An important part of John Quiggin's book, previously discussed here, here and here, is the argument that stocks are excessively volatile. Unfortunately, he doesn't really lay out the argument he points to Shiller's work and quotes another economist vouching that the puzzle is unresolved and moves right along. In this post, I want to explain why Shiller thinks stocks are too volatile.
Below is a simple model that explains the value of a stock. It says, if you take the value of all of a stocks future dividends and discount that by the appropriate discount rate, that must equal the value of a stock.
Shiller measured changes in the growth of dividends as well as changes in the prices of stocks (Vo) and found that stock prices are much more volatile that dividends. Thus, he concluded that stock prices are irrationally volatile prone to large moves in investor sentiment.
Where's the rub? Why weren't all financial economists convinced markets are hugely inefficient? Shiller's argument assumes that k, the required return on stocks, is constant through time. In fact, that was the dominant paradigm at the time. This wasn't dictated by theory, and certainly many were aware of the possibility that k could change over time (see this Fama article), but it seems as though the dominant model in most people's head was that of a random walk and there was no strong evidence at the time that k changed.
Now, thirty-five years later, there is lots of evidence that returns change over time, and it is, in fact, hard to imagine that they wouldn't. Maybe in 2005 an 8% return that could easily be +50% or -50% though is likely to be between +30% or -30% seems like a pretty good bet. Especially if you are pretty well off with a lot of life left. That same bet in the depths of the financial crisis probably seems a lot worse. No one wanted to hold risk at the depth of the financial crisis. Its likely k, the required rate of return on stocks rose very high, though there was tremendous risk associated with this return. Stocks did very well, and there expected return was likely very high, but in another world things could have gone very badly and returns could have been very low.
Perhaps, it is obvious now that it is difficult to differentiate between rational changes in the required rate of return on stocks and irrational swoons in sentiment. In regards to the decline in stocks during the financial crisis, Shiller might argue that this was a swoon in investor sentiment, an irrational pessimism, whereas an efficient markets advocate might attribute it to the dismal state of the economy and the uncertain state of the future. Is it possible to tell the two views apart? In order to do so, we need a model of what k should be.
Shiller has to some extent worked on that problem. He's tried to argue that k moves in response to things it shouldn't. For instance, one that I like is legalized gambling. Shiller argues that periods of lax gambling laws may cause asset price bubbles. Maybe the roaring 2000s were really caused by the rise in internet poker and collapsed due to the anti-poker legislation. Nevermind that the dates don't work too well, sentiment is a funny thing and there are other causes. I'm still skeptical, but Shiller works hard and attempts to address the problems in a rigorous way.