Deflation and the Equity Risk Premium

I posted this piece earlier today at the Asia Times “Inner Workings” blog. Normally I don’t double post, but this is a cool piece of analysis. Pardon me for repeating myself. 

In the crudest version of the dividend discount model, the stock price P is a function of earnings and the discount rate, such that P = E/r. That this is an inadequate model goes without saying, but it is not entirely misleading for comparative statistics in a short time horizon. 

Below I present a simple analysis indicating that the shrinking equity risk premium is due largely to disinflation or prospective deflation. 

The S&P 500 has fallen by a bit over 12% from its April peak. But the 10-year Treasury yield, the usual proxy for the discount rate in the equation, has fallen much father, from 3.85% to 2.57%. Plugging this into the model, the equation tells us that expected earnings have fallen by 42% between April and August. 

Of course, I don’t take that number at face value, but it seems intuitively clear that with the implosion of Treasury yields, stocks have become more attractive. The fact that stocks have fallen so far despite the decreasing attraction of alternative investments suggests that earnings expectations have fallen a good deal farther than the S&P index. 

Something else is happening, though, to the relative attractiveness of stocks and bonds. This explains one of the reasons not to take the dividend discount model at face value. In the chart below, the blue line is the difference between the corporate Baa yield and the dividend yield on the S&P 500. The red line is the annual inflation rate over the preceding five years. 

 

More than half of the change in the stock-bond relationship can be explained by inflation. I have kicked the econometric tires on this relationship, and it holds up under close scrutiny. 

Inflation is bad both for stocks and bonds, but it is much worse for bonds, because even though inflation introduces inefficiencies into the economy, corporate earnings have a chance to keep up with inflation and fixed interest payments do not. If we actually move into deflation, as in the 1930s, we observe periods in which the dividend yield on equities exceeds the yield on bonds (which makes sense; why buy physical assets when they are likely to be cheaper next year?). 

If we enter into a period of deflation, or extremely low inflation, the huge advantage of stocks over bonds may disappear; companies may have to pay a dividend yield comparable to their bond yield in order to keep equity investors in the game. And that would portend a prolonged period of low equity valuations. 

In short, two things changed between April and August: lower earnings expectations, and the fear of deflation. There simply is no way to parse the data closely enough to quantify which of these two effects was more important for the stock market. The point is that neither of them are good, and neither point to a particularly rosy outlook for stocks going forward.