Corporate Transparency, Financial Development and the Allocation of Capital: Empirical Evidence
Ashan Habib ([email protected]) is a Senior Lecturer in the Department of Accounting, Auckland University of Technology.
I appreciate the helpful comments from three anonymous reviewers and the editor, Graeme Dean.
Abstract
There is controversy regarding the role of financial development in promoting economic growth. Lucas (1988) suggests that the role of financial intermediation in economic growth has been very badly over-stressed in the popular and professional discussion. Levine et al. (2000), on the other hand, show that in a cross-country setting the exogenous component of financial intermediary development is positively and robustly linked to economic growth. Although empirical methodologies to investigate the finance-growth nexus have been refined, there is a lack of understanding about the exact mechanisms through which the financial system could affect economic performance in the real sector. Wurgler (2000) investigates one such mechanism of economic growth: whether capital is allocated efficiently. He then empirically shows that countries with well-developed financial architecture improve capital allocation. This article extends Wurgler (2000) by investigating the role of an important economic institution, the financial reporting system, on the efficiency of capital allocation. Financial reporting provides the primary source of independently verified information to the capital providers about the performance of managers and facilitates efficient resource allocation decisions. Results show that financial transparency is positively and significantly related to capital allocation efficiency. Further, this result holds after controlling for the impact of stock price synchronicity, state-owned enterprises and investor protection rights.
A fundamental factor in wealth creation in an economy is the efficiency with which scarce capital is allocated to profitable investment opportunities. This requires firms’ managers to allocate capital to positive net present value (NPV) projects, avoid projects that generate negative NPVs, and withdraw capital from projects found to be losers at some point after project initiation when such expected losses are confirmed.
Economic theories posit that formal financial markets and associated institutions improve the capital allocation process. Diamond (1984) develops a theory of financial intermediation where a financial intermediary (like a bank) has a net cost advantage in monitoring loan contracts (also Boyd and Prescott, 1986). Another theory is that efficient market prices help investors distinguish good investments from bad ones through a mechanism like Tobin's Q. Agency theory postulates that pressures from external investors (primarily institutional investors), as well as increased managerial ownership, encourages managers to pursue value-maximizing investment projects (Jensen, 1986), and strong investor protections provided by enforced laws facilitate the flow of finance to good projects (La Porta et al., 1997). Because countries around the world differ along these dimensions, allocation efficiency and hence economic growth are expected to vary accordingly.
Wurgler (2000) directly examines the relationship between the characteristics of a country's financial development and the efficiency with which capital is allocated to investment projects. Wurgler finds that generally countries with developed financial sectors increase investment more in their growing industries, and decrease investment more in their declining industries, than those with undeveloped financial sectors. The efficiency of capital allocation is improved with stock markets efficiently impounding more firm-specific information into individual stock prices, with less state ownership and strong minority investor rights. Wurgler's work is an important contribution to the debate concerning the role played by the development of financial intermediaries in economic growth because it focuses on an important channel through which economic development is affected. Although there now exists a large literature on the role of financial development in economic growth, little attention has been paid to understanding the channels through which financial development affects economic development (Zingales, 2003).
However, an important economic institution that is also expected to affect the capital allocation process is the financial reporting system. Financial reporting provides the primary source of independently verified information to the capital providers about the performance of managers (Sloan, 2001). This facilitates efficient resource allocation decisions by signalling changing investment opportunities to managers and outside investors, disciplining self-interested managers to invest in value-maximizing projects, and reducing firms’ cost of capital (Bushman et al., 2006). Bushman and Smith (2001, p. 304) argue that the efficiency of capital allocation depends upon ‘the extent to which managers identify value creating and destroying opportunities, the extent to which managers are motivated to allocate capital to value-creating investments and withdraw capital from value-destroying investments, and the extent to which capital is available to invest in value creating opportunities’. The financial reporting system, particularly financial accounting information, is expected to facilitate capital allocation decisions through any of these channels.1
Extending Wurgler (2000), this article investigates the role of the financial reporting system on capital allocation efficiency after controlling for financial development and other institutional characteristics. The quality of the financial reporting regime is broadly conceptualized as an output from multifaceted systems whose components collectively produce, gather, validate and disseminate information (Bushman et al., 2004). Three factors are hypothesized (Bushman et al., 2004) to affect the corporate transparency framework: (a) the corporate reporting regime, including measures of intensity, measurement principles, timeliness, audit quality of financial disclosures and the intensity of governance disclosures (i.e., identity, remuneration, and shareholdings of officers and directors, and identity and holdings of other major shareholders); (b) the intensity of private information acquisition, including measures of analyst following, and the prevalence of pooled investment schemes and insider trading activities; and (c) information dissemination, including a measure of the extent of media penetration in an economy. Factor analysis reveals two major components of corporate transparency, namely, financial transparency and governance transparency. Financial transparency captures the intensity and timeliness of financial disclosures, and their interpretation and dissemination by analysts and the media. Governance transparency, on the other hand, captures the intensity of governance disclosures.
