The NDX Can't Rally, And That's Not Good

08/20/08 3:20 PM EST

As of: SPX 1251 HELP  ARCHIVE

One of the signs I concentrate on is how markets respond to short-term overbought and oversold signals.  Generally, healthy markets that have future gains in store respond very well and almost immediately by rallying after hitting short-term oversold conditions, and either gently pause or don't react at all to overbought ones.

The opposite is also true - weak markets don't respond well to oversold conditions, and fail soon after hitting overbought.  This was one of the signs we discussed back in January, and was a big reason why I didn't want to keep buying into oversold conditions after we couldn't rally during the first few days of the new year.

We're seeing a similar test of that theory now on a smaller scale.  The STEM.MR Model for the Nasdaq 100 has been oversold for 39 straight half-hourly periods, equal to three full trading days.  Part of the reason is that those models are constructed so that they become oversold more easily in an uptrending market, and with the NDX's recent spurt, it isn't taking much of an extreme to push us into oversold.

Still, for the model to remain oversold for the equivalent of three days is very rare.  The last time it did so was into the early hours of March 11th as the index was carving out its low for the year.

Going back to January 2003, I could find 9 other instances of the model remaining oversold for 39 straight half-hourly periods.  Over the next week, the NDX showed a positive return 7 of those times, with the winners averaging a respectable +1.4% and an average maximum gain (+2.9%) more than twice as large as the average maximum loss (-1.3%).

Also notable is the two failures - January 21, 2003 and July 27, 2007.  Both instances led to negative five-day returns, and they both signaled a market that was weak and about to get even weaker.  Watching how the market behaved in the short-term gave a heads-up to the risk/reward that existed in the intermediate-term.

This circles back to what I've mentioned the past couple of days - we have something of an "oversold on support" setup here in the S&P, and if we can't rally given these conditions, and especially if we can't hold above the 1250ish area, then we'll have a pretty good sign that any intermediate-term goodwill from mid-July has been erased.  An inability of the NDX, which has helped to lead this rally, to bounce from such extended oversold model readings would be giving a sign similar to its last two failures.

I don't like to wait for technical failure to get out of my positions - I prefer to be proactive and sell when I want to, not when I believe I have to.  We got that opportunity this spring after the March rally gave us a bountiful number of excuses to sell during April and May.  But we're not getting the same kind of intermediate-term negatives here from a breadth- or sentiment-based perspective, so I haven't seen the logic in selling out of positives from July.  A technical failure and inability to rally from short-term oversold readings would change that.

We're clinging to some potential support here, but with thin trading volume and traders basing decisions off every $1 move in Oil, it's hard to have a lot of faith in any short-term trades.  A move back over 1280ish would be a good sign, a move under 1250ish would not, and in between I don't see much to do.

Watching For The Mid-Week Reversal Effect

08/20/08 11:00 AM EST

Good Wednesday morning...We begin the day with a modest recovery in the equity indices, after a brief foray below yesterday's lows.  Oil started out strong but has since pulled back after a higher-than-expected inventory build.  As we went over last Friday, Oil has had a high (negative) correlation to stocks over the past month, but it has not been the main driver.

Yesterday afternoon we went over the implications of consecutively bad breadth days, when it occurs early in a trading week.  Over the past decade, such setups have been almost impeccable at preceding short-term rallies, and it has still been very good over the past 20 years.  Prior to that, we ran into the mid-1970's which did not respond nearly as well to oversold breadth conditions.

Along with breadth, there were a smattering of extremes among our sentiment guides, most notably the short-term STEM.MR Models for both the S&P 500 and Nasdaq 100.  Overall, we still see little difference between the number of indicators in bullish or bearish territory when looking at the Indicators At Extremes section on the Daily Overview page, something that has been in force for most of the past few weeks of choppy market conditions.

Other than some short-term indicators that flit between one extreme or the other, most of the indicators on that list have been there since mid-July, like the net short positioning among hedge funds and the still-pessimistic nature of the major sentiment surveys.

We're seeing a modest jump in bullishness among those surveys, but we should keep in mind just how much of an extreme they had reached.  Bullishness in the Investor's Intelligence poll of newsletter writers, for example, jumped 10% this week - but even so, it's still at one of the lowest levels we've seen in the past decade.

Partly because of the depth of extremes we'd reached around mid-July, I've been giving the upside the benefit of the doubt, meaning that I've been less interested in selling or shorting overbought conditions, and more interested in buying into short-term oversold ones.  We're bumping up against the minimum time frame and percentage rally from the studies we went over in July, but so far I haven't seen much that suggests the intermediate-term recovery is over, unlike May when we were able to discuss a plethora of studies suggesting trouble ahead.

The recent break of the uptrend lines from the July low in the S&P 500 and DJIA are a caution sign, but they're so obvious that it's perhaps what we needed to see in order to wash out some weak holders and late-comers to the rally.  As long as we're able to hold above 1250ish, I'm still most interested in establishing or adding to long positions into oversold conditions like we've generated over the past day or so, and given what we went over yesterday related to mid-week rallies, we really should be seeing that upside kick in now.  We're seeing a bit of a recovery this morning as Oil pulls back and the Banking sector finally gets a bid, and again we should see this hold given the oversold conditions.  It would be troubling to say the least if we can't hold up here, and all bullish bets would be off, at least temporarily, if we can't hold 1250ish.