The Relative Importance of Dividend News and Discount Factors in the 2008 Stock Market

Here is my back-of-the-envelope version. Let me know if you have a more precise decomposition.

It’s easy to see how the stock market might fall by a fifth when it is learned (or market participants think they have learned -- see my earlier blog post) that a one year recession will occur that would not have occurred otherwise. Suppose that, absent a recession, corporate earnings grow perpetually at rate g. A recession means that earnings drop by 15% for one year, after which they resume their previous growth rate g. A 10-15% drop in earnings is typical of postwar recessions. Relative to the counterfactual, the 15% drop lasts forever, so this by itself reduces equity values by 15%. In addition, growth does not occur for twelve months, which reduces earnings in all years after the recession by exp(-g). So values are hurt by almost 15% + g, which is somewhere between 15 and 20 percent. If the recession lasted 2 years instead of one, values would be hurt by less than 15% + 2g, which is still not much beyond 20%.

In fact, the U.S. stock market has fallen by about third. Even the nonfinancial components of the stock market have fallen 30%. This seems hard to explain merely with bad news about earnings. Rather, it likely reflects a change in the valuation of a given earnings stream. In other words, you can buy the same earnings stream cheaper now than you could last year. I am not the first one to suggest that time-varying discount factors (as opposed to news about “fundamentals” or earnings streams) are an important reason for stock market fluctuations – see the last couple of decades of asset pricing research.