- Stocks are discounted cash flows. When you buy a stock, you get a claim to future cash flows, mostly dividends. We think of its value as reflecting these future cash flows and the discount rate we apply to them. If stock prices fall, it could reflect either source: either our assessment of dividends fell, or the rate at which we discount them went up.
- Stock price changes mostly reflect changes in discount rates. How do we know this? One reason is that stock prices vary a lot, and dividends do not. Robert Shiller noticed this a long time ago (see Figure I). The other is that changes in prices are not generally followed by similar changes in dividends, which is what we’d need if future dividends were driving prices. This is a moderately technical point, but John Cochrane has a nice summary.
- Changes in discount rates often reflect changes in risk. Digging further, we might divide the discount rates applied to dividends into two components, the discount rates we’d apply to low-risk assets like US Treasuries and the risk adjustment we add for dividends. Since stock prices and Treasury yields aren’t very closely related, risk would seem to be the central component.
So what happened yesterday? If we take history as a guide — and why not? — then it’s mostly discount rates. Treasury yields didn’t change much (in fact they fell a little), so that leaves us with risk.
We like to call that progress, but it leaves us with some unanswered questions. What’s the risk? And why did risk assessments go up yesterday and not the day before? Analysts refer to China and Europe, which is fine, but what makes this anything but a story we made up after the fact? And it doesn’t tell us why this happened yesterday rather than the day before or the day after.
My colleagues add: All the high-risk investors were in the Hamptons, leaving us at the mercy of short sellers. (Trolls are everywhere.)
Update (Aug 24): Ben Casselman adds: Don’t sell, you’ve already taken your lumps.