Not Looking Good At All

09/04/08 4:25 PM EST

As of: SPX 1251 HELP  ARCHIVE

Well, that was a wretched day.  As we went over earlier this afternoon, the market was showing all the classic signs of a trend day, making it more than likely we'd end the day at or near the day's lows - and it followed through on that pattern to a "t".

This is the first ~3% loss for the S&P since June, and only the third one in the past five years.  Since the last bear market, these large losses have led to a marginally consistent pattern, that being a one- or two-day reflex bounce, then another two- or three-day dive lower that leads to a better rally.

This was also the worst day for NYSE breadth since June, with 2300 more stocks declining on the day than rising.  That kind of exceptionally poor ratio could possibly be a signal that we've hit the "wash out" phase of the decline, as it triggered after the S&P has already declined for three straight days.

Over the past 30 years, when the S&P has dropped for three consecutive days and then suffers a day with an Up Issues Ratio under 15% (like we saw today), then it bounced back the next day 79% of the time (19 out of 24 instances) by an average of +1.2%.  If we look out longer than a day, or more than 30 years in the past, then the figures revert closer to random, suggesting that this is a purely short-term mean-reversion kind of thing.

Earlier today we touched on the Nasdaq 100, and its streak of four straight 1% declines (yesterday wasn't quite a 1% loss, but we'll just wink and say it is).  Again, we typically saw mean-reversion kick in over the short-term, with the index being higher three days later 80% of the time, with very positive (but volatile) average returns.

I'm sure there will be more aspects of today's decline that stand out after I have all the closing figures and a chance to take a look more in-depth at some others.  But generally what we see after days like today is no surprise and something we've discussed many times - a very short-term bounce is typical, but after that it becomes more of a toss-up. 

Over the past couple of weeks, my preference for long-side trades has dwindled significantly due to the reasons we went over in late August.  Now with the intermediate-term trendlines broken to the downside, support levels getting breached without a pause, and negative seasonality (during the month of September), I'm even less inclined to try to step in on the long side with anything other than small token trades unless we see some very solid edges emerge from historical levels of oversold readings, which I'm not really seeing yet.

We have the jobs report tomorrow morning, so that is another wild card - a large gap down open based off that report could trigger a stab at a long trade for a very, very short-term bounce only, but we'll cross that bridge if we come to it.

Watching For Four In A Row For The NDX

09/04/08 1:20 PM EST

As of: SPX 1251 HELP  ARCHIVE

This morning we went over some data that suggested the recent streak of up days in the BKX Banking Index should soon come to a halt, and if the market indices couldn't follow through on the potentially bullish price patterns by holding its first-hour low, then the long side didn't look good at all.

The indices have indeed failed, and in a major way, with most of them down 2% or more.  The S&P 500 spent most of the first hour chopping in a trading range, which could have - and should have - led to more of a bounce.  After the first half hour, I placed a (small) buy stop order over the morning's high in the S&P 500 tracking fund, SPY, and unfortunately it got tripped within $0.03 of the day's high, which is now being threatened to get stopped out.  Nothing like top-ticking the market to ruin one's day...chalk that one up to s**t happens, I guess.

Something else we discussed this morning was the S&P's recent out-performance in relation to the Nasdaq 100, which had reached an extreme according to the 5-day Relative Strength Index.  Historically when we've hit that kind of extreme, the ratio between the two reversed in the short-term, and we're seeing hints of that today with the NDX holding up a bit better than the S&P.

Still, the NDX is getting hit for another 1%+ loss.  Yesterday wasn't quite a 1% loser, but it was close enough that we can pretend it was.  If we look back over that index's history to see what's happened when it's lost 1% or more four days in a row, we see 25 instances (including overlapping days).

The next day, the NDX was higher 60% of the time by an average of +1.7%, but within three days it was up 80% of the time with a +2.4% average.  The returns were extremely volatile, with an average maximum gain during the three days of +6.0% and an average maximum loss of -3.4%.  The last eight occurrences were all positive, dating back to September 2001, with the only recent occurrence being 11/12/07.

So far today we're seeing all the classic signs of a trend day down, which is rare.  When it lasts this long into the day, we tend to close at or near the day's low, so I don't have a lot of hope that we'll be seeing a major upside reversal this afternoon.  The S&P lost that 1265ish support area and it's now looking to take out 1250 as well.  The NDX level that most are watching is 1800, with which it's flirting as I type.  By the end of the day, perhaps we'll be "so bad it's good" giving us a better upside edge for the coming days, but for now I'm even less optimistic about our prospects than I have been. 

