RealTime IT News

A Broadening Concern

A rare chart pattern that predicted huge declines in Juniper Networks and Nokia last fall is showing up again in some stocks.

The pattern is called the Broadening pattern, and it is among the most bearish of chart patterns. The conditions required to create the pattern almost never occur at market bottoms.

The pattern in its orthodox form is a series of three higher highs and two lower lows, each swing occurring no more than two months apart. The pattern can also have a flat top or bottom. Juniper's top last fall was a perfect example of the orthodox pattern:

At this point, we're going to turn the column over to Robert Edwards and John Magee. In their 1948 seminal work on chart patterns, "Technical Analysis of Stock Trends," they wrote:

"The Broadening Formation may be said to suggest a market lacking intelligent sponsorship and out of control - a situation in which the 'public' is excitedly committed and which is being whipped around by wild rumors. ... There are times when it is obvious that those are precisely the conditions which create a Broadening Pattern in prices, and there are other times when the reasons for it are obscure or undiscoverable. Nevertheless, the very fact that chart patterns of this type make their appearance, as a rule, only at the end or in the final phases of a long Bull Market lends credence to our characterization of them.

"Hence, after studying the charts for some twenty years and watching what market action has followed the appearance of Broadening Price Patterns, we have come to the conclusion that they are definitely Bearish in purport - that, while further advance in price is not ruled out, the situation is, nevertheless, approaching a dangerous stage."

That about says it all. So where are Broadening patterns appearing now? We wrote about two recently: Amazon.com formed a flat-topped Broadening pattern before breaking down yesterday. And Microsoft , which represents 10% of the Nasdaq indexes, is on the verge of breaking down out of one. A close below $64.20 would just about do it, but even at that point, a retracement to $70 wouldn't be out of the question. Broadening patterns are nothing if not volatile.

The third one we noticed was the one that made us think there is something going on here. PeopleSoft , one of the few technology companies that has posted strong growth throughout the market decline (they did it again last night, and the stock is up 10% this morning), broke down out of a Broadening pattern at the $38 level last week (see chart below), the day before Microsoft warned. Even with the breakdown, a retracement to $40-$45 is not out of the question.

A word on measuring predicted declines out of Broadening patterns. The minimum predicted move is from the top (5) to the bottom (4) of the pattern, below the bottom of the pattern. In PeopleSoft's case, that 12-point pattern predicts an eventual move to $26 or lower. It may take some time, and the reasons for it are not now understood, but the decline should eventually exceed that level.

A few caveats: Broadening patterns have been known to break to the upside (a close 3% above the top of the pattern - "5" - is required to negate the pattern), but those are rare events. The pattern is among the most reliable of chart patterns because of the conditions required for its formation. But the market is short-term oversold, and due to bounce in the next day or two, so any decline might be temporarily halted. And July 30-August 9 is shaping up to be a very important cycle turn for the market. A good earnings report from Cisco , which some think is possible, could send the market higher.

But the charts look bearish to us here, as we have been saying since mid-June. Another leg down in this bear market would not come as a surprise here. But when the market is at its darkest is when it always seems to turn. We're not quite there yet, but close.

And one final warning sign: In this chart from the Market Technicians Association Web site, note the Nasdaq 100's action since early 1999 - and which side of big moves the commercial futures traders have been on, and which side the non-commercial traders have been on (both in the bottom of the chart). No bottom is likely to occur as long as the commercials are short the NDX futures, and the non-commercials are long. Note how the opposite was the case at the October 1998 bottom. The market will find it tough going without the smart money on board.