Chart:  Traders In S&P 500, Nasdaq 100 and DJIA Futures

Large and small speculators in the major stock indexes bumped up their exposure last week by a sizable amount.

The nominal dollar value of speculators' net position in both the full contracts and e-minis for the S&P 500, Nasdaq 100 and DJIA jumped to $29.6 billion.  That's the highest since December 2008 and one of the highest all-time.

The breakdown of the positions are $16.1 billion in S&P futures, $7.3 billion in the Nasdaq 100 and $6.2 billion in the DJIA.

There have been four other times in history that have matched or exceeded this kind of exposure for speculators.  None of them were particularly appealing.

*  Late November 2000 through mid-March 2001 (stocks tumbled for the next few months)

*  Mid-November 2004 through mid-January 2005 (stocks chopped lower for the next few months)

*  Late October 2006 through late February 2007 (short-term gains were given back in late February)

*  Most of December 2008  (stocks tumbled for the next few months)

I haven't been a huge fan of this Commitments of Traders data for stock indexes for a few years, especially with regard to the S&P 500 (Nasdaq 100 and DJIA have been better).  And a better contrary signal tends to be when speculators go net short, or close to it, as a buy signal for stocks.  But this latest surge is unusual, and at least worth a mention based on the precedents.

Study:  S&P 500 Weekly Price Pattern

We went over some troubling price patterns in the S&P 500 last week, but those were very short-term.

The weakness during the week triggered another one, this time on a weekly time frame.  The S&P 500 SPDR (SPY) nearly put in a "bearish engulfing" candle pattern, which has consistently led to negative returns going forward.

Friday's weakness wasn't quite enough to trigger that pattern, but still we have essentially the same thing - a market that rallied the week before, to a 52-week high, then this week it carved out an outside bar (higher high and lower low) and sold off during the week.

This hasn't been a frequent occurrence, but the ones that have triggered have led to more weakness than strength during the next month.

Every one of the occurrences led to at least a -3% selloff at some point during the next month.  None of them were able to rally more than +1.6% at any point during the month.

If there's a good side to this, it's that much of the weakness was contained to the next 4-6 weeks, and after that any further losses were made up in the months ahead.

For example, if you had waited five weeks after this pattern triggered, then bought the S&P and held it for three months, you would have ended up with this:

The drawdown was still scary with a median of -7%, but by the end of the period the S&P sported a gain each time.