Time For The Nasdaq To Lead (Or At Least Fall Less)?

09/04/08 9:20 AM EST

Good Thursday morning...We begin the day with a fairly large gap down in the pre-market futures, as we saw a steady sell-off following the morning's economic releases.  Tomorrow morning we have the jobs report, which is almost always a market-mover.  Oil is bouncing back a bit this morning (lower Oil hasn't helped much the past couple of days) and foreign markets are modestly weak on average.

A subscriber pointed out the recent action between the S&P 500 and Nasdaq 100 over the past week, with the ratio between the two spiking to a unique level.   His question was whether the one-week Relative Strength Index between the two was extreme enough to be meaningful.

The chart below shows this ratio, along with the 5-day RSI:

The 5-day RSI hasn't been this high in 13 years, so we have seen an extreme move over the past week.  Going back to 1990, the RSI of the ratio has at least been above 90 on 29 days.  Over the next week, the S&P was up 59% of the time by an average of +0.5%, and the NDX was up 62% of the time by an average of +1.8%.

The ratio between the two was positive only 32% of the time by an average of -1.2%...meaning of course that we saw mean-reversion with the NDX catching up to the S&P over the next week.  Shorter-term, it was even more pronounced - over the next three trading sessions, the NDX out-performed the S&P 82% of the time by an average of 1.1%.  Going out further than a week, we see the NDX still out-performing the S&P, but it becomes much closer to random.  With the NDX down more than 1% each of the last three sessions, it looks like that index may be nearing a reprieve, at least in relation to the broader market.

One of the reasons the S&P has been stuck in a fairly tight range while technology gets hammered is because other sectors have been holding very well.  Retail and Finance in particular have continued to climb over the past week in spite of other trouble areas.

The BKX Banking Index in particular has shown a strong trend, rising each of the past six days (my quote vendor shows a slight gain on 8/29, others may show a tiny loss).  The last time the BKX rose six days in a row was in February 2007...right before it fell off a cliff.

This kind of trend persistency has not led to much that was positive in the past, with the index showing a negative bias going forward.  It went on to score seven straight up days only 40% of the time, and three days later it was higher only 38% of the time with an average return of -0.8%.  That seemed to have some impact on the S&P 500 as well - three days after the BKX climbed six days in a row, the S&P was positive 45% of the time with an average of -0.1%.  That's quite a bit weaker than its average performance over the past 15 years.

This morning's gap down is just adding to the trouble we've seen over the past several days.  The last time the S&P 500 showed three straight daily losses, then gapped down at least 0.5% the next morning was March 4th of this year, so it's been awhile since we've seen this kind of weakness.  Over the history of the S&P 500 tracking fund, SPY, there have been 41 such occasions.

Buying the S&P on the open of the day it gapped down and holding 'til the close resulted in winning trades 56% of the time with an average return of +0.6%.  That's not all that great, but holding for an extra day boosted the winning percentage to 61% and the average return to +1.2%.  Even better was holding yet another day, which put the winning trades at 73% and the return at +1.4%.  The average maximum gain was +2.9% and the average maximum loss was -1.8%, which is clearly skewed to the positive side but doesn't quite meet the 2-to-1 reward to risk ratio that I prefer to see.

We've looked at a few price patterns over the past couple of days that suggested another down day yesterday should lead to a short-term bounce.  Our most sensitive guides were on their way to carving out solid oversold readings, but the rebound yesterday derailed that and they're now mostly neutral again, so that support is no longer active.

Over the last couple of weeks, I've outlined why I was no longer interested in solely looking for short-term oversold conditions to buy, as our intermediate-term prospects began to dim.  More than anything, I was looking to sell short as much as go long, depending on whether we were looking overbought or oversold.  The technical picture in the indices is seriously questionable, with the S&P 500 neutral at the very best, and moderately bearish more objectively.

That questionable trend makes me less interested in trying to trade only marginal long-side setups, which it still looks like we have here.  I mentioned yesterday afternoon that it appears as though we have a weak "oversold on support" setup in the S&P given its ability to hold above 1265, and that's still my take.  If we trade under 1260ish, then I'd have no interest in long-side trades unless we either whoosh lower to create some major oversold readings, or we see a "false" breakdown that then snaps right back above 1265.  This morning's open will take us right towards that support level, so the first hour or so of trading should determine which way we're going to tip here.  If we hit lower intraday lows after the first hour of trading, and fall beneath yesterday's lows, then the probability of success of a long-side trade will not look good at all.