CHAPTER 6 Corporate Governance and Leverage Trends

Thomas Rumbaugh
Published Date:
January 2012
  • ShareShare
Show Summary Details

Serious corporate governance problems may lead to overborrowing by firms in a weak institutional environment. However, improvements in governance can help increase leverage when access to credit has been constrained. The relationship between governance and leverage also has a medium-term dimension: firms facing a significant debt overhang because of past overborrowing pass up profitable investment opportunities (Bris and Koskinen, 2002). The importance of governance to leverage decisions has macroeconomic implications—and this relationship may have an important impact on growth, particularly for industries that are most dependent on external finance (De Nicolo, Laeven, and Ueda, 2006). Moreover, economies with weak institutions may be prone to bigger contractions in output during crisis periods (Shimpalee and Breuer, 2006).

The 1997–98 Asian financial crisis illustrated the dramatic impact of governance deficiencies on leverage. In Indonesia, corporate governance problems led to major risk-management failures. Before the crisis, high GDP growth rates in Indonesia were spurred by rapid corporate borrowing, in many cases by groups with ownership links to banks providing the financial resources. Large corporate groups accessed low-cost financing beyond what was economically justified. 1 Once the crisis hit, the impact on the cost of borrowing for the corporate sector was substantial and access to financing was drastically curtailed.

This chapter assesses whether improvements in corporate governance have translated into improved access to credit by Indonesian corporate groups. Based on Rajan and Zingales (1995), sequential regressions are used to analyze the impact of changes in corporate governance on leverage. In particular, the low significance of firms’ size in explaining leverage ratios would reflect creditors’ lingering concerns about corporate governance problems in large firms. Also, the sign and significance of Tobin’s Q in explaining leverage would reflect shareholders’ preference for either issuing equity or borrowing in the face of asset valuation increases, with a negative sign expected in an environment of good corporate governance. Finally, an alternative specification of Rajan and Zingales (1995) is used to assess the role of debt overhang problems arising from past governance deficiencies in explaining leverage decisions. The analysis finds that creditors seem to have had concerns about corporate governance in large firms up to about 2005, but these concerns have subsided since then. In fact, the governance environment has improved markedly since the Asian crisis. Further improvements in the legal framework and protection of shareholder rights will be important for supporting higher investment in the future.

The chapter is organized as follows: The next section discusses governance-leverage links and is followed by a section highlighting the corporate governance environment and corporate leverage patterns in Indonesia. The subsequent section reviews a regression analysis based on Rajan-Zingales (R-Z) employing the sequential regressions and the alternative specifications described above.


Corporate governance and corporate financing are linked in multiple ways. At the country level, a better legal and institutional framework will ensure the enforcement of contracts. In that environment, creditors will be more willing to provide financing to economic agents. Better transparency and disclosure allow for better-informed decisions by firms, including on lending and borrowing. Protection of minority shareholders reduces the scope for internal conflicts within the firm, including on leverage choices. The channels through which governance and financing are interrelated can be classified as follows:

  • Risk taking. Firms are more willing to take risks in a better corporate governance environment. Corporate risk taking and firm growth seem to be positively related to the quality of investor protection (John, Litov, and Yeung, 2008). By contrast, investment will be discouraged in governance systems that are relationship-based, dominated by ownership structures that allow control by insiders, and that show high uncertainty with regard to property rights and lack of protection of minority shareholders’ rights (McGee, 2009).

  • Firm valuation. Good governance practices have a positive impact on firms’ market valuation. Evidence shows a positive and statistically significant relationship between firm value and the percentage of the board made up of directors not affiliated with the dominant shareholders, especially in countries with weak shareholder protection (Dahya, Dimitrov, and McConnell, 2008). Weak governance and low valuation may have an impact on overborrowing: Nielsen (2006) finds that companies with weaker shareholders are more highly leveraged and more likely to pay higher dividends.

  • Financing costs. Increased transparency and disclosure will tend to reduce risk premiums in credit markets. By the same token, weaknesses in the rule of law and problems in the exercise of creditor rights have an impact on firms’ borrowing costs. Creditors may be less willing to lend to companies when governance is a concern, which may lead to financial constraints for firms, with an impact on financing costs (Mahendra, 2009).