Using Wurgler's (2000) country-level ‘efficiency of capital allocation’ measure and ‘corporate transparency’ data from Bushman et al. (2004), this article provides evidence that the financial reporting system, particularly its financial transparency component, has a positive impact on capital allocation. The result is robust to possible endogeneity problems between financial transparency and financial development measures. Furthermore, the financial transparency measure remains positive and marginally significant after controlling for the effects of stock price synchronicity, state ownership, and investor rights. This is important because Wurgler (2000) finds that the efficiency of capital allocation is positively (negatively) correlated with strong investor rights (high state ownership). Also, when stock price synchronicity is high, efficiency of capital allocation is low.2 However, the relationship between capital allocation efficiency and these variables is most likely to be endogenous in nature and after controlling for the effect of endogeneity, the financial transparency measure becomes insignificant.
CONCEPTUAL FRAMEWORK
Although there now exists a large literature on the role of financial development in economic growth (Levine, 1997), the debate concerning the role played by the development of financial intermediaries in economic growth is contestable. The debate concerns the possible causality running from financial development to economic growth as well as the channels through which financial development affects growth (Bloch and Tang, 2003). Levine (1997), although warning about the causal relationship between financial development and economic growth,3 strongly argues in favour of the positive role of financial development. Theoretical models (Greenwood and Jovanovic, 1990; Bencivenga and Smith, 1991) show the positive role of financial intermediary development. On the empirical front, Rajan and Zinagles (1998) find that industries depending on external financing grow faster in financially developed countries.
Later studies using refined econometric techniques and panel data sets have found that the causal relationship runs from financial development to economic growth (Beck et al., 2000; Rousseau and Wachtel, 2000; Beck and Levine, 2004). However, opponents of this view abound. Robinson (1952) suggests, ‘Financial development follows economic growth’. Lucas (1988) observes that the role of financial intermediation in economic growth has been very badly over-stressed in both popular and professional discussions. Demetriades and Hussein (1996) do not find that financial development causes economic development; rather, their results provide considerable evidence of reverse causation. Arguably, debate on the role of financial development in economic growth has intensified due to lack of research on, and evidence of, the channels through which financial development affects growth.
Financial Intermediaries and Efficient Allocation of Capital
Despite the theoretical literature on the role of financial intermediation in improving growth through its effect on productivity, there is little direct evidence on whether and how financial markets improve the efficient allocation of capital. Theory suggests that financial intermediaries can improve the efficiency of capital allocation due to their capacity effectively to acquire and process information about the innovative activities of the entrepreneurs (King and Levine, 1993) or about the aggregate state of technology (Greenwood and Jovanovic, 1990). Wurgler (2000) provides empirical evidence on how the role of financial development helps in the efficient allocation of capital in a large number of countries. Strong support is found for the positive role of financial development in efficient allocation of capital. He also finds evidence that capital allocation is improved through at least three mechanisms. First, in countries with stock markets that have less stock price synchronicity, more firm-specific information is impounded into stock price and the allocation of capital is more efficient. This, in turn, makes price more informative in distinguishing between good and bad investments. Second, capital allocation improves as state ownership declines. This is consistent with the hypothesis that in state-owned firms, resource allocation is likely to be guided less by value-maximization than by political moves.4 Last but not least, strong minority investor rights are associated with improved allocation of capital. In countries with ‘strong minority shareholder protection’, managers cannot channel free cash flow into unprofitable investments for empire building. Given the significant variation in the efficiency of corporate investment decisions and resulting control problems documented by Wurgler (2000), the question arises as to whether financial reporting regimes play a role in the efficient allocation of capital after controlling for other aspects of financial development. This is important because, as explained later, comprehensive financial reporting provides value-relevant information to facilitate the efficient allocation of capital.
A very important component of the financial reporting system is the accounting information produced and disclosed by corporations aimed at informing stakeholders about the performance of entities. Investors seeking to invest their money are likely to invest their capital in companies that provide timely and transparent financial information. Levine et al. (2000) argue that ‘information about corporations is critical for identifying the best investments. Accounting standards that simplify the interpretability and comparability of information across corporations will simplify financial contracting. Furthermore, financial contracts that use accounting measures to trigger particular actions can only be enforced if accounting measures are sufficiently clear’ (p. 59). La Porta et al. (1998) suggest:
Accounting plays a potentially crucial role in corporate governance. For investors to know anything about the company they invest in, basic accounting standards are needed to render company disclosures interpretable. Even more important, contracts between managers and investors typically rely on the verifiability in courts of some measures of firms’ income or assets. If a bond covenant stipulates immediate repayment when income falls below a certain level, this level of income must be verifiable for the bond contract to be enforceable even in court in principle. Accounting standards might then be necessary for financial contracting, especially if investors’ rights are weak. (p. 1140)
Bushman and Smith (2001) propose a theoretical link between so called ‘high quality’ financial accounting information and economic performance. Figure 1 presents an adapted version of the theoretical link proposed by Bushman and Smith (2001). Financial accounting systems can directly affect economic performance through three channels. First, financial accounting information allows mangers to identify new investment opportunities based on profit numbers reported by other firms. Financial accounting information can also indirectly affect economic performance by contributing to the efficient price formation process.5 Extensive capital market research has produced convincing evidence that financial accounting information is impounded in the stock price (Kothari, 2001, provides an extensive survey of capital market research in accounting) and stock price, in turn, guides managers and investors in making efficient resource allocation decisions.
