Why P/E Multiples Point to Optimism

by: Caleb Sevian

The way I see it there are two ways to look at economic data. One way is to pick a select number of indicators, anoint them with royalty and focus exclusively on their message. Another is to try to take a 30,000 foot view of things and realize that some of the data is noise and some of it is genuine and rather than panic upon data releases, look for trends as a whole as opposed to any one individual indicator.

Last week was a confusing week because the Chicago Fed Survey came in higher than expected, diametrically opposite of the Philadelphia Fed Survey of the previous week.

At the same time GDP for Q2 was revised down from 2.9% to 2.6%, durable goods orders were down, and the National Association for Realtors reported that housing prices saw their first year over year decline in eleven years. Meanwhile the futures market has solidly begun to price in a Fed rate cut in Q1 of 2007.

The trend seems to be pretty clear: the economy is slowing. The Fed is banking on the premise that this slowing will cause a drop in demand that will in turn eventually lead to a down tick in inflation. The risk is that inflation is already entrenched and it will take more suffering to cool demand. I don’t think this suffering would play out as more rate hikes, more likely just a longer time before we actually see the Fed start to ease.

The major debate continues over whether we will experience the illusive soft-landing that has only happened one other time in modern Fed history or whether we are head for a more pronounced downturn.

It seems clear the economy will slow and earnings growth will diminish, but it is not clear what this means for the stock market.

One measure we could look at to evaluate this would be Price to Earnings multiples. Campbell and Sheller published a piece that is widely quoted that looks at historical P/E’s over the last 104 years. Over this time they calculated P/E’s to average approximately 14.5. Today’s S&P500 (NYSEARCA:SPY) P/E is about 17.4, so by historical standards we are overvalued.

But is the economy today the same as it was in 1920? If we take a more shallow look back to 1960 the average is 17.5. If we take an even more shallow look back to 1996 the average is 23.6. This data says two things: (1) the trend in P/E multiples is expansion if you look at rolling periods and (2) the stock market today is actually significantly undervalued versus its most recent standards.

Given that bond yields are still not offering investors a much better opportunity, the risk-reward favors equity exposure. This doesn’t mean the market can’t go down from here, but what it does mean is that the probability of a severe sell off is not likely.

One last note on Q4 performance: bring in the Quants. Quantinvestor Blogspot ran an interesting piece last week on the probability of a Q4 rally in a mid-term election year cycle, something I have been discussing for months. According to their calculations, from 1950 through 1995, the average returns of Oct, Nov and Dec during this cycle were 2.51%, 2.46%, 1.97%, respectively, with the two most recent periods of 1998 and 2002 turning in 8.03% and 8.64% in the month of October alone. That sounds sort of compelling to me.

If any of you remember, one of the last quant pieces I brought up was the adage of “Sell in May and Go Away”. I suggested that the data indicated that when we have a very bad May the rest of the adage goes out the window and quant results show a stronger likelihood of rally June through September.

By the way, did I mention that we just finished the best Q3 the stock market has seen in nine years?

More by Caleb Sevian.