  • Access to capital. More sources of financing are available to firms operating in an environment of good governance. Also, the impact of improvements in accounting disclosure and transparency on real economic activity appears particularly pronounced for industries that are most dependent on external finance (De Nicolo, Laeven, and Ueda, 2006). By contrast, in countries with weakly enforceable minority shareholder rights, a small loss of confidence may lead to a sharper decline in the amount of capital that investors are willing to provide, which seems to explain the sharp variation across countries in exchange rates and stock prices during the Asian crisis (Johnson and others, 2000).

In sum, improvements in governance will lead to higher demand for and more access to capital, associated with more confident risk taking, lower borrowing costs, and lower incentives to overborrow. The complexity of assessing these channels requires the use of more specific variables to try to explain the impact of governance on leverage. This analysis uses a set of tractable variables to identify the determinants of capital structure used by R-Z for the Group of Seven (G-7) industrial countries, based on Harris and Raviv (1991). The variables selected by R-Z are tangibility of assets, firm profitability, the market-to-book-value of assets ratio (a proxy for investment opportunities), and firm size. These variables are closely linked to governance features, underscoring the key role of governance in the determination of firms’ capital structure.

  • The tangibility ratio (fixed assets divided by total assets) as a proxy for tangible collateral will be important when minimum creditor rights standards are in place.

  • Firms showing higher profitability will prefer to finance their activities with internal funds when leverage is driven by firms’ decisions rather than creditors’ rationing in an environment of relatively unconstrained access to financing. 2

  • Firms expecting high future growth would use a greater amount of equity finance when shareholders give a higher priority to taking advantage of profitable investment opportunities than to maintaining control (negative impact on leverage).

  • Larger firms tend to be more diversified and fail less often, suggesting that size could be seen as an inverse proxy for the probability of bankruptcy. Other empirical work applying the R-Z framework, especially in developing countries, assumes that size provides an advantage when it comes to accessing financing.

This chapter uses the R-Z framework to assess the impact of governance on leverage over time. It assumes that differences between the signs and level of significance of the explanatory variables relative to the ones found by R-Z for G-7 countries are driven by governance differences. The analysis uses tangibility and profitability as control variables, showing that minimum creditor rights standards are in place and firms do not have major constraints in accessing financing.

The hypothesis is that weak governance is a significant factor in explaining leverage patterns. In particular, (a) weak governance will be reflected in a weak (or negative) relationship between firm size and leverage as the result of a combination of constrained access and higher borrowing costs; (b) weak governance will be reflected in a weak (or positive) correlation between market-to-book-value of assets and leverage, reflecting some overborrowing explained by a preference to maintain control even when firm valuation improves; (c) improvements in governance will bring significance and signs of these explanatory variables in line with R-Z findings; and (d) the impact of explanatory variables is influenced by debt-overhang problems inherited from governance deficiencies associated with the 1997–98 crisis that constrain risk taking.

The rationale for the above-mentioned hypotheses can be explained as follows: (a) it would take time for large firms to regain their advantage in borrowing following a financial crisis; (b) dominant shareholders will resist issuing equity if they perceive that keeping control of the firm provides information advantages relative to external creditors; (c) leverage decisions are actually conditioned by debt overhang problems arising from governance deficiencies from the Asian crisis period, regardless of governance improvements.


Immediately after the 1997–98 Asian crisis, progress in shareholder diversification was moderate. The share of family-based ownership structures declined and widely held ownership became increasingly more important while the presence of foreign companies expanded. However, companies still favored ownership structures that preserved the dominant position of traditional shareholders. An analysis using a methodology based on Sato (2003) found that 51 percent of the largest 100 companies were private domestic corporations for which a single shareholder or group held more than 40 percent of equity in 2005 compared with 60 percent in 2000. More important, sharp differences are found in the financial performance of “traditional” corporations and companies with more diversified ownership structures, with the former showing higher leverage and weaker financial positions (Figure 6.1).

Figure 6.1Indonesia: Top 100 Corporate Financial Indicators, 2000 and 2005

Source: Thomson Reuters, Worldscope database.

From 2000 to 2010, Indonesia benefited from the wide access to international financing available to emerging markets. In Indonesia, higher access to credit had been accompanied by lower leverage ratios. This apparent paradox simply reflected that asset growth had exceeded liability growth over time, for three reasons: (a) the share of retained earnings and equity issuance to total financing increased relative to borrowing; (b) capital expenditure increased faster than borrowing; and (c) the share of nontangible assets in total assets increased as firm valuation improved. This is consistent with trends observed during 2003–08 for leverage, profitability, and asset growth for a sample of 214 listed firms listed on the Indonesian Stock Exchange (Figure 6.2).