FACTORS AFFECTING FINANCIAL ACCOUNTING QUALITY AND CHANNELS THROUGH WHICH HIGH QUALITY ACCOUNTING INFORMATION AFFECTS ECONOMIC PERFORMANCESource: Adapted from Bushman and Smith (2001, Figure 1, p. 294 and Figure 2, p. 306).
Financial accounting information, to the extent that it provides value-relevant information, plays a critical role as a corporate governance device in disciplining managers to invest in profitable projects and refrain from investing initially in value-destroying projects and to withdraw from negative NPVs once that becomes apparent.6 Managerial compensation contracts tied to accounting performance measures and informed stock prices induce managers to create value. Also, accounting information facilitates corporate takeovers that are designed to replace inefficient management and hence poses a credible threat to managers who are not creating value.
The third channel through which financial accounting information affects economic performance is by reducing adverse selection and liquidity risk. Greater disclosure enhances stock market liquidity and reduces transaction costs or increases demand for a firm's securities. Diamond and Verrecchia (1991) and Kim and Verrecchia (1994) suggest voluntary disclosure reduces information asymmetry among informed and uninformed investors. Consequently, investors can be relatively confident that stock transactions occur at a ‘fair price’ for firms with high level of voluntary disclosures. This attracts more funds into the market and reduces investors’ liquidity risk. Well-developed liquid capital markets are expected to enhance economic growth by facilitating high-risk, high-return, long-term innovation-oriented corporate investments, and financial accounting information needs to provide important support for this capital market function. However, a high quality accounting regime has to be created and Figure 1 illustrates some of the many socio-political and economic variables that could affect the development of a high quality accounting regime (Bushman and Smith, 2001, pp. 293–6).
The role of high quality accounting information in efficient resource allocation decisions as outlined by Bushman and Smith (2001) arguably is observed primarily in market-based economies. Market-based economies are characterized by diffused ownership structure leading to an information asymmetry problem between shareholders and corporate managers. Because information asymmetry leads to higher costs of external capital, mangers in market-based economies have incentives to disclose financial information to investors to reduce such asymmetry. Thus in market-based economies corporate disclosures play a critical role in the efficient operation of the stock market (Healy and Palepu, 2001). Ali and Hwang (2000, p. 4) suggest that, ‘Market-oriented system contains numerous diverse investors without direct access to company information. Investors are likely to rely heavily on financial accounting disclosures to obtain information to be used in security valuation and monitoring management.’ Accounting information guides profitable investment decisions as modelled by Zhang (2000) and Biddle et al. (2001) among others. Zhang (2000) models investment activity based on current operating efficiency and then valuing the cash flows to be produced from operations. Zhang (2000) relaxes the assumption of Feltham and Ohlson (1996) that intertemporal investment follows a predetermined linear stochastic process. Biddle et al. (2001) extend linear information dynamics (LID) by introducing ‘capital investment dynamics’ that allow net capital investment to be informed by current profitability and, in turn, influence future projects.
Another important characteristic of the market-based economies is the active role played by financial analysts in forming efficient stock prices. The relationship between financial reporting and analysts’ information searching is a complex phenomenon. At one level, publicly available financial information acts as an important source of input for analysts. However, because analysts engage in comprehensive and costly information searching, financial reporting could act as a competing source of information. Either way, the competition among analysts for disclosures and for the interpretation of disclosures may result in stock prices that reflect a broad information system (Beaver, 1998).
A clear advantage of publicly available financial statements is the transparency of investment opportunities in the market (Sadka, 2004). By providing information on prior investments, financial reporting can improve the resource allocation and the investment decisions made by firms which would lead to increased productivity and growth. Empirical evidence supports the positive role of accounting information on growth. Carlin and Mayer (2002) find that the provision of high quality accounting disclosures (proxied by CIFAR score)7 is associated with higher average growth and share of research and development in value added industries that are naturally dependent on equity finance. However, their proxy of the disclosure score is arguably narrow and they do not examine the role of accounting information in the efficient allocation of capital. Guenther and Young (2000) show that the association between aggregate return on assets and the percentage change in GDP is high in the U.K. and the U.S.A. and low in France and Germany, supporting the strong demand for accounting information in market-based economies. However, they do not look at the information processing and dissemination aspects of accounting, and they also fail to link their performance measure with the allocative efficiency channel.
Sadka (2004) considers financial disclosure as a direct contributor to growth8 in total factor productivity (TFP) and GDP instead of a mere proxy for financial development or as facilitators of financial intermediation. He measures corporate transparency as the product of market capitalization over GDP (MKTCAP/GDP) and CIFAR disclosure score. Using data from thirty countries over the period of 1985–99, Sadka provides empirical evidence consistent with the hypothesis that corporate transparency has a first-order positive effect on TFP and GDP. Francis et al. (2003) find that higher quality accounting and auditing are positively associated with financial market development in countries whose legal systems are conducive to the protection of investors. However, they fail to find systematic evidence that higher quality accounting and auditing alone, independent of a country's underlying investor protection regime, affects the development of financial markets.
