New Highs and A New Three Peaks & Domed House Pattern
By: Jeffrey Hirsch & Christopher Mistal
August 11, 2016
Earlier this week updated Eighth Year of Presidential terms seasonal pattern charts for DJIA, S&P 500 and NASDAQ were presented. At that time it was noted that the market had most likely shrugged of the 8th year curse and were once again tracking the typical election year course which includes strength in July and August before some weakness in September and October.
In response to these charts inquiries were made regarding the impact that 2008 has on the election year pattern. This is quite reasonable given the magnitude of market declines in that year. DJIA shed 33.8% in 2008, S&P 500 plummeted 38.5% and NASDAQ plunged 40.5%. Losses were steep and rather consistent from start to finish in 2008 so it would seem that year could easily pull the overall election year performance lower.
[DJIA 8th Year Chart]
[S&P 500 8th Year Chart]
[NASDAQ 8th Year Chart]
In the charts above, a fourth line has been added representing all election years excluding 2008. With 2008 removed there is virtually no change in the shape of the pattern and major rallies and pullbacks on the pattern remain intact. For DJIA and S&P 500, there was only an improvement of about 2% near the end. NASDAQ improved around 5%. The simple reason why there is just a modest improvement excluding 2008 is the number of election years since 1900. Our database has 28 election years since 1901 for DJIA, 21 since 1930 for S&P 500 and 11 since 1971 for NASDAQ. Even excluding 2008, the typical election year pattern still exhibits weakness in August and September, just a bit more on the mild side.
What Goes Up Must Come Down
As the market continues its push higher off the February and Brexit lows, we continue to evaluate when the next pullback may occur and how deep the next correction might be. As noted in the August Almanac and the election year chart above, August is notoriously much better in election years, while September/October corrections are still prevalent.
The new Three Peaks and a Domed House Top Pattern (3PDH) that may be developing can provide some guidance. The market’s recent breakout lines up quite well with the current Three Peaks lasting just 8 months from Peak 1 (point 3) in February/March 2015 to Peak 3 (point 7) in October/November 2015, falling right in the 8-10 time span of Lindsay’s Basic Model.
The Domed House top timeframe in the Basic Model is about 7 months from the end of the Separating Decline at point 14 to the top at point 23. From the Separating Decline low in February 2016 that would put the top some around or after September. But we have seen this phase of the pattern transpire in much shorter and longer time frames.
[Current 3PDH]
[Basic Model]
Whenever that top occurs, the pattern resolves at point 28, somewhere back around points 10 and 14 at a minimum. It can go substantially lower before a new high, which could culminate in a bear market. If the top occurs soon, I would expect another correction and not the 20%+ type of bear market, which from current levels is well below the February low. If this 3PDH top does not occur until September or later, perhaps around the election, then this 3PDH is more likely to resolve in a bear.
In some instances Point 27 has been higher than Point 23, so we could alternatively see a correction to Point 26 in September/October back around the May/June lows in 5-10% correction before the market rallies into the election and yearend. Furthermore the Point 23 3PDH top may not be transpiring at the moment and will occur later this year or early next, culminating in that aforementioned bear market.
It is also possible that this pattern fails to materialize and other analysis and indicators will be more instructive. But for now, with the market richly valued and sentiment high, caution is in order as a pullback of some sort is on the near-term horizon.