Commodity Trading Advisors Overlook Recent Market Weakness

Institutions were consistent sellers on Friday, leading to a -34 closing TICKscore and the fourth consecutive session of lower lows for the S&P500. Interesting to note that with the S&P trading more than 0.5% below Thursday’s low, new 20-day lows remained below Thursday’s level. Typically you’ll see an expansion of new 20-day lows when the SPX trades below the previous day’s low given that this is a much more sensitive measure than 52-week highs/lows. In the five-year history we have on this data, there have only been 27 instances in which the S&P traded 0.5%+ below the previous day’s low and new 20-day lows contracted, 19 of which led to a bounce the following session. In only four cases did the S&P end down 1%+ the next day, while it rallied 1%+ seventeen times…

S&P500 Trades 0.5%+ Below Yesterday’s Low, New 20-day Lows Contract
01/29/10… S&P500 ??? next day
12/31/09… S&P500 +1.6% next day
10/30/09… S&P500 +0.7% next day
07/07/09… S&P500 -0.2% next day
06/16/09… S&P500 -0.1% next day
06/03/09… S&P500 +1.2% next day
04/28/09… S&P500 +2.2% next day
04/14/09… S&P500 +1.3% next day
04/01/09… S&P500 +2.9% next day
03/20/09… S&P500 +7.1% next day
03/03/09… S&P500 +2.4% next day
02/23/09… S&P500 +4.0% next day
12/18/08… S&P500 +0.3% next day
11/21/08… S&P500 +6.5% next day
10/27/08… S&P500 +10.8% next day
10/09/08… S&P500 -1.2% next day
10/07/08… S&P500 -1.1% next day
09/22/08… S&P500 -1.6% next day
09/18/08… S&P500 +4.0% next day
09/17/08… S&P500 +4.3% next day
08/07/08… S&P500 +2.4% next day
07/28/08… S&P500 +2.3% next day
03/27/08… S&P500 -0.8% next day
01/25/08… S&P500 +1.8% next day
01/08/08… S&P500 +1.4% next day
01/07/08… S&P500 -1.8% next day
12/18/07… S&P500 -0.1% next day
11/12/07… S&P500 +2.9% next day

I’d also note that with the NYSE McClellan Oscillator holding in extreme oversold territory (<-200), the NYSE Down Volume percentage fell from 69.5% on Thursday to 67.3% on Friday. This triggers another two-day buy signal similar to the one triggered last Wednesday (see this intraday update from January 27th) . While that signal proved to be incorrect, the overall track record still maintains a significant edge over random.

Interesting to note that despite a third down week for the S&P500, the Market Vane survey of commodity trading advisors hit a new 52-week high for the second consecutive week. When this survey hits a new annual high we can infer CTAs are generally bullish on stocks, and history tells us this group is generally correct in their thinking. The last thirty instances in which the Bullish Consensus hit a 52-week high are listed in the table below, along with the performance of the S&P500 over the next two weeks…

Market Vane Bullish Consensus Hits 52-week High
01/29/10… S&P500 ??? two weeks later
01/22/10… S&P500 ??? two weeks later
12/31/09… S&P500 +1.9% two weeks later
12/24/09… S&P500 +1.6% two weeks later
11/20/09… S&P500 +1.3% two weeks later
10/23/09… S&P500 -1.0% two weeks later
10/02/09… S&P500 +6.1% two weeks later
09/25/09… S&P500 +2.6% two weeks later
09/18/09… S&P500 -4.0% two weeks later (*)
08/28/09… S&P500 +1.3% two weeks later
08/07/09… S&P500 +1.6% two weeks later
04/20/07… S&P500 +1.4% two weeks later
01/13/06… S&P500 -0.3% two weeks later
02/20/04… S&P500 +1.1% two weeks later
01/30/04… S&P500 +1.3% two weeks later
01/23/04… S&P500 +0.1% two weeks later
01/09/04… S&P500 +1.8% two weeks later
12/26/03… S&P500 +2.4% two weeks later
12/19/03… S&P500 +1.8% two weeks later
12/05/03… S&P500 +2.6% two weeks later
10/17/03… S&P500 +1.1% two weeks later
09/05/03… S&P500 +1.5% two weeks later
06/20/03… S&P500 -1.0% two weeks later
06/13/03… S&P500 -1.3% two weeks later
06/06/03… S&P500 +0.8% two weeks later
05/30/03… S&P500 +2.6% two weeks later
05/16/03… S&P500 +2.0% two weeks later
05/09/03… S&P500 -0.0% two weeks later
05/02/03… S&P500 +1.5% two weeks later
04/25/03… S&P500 +3.9% two weeks later
01/04/02… S&P500 -3.8% two weeks later (*)
12/21/01… S&P500 +2.4% two weeks later

Note that in 23 out of 30 cases, or 77% of the time, the S&P was trading at a higher level ten trading days later. That’s significantly above the 55% at-any-time odds for a higher S&P two weeks later in the same period of time. Also noteworthy was that only two cases led to an S&P down 1.5% or more two weeks later, while fifteen cases led to an S&P up 1.5% or more two weeks later. Not only was this signal triggered as of Friday’s close, it was also triggered the previous week (ending 1/22). Since then, the S&P is down 1.6%. If the market follows historical tendencies, we should see prices rebound over the coming week.

From a longer-term perspective, we don’t want to overlook the fact that the S&P closed out January with a loss, a seasonal pattern that has a strong track record of leading to subsequent weakness. In 28 out of 29 cases since 1932, or 97% of the time, the S&P posted a subsequently lower monthly close within the next seven months (usually within three – see this January 4^th column for the track record.)  That’s significantly greater than the 66% random chance for a lower monthly S&P close (below the setup month’s close) within the next seven months. Intermediate-term studies suggest the potential for higher prices over the next 1-2 weeks, so for now this will take on more significance once those bullish setups fall off the board.

Categories : Archived EOD Commentary, End-of-Day Commentary

 Print This Post