The good news is:
• Next week is seasonally strong.
From time to time I have alluded to "normal" markets. For many (70 + that we have data for) years the market usually followed a cyclical pattern with the following, bottom to bottom characteristics:
In the beginning, new lows diminish rapidly while the NYSE advance - decline line (AD line) and small caps lead the way up. Approaching a top, new highs diminish while the small caps and AD line under perform the blue chips. About 2-6 weeks after a cycle high for the small caps and AD line a final top is made by the blue chips (DJIA). The high is accompanied by an increase in the number of new lows. After a final top new lows increase rapidly and all of the indices decline. If the decline was severe (more than 250 max new lows) there is usually a re-test (in 2-6 weeks) of the low. This pattern failed from 1999 to 2002 when the major averages were dominated by a few tech issues that inflated from October 1998 to March 2000 then deflated. The severity of that distortion was addressed in an article in Barrons in March of 2000 where the writer pointed out that there had never before been a large cap issue that traded at over 100 times earnings and at that time there were twelve. Since March of 2003 the market has appeared normal, that is, it has followed the pattern outlined above where the secondaries lead both up and down. I keep referring to "normal" because I am still a little skeptical.
Assuming the market has returned to normal, where are we now?
The chart below shows DJIA (in red) making a cycle high on February 11, 2004 confirmed (confirmed means both the DJIA and the AD line made a high on the same day) by the NYSE AD line (in blue).
The next chart shows the AD line making a new high on April 2, 2004 while the DJIA is not making a new high. The implication of the AD line high is a cycle high in the DJIA is yet to come.
The chart below is a little messy because it has three lines, the DJIA in red, the NYSE AD line in blue and the Russell 2000 (RUT or R2K) in green. The chart begins with the first of this year. It shows the February 11 high in the DJIA, the April 2 high in the NYSE AD line and the April 3 high in the R2K (in green). The secondaries (small caps represented by the R2K) were leading the up move until early April. This chart show that in spite of contamination of the NYSE breadth numbers by fixed income related issues it still correlates pretty well to a broad measure of the market, the R2K.
If the market is behaving "normally" the charts above suggest there should be a new high in the DJIA in the next few weeks.
Some times the pattern fails and there is a case to be made that the market is on its way to a cycle low.
The chart below shows the DJIA in red, a 10% trend of NYSE new highs in green and a 10% trend of NYSE new lows in purple. New lows are plotted on an inverted Y axis so an increasing number of new lows moves the indicator downward. Monday new lows hit a high for this cycle at 211 on Friday both new highs and new lows hit lows for the week at 26 new highs and 119 new lows. The new high indicator made its cycle high in mid January and is now down to the level it reached near the low last August. The value of the new high indicator as of the close Friday was 108. A new high indicator value is used in many timing programs as a No Sell Filter, that is when the value of the indicator is above the level specified in the filter, no selling is allowed. The values for these filters usually range between 100 for the most aggressive to 130 for the most conservative so the current value would allow a sell on all but the most aggressive programs. New lows exceeding 200 is very scary, but, most of those have been interest rate sensitive issues such as preferred's and bond ETF's.
The chart below is similar to the one above except it has been constructed from NASDAQ data. The chart looks a little more positive because the new high indicator did not drop below the March low, but the data was not much better. On Friday new lows at 77 hit their highest level since March of 2003 while new highs hit their lowest level since the recent March 2004 low.
The market is at a critical point:
All of the technical ingredients are in place for significant decline to a cycle low which would require all of the averages to drop below their March lows. The pattern for a developing top is also in place. This would be fulfilled if the DJIA rose to a new high unconfirmed by the breadth indicators and small cap indices. It could turn out to be a consolidation period following the rapid gains last year. I am guessing it will be a consolidation period because the advance into early this year was extremely broad and "normally" there would be period of narrowing leadership before a final top was reached.
Last weeks decline was the worst last 5 trading days May for both S&P 500 (SPX) and R2K in the 15 years covered by the FastTrack database. For the R2K it was more than double the previous worst decline.
Seasonally the first week in May has been pretty good, especially for the small caps as can be seen in the tables below.
First 5 days of May. The number following the daily return represents the day of the week; 1 = Monday, 2 = Tuesday etc.
