BULLISH PERCENT INDEX GIVES ANOTHER BEARISH DIVERGENCE -- A BREAK OF APRIL LOWS WOULD PROVIDE INTERMEDIATE SELL SIGNAL ON THE S&P 500

A HISTORICAL LOOK AT THE BPI... One of the themes that I keep coming back to is that the 2006 bull market advance into new multi-year highs isn't being confirmed by a lot of technical indicators that measure market breadth. I've written articles showing negative divergence between the market and the NYSE Advance-Decline line and the McClellan Summation Index. Today I'm going to show the Bullish Percent Index (BPI) for the S&P 500. [The Bullish Percent Indices can be found on the Market Summary page for all of the major stock indexes and most of the major market sectors]. You can see a line version or a point & figure version of those charts. I'm using the line version here. The BPI is the percentage of stocks in a given index that are on point & figure buy signals. Chart 1 shows the BPI for the S&P 500. The general guidelines are these. Readings below 30 identify a very oversold market and usually occur near major market bottoms (green line). The last time we had a reading under 30 was in the spring of 2003 just before the current bull market began. The 50 line divides bull and bear markets (purple line). Readings over 50 identify a bull market. The BPI crossed over 50 in April of 2003 (purple circle) and signaled that a new bull market was in effect. Once the 50 line is exceeded, subsequent bull market corrections should find support near the line. That happened three times during 2004 and 2005 (purple arrows). Pullbacks near 50 usually provide good buying opportunities. The red line at 70 is the start of major overbought territory. Readings over that line mean that the market is overextended and vulnerable to a setback. At the start of 2004, the market reached an unusually overbought reading near 90. During 2005, however, the market pulled back twice from more normal overbought readings in the 70's. The red arrows in Chart 2 show the last three pullbacks in the S&P 500 that coincided with peaks in the BPI. The purple arrows show the bottoms that occurred around the BPI 50 line. Notice that each of the successive BPI peaks is lower than the previous one. That's a sign that the market advance is being supported by fewer stocks in uptrends.

Chart 1

Chart 2

A STUDY OF DIVERGENCES SINCE 2005... Chart 3 is a closer comparison of the interplay between the BPI and the S&P since the start of 2005. In this chart, I'm focusing more on the last three negative divergences that have occurred between the BPI line (black line) and the S&P 500 (green line) since the start of 2005. To me, a negative divergence exists when the S&P hits a new high and the BPI doesn't. The first red arrow shows that happening in March of 2005, which was followed by a downside correction. The second arrow shows another negative divergence in August of 2005, which was followed by another correction. In those two instances, drops in the BPI below 70 signaled that a correction was in progress. Which brings us to 2006. The lines speak for themselves. After peaking in the low 70's in January, the BPI has fallen to a new 2006 low. The third arrow shows an obvious negative divergence between a falling BPI line and a rising S&P line. One of the two lines is going in the wrong direction. I'm not sure which one it is, but history would seem to suggest that it's the S&P 500. The relatively old age of this bull market, a climate of rising inflation and rising interest rates, and a negative turn in the seasonal pattern all add to the fact that the market is beginning to look increasingly vulnerable to a downturn.

Chart 3

SUPPORT LEVELS TO WATCH FOR ... It's one thing to point out negative divergences. It's another to spot an actual market downturn. Until an actual price downturn occurs, the divergences are just warning signs. Sometimes those warning signs are early and the market keeps advancing. That's what's been happening so far. The difficult choice at the moment is whether to pay more attention to the price action or the deteriorating market indicators. Chart 4 shows more deterioration in a couple of those short-term indicators. The 20-day Commodity Channel (CCI) is dropping back below 100. The daily MACD lines are still positive, but just barely. The late April high by the S&P wasn't confirmed by the MACD lines which remained well below their March highs. Another short-term MACD sell signal could be imminent. [The next chart shows the weekly MACD lines even closer to turning negative]. A more convincing sell signal, however, would be a close below the mid-April low at 1280. That happens to coincide with the (green) 100-day moving average that I wrote about last Friday in my article on Bollinger Bands that has acted as an important support line.

Chart 4

WEEKLY MACD NEARS SELL TERRITORY... The weekly S&P bars put the current trend in better perspective. Of primary concern is the fact that the weekly MACD histogram has already slipped slightly into negative territory. It wouldn't take much to give a weekly sell signal (which would be a Friday close below the zero line). The green line is the 20-week moving average (which is the 100-day average shown in the previous chart). A break of that line would be an intermediate-term sell signal. The three blue lines are Bollinger Bands using an 80-week span. I got that by converting the 20-month lines into weekly values. The upper band has provided resistance over the market since the start of 2005. The market has pulled back each previous time that it's touched that resistance line. The fact that the current touch has been accompanied by so many negative divergences is an added caution sign. If the 20-week average (green line) and the mid-April low at 1280 are broken, a fall toward the 80-week average would become a possibility. From current levels, that would be a correction of approximately 6%. That would result in a test of its 80-week moving average which acted as support during the 2005 pullbacks. Seasonal trends are turning less friendly between now and October. If nervous investors start selling in May (as the Wall Street axiom advises), and if the market starts to break underlying support levels, the next few months could be difficult ones.

Chart 5