On February 23, 2009 JP Morgan slashed its dividend from $0.38 to $0.05 per share—a whopping 87% cut. The decision was made in the midst of a global financial crisis, and it was described by JP Morgan as a “precautionary measure to help ensure that our fortress balance sheet remains intact.”
Did the stock price of JP Morgan fall through the floor? Not at all. The stock price went up by roughly 5%.
The experience of JP Morgan begs a natural question: Why did the market react so positively, given that dividend cuts are usually met with a large stock price decline?
In recent research, I propose one potential answer, arguing that the informational content of dividends varies with leverage. The classic view that higher dividends convey good news is only valid when firms do not have too much debt. When firms are highly indebted, higher dividends become bad news, and a dividend cut can send a positive message.
The basic story is straightforward. Consider a company that is close to default. Although the company’s fortunes look bleak, it still has some cash in its vaults. It does not take much thought to see that the firm’s shareholders face a strong monetary incentive to “cash out.” Any cash that is left inside the firm will be seized by its creditors once the firm enters bankruptcy. Why leave the money to creditors when you can put it in your own pocket?
(Morally, such behavior is dubious at best. In some jurisdictions it is also illegal. However, proving that a firm paid high dividends in an effort to hurt its creditors can be difficult in practice.)
The monetary incentive to cash out becomes stronger as the firm drifts closer to default. As a result, when the stock market observes high dividends, it need not attribute that to financial strength. If the firm has a lot of debt, the market may instead see a firm that is running for the exit. In these circumstances, a dividend cut conveys a reassuring message: “We are here for the long haul.”
When it comes to JP Morgan, they have certainly stuck around.
* * * * *
Complete paper—which includes a quantitative evaluation, discusses how well the results fit existing evidence and has really cool graphs—can be found here, along with the replication code. Or take a look at the slides for a quick tour.