Looking to Reach Overbought by Expiration

Bottom Line: “Everyone” knows that September has an historical tendency to be weak, but that has not been the case so far this year. If this strength should hold out, it bodes well for the following month as well.

So far, September has done a good job at shaking its reputation as the worst month of the year. Many seasoned technical analysts will tell you that when a consistent pattern fails, it often tells you more than if it had not failed in the first place. My take on that is that if September bucks its bad reputation and closes higher, perhaps that tells us the market is strong enough to overcome whatever negative influences this time of year tends to exert, and could lead to even better performance the next month as well.

To test that, I checked each September in the S&P 500 since 1950 and what its performance predicted for the following October. The table below outlines the results.

October Performance in the S&P 500 1950 - 2003
	After Positive September	After Negative September
Average Return	1.0%	1.1%
Chances of Being Positive	57%	63%
Biggest Gain	11%	16%
Biggest Loss	3.5%	22%
Average Intra-month Gain	3.4%	3.8%
Average Intra-month Loss	1.7%	3.5%
Biggest intra-month Gain	15.6%	16.9%
Biggest Intra-month Loss	7.4%	30.1%

From the table above, we see that if September was a positive month, it did not lead to extraordinary performance in October, but in general it was an indication that we were not going to fall apart in that month either. Out of the 10 worst intra-month drawdowns in October, 7 of them were preceded by losses in September. Perhaps more notably, out of the 21 years when September was positive, only 3 times did we see an intra-month drawdown of more than 5% during the following October. We can see from the table that the average drawdown was a rather small 1.7%, while the average maximum gain was twice that, at 3.4%.

So 70% of the worst Octobers, including every one of the 5 worst ones, were preceded by negative Septembers. However, three out of the top five best intra-month gains in Octobers were also preceded by negative Septembers. In fact, looking at the 10 best intra-month performances in Octobers, 6 of them were preceded by negative Septembers. It appears as though a loss in September had a certain tendency to lead to volatile performance in the following month, either positive or negative.

If we can maintain the positive performance that September has generated so far, it looks like the potential for a good October improves as well. Most of the really bad Octobers have been preceded by negative Septembers, and while a good performance by the S&P for the rest of this month does not guarantee we will not fall apart next month, the odds do seem to support a better month than if this month ends negatively.

Bottom Line: Analyzing the futures positions in the S&P 500 has been difficult, but below we outline a way that has still been relatively effective even over the past year.

I have not mentioned the information from the Commitments of Traders report for a couple of months, mainly because of the problem I mention nearly every time – the increasing acceptance of the e-mini contract as an alternative to the full contract has made analyzing the positions among the various traders much more difficult. The last time I mentioned the data, in June, I showed how using the total dollar value of all positions, for both full contracts and the e-minis, and for all index products (S&P 500, Nasdaq 100 and the Dow) could maybe give us a little better picture than the traditional view.

I think there’s still a better way to view the data than I showed the last time. On the site, we post stochastic indicators for each of the S&P 500 positions, so let’s pursue that type of analysis a bit further. Recall that a stochastic indicator simply compares the current reading to all other readings over the past year (or two years, or whatever other time frame you choose). If the stochastic is 100, then the current reading matches or exceeds the highest over the look-back period; if it is 0, then it is the lowest.

The chart below shows the difference in the stochastic between the total dollar value of the commitment from commercial traders minus that of the small speculators. If the indicator is 100, then that means that the commercial stochastic is 100 and the small speculator stochastic is 0. In other words, commercial traders are the net longest they have been in two years at the same time small specs are the most net short. This should be bullish for the market. If the indicator is -100, then the commercial stochastic is 0 and the small spec stochastic is 100 – which means commercial traders are the net shortest they have been while the small specs are the most net long. This should be bearish for the market going forward.

We can see from the chart that since 2003, whenever this indicator reached 50 or above, the market tended to do well afterwards. Conversely, when it hit -50 or below, it often struggled. If we tighten the parameters a bit and look out over a longer time period, this type of relationship holds up.

Since 1992, when the CFTC began reporting this data weekly, the market has performed fairly consistently after extreme readings in this indicator. The table below outlines the future performance of the S&P 500 after the indicator hit the +80 and -80 thresholds, giving the average return the given number of months later, along with the percentage of time the S&P was positive (in parentheses).

S&P 500 Performance After Stochastic Extremes

1992 – 2004

3 Months Later

6 Months Later

12 Months Later

+80 or Above

4.3%

(79%)

6.1%

(69%)

13.6%

(89%)

-80 or Below

-1.5%

(47%)

-2.9%

(44%)

-4.3%

(36%)

From the table, we can see that the S&P was higher after 3 months nearly 80% of the time when the indicator was a positive 80 or higher. Conversely, it showed a negative return and was higher less than half the time when it reached the opposite extreme. One year later, the differences were even more dramatic, as the S&P sported an average return of over 13% with a nearly 90% success rate after very positive readings, but was higher only 36% of the time with an average return of more than minus 4% after very negative readings.

I have cautioned against using the traditional interpretation of the Commitments of Traders data for many months, but I think viewing the data as described above can give us some useful information. Using the total dollar value of all contracts for all index products, and viewing it on a relative basis (via the stochastic indicator), we can gauge the extremes between the groups of traders that tend to be wrong and those that tend to be right. The data in the table above is compelling when looking at one-year returns after seeing positive and negative extremes, and I think it should continue to be watched. Currently, the indicator is slightly negative at -34; if we see it drop further next week, pushing it below -50 or especially -80, it will be additional evidence that the rally may be nearing at least a temporary end.

In the comment this weekend I noted that there were still some traders holding onto the idea that this rally is nothing more than an oversold bounce that’s due to fail. The Rydex traders were part of that group, as they had not yet embraced this rally as they had the other two 5% uptrends this year. That began to change a little with Monday’s performance, as the bull funds saw a $60 million inflow while the bear funds suffered a $75 million withdrawal. That activity pushed the Beta Chase Index higher (one of the warning signs I noted), though we should see it spike up over 8 or so before becoming too worried about that particular group of traders.

I also mentioned the SPY Liquidity Premium last time, and today the figure was again very low at -33%. The 10-day average has now dropped to -10%, which again is not quite extreme but another day like today will make it so. This is another sign that traders are close to becoming excessively confident in their stock ownership and do not feel a need for the relative safety that ETFs can provide.

Overall, then, the past two days have not changed the bigger picture much at all. It still appears as though shorting into ever-higher prices is an exercise in futility (and losses), thus obviously making that an unattractive option. If we see another couple of days like we have seen so far this week, it should make the short side look considerably more attractive, but so far we are not seeing enough extremes in our measures to be confident enough in a decline to risk fighting this kind of uptrend. Option expirations, such as occurs this Friday, often coincide with market turning points, so perhaps we will chug higher a few more days, giving us more solid overbought readings along with that kind of consistent seasonal influence.

This disclosure is not intended as trading advice in any form. It is meant as a note to subscribers that the author may have a position directly affected by the market outlook reflected in the commentary. Although the author takes great pains to remain objective in any commentaries, it is only fair that readers should know that the author may have taken positions in accordance with his market outlook. Positions can and do change at any time, without notice to the reader.