Finally, the S&P rallied, rising 5.2% last week. Although overdue for a bounce, the market climbed despite being hit with some of the worst economic and corporate news this year, so far.
Auto sales for the month of January were abysmal, with GM sales dropping a staggering 51% year over year. Weekly claims surged to 626,000, the highest level since 1982. Factory orders fell 3.9%, more than expected. And the big number–payrolls for January–was disastrous: down 598,000 jobs, the worst job loss figure since 1974. And finally, the unemployment rate jumped to 7.6%, also more than forecast.
Corporate earnings continued to disappoint: Dow Chemical, Disney, Kraft, Cisco, Time Warner, News Corp., Biogen Idec, and many others reported fourth quarter 2008 results that were below consensus estimates. To add insult to injury, many provided weak outlooks or withdrew guidance completely.
So the S&P closed the week at 869, about 45% below its peak in October 2007. At a 45% discount, many market analysts are convinced that investors should “take advantage” of such depressed prices and buy.
But let’s take a look at what constitutes depressed prices, using P/E ratios over the last 70 years, as provided by Standard and Poors. Let’s also use trailing 12 month actual, not projected, earnings to remove the guesswork of estimating next year’s earnings.
What does history say? Over the last 280 quarters (70 years), the percent of time the S&P500 was valued below a P/E ratio of 10 was 22%. Even as recently as the 1970’s, the S&P500 was valued at a single digit P/E 40% of the time.
By comparison, where are we now? As of last week, with about 65% of the Q408 reporting submitted, the P/E stood at a whopping 25.
So it really doesn’t matter exactly how accurate your 2009 earnings estimate is–take $40, $50, or $60. If our recession continues to deepen and lengthen, the current 25 multiple will come crashing down. Let’s use the $50 estimate (which is much more optimistic than the current top-down estimate–$42–published by Standard and Poors) and apply a multiple of 10 (possibly generous, given historical data) which gets us to an estimated S&P value of 500.
Now that would be cheap…..really!