Well the stock market just seems to keep chugging along. Despite the shaky economy, the national debt, the political discord, the terrorists on the march and the price of – well, just about everything made in China. And as a dutiful trend-follower I continue to “ride the ride” all the while repeating the ever apt phrase, “What, me worry?”
Three things for the record:
1) The stock market is in a clearly established uptrend and who am I to say otherwise?
2) Personally I still expect a serious “something” between now and the end of September
3) Arguably the most important bullish seasonal trend of all kicks in at the close on September 30, 2014.
Huh?
Yes folks, as of the close on September 30th it will be time for “the Mid Decade Rally”
Defining the Mid-Decade Rally
For the purposes of this article we will break each decade into two periods:
*Period 1 (i.e., the “Bullish 18”) = the 18 months extending from the end of September of Year “4” (2014, 2004, 1994, etc.)
*Period 2 (i.e., the “Other 102”) = the other 102 months of every decade.
So in a nutshell, Period 1 comprises 15% of each decade while Period 2 comprises 85% of each decade. Given that Period 2 is 5.67 times as long as Period 1 it would seem likely that an investor would make more money being in the stock market during Period 2 than during Period 1.
I mean if you had to choose – particularly given the long-term overall upward bias of the stock market – would you choose to be in the market 85% of the time or only 15% of the time. Safe to say most investors would answer the former. But if you are among those who chose this answer, perhaps you should read a little further.
Period 1: The “Bullish 18 Months” of the Decade
In Figure 1 you see the growth of $1,000 invested in the Dow Jones Industrials Average only during the middle 18 months (Sep. 30, Year 4 through Mar. 31, Year 6) of every decade starting in 1900. See if anything at all jumps out at you from the chart.
Figure 1 – Growth of $1,000 invested in Dow only during “Bullish 18” months of each decade; 1900-present
Does the word “consistency” come to mind? For the record, $1,000 invested only during this 18-month period grew to $40,948, or +3,994%. Now this all looks and sounds pretty good, but surely an investor would have made a lot more money investing in the stock market during the other 102 months of each decade, right? Well, not exactly. In fact, a more accurate statement might be, “not at all.”
Period 2: The “Other 102” Months of the Decade
Figure 2 displays the growth of $1,000 invested in the Dow Jones Industrials Average only during the other 102 months of each decade. $1,000 invested in the market 85% of the time – but excluding the mid-decade bullish period – would have grown to $6,166, or +517%.
Figure 2 – Growth of $1,000 invested in the Dow only during the “Other 102” months of each decade; 1900-present
Now for the record, +517% is +517% and no one is saying that you should simply mechanically sit out the 102 “other” months. But the point is simply that the “Bullish 18” made 7.73 times as much money – while invested only 15% of the time – as the “Other 102” – which was invested in stocks 85% of the time.
The difference is even clearer when the two lines are drawn on the same chart as shown in Figure 3.
Figure 3 – Growth of $1,000 during “Bullish 18” month (blue line) versus “Other 102” months (red line); 1900-present
Mid Decade “Bullish 18” Performance
| Period | % +(-) during “Bullish 18” |
| 9/30/1904-3/31/1906 | +68.3% |
| 9/30/1914-3/31/1906 | +30.6% |
| 9/30/1924-3/31/1906 | +36.2% |
| 9/30/1934-3/31/1906 | +68.8% |
| 9/30/1944-3/31/1906 | +36.1% |
| 9/30/1954-3/31/1906 | +42.0% |
| 9/30/1964-3/31/1906 | +5.6% |
| 9/30/1974-3/31/1906 | +64.4% |
| 9/30/1984-3/31/1906 | +50.7% |
| 9/30/1994-3/31/1906 | +45.4% |
| 9/30/2004-3/31/1906 | +10.2% |
Figure 3 – “Bullish 18” month performance
For the record:
*The average % gain for the Bullish 18 was +41.7%
*The median % gain for the Bullish 18 was +42.5%
On the other hand:
*The average % gain for the Other 102 was +34.3%
*The median % gain for the Bullish 18 was -2.2%
So you see why I am marking September 30, 2014 on my calendar.
By the way, if you like this one, in the immortal words of Jimmy Durante, “I got a million of ’em.”
Jay Kaeppel
Please show the buy-and-hold return in the chart with blue and red (18 vs 102) so we see how 18 trumps total returns. Thank you!
Jon, for the record just holding during the “bullish 18” does NOT outperform buy-and-hold. As shown in Figure 2 the market has made money during the “other 102” months. The real point is the magnitude and consistency of the gains during the “bullish 18” and the importance of giving the bullish case every benefit of the doubt during that 18 month period. Jay
You’ve got a million of ’em? Bring ’em on. We’re ready!
Dear Sir,
How can i play this using options?
Is there like a 18month option i could buy for the dow jones?
Lets say i wanted to invest 10K into an option, then what would this have been worth like last decade if i played it?