Public financial reporting and efficient capital markets thus play an important role in efficient resource allocation decision in the market-based economies. However, this is not to suggest that countries with primarily bank-based financing lack growth. There are theoretical arguments suggesting that banks are a rather more efficient monitor of corporate managers than the stock market. Boyd and Prescott (1986) model the banks’ critical role in reducing information frictions while Stulz (2002) argues that banks are more efficient in providing external resources to new and innovative activities. Bhide (1993) argues that a high level of stock market liquidity discourages ‘active’ investors, who provide valuable internal monitoring. Dow and Gorton (1997, p. 1090) show that the ‘two tasks of investment appraisal and monitoring management that stock market information is used for [are] precisely the functions that banks are supposed to perform in making loans’. Bencivenga and Smith (1991, p. 196) model the growth promoting role of financial intermediaries by showing that ‘an intermediation industry permits an economy to reduce the fraction of its savings held in the form of unproductive liquid assets, and to prevent misallocation of invested capital due to liquidity needs’.
Empirically, Wurgler (2000) also does not find any differential effect of stock market and banking sector development in ensuring the efficient allocation of capital. Tadesse (2004) reaches a similar conclusion. Tadesse's result suggests that the financial system (both stock market activities and banking sector development) positively affects growth and factor productivity by improving efficiencies through enabling technological inventions and innovations. Beck and Levine (2004) use both stock market development and bank development data averaged over five-year intervals from 1976 to 1998, and find strong support for the independent effect of banks and stock markets. Bank and stock market development always enter jointly and are significant in all the system panel estimators.
DATA AND DESCRIPTIVE STATISTICS
The present study uses Wurgler's (2000) sensitivity of capital investments to value-added (η measure) as the dependent variable and corporate transparency framework developed by Bushman et al. (2004) as the primary independent variable.
Efficiency of Capital Allocation Measure
To measure the efficiency of capital allocation, Wurgler (2000) uses the United Nations’ General Industrial Statistics (the INDSTAT-3) as a source of basic manufacturing statistics. It contains gross fixed capital formation, value added, and output for up to twenty-eight three-digit ISIC manufacturing industries. Value added has been defined as the value of shipments of goods produced (output) minus the cost of intermediate goods and services required (but not including labour). Thus, this value added measure reflects value added by labour as well as capital. Gross fixed capital formation has been defined as the cost of new and used fixed assets minus the value of sales of used fixed assets. Wurgler assumes that optimal investment implies increasing investment in industries that are ‘growing’ and decreasing investment in industries that are ‘declining’. He uses the following simple regression (p. 194) estimate to determine the country-specific elasticity measure:
lnIict/Iict-1=αc+ηclnVict/Vict-1+ɛict,
()where I is gross fixed capital formation, V is value added, i indexes manufacturing industry, c indexes country, t indexes year and ln represents log transformation. The slope estimate in equation (1) (η) measures the extent to which country C increases investment in its growing industries and decreases investment in its declining industries (investment-value added sensitivities).9
Corporate Transparency Framework
Corporate transparency data are from Bushman et al. (2004). That work isolates two factors from the array of country-level measures of the firm-specific information environment based on factor analysis. The first, interpreted as financial transparency, captures the intensity and timeliness of financial disclosures, and their interpretation and dissemination by analysts and the media (FACTOR_1). The second, interpreted as governance transparency, captures the intensity of governance disclosures (FACTOR_2).
CONTROL VARIABLES
- (a)
Overall financial development (FD)
-
Financial development is computed as the sum of stock market capitalization to GDP and private and non-financial public domestic credit to GDP.
-
- (b)
Stock market variable (STOCK)
-
Stock market capitalization to GDP.
-
- (c)
Banking activities (BANK)
-
Private and non-financial public domestic credit to GDP.
-
- (d)
Stock price synchronicity (SYNCH)
-
Morck et al. (2000) measured the synchronicity of stock prices in a few dozen stock markets in 1995. This synchronicity measure is defined as the fraction of stocks that move in the same direction in a given week in the first half of 1995. The higher the co-movement the higher the synchronicity and lower the firm-specific information impounded in the stock price.
-
- (e)
Prevalence of state-owned enterprises (SOE)
-
Share of country-level output supplied by state-owned enterprises (SOE), where countries with less SOE investments receive higher ratings. Ratings range from 0 to 10. Derived from the Economic Freedom of the World (2002) database.
-
- (f)
Legal rights (RIGHTS)
-
Legal rights of external investors, creditor protection scores and rule of law scores are collected from La Porta et al. (1998).
-
- (g)
Initial wealth (GDP)
-
Log of per capita GDP, 1960 ($000).
