Short-term Outlook:  Neutral  From Jan 27, 1090 SPX

Equity Indicators - Updates and Extremes

Investor's Intelligence Correction %

A popular topic on trading desks has been the spike in the percentage of newsletter writers expecting a stock market correction, according to Investor's Intelligence.  The figure has been climbing for months, and currently stands at its highest level since 1983.

By "correction", the newsletters would be looking for a drop of 10% or so in the near-term, but ultimately a resumption of the uptrend.  So they're not quite bullish (short-term) and not quite bearish (long-term).

The chart below shows a long-term view of the Correction camp.  I've drawn some rough trend lines from the early 1970's through 2009, when it was clear that these letters were more likely to take a hard stand one way or the other (either bullish or bearish), and on an increasing basis.  Decade after decade, the Correction percentage coiled up into a tighter and tighter group...until the past few months.

I'm not exactly sure what this means going forward.  Perhaps it's a sign that we're heading into a time where letters are less likely to take a firm stand after being burned so many times, and will now just throw their hat into the Correction camp in order to be somewhat protected no matter what the market does.  Or, more likely, it's just a fluke.

The folks at I.I. seem to think that the current reading means there will be no correction, and instead, in accordance with contrary theory, stocks will simply shoot higher.  I'm not quite sure how they come to that conclusion (stocks could just as easily sail right past a correction and into a bear market, according to contrary theory), but I've never been much of a fan of theory anyway.  So let's look at some hard numbers.

There isn't any recent precedents for this, but let's go back to 1970 and find any other time the Correction percentage made up the majority (i.e. greater than both Bulls and Bears).  We're only looking at unique instances here, so any time weeks were bunched together only the first week is counted.

There isn't much of an edge here.  While performance going forward was weaker than random, it's not strong enough to trade or even form a conviction.  The maximum loss over the next three months was a tiny bit larger than random, and the maximum gain was smaller than random, but again there isn't much there.

Only once (02/08/80), the letters were absolutely correct in that the S&P corrected about 10% over the next few months, then headed higher.  But there were seven times when it came close, and with a very heavy skew towards risk as opposed to reward.

Overall, I think it's much ado about nothing.  I certainly would not jump to the conclusion that because of this Correction majority, stocks will just automatically head higher.  If anything, there's a slight historical precedence for the opposite.  But due to the lack of any recent occurrences, I don't have a lot of faith in that conclusion, either.