The S&P's P/E Ratio

I'm reading The Intelligent Investor again. If you are interested in investing and haven't read this, you can pick it up for $14 from Amazon.com (click here). Don't worry, there's no kickback to me for that, I just think it's a great read that all investors should have at the top of their collection.

In one section of the book Graham gets into valuing the overall market. He looks at a number of metrics including price to prior year's earnings and price to prior 3-years' average earnings. In the commentary section (which Jason Zweig wrote) there is an additional look at more recent times (the book was published in 2003) and comparisons to the past using some data generated by Robert Shiller.

Not surprisingly, it inspired me to look at where we are today. Shiller's preferred metric is current S&P price versus average earnings over the prior ten years. His site shows us at a bit over 24x as of January. Looking back at his numbers (which go back to 1871), 24x isn't all that bad if you just look at post-1990, but it's still pretty heady if you consider it in the context of the entire history that Shiller has laid out. In fact, prior to 1990, the last time Shiller shows us at over 24x was in 1966, and the period from 1966 to 1980 was not too pretty for US equities.

Now if you flip to the tab that he has comparing P/E to interest rates (figure 1.3 at the bottom) it seems relatively obvious why we've held these higher-than-average earnings multiples -- low interest rates. I wouldn't give this chart the last word, but it shows P/E ratios and long term interest rates moving in fairly opposite directions. This is also something that jibes with theory.

But what happens when interest rates start moving back up?

After all, this credit mess isn't going to last forever, and when something happens to get us out of defcon 5, the Federal Reserve will have to get back to fighting inflation (much to many people's chagrin). I'm not expecting hyperinflation or anything like that, but 3% rates are hardly sustainable. And when that happens, we'll probably -- or we should at least -- see a contraction of P/E multiples.

I wouldn't argue that this means to stay out of US equities altogether. Particularly if you're Graham's classic defensive investor, investing on schedule is still the best idea (if you don't get that reference, see above and buy the book). Even if you're a bit more on the enterprising side, I don't think it means bail out completely.

What I do think it means is that investors should be avoiding stocks that are at the high end of the valuation scale. I know, I know, new and shocking info... As unexciting of a result as that might be, investors that aren't chasing high P/E momentum stocks when overall P/Es start to compress will have fewer wounds (if any) to lick than those that are caught with stocks sporting 40x and 50x multiples.

-AvgJoe