Figure 6.2Indonesia: Leverage, Profitability, and Asset Growth for a Sample of Listed Firms, 2003–08

Sources: Indonesia Stock Exchange; and authors’ calculations.

The governance environment in Indonesia improved measurably following the Asian crisis, both at the country level and at the firm level. Widespread bank-corporate ownership ties were broken as a result of bank restructuring; financial reporting became more transparent following the adoption of basic international financial reporting standards; and the excessive degree of ownership concentration observed before the crisis moderated markedly. Then, before the 2007–09 global financial crisis, improvements in governance showed a degree of international convergence over time. Many countries, especially emerging markets, went through a process of bringing their domestic governance frameworks up to international standards. Therefore, improvements in governance quality may not give a country an advantage in accessing capital markets beyond a certain threshold (De Nicolo, Laeven, and Ueda, 2006). In fact, Indonesia still compares unfavorably with other countries in the region on compliance with the rule of law (Figure 6.3).

Figure 6.3Governance Indicators in Selected Asian Countries, 2008

Sources: World Bank, World Governance Indicators,

In Indonesia, limited openness of ownership structures remains a problem. The World Bank’s 2004 “Report on the Observance of Standards and Codes (ROSC)” found widespread pyramidal structures and family-based groups to be the preferred ownership structure. Although disclosure is required by the Capital Market Supervisory for ownership shares of 5 percent or more for listed companies, detection of cross ownership is difficult because of deficient reporting procedures. Thus, controlling shareholders often pursue group or family interests rather than those of the firms.


The analysis uses annual data from nonfinancial companies listed on the stock exchange available through Datastream. Some caveats are worth noting: The results may not be fully comparable with other empirical work because this analysis uses financial indicators as reported by firms, whereas other cross-country applications (including R-Z) make adjustments to account for differences in treatment between countries. Also, using information from listed firms introduces a sample bias—listed companies are normally representative of high quality firms. As shown by Rauh and Sufi (2008), “credit-quality firms” may enjoy access to discretionary, flexible sources of financing whereas other firms rely on tightly monitored, secured bank loans for liquidity.

The basic regression model is the R-Z empirical model based on Harris and Raviv (1991):

An alternative specification tries to capture the postcrisis process of deleveraging while controlling for sector-specific effects:

This alternative specification measures the impact of the same variables as in R-Z on corporate leverage, but adds as an explanatory variable the initial deviation from the sectoral mean of every individual firm’s leverage ratio in the initial year of the data sample. This variable is expected to capture industry-specific debt overhang problems. Firms with high leverage relative to their sectoral mean at the beginning period are more likely to experience debt overhang problems inherited from the crisis period. These firms will embark on a deleveraging process independently of other variables.

Tangibility and profitability are treated as control variables. The high significance of these two factors in the sample period (2003–08) provides evidence that (a) Indonesian firms generally enjoyed access to financing associated with the strength of their collateral and (b) firms had sufficient discretion to decide on their borrowing based on their capital structure strategy. As for the other explanatory variables, size shows a steady increase and the market-to-book ratio responds to market conditions, specifically declining in 2005 and 2008 when stock markets were under stress (Figure 6.4).

Figure 6.4Indonesia: Average Sales and Market-to-Book Ratio for a Sample of Listed Firms, 2003–08

Sources: Indonesia Stock Exchange; and authors’ calculations.

Regressions were run with debt-to-capital and debt-to-assets as dependent variables, for 2002–05 and 2004–07. 3 The results are generally consistent with R-Z (Table 6.1), with better results for debt-to-assets as a measure of leverage. The high significance of tangibility and profitability confirms that access to financing has been generally appropriate for Indonesian firms, with collateral being important to creditors and retained earnings a key source of financing for firms. Significance seems to improve when sectoral dummies are introduced.

TABLE 6.1Determinants of Corporate Leverage: Basic Regressions
Dependent variable: Debt-to-capitalDependent variable: Debt-to-assets
Independent variablesUnited Statesa 1987-90Canada 1987-90Japana 1987-90Indonesia 2002-05 (1)Indonesia 2002-05 (2)Indonesia 2004-07 (3)Indonesia 2004-07 (4)
(0.01)(0.01)(0.02)(2.45)(1 -45)(1.08)(1 -03)
Sectoral dummiesNoYesNoYes
Number of observations2,079264316206213142142
Source: IMF staff estimates.Note: Robust standard errors are in parentheses. Equations are run using ordinary least squares.

significant at 10 percent.

significant at 5 percent.

significant at 1 percent.