-
EMPIRICAL RESULTS
Panel A of Table 1 provides descriptive statistics of the variables used. Panel B presents correlation analysis. The investment-value added sensitivity measure is significantly positively correlated with the financial transparency measure but not with governance transparency measure. The cross-correlation between the two transparency measures is statistically indistinguishable from zero, reflecting these two factors capturing two different dimensions. As predicted, the correlation between the efficiency measure and the overall financial development is significantly positively correlated consistent with Wurgler (2000). This implies that developed financial markets, as measured by the size of the domestic stock and credit market to GDP, are associated with an efficient allocation of capital. The correlation between overall financial development and two transparency measures is positive and statistically significant. Both financial and governance transparency promote stock market development, but financial transparency alone is significantly related to the banking sector development.
|
Panel A: Descriptive statistics of variables used |
|||||||
|---|---|---|---|---|---|---|---|
| Variables | N | Mean | Median | S.D. | Maximum | Minimum | |
| Dependent variable | ηScore | 39 | 0.56 | 0.59 | 0.26 | 0.98 | 0.068 |
| Primary independent variables | FACTOR1 | 39 | 0.0038 | 0.093 | 0.94 | 1.61 | −1.70 |
| FACTOR2 | 39 | 0.039 | 0.039 | 0.88 | 1.34 | −2.87 | |
| Control variables | FD | 38 | 0.86 | 0.80 | 0.56 | 2.67 | 0.13 |
| BANK | 38 | 0.61 | 0.55 | 0.37 | 2.00 | 0.09 | |
| STOCK | 39 | 0.28 | 0.17 | 0.30 | 1.23 | 0.00 | |
| GDP | 39 | 3.81 | 3.13 | 2.58 | 9.91 | 0.56 | |
| SYNCH | 30 | 0.66 | 0.67 | 0.04 | 0.75 | 0.58 | |
| SOE | 39 | 4.62 | 4.00 | 2.09 | 10.00 | 0.4 | |
| RIGHTS | 38 | 3.78 | 3.72 | 2.06 | 8.00 | 0.54 | |
|
Panel B: Correlation analysis |
|||||||||
|---|---|---|---|---|---|---|---|---|---|
| η SCORE | FACTOR1 | FACTOR2 | FD | STOCK | BANK | GDP | RIGHTS | SOE | |
| ηSCORE | 1.0000 | ||||||||
| FACTOR1 | 0.6039* | 1.0000 | |||||||
| FACTOR2 | 0.2098 | 0.0581 | 1.0000 | ||||||
| FD | 0.4527* | 0.6189* | 0.3824** | 1.0000 | |||||
| STOCK | 0.2474 | 0.4228* | 0.5096* | 0.8314* | 1.0000 | ||||
| BANK | 0.5066* | 0.6329* | 0.2194 | 0.9205* | 0.5482* | 1.0000 | |||
| GDP | 0.6407* | 0.7486* | 0.0757 | 0.3582** | 0.2225 | 0.3819** | 1.0000 | ||
| RIGHTS | 0.4212* | 0.5200* | 0.5421* | 0.7207* | 0.7122* | 0.5830* | 0.5447* | 1.0000 | |
| SOE | 0.2686 | 0.4078** | 0.1209 | 0.5559* | 0.5789* | 0.4297* | 0.3725** | 0.4052** | 1.0000 |
- Notes: η SCORE is the investment-value added elasticity measure derived from Wurgler (2000), Table 2. Wurgler (2000) uses a simple regression to estimate the country-specific elasticity measure: lnIict/Iict-1=αc+ηc lnVict/Vict-1+ɛict, where I is gross fixed capital formation, V is value added, i indexes manufacturing industry, c indexes country, t indexes year and ln represents log transformation.
- FACTOR1 and FACTOR2 represent financial transparency scores and governance transparency scores respectively derived from Appendix B of Bushman et al. (2004, p. 284).
- FD represents overall financial development data computed as the sum of stock market capitalization to GDP (STOCK/GDP) and private and nonfinancial public domestic credit to GDP (CREDIT/GDP).
- STOCK is STOCK/GDP and BANK is CREDIT/GDP. Data are averaged over the period of 1980 to 1990.
- GDP is the 1960 value of log per capita GDP. All these data come from Wurgler (2000, Appendix A).
- SOE is share of country-level output supplied by state-owned enterprises (SOE) derived from Economic Freedom of the World (2001).
- RIGHTS is the summary measure of the effective legal rights. First antidirector rights and creditor rights scores from La Porta et al. (1998) are summed to get an aggregate score out of 10. (Six anti-director provisions and four creditor rights provisions). This score is then multiplied by measure of the domestic ‘rule of law’ (continuous measure from 0 to 1) also derived from LaPorta et al. (1998).
- SYNCH is the stock price ‘synchronicity’ data derived from Morck et al. (2000), Table 2. This represents the fraction of stocks that move in the same direction in a given week in the first half of 1995 with higher score implies less firm-specific information impounded into the stock price.
- * ,
- ** ** and
- *** represent statistical significance at 1%, 5% and 10% level, respectively (two-tailed test).
Table 2 presents the regression result of the efficiency score on corporate transparency and on alternative measures of financial development. Specifications (1) to (6) present individual regression results of two transparency measures, overall financial development and its components (STOCK and BANK) and national wealth as proxied by GDP. Specification (1) shows that the coefficient on financial transparency is positive and statistically highly significant. Also, this parsimonious model explains about 33 per cent of the variation in the investment efficiency measure.
