This piece in the NYT from the weekend contrasts two views of the U.S. stock market. In one corner is Yale professor Robert Shiller, who says that the market is valued meaningfully higher than it normally would be, i.e., higher than it should be. In the other, Wharton professor Jeremy Siegel, saying that it is fairly valued.
These guys are both smart. Who's right? Probably Shiller has the better argument; the market is high.
As the NYT piece suggests, it is hard to resolve the argument using only the Shiller 10-year P/E (also known as the cyclically-adjusted price-to-earnings ratio or "CAPE"). The market looks high on Shiller's number, but it looks fairly valued based on a trailing 12-month P/E ratio. Plausibly, this is because the two recessions of the past decade have made average 10-year earnings figure too low. I mean, that's the argument the bulls make. On its face, it seems like it might be right.
The bears, in Shiller's camp, argue that the current earnings number is above trend, and at risk of reverting. In the end, that seems to be the stronger case. Shiller's 10-year P/E number looks about right, even with the recessionary earnings in as part of the history. I think the easiest way to get a sense of this is to use the "q" ratio from Andrew Smithers (Tobin's q, basically) that I blogged about yesterday. When you click through to the data in yesterday's post, be sure to look at how very well correlated q and "CAPE" are over time. The fact that they say the same thing is a pretty decent indicator that the "high market, high profit margin" camp, represented by Shiller, has the stronger argument.
As Bob Arnott says in the same NYT piece, there are other bits of evidence that help Shiller's case, like the fact that wages as a share of GDP are depressed. But I think you may find that using the q chart I linked to yesterday is a particularly intuitive way to see it.
Shiller opines in the linked article that stocks are high, but not crazily high. Long-term returns from these levels will have to be lower than they have been historically.
[Edited 10/16/12 and 10/17/12.]