The relation between capital structure and expected returns

A basic tenet of capital structure theory is that the firm should choose its capital structure so as to maximize the value of the firm. This thesis explores the relationship between capital structure and expected returns by way of three studies. The objective is to increase understanding of how capital structure is related to the value of the firm under changing expectations.The second study is an empirical paper that examines the value that debt contributes to the value of the firm in the context of the book-to-market equity effect. The Kemsley and Nissim (2002) debt valuation model is used to proxy the value of debt and the capitalization rate1 , and the book-to-market equity ratio is used to proxy for expected profitability. The results show a leverage effect where the proportional value that debt contributes to firm value is negatively related to the book-to-market equity ratio2 (BE/ME), equivalent to a positive relation with expected growth prospects for cash flows (profitability) this leverage effect will influence the firm to use more debt and less equity as the BE/ME ratio decreases. There is also a concurrent capitalization rate effect where the capitalization rate is positively related to the BE/ME ratio, reflecting the cost of equity capital this capitalization rate effect will influence the firm to use less debt and more equity as the BE/ME ratio decreases. These opposing forces on the leverage level of the firm explain the puzzle of the low covariance between the leverage ratio and changes in the BE/ME ratio.The third study is an empirical analysis of the impact of market timing on capital structure for seasoned equity offerings (SEOs), using the hot-market timing measure of Alti (2006). In a sample of SEO events organized in event time, the results show that firms issuing equity in "hot markets" issue higher proportions of equity compared to their "cold-market" counterparts, but this market timing effect becomes insignificant within three years of the SEO year. This market timing effect is driven largely by "hot-market" firms with high market-to-book (M/B) asset ratios3,4. In a window of SEO events organized on the M/B mean-reversion cycle, the results show market timing increases with the M/B ratio, suggesting that market-timing reflects expected growth opportunities. These results are consistent with a dynamic form of the trade-off theory of capital structure.Overall, this dissertation contributes to the existing literature on capital structure and shows that the leverage level of firms is influenced by changing expectations.The first study is a theoretical paper that presents a mean-variance trade-off model of capital structure for the aggregate economy, examining capital structure in a mean-variance environment similar to the CAPM (capital asset pricing model). Whereas the standard trade-off model makes the counter-factual prediction that leverage should be positively related to expected returns, the mean-variance trade-off model shows that optimal leverage is negatively related to expected returns under conditions of leverage costs and investor portfolio choice. The intuition of this result is that average investors hold a portfolio containing both stock and bonds issued by firms, and the excess return per unit of risk rises more rapidly for unlevered equity than for levered equity with increasing expected return, inducing a shift in portfolio allocation toward lower leverage.1The capitalization rate is the discount rate, or cost of capital for unlevered equity. 2BE/ME = (book value of issued equity shares) / (market value of issued equity). 3M/B = (market value of equity + book value of debt) / (book value of equity + book value of debt). 4The BE/ME and M/B ratios are very similar in definition and implications, but different ratios are used in studies two and three to be consistent with practice in the previous literature that motivates these studies.