No replies to this topic

[TTHQ Staff]

Is it real or is it Memorex? This question is in regards to the huge surge in breadth and more so the huge surge in breadth based momentum indicators. To review, the McClellan oscillator has moved from a near record low to a record high in a matter of 13 trading days. Although not a record, our breadth oscillator has nonetheless moved to a significant overbought reading from near record oversold levels in the same period. This is no doubt a breadth thrust or initiation signal at least as it relates to these indicators. However, while breadth and breadth related indicators have surged the averages have not come close to showing a commensurate move in price. The S&P for example has rallied 47 points or 4.3% while the DJIA has added 364 points or 3.7%. Since and including August 1982 there have been twelve occurrences of what can be viewed as a breadth thrust. Those were August 9 to August 26 1982, July 24 to August 9 1984, December 31 1986-January 15, 1987, January 15 to February 6 1991, November 29, 1991 to January 6 1992, July 24- August 6 1996, April 14-May 12 1997, January 9-February 9 1998, October 8- October 22 1998, September 21-October 5, 2001, July 23-August 19 2002 and March 13-June 4 2003. These dates are based on the price low of the beginning of the move and the peak in momentum. We are using a simple 10 day moving average of breadth. Including the current signal there are now 13 occurrences. According to Marty Zweig who is credited with developing the signal, the shorter the time span between the initiation of the signal and the peak momentum reading the stronger that signal is. In some cases above the peak in momentum occurred in less than 15 trading days. In some, such as October 1998 the peak occurred in less than 10 days. However, in looking over the above dates most of these were indeed were the initiation signals. The one true exception was July-August 2002. There are a number of common threads between these signals and price. In every case, even in August 2002 the S&P has rallied strongly into the momentum peak. The average gain for the S&P on the previous 12 signals is 14.42%. The lowest gain was 9.4% in July 1996 and the highest was 22.4% in June 2003. If we eliminate these two we still have an average gain of 14.1% for the ten remaining signals. Of those 12 there are seven that saw the peak in momentum occur within three weeks of the price low. Those dates are August 1982, July 1984, January 1991, July 1996, October 1998 and September 2001.The average price gain for the S&P on these seven occasions was 14%. The average tine from low to peak in momentum is 25 calendar days. If we drop off the shortest (13) and the longest (91) we get 20.5 calendar days. Another common thread of the previous successful signals is that price usually continued to move higher well after the peak in momentum and if it did not move higher it corrected very little. Moreover, the initial reaction in the indicators was to correct to a less overbought condition or to neutral but they never moved back to oversold levels on the first correction off the peak. However, this is not as important as the reaction in price. The one failure of the above mentioned signals was in August 2002. This is exactly what occurred then as price peaked commensurate with momentum and declined rapidly as did the indicators as well. We are now at day 21 and near the average time of the previous 12 signals. As mentioned above, the S&P has so far rallied a fairly unimpressive 4.3%. This well below the averages of the previous 12 signals. This is by far and away the least impressive move given one of the, if not the strongest momentum surge on record. One last common thread of the previous 12 signals was that there was a fairly substantial divergence between price and momentum where as momentum bottomed well in advance of the price low. The two exceptions to this was September 2001 and July 2002. In both cases those price lows were taken out and in the case of the former by a fairly substantial amount. At the May low we did not have a divergence. It does seem logical that this rally and momentum surge is clearly lacking in comparison to the previous signals over the past 22 years and is flawed. However, we cannot yet confirm that the peak in momentum is in so that does buy it some more time. Moreover, since we have nothing historically to compare this with we really cannot draw any true conclusions at least not yet. This should become more evident over the next week or two but at this point we need to be a bit open and flexible in regards what it may or may not mean. The failure in August 2002 not withstanding, the strong surge in momentum even if not a bona fide breadth thrust still needs to be respected as it does suggest some decent momentum if only for the short-term. At the same time, the volume related indicators, while overbought are not nearly as strong as their breadth counter parts. In all but one of the previous 12 signals did the 10-day moving average of net volume fail to confirm. Here too we cannot confirm that peak momentum is in so we cannot yet draw anything conclusive as yet but it is another warning sign and one that needs to be watched closely. Short-term we are going to remain neutral but with the idea that the rally from mid May is not over. Medium-term, we are still of the view that the rally is a counter trend affair and that the decline from the March high on the S&P is not complete. This is supported by the majority of our medium-term indicators. However, we do see the need to be flexible and we will be watching closely to see how the market does in regards to the strong surge in the breadth related momentum. Long-term we are bearish.