On the plane Thursday we watched Madagascar 2 and there was a point in the movie where King Julian lamented that "the science seemed so solid" after a sacrifice into the volcano did not work as expected.
Professor Jeremy Siegel wrote an opinion piece in the Wall Street Journal making a case for calculating earnings for the S&P 500 by market cap as opposed to allowing the largest company losing $1 to have the same impact on the earnings as a $1 loss when it comes from the smallest company in the index.
By his reckoning, in 2008 the S&P 500 would have earned $71.10, which he says is 80% above the earnings number calculated by the S&P's method. Felix Salmon and Paul Kedrosky each jumped on this as not making sense. Felix called it astonishing, and Paul said the professor was a "little nutty."
I don't think there is anything wrong with the exploration - throw it on the wall and it might stick - but I don't see where this one does stick. That Jones Apparel (NYSE:JNY) should have less impact on SPX earnings than Exxon Mobil (NYSE:XOM) could make some sense, but it seems that then the index's earnings would not feel the full effect of AIG (NYSE:AIG) going from $180 billion down to $1.4 billion or Fannie Mae (FNM) dropping from $65 billion down to half a bil, not even close.
If Siegel is somehow right, then I suspect that if they went back and refigured everthang what needed refigurin' , the current 9.4 P/E for the index that he calculated would not be anywhere near as cheap as he thinks.
I don't talk a whole lot about the P/E ratio for the market because it offers no useful predictive value. The market can be cheap or expensive based on P/E for a very long time and get cheaper or more expensive.