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | (10) | |
|---|---|---|---|---|---|---|---|---|---|---|
| FACTOR1 | 0.16* | – | – | – | – | – | 0.14* | 0.10** | – | 0.03 |
| (4.41) | (2.82) | (2.06) | (0.57) | |||||||
| FACTOR2 | – | 0.07 | – | – | – | – | – | – | – | – |
| (1.37) | ||||||||||
| FD | – | – | 0.21* | – | – | – | 0.06 | – | – | 0.10 |
| (3.00) | (0.80) | (1.31) | ||||||||
| STOCK | – | – | – | 0.28** | – | – | – | – | −0.05 | – |
| (2.10) | (−0.35) | |||||||||
| BANK | – | – | – | – | 0.35* | – | – | – | 0.24** | – |
| (3.48) | (2.21) | |||||||||
| GDP | – | – | – | – | – | 0.06* | – | 0.04** | 0.05* | 0.05* |
| (4.45) | (2.13) | (3.99) | (2.82) | |||||||
| Constant | 0.56* | 0.56* | 0.38* | 0.48* | 0.34* | 0.31* | 0.50* | 0.42* | 0.22* | 0.28** |
| (16.31) | (14.31) | (5.36) | (8.82) | (4.78) | (5.23) | (6.43) | (5.76) | (3.27) | (2.72) | |
| Adjusted R 2 | 0.33 | 0.02 | 0.18 | 0.08 | 0.23 | 0.33 | 0.31 | 0.39 | 0.45 | 0.43 |
| Observations | 39 | 39 | 38 | 39 | 38 | 39 | 38 | 39 | 38 | 38 |
- * and ** represent statistical significance at 1% and 5%, respectively (two-tailed test).
- Note: All the variables are defined in the note to Table 1. Missing observations on the value of financial development for Hong Kong reduce the sample size to 38 in some specifications.
However, governance transparency, though positive, is not statistically significant. A possible explanation is that the governance transparency may prevent direct stealing (e.g., excessive executive pay) but may fail to discipline managers from over-investing in negative NPVs.10Richardson (2006), for example, provides evidence that among many corporate governance mechanisms, only anti-takeover devices and the presence of activist shareholders constrain managerial propensity to over-invest. It is to be noted that the governance transparency score of Bushman et al. (2004) primarily consists of information regarding identity of managers, identity of board members and their affiliation, remuneration of officers and directors and the like.
Thus, on a stand-alone basis, it seems financial transparency elements provide more relevant information for efficient resource allocation decisions. As expected, the FD variable is positive and statistically significant, but the explanatory power of specification (3) is nearly half of that of specification (1). The coefficients on both the components of financial development, STOCK and BANK, are positive and statistically significant individually although the effect of BANK is stronger than that of STOCK.11 Finally, specification (6) shows that the impact of national wealth on the efficiency of capital allocation is both positive and statistically highly significant.
Specifications (7)–(10) include many of the explanatory variables in the regression. Specification (7) shows that when financial transparency and overall financial development measures are jointly entered into the regression model, only the financial transparency variable is positive and statistically significant. This is an important piece of evidence regarding the positive role of a financial reporting regime in the better allocation of capital. Specification (8) shows that the financial transparency measure remains significant even in the presence of GDP. When the components of FD, the stock market development and banking sector development are used with GDP as the independent variables in specification (9), STOCK turns out insignificant, but BANK has a positive coefficient that is statistically significant at the 5 per cent level. Wurgler (2000) finds a similar result. This could be due to the fact that some countries with a high financial transparency score also have banking sector dominance (U.S.A. is a good example). Finally, specification (10) shows that only GDP is significant and both financial transparency and FD measures are insignificant when these variables are used in the same regression model. Overall, there is some support for financial transparency to be positively associated with capital allocation efficiency.12 But the results calculated using FD as well as the components of FD seem to better explain the dependent variable. However, the result should be interpreted with caution due to the high degree of collinearity between some of the independent variables.
Do the Exogenous Components of Corporate Transparency and Overall Financial Development Impact Allocational Efficiency?
A major problem that cross-country empirical research often has to confront is the problem of endogeneity. If allocation efficiency, overall financial development, and corporate transparency are all endogenously determined, then a simple OLS regression will produce biased and inconsistent coefficient estimates as the regressors and the error term are correlated. To alleviate this concern, the standard econometric solution is to run two-stage least square regressions (2SLS) with instrumental variables.
Table 3 presents the 2SLS regression results. The log of per capita GDP, Common, German, and French legal origin dummies are used as instrumental variables. The choice of legal origin dummies as instrumental variables is appropriate because these legal protections are determined to a large extent by the colonial history of the country (La Porta et al., 1997). Also the wealth of the country represented by log of per capita GDP in 1960 precedes the dataset used for corporate transparency and financial development measures. The 2SLS regression result shows that the component of financial transparency predetermined by legal origin and wealth is positive and statistically significant, thus mitigating the concern of endogeneity. Therefore, the independent effect of financial transparency suggests that the basic relation reflects, at least in part, the influence of the financial reporting regime on allocation efficiency. The same result holds for FD regression specification. The adjusted R2 on the financial transparency model and FD model is almost identical.