| R2K | Day1 | Day2 | Day3 | Day4 | Day5 | Totals |
| 1989-1 | -0.24% 1 | 0.15% 2 | 0.15% 3 | 0.24% 4 | 0.33% 5 | 0.63% |
| 1990-2 | 0.25% 2 | 0.40% 3 | 0.37% 4 | 0.42% 5 | 0.21% 1 | 1.66% |
| 1991-3 | 0.91% 3 | 0.84% 4 | 0.18% 5 | -0.10% 1 | 0.10% 2 | 1.93% |
| 1992-4 | 0.15% 5 | 0.62% 1 | 0.36% 2 | 0.33% 3 | -0.10% 4 | 1.36% |
| 1993-1 | 0.44% 1 | 1.12% 2 | 0.69% 3 | 0.10% 4 | 0.01% 5 | 2.36% |
| 1994-2 | 0.59% 1 | 0.18% 2 | -0.02% 3 | -0.14% 4 | -0.96% 5 | -0.35% |
| 1995-3 | -0.09% 1 | -0.02% 2 | 0.35% 3 | -0.44% 4 | -0.02% 5 | -0.21% |
| 1996-4 | 0.57% 3 | -1.24% 4 | 0.26% 5 | 0.01% 1 | -0.35% 2 | -0.74% |
| 1997-1 | 0.78% 4 | 2.41% 5 | 2.39% 1 | -0.19% 2 | -0.42% 3 | 4.96% |
| 1998-2 | 0.42% 5 | 0.11% 1 | -0.77% 2 | -0.49% 3 | -0.71% 4 | -1.44% |
| 1999-3 | 0.11% 1 | -0.16% 2 | 0.39% 3 | -0.20% 4 | 0.63% 5 | 0.76% |
| 2000-4 | 2.50% 1 | -2.62% 2 | -1.93% 3 | 1.28% 4 | 2.18% 5 | 1.41% |
| 2001-1 | 1.06% 2 | 0.24% 3 | -1.22% 4 | 1.49% 5 | -0.66% 1 | 0.91% |
| 2002-2 | 0.03% 3 | 0.50% 4 | -0.20% 5 | -1.84% 1 | -0.78% 2 | -2.29% |
| 2003-3 | 0.04% 4 | 2.22% 5 | 0.52% 1 | 0.72% 2 | -0.61% 3 | 2.89% |
| Averages | 0.50% | 0.32% | 0.10% | 0.08% | -0.08% | 0.92% |
| Winners | 87% | 73% | 67% | 53% | 40% | |
| SPX | Day1 | Day2 | Day3 | Day4 | Day5 | Totals |
| 1989-1 | -0.17% 1 | -0.32% 2 | 0.01% 3 | -0.13% 4 | -0.05% 5 | -0.66% |
| 1990-2 | 0.44% 2 | 0.67% 3 | 0.33% 4 | 0.84% 5 | 0.63% 1 | 2.91% |
| 1991-3 | 1.32% 3 | 0.06% 4 | 0.07% 5 | -0.19% 1 | -0.73% 2 | 0.53% |
| 1992-4 | -0.58% 5 | 1.06% 1 | -0.02% 2 | -0.01% 3 | -0.23% 4 | 0.22% |
| 1993-1 | 0.52% 1 | 0.36% 2 | 0.11% 3 | -0.28% 4 | -0.21% 5 | 0.48% |
| 1994-2 | 0.47% 1 | 0.00% 2 | -0.29% 3 | -0.08% 4 | -0.79% 5 | -0.68% |
| 1995-3 | -0.09% 1 | 0.12% 2 | 1.09% 3 | 0.01% 4 | -0.08% 5 | 1.05% |
| 1996-4 | 0.06% 3 | -1.71% 4 | -0.27% 5 | -0.13% 1 | -0.40% 2 | -2.45% |
| 1997-1 | -0.35% 4 | 1.81% 5 | 2.13% 1 | -0.30% 2 | -1.47% 3 | 1.82% |
| 1998-2 | 0.83% 5 | 0.10% 1 | -0.59% 2 | -0.95% 3 | -0.89% 4 | -1.49% |
| 1999-3 | 1.46% 1 | -1.67% 2 | 1.15% 3 | -1.13% 4 | 0.97% 5 | 0.78% |
| 2000-4 | 1.09% 1 | -1.50% 2 | -2.16% 3 | -0.39% 4 | 1.64% 5 | -1.32% |
| 2001-1 | 1.36% 2 | 0.08% 3 | -1.49% 4 | 1.44% 5 | -0.24% 1 | 1.15% |
| 2002-2 | 0.89% 3 | -0.17% 4 | -1.03% 5 | -1.93% 1 | -0.30% 2 | -2.55% |
| 2003-3 | -0.07% 4 | 1.50% 5 | -0.38% 1 | 0.85% 2 | -0.51% 3 | 1.39% |
| Averages | 0.48% | 0.03% | -0.09% | -0.16% | -0.18% | 0.08% |
| Winners | 67% | 67% | 47% | 27% | 20% | |
As of Friday's close the NASDAQ composite was 1% above its March low while the R2K was only 0.4% above its March low. The R2K was down the last 3 days of last week. The R2K has exceeded 3 consecutive down days only 5 times in the past year and each of those preceded strong rallies.
I expect the major indices will be higher on Friday May 7 than they were on Friday April 30.
Last weeks forecast was a spectacular miss. When the indicators are unanimous they can be unanimously wrong.
