|
The pattern goes something like this: Leading up to a top, the # of New Highs far exceeds the # of New Lows. This, of course, is common sense. Signs of trouble occur when the # of NYSE New Highs becomes approximately equal to the # of 52 week New Lows, with both figures above 80 issues. This equality sometimes creates what my good friend Kennedy Gammage has dubbed the “Hindenburg Omen.” This indicator is a reliable “crash” predictor. It occurs when both New Highs and New Lows on the NYSE each exceed 2.2% of the issues traded that day, the 10 week moving average of the NYSE is still moving up, the McClellan Oscillator is below zero, NYSE New Highs are not more than twice New Lows, and there are a cluster of them (more than one) within a 36 day period. I have modified this crash signal slightly from the one Jim Miekka (The Sudbury Bull & Bear Report) is largely given credit for finding, by adding the last two requirements, based upon higher correlation with crashes going back to 1985 (see research in the Guest Articles section at www.technicalindicatoridex.com). The timing of the Hindenburg Omen is usually soon after the start of the multi-week stock market crash, telling us a top is in and we should expect panic selling into a bottom, on average within 41 days, but even possibly as far away as three months. Consider it the starting point identifier. But how do we know when that bottom is in, when the crash is over, and when it is safe to go back into the investing waters? The pattern of new highs and new lows proceeds from a point of near equality, above 80 each, to a point of reversal where the number of new lows begins to exceed the number of new 52 week highs in a day — but before the worst of the crash takes place. The margin of difference usually jumps to a pretty wide number right from the get-go of this reversal. What makes this aspect of the pattern a pretty good clue of an impending crash is that when you think about it, the averages are already near their 52 week highs (since they recently topped), yet many stocks individually have experienced severe price deflation to quickly reach 52 week lows. This suggests serious deterioration in underlying market conditions, invisible to index watchers. Continuing the crash high/low pattern development, as the freefall hits its stride, you typically see the number of new lows start to exceed the number of new highs by about 150 to 200 issues, with the number of new highs coming in below 50. The bottom of the crash is near when the # of new lows exceeds 300 to 500 or more. That sort of spike, followed by a sharp drop well under 100 gives us our signal the bottom is likely in, the worst is over. During Crash # 1, which began at the all-time top for the Dow Industrials on January 14th, 2000 and ended 36 trading days later on March 8th, 2000, the Dow fell 2,296.75 points, or 19.2 percent. We received New Lows warnings that a bottom was approaching on 2/24/00 (348 New NYSE 52 week Lows), 3/7/00 (303), and then they turned up on 3/8/00, when the intraday crash bottom of 9,611.75 occurred, diverging Bullishly with price. And yes, we had the presence of a Hindenburg Omen. Interestingly, on February 25th, the DJIA closed at 9,862.12 (a day after getting 348 new lows), not far from the eventual closing bottom for this crash of 9,796.03 on March 7th (303 new lows). By 3/16/00, they were below 100, confirming the worst was over. During Crash # 2, which started on 9/6/00 at 11,518.83, and lasted 30 trading days, and fell to 9,571.40 on 10/18/00, dropping the DJIA 1,947.43 points, or 16.9 percent, we saw a spike to 295 New 52 Week Lows the last day, 10/18/00. Early warnings that a bottom was approaching came on 10/12/05 (229 New Lows) and again on 10/17/00 (230 New NYSE Lows). And yes, there was a Hindenburg Omen present. The next day they dropped below 100 indicating the storm was over. During Crash # 3, which started on 3/8/01 and lasted 10 trading days, the Dow fell 1,893 points, or 17.3 percent. The most New Lows spiked to 258 on the last day, March 22nd. The next day they fell to 55, far below 100, the sign the worst was over. Yes, it had a Hindenburg Omen. Crash # 4 started on 8/27/01 and lasted 14 trading days. This crash included 9/11, and saw the DJIA fall from 10,498.03 to 7,926.93 on 9/21/01, a 2,571.10 point, 24.5 percent plunge. A bottom was indicated when 508 New NYSE 52 week lows registered on 9/17, 332 on 9/18, 539 on 9/19, 578 on 9/20, and then a final spike to 784 at the bottom on 9/21/01. On October 1st, New Lows fell below 100, giving the all-clear sign. Yes, it had a Hindenburg Omen. Crash # 5 began on 5/17/02 and lasted 46 trading days. The decline from 5/17/02’s 10,400.62 to 7/24/02’s 7,489.53 took the Dow Industrials down 2,911 points, or 28 percent. We received early warnings of an approaching bottom on 7/19/02 with 386 NYSE New 52 week Lows, 513 the next trading day on 7/22, 685 on 7/23, and a spike to 917 at the bottom on 7/24/02. On 7/29/02 New Lows fell under 100, indicating the worst was over. Yes, there was a Hindenburg Omen present. Crash # 6 began on 8/23/02, lasting 33 trading days, and the DJIA fell 1,870 points, or 20.6 percent into its 10/10/02, 7,181.47 low. Early warnings of an approaching bottom came with 307 NYSE New 52 week Lows on 9/30/02, 376 on 10/4, 477 on 10/7, 474 on 10/8, 604 on 10/9, and 559 the day of the bottom on 10/10/2002. The next day, on October 11th, New Lows dropped to 72, and an all-clear signal was given. Yes, there was a Hindenburg Omen on the meter. Crash # 7 began on January 13, 2003 and lasted 38 trading days. This event shaved 1,498 points off the DJIA, or 16.8 percent, to 3/12/03’s 7,397.31. Early warning of an impending bottom came on March 10th, 2003 when 249 NYSE New 52 week Lows were registered, 275 on 3/11/03, and a final spike to 319 on March 12th, 2005. The next day, on 3/13/03, New Lows fell to 78, and an all-clear signal was given. This was probably the toughest crash to call a bottom for because the number of new lows did not come close to exceeding 500. The Iraq War and massive amounts of liquidity likely served to cut this crash short. New Lows dropping below 100 was a helpful indication the crash had bottomed. Which most importantly brings us in long-winded fashion to now, the potential for another panic selling event, which we believe could come in the next few weeks — and may already be underway. If a crash is underway, I would label its start from 9/12/05, at 10,701. So far, here’s the most number of NYSE New 52 week Lows we’ve had: 319 on 10/12/2005 and 321 on 10/13/2005. But after bottoms, the number of New Lows subsides fast, dropping well below 100. Wednesday, October 19th, we got an increase to 215. While we did see a drop to 75 on October 24th,and 80 on October 25th (which in themselves are not “well under” 100, per se, since we have not had a crash yet, any drop in new lows below 100 is not meaningful as an all-clear sign yet. This analysis is relevant once a crash has occurred, to know when the worst of the storm has likely passed. Since we finished a huge three-year rally over the past several months in all the major indices, with prices hitting all time highs in many, including the NYSE, we should expect the signal for the bottom of the next crash — if in fact we are going to have one — to be a very high # of new lows, over 500, probably approaching 700 or more, followed by a drop well under 100. We cannot be sure another panic selling event has started, but if it has, the # of new lows should help us identify the bottom, both as to a spike, and then subsequent denouement as they drop well under 100.
CONTACT
INFORMATION The opinions of FSU contributors do not necessarily reflect those of Financial Sense. |
