Snapshot Notes - November 8, 2007

As of:

SPX 1457

This morning we went over a couple of things that suggested we were perhaps in for more short-term weakness. Despite a smattering of oversold readings, taking a look at the data from a couple of different perspectives showed that yesterday probably wasn't the last of the selling.

One of those looks was via the AAII sentiment survey. That poll of individual investors showed more than 50% of them with a bearish point of view for the latest week, which was the highest amount in more than six months.

Over the life of that survey, readings over 50% were a perfect contrary indicator a handful of times, but mostly the market had a bit more work to do on the downside before bottoming. Waiting a couple of weeks, then buying and holding for a couple of months, led to 83% winning trades, so there was some value in such an extreme reading.

With stiff selling pressure again today, we registered a few more extremes, mostly related to the percentage declines we were seeing in the indices. That was particularly true in the Nasdaq 100, which was in danger of dropping more than 6% over just a couple of sessions. When that has happened during past Novembers, and/or when the index was in a general uptrend at the time, it had a tendency to precede short- and intermediate-term rallies.

There's nothing shocking in that, of course. Buying into price panics during uptrends, with positive seasonality as a kicker, is a consistently successful strategy. Curiously, even with the selling pressure prior to today, we weren't getting too many oversold types of readings, but that changed somewhat today - our short-term STEM.MR Models for both the S&P 500 and Nasdaq 100 have given "buy" signals by reaching "excessive pessimism" territory and reversing higher.

The technical condition of the S&P has deteriorated significantly over the past couple of days, so I'm not too excited about oversold conditions there. The NDX is still in good condition, though honestly I would have preferred to see this kind of damage spread out over a week or so instead of concentrated in just two days. Be that as it may, the bottom line is that we're oversold, during a generally good time of the year, and with the NDX at least still in a defined uptrend. That should result in at least a bounce, so we'll see if the bulls can manage it. If not, it speaks volumes about just how weak we really are here.

As of:

SPX 1457

HELP ARCHIVE

One of the reasons I've been citing why I want to look for oversold conditions is simply because of the time of year. Oversold conditions during the fourth quarter tend to be good intermediate-term buying opportunities.

As an example, I checked for any time since 1950 that the S&P 500 dropped 2.5% in one day during November. It was relatively rare, with 10 occurrences. Buying that close and holding for two months (taking us through early to mid January of the following year) resulted in 9 out of 10 positive trades, with an average return of +5.9%. They weren't without some volatility though - the average drawdown during the two months was -4.4%, while the average maximum gain was +8.4%.

If you would try that same stunt in August, by contrast, you would have only had 6 winners out of 11 tries with an average return of just +1.4%. For the record, doing it in September would have turned out the best, with 12 out of 12 winners and an average return of an impressive +9.8%.

The Nasdaq 100 is also showing a rare performance here, with a two-day decline of greater than 5%. This has only happened twice to that index in November while it was still in a long-term uptrend at the time (defined as a rising 200-day moving average). Buying into that mess was a good idea both times in the short- and intermediate-term, but the usual caveats apply about reading too much into just two occurrences (11/15/91 and 11/18/03).

If we just look at one-day declines of more than 4% in that index, while it's in an intermediate- and long-term uptrend (i.e. rising 50-day and 200-day moving averages), regardless of the month, then we get 20 occurrences. It bounced back the next day 12 of those times by an average of +1.1%, and over the next week 14 times by an average of +2.5%.

None of this is terribly surprising. Large price panics and/or oversold conditions during up-trending markets, especially during a seasonally positive time of the year, tend to lead to rising prices over the short- and intermediate-term.

I mentioned something earlier today about the possibility of more short-term pain, and I'm not ignoring that now, but with what we've gone over the past couple of days, and our short-term models beginning to flatten out in deeply oversold territory, the risk/reward is finally beginning to look more positive on the long side. I don't at all like the fact that the S&P 500 is back to what I would consider a down-trend, so the risk of trying to buy into oversold conditions is higher than if it was still in a technically positive condition, but still maybe worth it for a small trade.

As of:

SPX 1511

HELP ARCHIVE

Good Thursday morning...we begin the day mixed futures in the pre-market, with tech being taken down a notch and most of the other indices creeping into positive territory.

Many times over the past week, I've lamented the fact that I couldn't find much evidence that we were becoming oversold, despite the stiff selling pressure. I was most interested in looking for that kind of condition, since the overall price trend of the market still appeared healthy, and we are in a generally positive seasonal time frame.

Because of that, I hadn't been doing much trading-wise. The chop we'd seen quickly flipped on traders trying to be aggressive one way or the other, but finally yesterday the bears were able to profit handsomely on the break of support in the S&P.

