ONE OF THE MOST contentious issues in the theory of finance during the past quarter century has been the theory of capital structure. The geneses of this controversy were the seminal contributions by Modigliani and Miller [18,19]. The general academic view by the mid-1970s, although not a consensus, was that the optimal capital structure involves balancing the tax advantage of debt against the present value of bankruptcy costs. No sooner did this general view become prevalent in the profession than Miller [16] presented a new challenge by showing that under certain conditions the tax advantage of debt financing at the firm level is exactly offset by the tax disadvantage of debt at the personal level. Since then there has developed a burgeoning theoretical literature attempting to reconcile Miller's model with the balancing theory of optimal capital structure [e.g., DeAngelo and Masulis [5], Kim [12], and Modigliani [17]. The general result of this work is that if there are significant "leverage-related" costs, such as bankruptcy costs, agency costs of debt, and loss of non-debt tax shields, and if the income from equity is untaxed, then the marginal bondholder's tax rate will be less than the corporate rate and there will be a positive net tax advantage to corporate debt financing. The firm's optimal capital structure will involve the trade off between the tax advantage of debt and various leverage-related costs. The upshot of these extensions of Miller's model is the recognition that the existence of an optimal capital structure is essentially an empirical issue as to whether or not the various leverage-related costs are economically significant enough to influence the costs of corporate borrowing.The Miller model and its theoretical extensions have inspired several timeseries studies which provide evidence on the existence of leverage-related costs. Trczinka [28] reports that from examining differences in average yields between taxable corporate bonds and tax-exempt municipal bonds, one cannot reject the Miller hypothesis that the marginal bondholder's tax rate is not different from the corporate tax rate. However, Trczinka is careful to point out that this finding does not necessarily imply that there is no tax advantage of corporate debt if the personal tax rate on equity is positive. Indeed, Buser and Hess [1], using a longer time series of data and more sophisticated econometric techniques, estimate that the average effective personal tax rate on equity is statistically positive and is not of a trivial magnitude. More importantly, they document evidence that is consistent with the existence of significant leverage-related costs in the economy.

We provide evidence on the covenant structure of corporate loan agreements. Building on the work of Jensen and Meckling (1976), Myers (1977) and Smith and Warner (1979), we summarize and test the implications for what we refer to as the Agency Theory of Covenants (ATC), using a large sample of privately placed corporate debt. Our results are consistent with many of the implications of the ATC, including a negative relation between the promised yield on corporate debt and the presence of covenants.We also find that borrower and lender characteristics, as well as macroeconomic factors, determine covenant structure. Loans are more likely to include protective covenants when the borrower is small, has high growth opportunities or is highly levered. Loans made to investment banks and syndicated loans are also more likely to include protective covenants, as are loans made during recessionary periods or when credit spreads are large. Finally, we show that consistent with the ATC, firms that elect to issue private rather than public debt are smaller, have greater growth opportunities, less long term debt, fewer tangible assets, more volatile cash flows and include more covenants in their debt agreements. An important byproduct of our analysis is to demonstrate empirically that covenant structure and the yield on corporate debt are determined simultaneously. AbstractWe provide evidence on the covenant structure of corporate loan agreements. Building on the work of Jensen and Meckling (1976), Myers (1977) and Smith and Warner (1979), we summarize and test the implications for what we refer to as the Agency Theory of Covenants (ATC), using a large sample of privately placed corporate debt. Our results are consistent with many of the implications of the ATC, including a negative relation between the promised yield on corporate debt and the presence of covenants.We also find that borrower and lender characteristics, as well as macroeconomic factors, determine covenant structure. Loans are more likely to include protective covenants when the borrower is small, has high growth opportunities or is highly levered. Loans made to investment banks and syndicated loans are also more likely to include protective covenants, as are loans made during recessionary periods or when credit spreads are large. Finally, we show that consistent with the ATC, firms that elect to issue private rather than public debt are smaller, have greater growth opportunities, less long term debt, fewer tangible assets, more volatile cash flows and include more covenants in their debt agreements. An important byproduct of our analysis is to demonstrate empirically that covenant structure and the yield on corporate debt are determined simultaneously.

We provide evidence on the covenant structure of corporate loan agreements. ]. We summarize and test the implications for what we refer to as the Agency Theory of Covenants (ATC), using a large sample of privately placed corporate debt. Our results are consistent with many of the implications of the ATC, including a negative relation between the promised yield on corporate debt and the presence of covenants. We also¯nd that borrower and lender characteristics, as well as macroeconomic factors, determine covenant structure. Loans are more likely to include *This paper previously circulated under the title \Are bond covenants priced?" and was last revised in May of 2004. We thank

We explore the role of expected cash‐flow volatility as a determinant of dividend policy both theoretically and empirically. Our simple one‐period model demonstrates that, given the existence of a stock‐price penalty associated with dividend cuts, managers rationally pay out lower levels of dividends when future cash flows are less certain. The empirical results use a sample of REITs from 1985 to 1992 and confirm that payout ratios are lower for firms with higher expected cash‐flow volatility as measured by leverage, size and property‐level diversification. These results are consistent with information‐based explanations of dividend policy but not with agency‐cost theories.

No abstract