Essays in financial economics

Abstract

The thesis consists of three essays dealing with corporate governance in an international context. The first essay is entitled "A Cross-Firm Analysis of the Impact of Corporate Governance on the East Asian Financial Crisis." In a sample of 399 firms from Indonesia, Korea, Malaysia, the Philippines, and Thailand, cross-firm differences in variables related to corporate governance had a significant impact on firm performance during the East Asian financial crisis of 1997- 1998. Higher outside ownership concentration led to significantly better stock price performance during the crisis, but higher managerial ownership concentration had no significant effect on performance. This may indicate that the presence of an outside lock holder can mitigate expropriation of minority shareholders, but that managers with significant holdings can resist this effect. Diversified firms performed significantly worse than focused firms during the crisis. On average, single-segment firms emerged from the crisis trading at a premium of over 20% relative to diversified firms with which they were equally valued prior to the crisis. The relative loss in value for diversified firms was due primarily to the performance of firms with high variation in investment opportunities across divisions, suggesting that cross-subsidization of divisions may have been a source of the value loss. Variables indicative of higher disclosure quality are associated with significantly better performance during the crisis. Having an ADR and having an auditor from a "Big Six" accounting firm had separate positive effects on firm performance. Firms with both indicators came out of the crisis valued at a premium of over 50%, on average, relative to firms without these indicators with which they were equally valued prior to the crisis. Taken as a whole, the results provide some evidence at the micro level that poor corporate governance contributed to the depth of the East Asian financial crisis. The second essay is entitled "The Performance of Politically Favored Firms in the East Asian Financial Crisis: Evidence from Malaysia." Malaysia presents an interesting opportunity to study the impact of political favoritism on firm performance during the East Asian crisis. Favoritism runs along two dimensions in Malaysia. Firms are favored based on the ethnicity of their owners as well as through personal relationships with key government officials. I find that the stock price performance of firms favored based on ethnicity was significantly better than the performance of non-favored firms during the crisis. However, the performance of firms favored through personal connections was significantly worse than the performance of non-connected firms. The evidence does not suggest that the relative loss for connected firms was driven by excessive leverage or inherent operating inefficiencies. Rather, the evidence suggests that the performance difference was driven by changes in the expected value of benefits for politically favored firms. The third essay is entitled "Do Agency Problems Affect Dividend Policy? Firm-Level Evidence from Around the World." The "outcome" agency model of dividends (La Porta, Lopez-de-Si lanes, Shleifer, and Vishny (LLSV (2000)) yields two key empirical predictions. First, dividend payouts will be higher among firms in which agency problems are less severe. Second, a negative relationship between growth opportunities and dividend payouts will be stronger among firms in which agency problems are less severe. LLSV (2000) use country-level measures of legal protection as a proxy for lower agency costs, and find empirical support for both predictions. I build on these findings by using firm-level measures indicative of the severity of agency problems. The proxies I employ are based on the cross-listing of firms in the U.S., the quality of accounting disclosure, diversification across industries, and the presence of a large outside shareholder. In a sample of 3,385 firms across 32 countries, I also find empirical support for both predictions of the outcome model.