Dr. P.V. Viswanath
The Debt-Equity Choice
Report 137 - March 22, 2006
Managers frequently issue equity in periods when stock prices are rising. The fact that managers sell equity when their company’s stock price is historically high is at odds with the prevailing paradigm of corporate finance theory. The authors look for an explanation of this puzzle. They also study why firms issue equity rather than debt, and how leverage ratios vary with stock market cycles.
The paper develops a new theory of corporate capital structure based on a game-theoretic methodology. It then proceeds to test it using data on more than 5,000 equity and debt issues which occurred between 1993 and 2002 on US stock exchanges. The issuing behavior of the firms is carefully described, and then the main claims of the authors’ theory are found to be consistent with the evidence.
Managers avoid issuing equity when investors disagree with their expectations on the profitability of new investment projects. Issuing equity under these conditions would drive the stock price down. If managers are interested in short-term stock price movements, they will issue equity when they have the same expectations as investors. This implies that equity issuance will prevail when stock prices are high, and debt issuance when they are low. It also implies that equity issuance is followed by increased capital expenditure. These predictions are confirmed by the recent issuing behavior of US companies.
Many theories try to explain why and when firms issue equity, but their predictions can only partially explain the observed behavior. There are, in fact, various unexplained empirical puzzles. This paper proposes a new theory based on potential disagreement between investors and managers, and uses innovative empirical techniques to show that it can explain the puzzles observed.
The “pecking order theory”, instead, states that firms prefer to avoid issuing equity because investors expect managers to issue equity when the firm is overvalued. As a consequence, investors are only prepared to buy equity at a discount. This in turn induces firms to avoid issuing equity, and to resort to debt whenever possible. There is empirical support for both theories, but there are also a few empirical facts that are not entirely consistent.
A puzzling circumstance is that firms tend to issue equity when stock prices are historically high. According to the trade-off theory, each firm should have an optimal leverage ratio - the ratio between debt and equity. When the value of equity goes up, the leverage ratio goes down. Firms should then increase debt. We may observe though that firms do exactly the opposite, issuing additional equity as its price goes up. The pecking order theory too is hard to reconcile with the evidence. We clearly do not observe firms issue equity only as a last resort. On the contrary, equity issues are quite frequent, and are not typically made by firms with no other options. How can we explain these facts?
One popular explanation invokes a kind of irrationality within capital markets. According to this view, managers who think that the market is pricing equity above its true value decide to issue equity in order to obtain funds at a low cost, with investors not realizing it. Without abandoning the realm of rationality, the authors propose a new theory focusing on a possible disagreement between managers and investors on the profitability of new investment.
The reasoning goes as follows. Suppose that managers can either invest in conservative projects - with a low return but also with small risk (variance, in jargon) - or in innovative projects with higher risk and higher expected return. If they issue debt they will only be able to finance the conservative project, since lenders will want to put covenants in the debt contract preventing managers from taking the risk. Issuing equity, on the other hand, does not restrict decisions. Debt also has well-known tax advantages, so managers would want to issue equity only if the extra flexibility is valuable.
The value of equity depends on whether managers and investors agree on the profitability of an innovative project. Managers are interested both in the long-term and short-term impact of their decision. When they recognize a good innovative project, they know that the long-term impact will be positive. On the other hand, the short-term impact depends on the way investors perceive the project. If they disagree with management, and believe that the investment is not profitable, they will sell shares as soon as the investment is announced, so that the short-term impact on the stock price will be negative.
The main point of the paper is that equity's extra flexibility is valuable only when the disagreement between investors and managers is small. When the disagreement is large, managers avoid innovative projects because they want to avoid the negative short-term impact on the stock price. If they choose the conservative project then debt financing is preferred because of its tax benefits. The theory therefore predicts that equity will be issued more frequently when prices are high, because high prices are associated to low disagreement. However, the real reason to issue equity is that disagreement is low, not that prices are high.
If we could measure the level of disagreement between managers and shareholders, we would observe more frequent equity issues when disagreement is low, even if stock prices are low. A second prediction is that equity issuances are more likely to lead to higher capital expenditures than debt issuances, since firm are likely to issue debt for tax-saving purposes rather than to finance investment.
But is it possible to find reliable empirical measures of disagreement? The authors propose a few alternative measures, but choose to focus on the difference between actual and forecasted earnings. When actual earnings beat expectations, this is taken as a signal that investors trust managers and therefore agree with their investment decisions. In other words, if managers can deliver earnings that consistently beat expectations then there is little reason for the market to question their strategy.
Based on a sample of over 5,000 debt and equity issuances between 1993 and 2002, the empirical results strongly support the theory. The probability of issuing equity for a firm with a high value for the agreement variable is 20% higher than the corresponding probability for a firm with a low value. It is noteworthy that high levels of agreement increase the probability of equity issuance even when stock prices are low.
The authors also show that capital expenditure increases more for firms issuing equity than for firms that issue debt. This is consistent with their theory, but not with other existing theories. If, for example, managers decide to issue equity just because stock prices are high then there is no reason why they should increase capital expenditures afterwards. If on the other hand equity is issued when there is low disagreement on the investment strategy, it is natural to expect an increase in capital expenditure.
According to these results, investors should not feel cheated when a company issues new shares, nor should managers believe they can exploit investors’ irrationality.