I am not sure that there is a really good option play for this. Essentially on Sep 30, 2014 you would need to buy a January 2017 LEAP (in order to be able to hold the position through March of 2016). There could be a lot of time premium involved and as a result if the Dow only advance 10-15% during the 18 bullish months, there might not be much profit on the option. Another alternative would be to buy deep-in-the-money calls 3 to 6 months out and to “roll out” hen they expire. Here too though there will be some time decay as well as commissions and slippage. But it is something to think more about between now and Sep. 30. Jay
Jay,
Love the site. Two quick questions: figure 3 shows each 18 month mid-decade period to 3/13/06, which I don’t understand, and “for the record” says that the bullish 18 average return was 41%, and the 102 period was 34%, which doesn’t seem to represent 7.7% outperformance over ten years.
I am not a trained statistician, so take it easy on me.
Thanks,
Rob
Rob, thanks for the comments. The blue line in Figures 1 and 3 simply show the cumulative growth assuming you bought and held the Dow only during the “Bullish 18 months” of each decade and stuck the money in a mattress the other 102 months (hence the “flat lines” between the “squiggles”). If you look at each 18 month period decade by decade, the average gain for the Dow within those “bullish” 18 months was +41%. If you look only at all the other months within each decade, the average total gain per decade registered by the Dow within the “other 102” months was (only) +34%. Hope that makes sense. Jay
On your spread sheet, I think you copy and pasted the years and did not ratchet up the second series of years.
9/30/1904-3/31/1906 +68.3% (ok here)
9/30/1914-3/31/1906 +30.6% (I think you mean 9/30/1914-3/31/1916)
9/30/1924-3/31/1906 +36.2% (I think you mean 9/30/1924-3/31/1926)
9/30/1934-3/31/1906 +68.8% (I think you mean 9/30/1934-3/31/1936)
9/30/1944-3/31/1906 +36.1% (I think you mean 9/30/1944-3/31/1946)
9/30/1954-3/31/1906 +42.0% (etc.)
9/30/1964-3/31/1906 +5.6% (etc.)
9/30/1974-3/31/1906 +64.4% (etc.)
9/30/1984-3/31/1906 +50.7% (etc.)
9/30/1994-3/31/1906 +45.4% (etc.)
9/30/2004-3/31/1906 +10.2% (etc.)
I am not sure the following wording is correct:
On the other hand:
*The average % gain for the Other 102 was +34.3%
*The median % gain for the Bullish 18 was -2.2% (Do you mean “The median % gain for the other 102 was -2.2% ???)
Thanks, I enjoyed the article. Another possible twist is when it is the 5th year of the decade and the 3rd year of the presidential cycle. If my math is correct, this would be every other decade or every 20 years.
Oops. Um, the editor here at JayOnTheMarkets sometimes misses stuff…….Nice catch though. Thanks, Jay
Hey Jay (kind of has a ring to it, don’t you think?),
A subscriber to my NewsLetter at TheDowTheory.com forwarded this piece to me recently thinking I would find it of interest. Indeed I did. My question for you is: Where did this come from originally, did you ‘discover’ the phenomena or should someone else get credit? I’d appreciate a little background if you know it. Thanks,
Best Regards,
Jack Schannep
Jack, I didn’t “invent” but I don’t remember exactly where I got the idea. In 1982 I heard that every 20 years from end of June Year 2 through end of December Year 5 was “bullish”. I though that was stupid. But June 1982 through Dec 1985 the Dow doubled. Year 5 was also widely heralded as bullish (probably learned that from Hirsch of Fosback). Somewhere along the way “extended” that middle portfolio of the decade 3 months in either direction and voila. Of course, Sep 30, 2014 was not such a great time to “get bullish” – figures. Jay
I would never trade this because it seems so random. You, yourself, have commented on this multiple times over the years. Problem is, a lot of your Seasonality observations seem to fit this mold. Indeed, I think you could come up with 1,000,001 of these, have hundreds of them fail miserably every year, and you wouldn’t have to worry about mentioning them because there would still be hundreds of thousands leftover for you to choose from and write about. What’s your take on the extent to which this is data snooping?
Mark, Thanks for the comment and you raise some excellent questions. Here is how I look at seasonal trends. I do think it is fair to say that seasonal trends are entirely the result of data snooping. There is no other way to find something like that. As a result, many traders are uncomfortable relying on these trends for actual trading. And that is completely understandable. The two counter arguments in favor of consider seasonal are, 1) 90%+ of trader rely on fundamentals and/or technicals, I would guess well less than 10% rely on seasonal. So if one is looking for an “edge”, one can argue that it makes sense to look where the vast majority of traders are not looking (i.e., seasonal trends). 2) The key to any seasonal trend is consistency and an understanding that even the most trend will fail from time (although the same can be said of any fundamental for technical tool as well). Seasonal trends are not meant to be thought of as some secret formula or the “hidden order in all markets”, etc. Bottom line, if seasonal trends make you uncomfortable as a trading tool then avoid them. If something looks compelling, 1) fell free to incorporate it in your analysis but, 2) don’t bet the ranch, 2) use a stop-loss, and 3) consider using some sort of technical confirmation to confirm the trend. Hope that helps. Jay