Rajan and Zingales (1995) results.

Source: IMF staff estimates.Note: Robust standard errors are in parentheses. Equations are run using ordinary least squares.

significant at 10 percent.

significant at 5 percent.

significant at 1 percent.

Rajan and Zingales (1995) results.

The following findings are worth highlighting:

  • Access to financing has been improving over time, as shown by the increasing significance of tangibility in explaining corporate leverage.

  • Resorting to retained earnings has always been significant in explaining leverage, but its coefficient seems to decline over time. This may reflect sizeable debt-equity conversions in the first half of the 2000s as profits increased, which may have magnified the reduction of leverage in this period. In recent years, the coefficients are closer to those normally observed in other countries.

  • The “normalization” of the significance of firm size over time confirms that large firms did not benefit from access advantages in obtaining credit soon after the crisis (2002–05), although this advantage is present and significant when moving away from the crisis period (2004–07).

  • The fact that Indonesian firms increased leverage as firm valuation improved rather than raising equity as would be expected is a common feature in emerging markets, which may reflect governance challenges in these economies (Booth and others, 2001). The interpretation is that the composition of financing becomes skewed to favor control by existing shareholders, reflecting a perceived “information advantage” by shareholders that has implications for the efficient allocation of resources.

To what extent has the behavior described above been driven by an autonomous deleveraging process? Table 6.2 shows the results for the alternative specification, using changes in leverage as the dependent variable.

TABLE 6.2Indonesia: Alternative Specification
Dependent variable: Change in debt-to-assets
Independent variables2001-05 (5)2003-07 (6)
(1 -42)(0.98)
Initial deviation of debt-to-assets-0.83**-0.74**
Number of observations211197
Source: IMF staff estimates.Note: Robust standard errors are in parentheses. Equations are run using ordinary least squares.

significant at 10 percent.

significant at 5 percent.

significant at 1 percent.

Source: IMF staff estimates.Note: Robust standard errors are in parentheses. Equations are run using ordinary least squares.

significant at 10 percent.

significant at 5 percent.

significant at 1 percent.

The debt overhang seems to be a very important variable explaining changes in leverage. Firms showing higher debt-to-assets ratios relative to their sectoral mean embarked on a deleveraging effort. For every percentage point that the debt-to-assets ratio was above the corresponding sectoral average in the initial period, firms reduced their debt-to-assets ratios by between 0.7 and 0.8 percentage points in the following five years.

Findings worth noting, taking into account that this specification is more restrictive because it explains changes in leverage ratios, follow:

  • Profitability is still significant, which seems to ratify that firms’ financing was relatively unconstrained, that is, the observed deleveraging process was driven by a decision by firms rather than by creditors.

  • Tangibility and size show the correct signs and increasing significance, consistent with the hypothesis of gradual convergence toward R-Z results.

  • The market-to-book ratio still shows the wrong sign, but a lower significance in 2003–07. It remains to be seen if in the future this translates into an increasing preference to issue equity as control advantages fade away.


The recent history of corporate financing in Indonesia provides a good illustration of the different dimensions of the impact of governance on leverage. During the crisis, governance problems led to overborrowing, followed by creditors’ reticence to lend when default problems materialized. Improvements in governance increased access to credit, but firms that overborrowed at the time of risk-management failures took a long time to reduce leverage to more reasonable levels. Still, the ample availability of liquidity allowed Indonesian firms to increase borrowing at the same time that asset growth exceeded liability growth supported by equity issuance and the use of retained earnings.

The complexity of the interaction between governance and leverage is impossible to capture with any linear formulation. This analysis used an approach based on a widely accepted representation of the determinants of capital structure to assess consistency with governance patterns observed in Indonesia following the crisis. The results confirm that creditors may have had lingering concerns about corporate governance problems in large firms up to 2005, but that those concerns seem to have subsided. However, shareholders prefer to borrow rather than issuing equity in the face of asset-valuation increases, to prevent loss of control. This effect seems to remain in place even well after the 2007–09 crisis, consistent with low protection of minority shareholders and remaining disadvantages in the governance environment relative to comparable countries in recent periods. However, once autonomous deleveraging is introduced, as driven by firms suffering from debt overhang problems arising from past governance deficiencies, this effect vanishes, although the sign remains the opposite of what is observed for G-7 countries.