That kind of break of obvious resistance looks horrid on a chart, and most technicians will now be staring at the pattern of lower highs and lower lows in the S&P that has formed over the past month. Combined with the numerous breadth divergences we were seeing, and some warning signs among sentiment like we went over this weekend, a heightened sense of caution doesn't seem like that bad of an idea.

A day like yesterday is sure to trigger at least a few extremes, and finally we're getting some. One of them is the Up Issues Ratio, which computes the percentage of stocks that closed up on the day relative to all stocks that traded. Over the past week, that ratio has averaged 32%, low enough to be considered extremely oversold. Even on a month-long basis, the average has dropped under 45%, enough to be considered oversold on a longer-term time frame.

Out of curiosity, I checked for any time those two averages were this low over the past 40 years, while the S&P 500 also lost at least 2.5% on the day, and the 200-day average of the S&P was rising. Basically, I was looking for price panics with oversold short- and long-term breadth measures during an up-trending market.

Surprisingly, only two occurrences popped up....10/14/87 and 8/27/98. That caught my eye, because afterwards it took the S&P two days to lose another 7% in the '87 case (before it then lost another 20% on Black Monday), and it lost 8% over the next two days after the '98 occurrence. Those are modestly intriguing examples because the setups were fairly similar to our current one - a summer high, followed by a quick correction in the fall, a rally back towards the high, then a break below support that triggered the oversold readings. Then *that's* when the real pain hit.

If one would have been able to wait until two or three days later, then began buying, those were great buying opportunities (hindsight is a wonderful thing, eh?). About the only trigger I see that would suggest we've seen the bulk of the selling is a big intraday or consecutive-day reversal. We saw the latter in both the '87 and '98 cases as equities rebounded big-time in one day after the low was put in, settled back over the next few days, then took off to the upside.

Some more signs of oversold are coming in, such as the AAII sentiment survey. This poll has given us some good signals lately, and has once again dipped into "excessive pessimism" territory, at least in terms of the percentage of folks saying that they're negative on the market.

That bearish percentage reached 51% this week, the highest number since late April. Here's an interesting thing...if during any time in that survey's history (about 20 years), we'd seen a bearish percentage over 50% and waited for two weeks before buying the S&P, and we held that trade for two months, then we'd have had 83% winning trades (24 out of 29 weeks), with an average return of +3.1%. The average drawdown (i.e. maximum loss) of -3.2% was dwarfed by the average maximum gain during the trades of +5.7%.

The key, though, was waiting for a week or two - we often saw the market get whacked in the short-term. After 17 of the 29 weeks that had this large of a bearish percentage in the survey, the S&P lost at least 2% during the next couple of weeks, and it lost more than 5% six times. While a few times it caught the exact low, the risk of it not doing so was high enough to be a concern.

Arguing that at least a short-term low has already been put in is the pattern from yesterday's trading. I checked for any time that the S&P 500 futures gapped down more than 1%, then ended up closing at least 2.5% lower than the previous day's close.

A total of six occurrences popped up, and the S&P managed to close the next day in positive territory in five of the six cases by an average of a very impressive +2.3%. By a week later, all six were positive by an average of +6.0%.

Those futures were pounded after the close yesterday, losing another 12 points on the S&P before a gradual recovery this morning. CSCO seemed to be the main culprit for the weakness, as it's being taken down by 7%+ in the pre-market.

I can find only three other times in the past decade that it's dropped this much on the open (09/28/98, 03/05/01 and 09/10/01), and on average the Nasdaq 100 opened lower by more than 3% on average - though those were much more volatile times. Currently the NDX is indicated to open lower only by -0.3%, reflecting CSCO's drop in influence as a bellwether. A probably useless tidbit...buying the open in the NDX on the days that CSCO gapped down by more than 7% and holding for two weeks resulted in positive returns all three times - and by an average of +22.2%(!).

Given the stuff we went over this morning and yesterday afternoon (regarding the Banking sector), it seems likely that we're within two to three weeks of a trade-able long situation heading into the end of the year. Perhaps we've already reached that point, but I want to see some price confirmation first - in too many cases in the precedents we researched, the market continued to drop over the next one to three weeks before stabilizing and shooting higher. So in exchange for the danger of missing out on catching the low, I'm going to excuse myself from the risk of suffering one of those "not quite yet" moments.

Forwarding or otherwise distributing this copyrighted material is a breach of your subscriber agreement. Violators are subject to termination of their subscription with any received subscription fees forfeited. Any references to historical performance are based on data we deem to be reliable, but are based upon feeds from third parties. We do not recommend subscribers take positions based on data presented here alone, but rather incorporate it into a comprehensive investment outlook.