A Key Test for Stocks

If the S&P 500 falls through support in the 1440 to 1465 range, things could get very ugly, very quickly

Intermediate-term pullbacks and corrections during bull markets seem to come in at least two flavors: A lot of pain all at once, or mild pain that lasts weeks or months. Obviously, the late-July slide is the first variety. In other words, the faster the drop, the quicker the bottom.

While technical analysis is our bread and butter, we think following the indicators we discuss each week becomes especially important in times of exaggerated market moves – like the ones we saw last week. Why? Technical analysis takes the emotion out of the entire investment decision. It also gives us definable targets to look for.

That said, while we made some cautious statements in our comment last week, we missed raising the red flag and while it is our hope and desire to call all the intermediate-term moves before they happen, not wait until certain supports are broken to shout sell, our crystal ball is dirty but hopefully not broken.

The S&P 500 completely obliterated all pieces of short-term technical support during its extended sell-off last week, very similar to the February pullback and the one last summer. On Tuesday, the index broke underneath its 50-day exponential moving average, and on Thursday, all hell broke loose. The index slid right through the 65-day exponential and 80-day simple moving averages for the first time since late February.

The "500" then sliced through the double bottom lows formed in June at the 1490 zone. The index then proceeded to plunge into an area with no real definable support, in our opinion, declining all the way to 1465.30 on an intraday basis.

Right below Thursday's intraday low is a stack of technical supports in the 1440 to 1465 range. We not only think this zone will be tested but also believe that this area of support must hold or things could get very ugly, very quickly. The S&P 500 bounced off the first piece of support Thursday, and that is the trendline drawn off the closes since last July. A similar trendline drawn off the intraday lows since last summer sits at 1450. The first piece of significant chart support, from the February high, is at 1460. A 50% retracement of the rally from March to July targets the 1463.60 level, while a 61.8% retracement comes in down at 1442.50.

It is likely that there will be a lot of talk about the 200-day moving average and its importance for providing support for the S&P 500. The 200-day simple average is at 1448 and the 200-day exponential average lies at 1450. We do not know why this average gets so much press but we think it is not a good indicator to use for timing the overall market.

For instance, since the bull market began, both the 200-day simple and exponential averages have been busted on four occasions, and the "500" then proceeded to reverse back to the upside. The 200-day exponential average did provide nice support during the March pullback, but it is just not a consistent winner at timing the longer-term moves of the market.

We prefer to use the 325-day exponential moving average, which correlates to a 65-week average. This line provided a floor for the S&P 500 in August 2003, August 2004, April 2005, and September 2005. It also acted as support, although not perfectly, in the summer of 2006.

In addition to the 325-day average, we also monitor a set of simple, moving average crossover systems that many times have done a very good job of keeping you in during bull markets and out during bear markets. We look at the 13-week exponential average vs. the 34-week exponential as well as the 17-week exponential average vs. the 43-week average. Thirteen and 34 just happen to be Fibonacci numbers.

We use these crossover systems in a number of ways. For instance, when the market has been rallying on an intermediate-term basis, and the gap between the two becomes wide, the likelihood of a pullback or correction rises. When the "500" drops below the 13-week or the 17-week, caution is advised from an intermediate-term standpoint.

But the real value for timing the market from a longer-term perspective is to watch for bullish and bearish crossovers.

Both these crossover systems have been on buy signals since May and June 2003, respectively, and although there have been some near misses; we have not yet seen a bearish cross in either of these.

We believe Thursday's action represented at least part of a selling capitulation or panic that might not have fully played itself out. Basically, we're waiting for another shoe(s) to drop before we can be more comfortable that the worst is over. Volume on both the NYSE and the Nasdaq were all-time highs on Thursday, as we believe market participants have gotten less rational and more emotional. These periods can be quite nerve wracking but many times represent a market that is closer to a bottom than one that has a lot further downside.

Market internals were horrible Thursday, and this many times is seen after a fairly decent-sized pullback and not usually this close to all-time highs. NYSE new lows/issues traded exploded to 23.5%, the highest reading since May 10, 2004, when they hit 24.2%. It is very rare to see this many lows, and it is even rarer to see this many lows with the market so close to an all-time high. Since 1990, there have been only 15 times when new lows/issues traded exceeded 20% and 4 of them occurred on a consecutive basis during the 1998 bear market and 5 of them occurred during the 1990 bear market. In general, they tend to occur towards the bottom of an extreme washout or where there is evidence of panic selling, when everything is thrown out.

Combining NYSE lows and Nasdaq lows was also striking on Thursday with a total of 1,197. This was the highest reading since July 24, 2002, or a day after the first low of the bear market. During the current bull market, readings over 400 issues have represented oversold territory. What we like to see from these charts before calling a bottom is a positive divergence or lower peak in the number of issues posting 52-week lows. This can take weeks or months, depending on the duration of the pullback or correction.

We think the intermediate-term trend of the stock market is bearish, and despite some of the evidence of panic selling and washed out internals, the market is not at a bottom until the price action tells us so. We like to see a big intraday reversal to the upside, although this is not necessary, followed by a pretty strong rally that fails, and then tests the initial low. We then like to see a strong upside move, preferably on strong volume that takes out the initial rally high.

What we described is a double bottom that many times acts as a bullish reversal formation. This is certainly not the only type of low, but seems to be most prevalent. Until we get some type of reversal formation, it is best to step aside and wait for Mr. Market to tell you when the bottom is here.

Before it's here, it's on the Bloomberg Terminal.