There is a rare bull market pattern that, once it ends, is one of the worst formations that can occur. It has two names. Some call it the megaphone pattern, as its shape resembles what many know as the voice amplifier that coaches and referees use to get large groups of people to pay attention.

Those who know its implication for the financial markets refer to it as the Jaws of Death pattern. In fact, both names have meaning for this ominous indicator of egregious crowd bullishness. Its megaphone analogy reminds us that when this pattern appears, investors need to pay attention that the party attitude of recent memory is on it's "last call." The Jaws of Death descriptor is clear. When the pattern ends, the jaws snap shut like those of an alligator chomping down on hapless prey that got too close to the water.

Click here to see the following chart in a new window

Image placeholder title

Above is the monthly bar chart of the Dow Jones Industrial Average I:DJI , from the 1987 crash low (far lower left of graph) near 1600. This is the origin of the pattern, too. The green box highlights the zone between that low and place where the lower jaw line (connecting the 2002 low and the 2009 low -- the bottoms of the two biggest market crashes since 1929) will intersect a dramatic crash if one occurs within the next three years (around 6000). However, since the pattern began at the 1987 crash low, a complete round trip to the beginning cannot be ruled out, although it's not required. The price highs of December 2014, as well as February, March, and May of this year all tested the upper jaw line within 0.5%. These achievements mark the second, third, fourth and fifth tests of this monster trend line in the past 186 months (15.5 years); about 2.7%. In other words, this is a rarity of irrational exuberance.

But the exuberance doesn't stop there. Click the link to expand the chart and observe the bearish divergence sell signals that occurred at the 2000 and 2007 peaks, labeled (A) and (C), where higher highs in price were met with lower highs in stochastics. The same signal was just triggered at the May high this year and will likely recur if the Dow sneaks a new high in price into history to test the upper jaw line again. Regardless, the last two times risk of portfolio wipeout was this high were 2000 and 2007, and we can see what happened next: 35% and 50% Dow crashes, respectively; and 50% and 50% S&P 500 crashes. The crashes were even worse for the Nasdaq and Russell indices). 

What about that awesome rally on Friday? Well, looking at the internals, we notice that Friday's 370-point rise occurred with a muted 65% of stocks advancing. This compares to Thursday's decline of 252 points, when 81% of stocks declined. So, although the "manipulators" got the headlines, the story was told by the crowd.  

To put the previous chart into perspective and open our minds to the possibility that Dow 6000 is within the realm of possibility, we offer the next chart, also monthly bars, going back to 1900.

Click here to see the following chart in a new window

Image placeholder title

Viewed from this satellite perspective, 6000 becomes nothing more than an expected wave IV-circled after the completion of wave III-circled that the first chart labels this summer's peak. In fact, as the 1929 peak would be labeled wave I-circled, and the 1932 low labeled wave II-circled, a decline to 6000 would only be about half the size of the 1929-1932 crash. In other words, 6000 would be proportionally too small, under both Elliott Wave and Fibonacci theories, each having guidelines of wave II's and IV's being equal in size but different in pattern complexity. Keeping with this equality expectation, wave IV-circled should decline more like 90%, as the Dow did during the wave II-circled (1929-1932 crash) tsunami, when it dropped from near 400 to near 40. However, 6000 would fulfill the Jaws of Death's "measured move," and tickle the upper edge of the green box. 

This brings up the decision support question that allows objectivity to determine one's actions in the market: "If I had no money in the market at this moment in time, would the current conditions support buying or selling actions with new money?" Our decision support engine answers the question by telling us that only selling actions are indicated here. This is because it rarely pays off to be buying anything when a bearish divergence sell signal is on the clock, like was triggered in May, and was on the clock at the 2007 and 2000 peaks. Buying is not ideal again until at least 25% is seen on the stochastics, which is a long way from the current 79%.  

Another echo of market peaks' past is the Conference Board Consumer Confidence Index. The all-time high was in 2000, when the index logged a whopping 144 reading in January, peaking simultaneously with the Dow and S&P. Stocks crashed, as did consumer confidence, reaching the 60s as the stock market finally bottomed. In 2007, Consumer confidence peaked at a lower high around 110, in July, while stocks stretched three additional months into their all-time highs, before joining confidence's crash to lower lows than 2002, this time reaching the 20s. This year, confidence peaked in January, again like in 2000, but stocks rose four more months to peak at their new all-time highs, as confidence reached another lower high, just above 100. So, here, too, storm clouds are forming for a rocky season of financial weather.  

Finally, for this analysis, but nowhere near the end of the list of clarion calls to exit the market before "stock-mageddon" arrives, guess what happened in 1929, as well as in 2000, that is likely going to happen at the coming Federal Reserve policy meeting? Yes, both of those dates in history mark times that the Fed turned off the party lights by beginning rate-hike sequences in the only interest rate it actually "controls."

The decision support engine warns that if your are hoping to perfectly exit stocks, before the next "surprise" decline, 18,000 +/-1000 is the place to do it. Therefore, use sell stops at 17,000 as your line in the sand, and allow no rationalizations to keep you from pulling the rip cord, or you could be joining stock prices in free fall.  

For more of this kind of analysis, try our many DSE-based services with a free week of full-access to our highest level membership. Special pricing is available for TheStreet.com readers who subscribe AFTER a no-obligation, complimentary trial week.

This article is commentary by an independent contributor. At the time of publication, the author was long Direxion Daily Small Cap Bear 3X Shares (TZA), which is three times short the Russell 2000 index.