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | |
|---|---|---|---|---|---|---|---|
| FACTOR1 | 0.27* | – | – | 0.21** | – | 0.19** | – |
| (4.43) | (2.43) | (2.14) | |||||
| FACTOR2 | – | −0.0032 | – | – | −0.15 | – | −0.13 |
| (−0.0.04) | (−1.45) | (−0.93) | |||||
| FD | – | – | 0.53* | 0.20 | 0.64* | – | – |
| (3.21) | (1.04) | (3.18) | |||||
| STOCK | – | – | – | – | – | −0.01 | 0.53 |
| (−0.04) | (0.81) | ||||||
| BANK | – | – | – | – | – | 0.34 | 0.67* |
| (1.35) | (2.77) | ||||||
| Constant | 0.56* | 0.56* | 0.10 | 0.38** | 0.0030 | 0.35** | 0.02 |
| (14.88) | (13.16) | (0.65) | (2.20) | (0.02) | (2.07) | (0.08) | |
| Adjusted R2 | 0.19 | −0.03 | 0.20 | 0.33 | 0.18 | 0.34 | 0.17 |
| N | 39 | 39 | 38 | 38 | 38 | 38 | 38 |
- * and ** represent statistical significance at 1% and 5%, respectively (two-tailed test).
- Note: Two-stage least square (2SLS) regression result. Instrumental variables are GDP, Common, German and French legal origin dummies. Scandinavian legal origin is used as the base case. All the variables are defined in the note to Table 1. Specifications (3)–(7) have 38 observations because of missing data on Hong Kong.
The predetermined component of governance transparency does not affect allocation efficiency, as is evident from Table 3, column 2. The unreported result shows that the correlation between FACTOR_2 and legal origin dummy (one for code law countries and zero for others) is significantly negative (the correlation coefficient is −0.53, statistically significant at better than 1 per cent level). However, the correlation between ηSCORE and legal origin dummy is positive but insignificant. This is contrary to expectation because if legal origin matters for capital allocation efficiency then code law countries representing weaker investor protection should have a lower ηSCORE. However, unreported result shows that ηSCORE is higher in code law countries compared to their common law counterparts (average ηSCORE of 0.60 versus 0.50, respectively). This confounding effect may be responsible for the insignificance of the exogenous component of governance transparency measure.
The primary regression result of interest comes from specification (4), which includes both financial transparency and financial development variables in a 2SLS setting. Results show that the component of financial transparency predetermined by legal origin and national wealth is positive and statistically significant, while that of FD is not. This result is insightful as it suggests that financial transparency plays a significant role in providing value-relevant information for channelling resources to their most productive use. Wurgler (2000) shows that the component of FD predermined by legal origin has a strong effect on the allocation efficiency score. However, this effect becomes insignificant in the presence of financial disclosure intensity. As expected, the exogenous component of FD is significant in the absence of financial transparency variable in specification (5). In specification (6), the exogenous component of financial transparency is again positive and statistically significant at the 5 per cent level but the components of FD, namely STOCK and BANK, are not. This implies that it is financial transparency that matters. Whether a country is market-based or bank-based does not provide additional insight. Beck and Levine (2004) also find a similar result in the context of the role of financial development in explaining growth. Both bank and stock market development variables always enter jointly and are significant in all the system panel estimators. Finally, in specification (7), the exogenous component of BANK is positive and statistically significant, but the exogenous component of governance transparency and stock market development are not. Overall the 2SLS regression result supports the view that financial transparency enhances capital allocation efficiency. Also, financial development is important but when both financial transparency and the overall FD are considered together, the former is more strongly related to the efficiency measure than FD.
Channels Through Which Corporate Transparency and Overall Financial Development Affect Allocation Efficiency?
Zingales's (2003) commentary emphasizes the channels through which finance affects economic development:
Much of the literature has focused on proving that the observed correlation between finance and growth is causal. Less attention has been focused on understanding the channels through which finance works. Establishing the main channel is important not only for instilling confidence in the theory of a causal link, but also from a policy point of view. (p. 48)
Wurgler (2000) finds that capital allocation is improved at least through three channels. First, the SYNCH measure captures the amount of firm-specific information impounded into stock prices. Morck et al. (2000) find that the higher the SYNCH measure, the lower is the amount of firm-specific information impounded into stock prices. Since the most frequently cited function of stock price is to provide public signals of investment opportunities, the higher SYNCH measure implies lower firm-specific information and accordingly stock price becomes less useful for investment decision making. Wurgler finds that the SYNCH measure is strongly negatively related to allocation efficiency. This is confirmed in Table 4. In five of the eight models, the variable SYNCH is negative and statistically significant, implying that higher SYNCH impounds less firm-specific information and makes stock price less informative with respect to capital allocation decisions.13 The financial transparency variable is positive and statistically significant in specifications (1) and (6), which also include SYNCH and other control variables.
