The DJIA 30 index closed at its record high on 11/17/06, 5.3% above its previous all-time high on 1/14/00. The day before, 11/16/06, was the first day during the rally from the summer lows that none of the 30 components made a record high along with the index, which we pointed out – to our private list recipients at the time was a negative precursor to the rally topping out. (To join see the left side panel at this website.)

We noted that because August, September and October were up months, both individually and cumulatively, which is contrary to usual seasonal trends -- especially in the mid-term Congressional election year -- this year would probably be another exception to the “average” pattern of the six-month strong period of the annual cycle starting from an autumn low and continuing through the winter and peaking in the spring, especially following the four-year election cycle lows.

We expect these recent highs will prove similar to the high four years ago on 12/2/02 where the gains from the mid-term Congressional election year lows were effectively wiped out during the following 3½ months later though mid-March the following year.

Also, we noted that 19 of 30, or 63%, of the 30 Dow components made their individual all-time highs more than five years earlier, and, on average, were 27% below those individual record highs – see the updated table below. We expect this diminishing leadership and breadth will also be bearish over the intermediate (weeks to months) and long term (quarters to years). This is because we expect that the six-year downtrend in smoothed stock market P/E ratios – see charts below and attached – will continue to decline along with corporate earnings during 2007 and well into 2008 as a result of the incipient severe recession. See our Stock Market and Economic Cycle Template Model.

We also pointed out on 11/17/06 that the magnitude of the primarily short-covering and safe-haven catch-up driven sucker’s rally taking the Dow to multiple record highs was very similar to Dow’s 5.2% rally of new all-time highs at yearend 1972 when the previous Supercycle bear market B-upleg made its “irrational complacency” echo-mania high.

The S&P 500 made its closing high nine trading days later on 12/5/07 and its intraday high this morning, 12/7/06, although it was less than 1% higher than its intraday high on 11/17/06 when the Dow peaked. This S&P 500 high is 63 trading days after both its and the Dow’s record-high breakouts on 10/3/06, which is only one day longer than the year-end 1972 echo-mania high analog.

The recent intraday and closing highs in the S&P 500 are still 9% below its all-time highs on 3/24/00, which is consistent with the pattern of the “irrational complacency” echo-mania highs in the four previous Supercycle bear markets of the past 125 years, as we explained and illustrated in our bearish call at the May 5 stock market highs. [Read]

Below the following table is an e-mail I sent to Mark Hulbert about his review of responses from Jeremy Siegel and Cliff Asness to stock market performance since Robert Shiller’s “irrational exuberance” call and about our P/E Predictor Study 1 following up Campbell-Shiller’s study about the predictability of stock market P/E ratios.

Bob Bronson December 7, 2006
Bronson Capital Markets Research

Here’s a typical attempt to technically chart the stock market’s quarterly P/E ratio on a current or trailing basis. We don’t believe any shorter term approach like this is very effective without understanding what fundamentally drives the stock market’s P/E ratio

But longer term, a properly smoothed P/E ratio does explain about 50% of the stock market’s Supercycle trends over the following 12 to 20 years. I explain this in the following e-mail, slightly modified, that I sent Mark Hulbert with four following charts:

From: Bob Bronson [mailto:Bob@Bronsons.com]
Sent: Thursday, December 07, 2006 9:35 PM
To: 'Hulbert, Mark'
Subject: RE: Hulbert, Siegel and Asness on Campbell Shiller's market P/E ratio forecasts

Mark,

Following Campbell-Shiller's July 1997 published work, we followed up with our P/E Predictor Study I, which I discussed with Robert Shiller in April 1999. It differed from their work in the following ways:

• While they simply used Ben Graham’s 10-year averaging horizon, we tested for the optimal forecasting horizon, which turns out to be a 50%-exponentially smoothed mean-reverting average (similar to a three-year moving average) of the P/E ratio itself.

• While they arrived at a 51% r-squared for 10-year averaging and 10-year forecasting horizons, we found that overstated the relationship because of multi-collinearity with dividends in both in total return and earnings, which include dividends.

• While they used a 10-year forecasting horizon, we found the optimal one ranged from 12 to 20 years, which was consistent with our ~16-year BAAC Supercycle Periods. These improvements correctly resulted in a ~50% r-squared.

• While Shiller introduced their 10-year forecast in a presentation to the Fed on Dec 3, 1996, which you referenced, our ~16-year forecast was first applicable when the small-caps peaked on Oct 10, 1997 and we, at the same time, made our Supercycle Bear Market Period call.

• Our study also referenced the history of Supercycle pattern alternation between small- and large-caps, using their price peaks of Jul '99 and Mar'00 for the start of the Supercycle Bear Market Period in large caps.

As a currently relevant consequence of our P/E Predictor Study I, which is available upon request, I’ve attached some .gif file charts you may find interesting.

Click to enlarge

Click to enlarge

Click to enlarge

Click to enlarge

© 2006 Bob Bronson
Editorial Archive

Contact Information
Bob Bronson
Bronson Capital Markets Research
Email