Market Analysis Posted To Market Mavins Web Site In Early January, 1999

1998 was certainly one of extremes as far as the stock market is concerned. The S&P 500 had a rally into July of nearly 23 percent. That was  followed by a decline of just over 22 percent  into the October 8 low which in turn was followed by rally of nearly 24 percent culminating in a net gain of  26.7 for the year as a whole. 1999 looks to have the potential for even bigger moves. 

The Elliott wave picture shows the October 8, 1998 low to be either all of a large fourth wave from the 1987 low, or just the A wave of a larger pattern. Currently, I am approaching the market with the idea that October 8 did complete wave 4 from the 1987 low with the post October 8 advance wave 5.

Under the Wave Principle only 1 wave of a five wave sequence can extend. If the third wave extends than waves 1 and 5 are usually related by a Fibonacci ratio or multiple. The rally from October of 1990 to the July 1998 high is best counted as a five wave structure all into itself. This rally is also best counted as a third wave from the October or December 1987 low. This wave (post 1987) is also a bigger third wave from August 1982.  

The 1990 to 1998 rally certainly fits the description of an extended wave as it stopped literally 6 points away from being a perfect 2.618 the length of wave 1 October 1987- July 1990) on logarithmic scale. Log scale measures movements and relationships between movements in percentage terms as opposed to just price. This is important and given the extent of the move in terms of both points gained and duration of the moves gives a better overall picture. The internal relationship between the sub waves from 1990 to 1998 also show a very neat Fibonacci relationship and in fact allowed us to forecast a peak last summer in the 1190 to 1210 area basis the S&P 500 (the S&P peaked on July 20 at 1190.58).

Since we have already determined that the October 1990-July 1998 rally is best counted as an extended third wave from 1987, wave 5 from October 8 1998 should then have some sort of Fibonacci relationship to wave 1. There are four separate Fibonacci relationships pointing to a fifth wave target for the S&P 500 in the 1525-1577 area. The post October 8 rally in the S&P would be 4.236 times the 1987-1990 rally (wave 1) at 1550 on a simple price scale. The rally would equal the 1987-1990 rally (wave 1) on log scale at 1577. Another measuring technique for fifth waves is that they travel about .618 the distance of the entire move of waves 1-3. The rally from October 8 would be .618 times the 1987-1998 rally at 1525.  Lastly, the rally from 1987 (which I am counting as wave 3 from 1982) would be 1.618 the length of the rally from 1982 to 1987 on log scale at 1550.

Thus we see the potential for a lot more upside to complete a major up leg in the still on going secular bull market. As discussed, the alternate count has the low of October 8 labeled as only the A wave of a larger pattern. While our preferred interpretation supports the bullish count we cannot as yet eliminate the possibility that the post October 8 rally is a B wave from the July 1998 peak. A guideline we have used for irregular patterns is that the B wave should not exceed a 1.383 multiple of wave A. That said, the 1293-1303 level for the S&P 500 takes on an important function as that is where the rally from October 8 would be 1.383 the July 20 to October 8 decline. Furthermore, the rally from December 14, which on the weekly chart is clearly a whole new wave from October 8, would be .618 the length of the rally from October 8 to November 27 at 1303. Thus we have two separate Fibonacci relationships pointing to this area as being very important. Until price moves above this level the B wave count cannot be eliminated. On the downside, our major focus is on the 1170-1175 level as that represents the .618 retracement of the rally from the December 14 low and what can be counted as wave 1 of 3 from the October 1998 low. The next level, and one I deem critical is the December low of 1136. A move below that level would turn the monthly chart negative

What we are facing then is two possible scenarios for 1999. The first and more bullish count, at least initially, shows the potential for a move to the 1550 area on the S&P 500 once the 1303 area is exceeded.  This is the good news, and from current levels at the beginning of 1999 suggests a move of over 25%. The bad news, however, is that this rally is a fifth wave and a larger degree fifth wave which will also be culminating a five wave advance going back over 11 years. Therefore, once complete the stage will be set for the biggest decline on a percentage basis since the August-October 1987 wave 2 decline. If indeed we get to the 1550 level we would then be looking to correct a move of over 1300 points from 1987. A normal Fibonacci .383 retracement of that advance points to a decline of about 500 to 600 points and could see prices back to the 950-1000 area.

Under the wave principle the fourth wave of previous degree is a natural support level, and coincidentally the October 8 low, which under this count is the fourth wave of previous degree was 923. That is not too far away from a hypothetical .383 retracement of the entire structure from 1987. If the alternate count turns out to be correct then we will be seeing a move back to or through the October 8 low. That decline, if it should occur should then be followed by a move back to new highs to complete wave 5 from 1987 and wave 3 from 1982.

Under either of these counts 1999 is shaping up to be quite a year.