor quality and 0 indicating high quality. We derive an index of corporate governance as the sum of the values for the three elements. Thus the index values are: Hong Kong: 10.50; Indonesia: 25.75; Korea: 23.85; Malaysia: 19.00; Singapore: 6.80; Taiwan: 19.00; Thailand: The correlation coefficient between ICG values and Economist index values is 0.88 (recall that the ICG are ordered such that high values indicate high quality and low values indicate poor quality). The correlation coefficients between the Economist index and the cash flow regression coefficient and with the Tobin s q regression coefficient are 0.61 and 0.87, respectively; the correlation coeffients between the ICG and the two regression coefficients are 0.19 and 0.60, respectively. Thus the two measures of the corporate governance environment reinforce the findings reported here. 10 The authors identify variations in company law across countries and economic development of the specific country as other factors affecting the degree of separation of ownership and control.

18 18 segmentation procedure. We identify firms belonging to the top, middle, and bottom third groups and estimate the investment regressions separately for these three segments for the two definitions of firm size. 11 Appendices 3 and 4 contain the details of the investment regressions and Table 2 summarizes these results. (Insert Table 2 here) Discussion of the firm size segmentation results The regression equation has been estimated separately for each of the three subgroups created for each definition of firm size (book value of total assets and market value of the firm), i.e., a total of six subgroups. An examination of the results from the Hong Kong sample across the two definitions of firm size reveals that coefficients for cash stock and Tobin s q are significantly positive in all six subgroups. Furthermore, the cash flow variable has significantly positive coefficients in the low and intermediate tertiles for both definitions. Thus the cash flow variable has significant and positive coefficients in four of the six subgroups. Finally, in the Hong Kong sample, the coefficients for lagged sales are not significant in any subgroup. A plausible interpretation is that while firms have access to external capital, they rely on liquid resources to finance their capital investments. This strategy may be driven by political considerations. The general uncertainties surrounding the accession of Hong Kong to PRC in 1998 may have encouraged firms to rely more on liquid resources. A similar examination of the results for Indonesia reveals that the cash flow variable has significantly positive coefficients in all six sub-groups. The cash stock variable has a significantly positive coefficient in two of the six subgroups. Tobin s q has one significantly negative coefficient and the lagged sales variable has no significant coefficient in any subgroup. These results do not support the null hypothesis that firms have ready access to external capital. On the contrary, the positive coefficients for the cash flow variable clearly indicate that corporate investments are driven by the availability of internal funds. The results for Korea indicate significantly positive cash flow coefficients in four of the six sub-groups (medium and high tertiles). Cash stock has significant coefficients 11 We chose to delineate sub-samples in tertiles in order to preserve adequate number of observations in each sub-group.

19 19 in three sub-groups; Tobin s q and lagged sales have two and three significant coefficients, respectively. We interpret these results as indicating that investments by Korean firms are determined by cash flow. In the case of Malaysia, the cash flow variable exhibits significantly positive coefficients in all six sub-groups. The lagged sales variable has significantly positive coefficients in four subgroups (low and intermediate tertiles). The cash stock variable has no significant coefficient in any subgroup and Tobin s q has four significant coefficients among the six subgroups. There is a consistency to the Malaysian sample in that only the cash flow variable has significantly positive coefficients in both definitions of firm size. Lagged sales have significant coefficients in the small and medium size subgroups. Thus, we conclude that Malaysian firms have some access to external capital but tend to rely on internal resources. The results for the Singapore sample indicate that the cash flow variable is not significant in any sub-group. The coefficient for the cash stock variable is significant in two subgroups. The coefficient for Tobin s q is significant in all subgroups and three subgroups exhibit significant coefficients for the lagged sales variable. Tobin s q is the only significant and consistent determinant of investment by firms in Singapore. Hence, it can be concluded that there is support for the null hypothesis that firms have ready access to external capital. Cash flow and cash stock coefficients are significant and positive in four subgroups in Taiwan. Tobin s q has a significant coefficient in one subgroup, whereas the coefficient for lagged sales is significant and positive in all the six subgroups. Investments by firms in Taiwan are driven primarily by lagged sales (sales accelerator hypothesis) and by cash flow, leading to the conclusion that firms do not have ready access to market capital. Finally, in the case of Thailand the cash flow variable has significant coefficients in four subgroups. Cash stock has no significant coefficient in any subgroup. Tobin s q and lagged sales have significant coefficients in four and three subgroups, respectively. We conclude that firms in Thailand have some access to capital markets but tend to rely on internal resources.