Several policy implications are pertinent:

  • A sustained increase in the investment ratio in Indonesia would require that firms take larger risks, which is associated with the quality of investor protection. Relationship-based structures that allow control by insiders will be a significant constraint, especially if compliance with the rule of law, raising uncertainty about property rights, is weaker than in neighboring countries.

  • The valuation of firms will be affected by the inability of investors to ensure a proper assessment of the firms’ assets and liabilities, leading to higher financing costs to compensate for the higher risk.

  • High priority should be given to synchronizing governance regulations and practices with other emerging-market countries and improving protection of minority shareholders. At the aggregate level, improvements in regulations and the legal framework need to be paired with significantly better enforcement that may improve investors’ perceptions of Indonesia’s business environment.


    BoothLaurenceVaroujAivazianAsliDemirguc-Kunt and VojislavMaksimovic2001Capital Structures in Developing Countries,Journal of Finance Vol. 56 No. 1 pp. 87130.

    ClaessensStijnSimeonDjankov and Larry H.P.Lang2000The Separation of Ownership and Control in East Asian Corporations,Journal of Financial Economics Vol. 58 pp. 81112.

    DahyaJayOrlinDimitrov and JohnMcConnell2008Dominant Shareholders, Corporate Boards and Corporate Value: A Cross-Country Analysis,Journal of Financial Economics Vol. 87 No. 1 pp. 73100.

    De NicoloG.L.Laeven and K.Ueda2006Corporate Governance Quality: Trends and Real Effects,IMF Working Paper No. 06/293 (Washington: International Monetary Fund).

    HarrisMilton and ArturRaviv1991The Theory of Capital Structure,Journal of Finance Vol. 46 No. 1 pp. 297355.

    JohnKoseLubomirLitov and BernardYeung2008Corporate Governance and Risk Taking,Journal of Finance Vol. 63 No. 4 pp. 1679728.

    JohnsonSimonPeterBooneAlasdairBreach and EricFriedman2000Corporate Governance in the Asian Financial Crisis,Journal of Financial Economics Vol. 58 No. 1–2 pp. 14186.

    KorajczykRobert and AmnonLevy2002Capital Structure Choice: Macroeconomic Conditions and Financial Constraints,Journal of Financial Economics Vol. 68 No. 1 pp. 75109.

    MahendraEdo2009Financial Dependence, Macroeconomic Stability, and Firm Growth: What Policy Implications for Indonesia?Economics and Finance in Indonesia Vol. 57 No. 2 pp. 21954.

    McGeeRobert2009Corporate Governance in Developing Economies (New York: Springer).

    NielsenAnders E.B.2006Corporate Governance, Leverage and Dividend Policy,” paper presented at the EFA 2006 Zurich Meetings August 11 2006 (ZurichSwitzerland: University of Zurich – Swiss Banking Institute). Available via the Internet:

    RauhJoshua and AmirSufi2008Capital Structure and Debt Structure,NBER Working Paper No. 14488 (CambridgeMassachusetts: National Bureau of Economic Research).

    RajanRaghuram and LuigiZingales1995What Do We Know about Capital Structure? Some Evidence from International Data,Journal of Finance Vol. 50 No. 5 pp. 142160.

    SatoYuri2003Corporate Governance in Indonesia: A Study on Governance of Business Groups,” in The Role of Governance in Asia ed. by Y.Shimomura (Singapore: Institute of Southeast Asian Studies).

    ShimpaleePattama and JaniceBoucher Breuer2006Currency Crises and Institutions,Journal of International Money and Finance Vol. 25 No. 1 pp. 12545.

    World Bank2004Report on the Observance of Standards and Codes (ROSC)” (Washington).

Some 417 firms and 17 percent of market capitalization were traced back to a single family, and 10 families controlled more than half of the corporate sector (Claessens, Djankov, and Lang, 2000).

This is consistent with Korajczyk and Levy (2002), who show that leveraging is procyclical for constrained firms and countercyclical for unconstrained firms. Creditors are unlikely to respond to lower profitability by increasing lending, which is the reason a negative sign must reflect the prevalence of borrowers’ financial decisions in the absence of financing constraints.

To address potential endogeneity problem, the analysis follows R-Z strategy by using four-year averages as the independent variables with one period of lag with respect to the dependent variable. For example, if the independent variable sample period is 2002–05, then the dependent variable sample period is 2006. See Harris and Raviv (1991) and Rajan and Zingales (1995) for further discussion.

    Other Resources Citing This Publication