I reviewed the Dow charts back to year 1900 looking for the following criteria:
1. a drop of 25%+ from top to bottom
2. an initial drop of 8%+ in 6 months or less within that 25%+ drop
3. the largest % retracement greater than a 20% retracement ending the initial downtrend of 8%+ as obvious on a chart before making a lower low
Admittedly the last criteria is a bit subjective since I was looking for an obvious significant short-term bottom like we had August 24th.
Here are the results.
# of cases = 22 (1901,1903,1906,1912,1916,1919,1929,1930,1937,1938,1940,1966,1968,1973,1976,1981,1987,1998,2000,2002,2007)
Average initial drop = 15.3%
Median initial drop = 12.9%
Average retracement = 56.9%
Median retracement = 52.7%
% of cases retracing >61.8% = 32%
% of cases with initial drops > 20% = 13.6%
Here is the current Dow drop & retracement
18351-->15370 = 16.2% drop vs average 15.3%
15370-->16933 = 52.4% retracement vs median 52.7%/average 56.9%
Obviously, the current situation fits very well with the average and median cases since year 1900, but what happened next?
You need to remember that all of my cases were hand-picked to assume a lower low after a multi-month 8%+ drop and a significant retracement. The initial drop was usually 10-20% with an ultimate drop of 25%+ over a few months or years. So.........there is a chance that this research does not apply if Dow does not make lower lows ultimately heading to a 25%+ correction. But, so far, the initial drop and initial retracement fall almost perfectly in line with the history of 25%+ drops and, assuming the Dow does ultimately drop 25%+ as I have proposed, it is good to have an idea what behavior to expect along the way, right?
Only 4 of the 22 cases above rallied back near or above the previous rally top after retesting the low including year 2000. A few others retraced a little more than 50% on the 2nd rally like year 2007. The vast majority of cases made significantly lower lows before rallying back to a far lower level than the initial rally retracement. Perhaps we should be looking at the most recent examples in 2002 and 2007 since I have been claiming we are still in the same secular bear market. If so, you should keep in mind that year 2000 bounced around the initial highs and lows for over a year after a preceding multi-year sharp rally. I could claim we already had a multi-year sharp rally followed by a similar 9-month consolidation period like that in 2015 so another 9-12 month consolidation would be virtually unprecedented. And, in year 2007, Dow initially fell 10% and retraced 72% but then fell 17.4% below its top and then fell 5% short of its initial rally level. That example would indicate much further pain for the Dow before another large rally. And, if you buy into my exponential parabolic theory posted weeks ago, then you might notice it fits this research as well in that the initial 2000 drop was met with a larger initial retracement and a larger second rally than year 2007. Will the parabolic trend continue meaning smaller, faster retracements? So far, Dow retraced 84% in 2000, 72% in 2007 and 52% in 2015.
What does that indicate for the next few months? There is still a chance Dow could retrace higher before making a lower low, but that prospect is looking dim as key levels are being broken this week. Assuming Dow makes a lower low within the next few days or weeks, the odds based on history and even parabolic recency suggest that it will be a much lower low. I have been arguing for a slightly lower low near SPX 1828 or the equivalent of Dow 15300-15400ish, but that was based partly on the candle analogy which I expelled last week after the uptrend lasted longer than the other 9 candle examples since year 2007. It was also based on the annual reversal level of Dow 15341 which I felt would be significant support, but the August 24th low already got close to it perhaps satisfying that support and now enough time has passed for the market to potentially build enough momentum to break below that level. So, my primary reasons for expecting a slightly lower low have disappeared.
There are some new reasons to argue for a slightly lower low including likely positive divergences, overly bearish sentiment and certain extreme technical indicators, but we all know extremes can get more extreme especially in this day and age of massive leverage, high-frequency electronic trading, globalism, unprecedented Fed intervention and Pavlovian buy-the-dip mentality being reversed.
So, I think the historical evidence above, despite bearish sentiment, indicates the Dow will make much lower lows in the coming weeks. Since I focus on studying SPX, I will use that for price targets. In previous weeks, I gave many reasons including the bull market uptrend line, Fib retracements and an Andrew's pitchfork for a target of SPX 1700ish in October. Another important reason for SPX 1700ish is that it equates to Dow 14000ish which is the year 2007 top. That would actually fall just shy of a 25% drop from the top. And, you might also remember my study months ago of charts since 1896 indicating that Dow/SPX would break (or at least significantly retrace) their year 2007 highs based on historical behavior of markets that zoom more than 10% beyond all-time highs without a 62%+ retracement mixed in. That makes common sense since markets need to cleanse periodically. Depending on how you calculate Fibs and draw the trend lines and depending on how Dow performs in relation to SPX, you can end up with a range of SPX 1675-1750. If SPX 1675-1750 is the October downtrend target, I will just use SPX 1700 for calculation purposes.
If we assume SPX will drop near 1700 in October, what should happen next? Well, I recently posted charts with my projections using the exponential parabolic theory as my guideline. I called for a drop to the 1800s in Aug/Sep, then a drop near 1700 in Oct, then a rally near 1900 between late October and late November when my discretionary spending indicator projects a significant top. That projection actually fits the historical analysis above in that the odds greatly favor a much lower low followed by a rally that will be fortunate to retrace 50-70% of the previous high. A 50-70% retracement of SPX 2020-->1700 would be 1860-1925, so I think a projection of 1900 +/- 2% is historically conservative depending on the actual price low. Then, my projection calls for bouncing around the SPX 1700-1900 level into year-end perhaps triangulating to 1900, 1775, 1875, 1800 and 1850 before collapsing in Q1 to 1100ish and perhaps lower.
In summary, the history of 25%+ bear market drops aligns well with Dow behavior in 2015 and, if that alignment continues, there is a strong chance that Dow will revisit its year 2007 high in the coming weeks or months. The equivalent for SPX is 1700ish. History further tells us that the odds favor that low will also be broken significantly with recent highs (2020 in SPX) not being approached for a long time. The historical research used allows for Dow to bounce above last week's high first (however unlikely) and it makes assumptions, but I'll let you be the judge if it is further reasonable evidence to suggest severe damage dead ahead. Personally, even though Dow theory, T-Theory and cycles as studied by Tim Wood and Peter Eliades indicate a severe bear market lies ahead, I'd like to see Dow 15341 broken to be more confident this bear market will likely fall in the 45%+ drop category based on my annual reversal research, exponential parabolic theory and the research above. Tony Caldaro and his Objective Elliott Waves are also expecting a 50%-type bear market albeit after 1 more high unless a long-term downtrend is confirmed with a much lower low. By the way, if SPX is on its way from 2020 to 1700ish, it would make sense for the half-way point at 1867 +/- 1% to serve as initial support and perhaps suck in the bulls waiting for a retest of the lows before going back to 2100+. Good fortune.
No comments:
Post a Comment