20 20 In summary, segmentation of the data by firm size indicates that corporate investments in Indonesia, Korea, Malaysia, and Taiwan are strongly influenced by cash flow. With the exception of Malaysia, these countries have low ICB scores. In contrast, firm investments in Singapore, which has the highest ICG score, are driven by Tobin s q. Corporate investments in Hong Kong and Thailand are determined by cash flow and by Tobin s q. Hong Kong has the second highest ICB score and Thailand has the lowest ICG score Test for a hierarchy among the firm size regression coefficients Do small (large) firms rely on their internal resources to finance their investments to a greater degree than large (small) firms do? We would expect a priori that asymmetric information is less in the case of large firms than in small firms and hence large firms should have better access to capital markets to finance their investments. The answer to this question may be revealed by comparative analyses of the results in the sample sub-groups. Segmentation of the sample into tertiles by the particular variable selected, i.e., book value of total assets and market value of the firm, is a useful procedure to establish whether a hierarchical pattern exists among the sub-groups. For example, if the coefficients for the cash flow variable were significantly positive (and the coefficients for Tobin s q were not significant), and in addition, the magnitude of these coefficients decreased from the low tertile to the high tertile, then we would conclude that small firms depend on internal resources to finance their investments to a greater extent than large firms do. Conversely, if the coefficients of Tobin s q were significantly positive (and the coefficients of the other independent variables were not significant), and in addition, the magnitude of this coefficient increased from the lowest tertile to the highest tertile, then the conclusion is that large firms have better access to external capital markets than small firms have. Table 3 presents the results of a test of hierarchical differences in the regression coefficients. (Insert Table 3 here) In the following discussion, we examine firm segments for hierarchies only in those variables whose coefficients are significant. The Hong Kong regression results indicate the existence of a hierarchy in the cash flow coefficients in both firm size segments. We find that, in both measures of firm size, investments of small and

21 21 intermediate firms are more cash flow-dependent than investments of large firms. While firm investments in all three segments are driven significantly by cash stock and Tobin s q, there is no distinct hierarchy among these coefficients. The Indonesian regressions lead to the conclusion that investments of large firms are more cash flow-dependent than investments of small or intermediate firms (particularly, in the case where size is measured as market value of equity). The Korean results suggest that while both cash flow and cash stock influence firm investments, there is a hierarchy in the cash flow coefficients in the book value of total assets measure. The impact of cash stock is the greatest in small firms and its impact in intermediate firms is greater than its impact in large firms. Among Malaysian firms, while the coefficients of cash flow are significant in the two size segments, there is no observable hierarchy among them. However, in the case where size is measured as market value of equity, there is a distinct hierarchy in the coefficients of Tobin s q. Investments in large and intermediate size firms are influenced by Tobin s q to a greater extent than in small firms. The Singapore results suggest that the influence of Tobin s q on firm investments is more significant in large firms than in small or intermediate firms. This effect is particularly noticeable when firm size is measured as book value of total assets. Among Taiwanese firms, investments are driven by cash flow and liquid assets to a greater extent in small and intermediate-size segments than in the large-size segment, particularly when size is measured by market value of equity. Finally, the results for Thailand indicate that the influence of cash flow on firm investments is greater in large and intermediate-size firms than in small firms in the market value of equity size measure. In summary, a distinct hierarchy in the firm investment-cash flow/cash stock dependency is observed among the firm size groups segmented by book value of total assets and/or market value of equity in Korea. The impact of cash flow/cash stock is the greatest in small firms, and the impact on intermediate firms is greater than the impact on large firms. In Hong Kong and Taiwan, there is a moderate hierarchy in the sense that the impact on small or intermediate firms is greater than the impact on large firms. In Indonesia, the hierarchy is reversed in the sense that cash flow dependency is greater in

22 22 large firms than in intermediate or small firms; whereas, in Thailand, cash flow dependency is greater in large and intermediate-size firms than in small firms. In Malaysia, there is a moderate hierarchy in the impact of Tobin s q in that the coefficients for large or intermediate-size firms are greater than the coefficient for small firms. A moderate hierarchy is also observed in Singapore where the coefficient for Tobin s q for large firms is greater than the coefficients for intermediate or small firms. 5.5 Does Firm Age Matter? Appendix 5 details the results of the regression analyses for the three sub-samples segmented by firm age. The Hong Kong data indicate that cash stock and Tobin s q are the only variables with positive, significant coefficients in all three firm age sub-groups. The firm size regressions had positive, significant coefficients for cash flow, in addition to cash stock and Tobin s q. The results for Indonesia preserve the patterns observed in the two firm size regressions in that only cash flow has significantly positive coefficients in all firm age sub-groups. The results for Korean firms signify that only cash flow has significant, positive coefficients in all three sub-groups. The Malaysian data follow the same pattern, indicating that cash flow is the only variable with significant, positive coefficients in all three sub-groups, thus replicating the results in both measures of firm size. Tobin s q is the only variable with positive, significant coefficients in all three subgroups among Singapore firms, thus following the pattern observed in both segmentations based on firm size. The Taiwan data indicate that cash flow has significant, positive coefficients in the low and intermediate sub-groups and lagged sales has positive, significant coefficients in all the three sub-groups. These results are similar to the results obtained for both segmentations by firm size. Finally, the results for Thailand signify that cash flow has significant, positive coefficients for the low and intermediate sub-groups. These results differ from those obtained with segmentation based on firm size, in which the intermediate and high sub-groups have significant, positive coefficients Test for a hierarchy among firm age regression coefficients Table 4 details the tests for hierarchies in the regression coefficients from the subsamples segmented by the age of the firm. The Hong Kong results reveal that while cash stock and Tobin s q have positive, significant coefficients, their pair-wise differences