| 2SLS | 2SLS | 2SLS | ||||||
|---|---|---|---|---|---|---|---|---|
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | |
| FACTOR1 | 0.17* | – | 0.37* | – | 0.37** | 0.14** | 0.132** | – |
| (3.22) | (2.28) | (2.12) | (2.34) | (2.05) | ||||
| FACTOR2 | – | – | – | – | – | – | −0.03 | |
| (−0.39) | ||||||||
| FD | – | 0.15** | – | 0.23 | 0.13 | – | 0.08 | 0.15 |
| (2.13) | (1.45) | (0.64) | (0.67) | (1.28) | ||||
| SYNCH | −1.01 | −2.25** | 1.81 | −3.83** | 1.95 | −0.99 | −1.35 | −2.12** |
| (−1.08) | (−2.53) | (0.59) | (−2.50) | (0.58) | (−1.03) | (−1.31) | (−1.98) | |
| SOE | – | – | – | – | – | 0.09 | −0.07 | 0.00 |
| (0.42) | (−0.28) | (−0.01) | ||||||
| RIGHTS | – | – | – | – | – | 0.01 | −0.002 | 0.007 |
| (0.60) | (−0.06) | (0.20) | ||||||
| Constant | 1.24** | 1.95* | −0.51 | 0.23 | −0.89 | 1.14** | 1.43** | 1.83** |
| (1.97) | (3.22) | (−0.24) | (1.45) | (−0.39) | (1.74) | (2.03) | (2.36) | |
| Adjusted R 2 | 0.36 | 0.27 | 0.29 | 0.30 | 0.26 | 0.33 | 0.30 | 0.18 |
| N | 30 | 29 | 30 | 29 | 29 | 30 | 29 | 29 |
- *, ** and *** represent statistical significance at 1%, 5% and 10% level, respectively (two-tailed test).
- Note: All the variables are defined in the note to Table 1. Specifications (3)–(5) provide 2SLS regression result. Instrumental variables are GDP, Common, German and French legal origin dummies. Scandinavian legal origin is used as the base case.
However, results based on 2SLS provide different results. The instrumental variables are the same as those used in Table 3. In specifications (3)–(5), the component of financial transparency predetermined by legal origin and GDP is positive but does not attain statistical significance at the conventional level. In specification (7), the allocation efficiency score is regressed on FACTOR_1, FD, SYNCH, SOE and RIGHTS. Results show that the coefficient on FACTOR_1 remains positive but loses statistical significance. Therefore Table 4 provides mixed evidence on the role of financial transparency in the efficient allocation of capital internationally.
IMPLICATIONS AND CONCLUDING REMARKS
A fundamental factor in wealth creation in an economy is the efficiency with which scarce capital is allocated to investment opportunities. Wurgler (2000) finds that the level of overall financial development is positively associated with capital allocation efficiency. However, an important institution that affects financial development is the corporate reporting regime, including measures of intensity, measurement principles, timeliness, audit quality of financial disclosures, and the intensity of governance disclosures (i.e., identity, remuneration, and shareholdings of officers and directors, and identity and holdings of other major shareholders). The empirical results presented here strongly support the view that financial transparency is a significant contributor to the allocation efficiency in some countries. This result is robust after controlling for overall financial development, the components of financial development and various institutional characteristics.
If a high quality corporate reporting regime indeed affects allocation efficiency which is a channel to economic growth, then why do not all countries develop a comprehensive corporate reporting framework? A plausible answer to this question is that different countries have different channels of promoting allocation efficiency. In market-based economies the information asymmetry problem is resolved through corporate disclosures of financial as well as governance information. The stock market is the primary mechanism for channelling investment funds to profitable opportunities and the stock price guides such a decision. On the other hand, bank-based economies are usually characterized by a dominant role of banks in allocating funds to investment opportunities. Investment appraisal and the task of monitoring management are carried out by the bank and banks rely on inside information to perform the tasks which potentially reduce the significance of public disclosure of corporate financial information. A stream of literature has started to unfold the impact of economic, legal, and political infrastructures on the evolution of accounting and disclosure infrastructures (e.g., Ball, 2001).
Ball et al. (2000, p. 2) find that financial statement transparency as captured by timeliness and conservatism is higher in countries where information asymmetry is resolved through public disclosures (primarily common law countries with a stock market dominance) compared to countries (code law-based with bank dominance) where ‘information asymmetry more likely is resolved . . . by institutional features other than timely and conservative financial statements, notably by closer relations with major stakeholders’. Using a different proxy for financial transparency, Bushman et al. (2004) find that the financial transparency measure is higher in countries with low state ownership of companies, low risk of state expropriation of firms’ wealth and low state ownership of banks. This evidence suggests that corporate transparency, particularly financial disclosure transparency, is strongly influenced by a country's institutional structure.14
This article, although it finds a significant relationship between financial transparency and the efficiency of capital allocation, is to be interpreted in light of the broadness of the transparency measure. More significant insight can be gained by analysing a particular property of accounting information, like conservatism, in the efficient allocation of capital. Bushman et al. (2006) is an important study in that direction.
