This past week the American Association of Individual Investors (AAII) this past week reported an extremely low 16% bulls. This is the lowest one week level in bulls since mid September of 1992. What we find most interesting and a bit disturbing is not the low levels of bulls but the fact that this indicator, which is one of the least followed sentiment gauges we are aware of, has gone from obscure to mainstream in a little over three days. We have even seen comments from a number of people who, to paraphrase do not put any reliability in sentiment surveys of any type. These are the same people who in the summer and fall of 2003, when the bullish percentage was off the charts, were falling over each other to be the first in line to point out how useless this survey is.
Please do not get us wrong, we are not saying that this is not a potential bullish development. However, what we are saying is that to make a market call based on one weeks worth of data, no matter how extreme it may seem, is just plain dangerous. For example, from mid August 2003 to mid February 2004 a span of 28 weeks the percentage of bulls from AAII were over 50%. And twelve of those 28 weeks saw over 60% bulls. During those same 28 weeks the percentage of bears did not once move over 30% while 17 of those 28 weeks the bearish percentage was below 20. We use a 10-week moving average of the bulls divided by the bulls plus the bears to both smooth out the weekly fluctuations and for a longer-term view. In the period discussed above this indicator not only moved to its most bearish level in its nearly 18 year history, surpassing the level seen in 2000, but it also stayed at extreme levels for an unprecedented period. With the risk of sounding like a broken record, we will once again point out that sentiment is not a good timing tool. What may look to be extreme can get a lot more extreme and as we can see from the discussion above it can stay that way for a lot longer than one might expect.
Market Vane is another example of excessive and persistent sentiment both bullish and bearish. From May 2002 to April 2003 the four week moving average of this survey remained below 34% with 31 of those 50 weeks below 30%. On the flip side, from early January 2004 to this past week, a span of 67 weeks, the 4-week moving average has been over 60%. Sentiment is improving across the board. Short-term measures such as the put to call ratios and the Rydex ratios are bullish but they have been bullish or near bullish for weeks and so far have lead to nothing but lower prices. This is another example of sentiment's weakness as a timing tool. However, as was the case in 2003, this is not meant to imply that these indicators should be ignored or that they have no value. On the contrary, they tell us a lot about investors and traders future expectations and that in the long-term is extremely important.
Technically a lot has occurred over the past two weeks. All of the key market averages have moved below their 200-day moving averages, with the S&P the last to do so late last week. The break of this moving average is so far only modest and far from conclusive but the fact that it is so widely followed is reason enough to be aware of it. Of more importance in our view is the fact that the averages have broken below rising trend lines drawn off the March 2003 and August and October 2004 lows. The averages have moved well below their early 2004 peaks suggesting strongly that the early to mid November breakout has failed. The S&P and DJIA look to have completed a head and shoulders top last week when they moved below their January and late March lows. However, this pattern is a bit suspect as the right shoulder was well out of proportion with the left shoulder. However, volume within the pattern wax text book as the head had lower volume than the left shoulder and the right shoulder saw lower volume than the head. Moreover, the neckline was broken last week on increasing volume, another classic characteristic of a true head and shoulders pattern.
Short-term momentum is oversold and we do have some potential divergences on such indicators as the McClellan oscillator and the breadth and volume oscillators as they are above their late March low. However, there are two important points in regards to the current position of these indicators. The first is that while they are showing potential divergences, the key word is potential as they are far from being confirmed. Secondly, and more importantly, both the depth of the late March oversold readings coupled with the pathetic rally that as that condition was relieved looked strikingly similar to the early November period following the initial momentum surge but obviously in reverse. The McClellan oscillator had five divergences following its November 4 peak into the January 3 top while the S&P gained 70 points (6.1%). Moreover, while these indicators are showing potential bullish divergences a great many more have confirmed the lower lows. This includes RSI, both daily and weekly along with other price based indicators. The new lows have also confirmed and have hit their highest one day total in over two years. In addition, these same indicators are coming off a series of lower highs beginning with their peak in mid to late 2003. This is also the case with a number of our longer term indicators such as the percentage of stocks above their 200 day moving average.
There are a small number of indicators that can support the idea that a rally of some sort is not far off but there are many more indicators that clearly point to lower prices whether we rally first or not. We have been approaching the post March 7 decline in the S&P as a correction of the August-March rally favoring the idea of one last rally above the March peak prior to the completion of the post 2002 counter trend rally. The S&P has not yet come close to our bottom line Fibonacci support level of 1090 and as long as that level holds we have to allow for this outcome. However, the decline is taking on impulsive characteristics while a lot of damage has been done to the markets already weakening technical structure over the past two weeks. Thus, while we do have to keep open the potential for one last rally the odds are moving up that the March high may well have completed the post 2002 cyclical bull market. Short-term we are bearish and we will lower our stop from 1175 to 1170. Medium and long-term we remain bearish.
Larry Katz
Market Summary and Forecast
3835R E. Thousand Oaks Blvd PMB 297
Westlake Village, Ca. 91362
(818) 889-0485
About Larry Katz
