I am the editor of the Stock Trader's Almanac & Almanac Investor Newsletter and Chief Market Strategist at Probabilities Fund Management, LLC. I use historical patterns and market seasonality in conjunction with fundamental and technical analysis...+ FOLLOW THIS TUMBLR
Since 2008, DJIA, S&P 500, NASDAQ and Russell 2000 have not fared all that well during the last five trading days of the third quarter, the end of September. 2010 was the last year where the major indices produced solid, across-the-board gains. S&P 500 and Russell 2000 have been the worst since 2008, down 7 of the last 9 years during the last five days of Q3. Since 2008, Russell 2000 has lost an average 1.1% during the last five days.
Post-election year October’s are neither great nor bad since 1953, ranking mid-pack across DJIA, S&P 500, NASDAQ and Russell 1000 with gains averaging from 0.7% (DJIA) to 1.2% (NASDAQ). DJIA has the best historical odds for gains having advanced in 11 of the last 16 post-election year Octobers. Despite the best average gain, NASDAQ actually has the worst record, declining in 6 of the last 11 post-election year Octobers. A 12.8% gain in 2001 boosts its average. Should a meaningful decline materialize in October it is likely to be an excellent buying opportunity, especially for any depressed technology and small-cap shares.
In mid-August we
noted that small-cap
stocks tended to outperform large-cap stocks from around the 18th
trading day of August until the 12th trading day of September. This
certainly was the situation this year as the Russell 2000 index surged 4.8%
from August 24 through yesterday’s close compared to a 2.8% gain by the Russell
1000 during the same time period. The Russell 2000 is still outperforming
today, but there are signs that this trend could be coming to an end.
The first sign is the historical seasonal pattern highlighted in the above chart is on the verge of heading lower once again indicating that small-cap underperformance will return. Other headwinds to the continuation of the small-cap rally are Stochastic, relative strength and MACD indicators that are all in overbought territory (in the following chart). Russell 2000 is also sitting right around projected monthly resistance (red-dashed line). Stretched technical indicators at the end of a favorable seasonal period suggest that the bulk of the Russell 2000’s latest advance is most likely done.
Some of you may remember the old saying on the Street, “Buy Rosh Hashanah, Sell Yom Kippur.” Though it had a good record at one time, it stopped working in the middle of the last century. It still gets tossed around every autumn when the “high holidays” are on the minds of traders as many of their Jewish colleagues take off to observe the Jewish New Year and Day of Atonement.
The basis for the new pattern is that with many traders and investors busy with religious observance and family, positions are closed out and volume fades creating a buying vacuum. Holiday seasonality around official market holidays is something we pay close attention to (page 88 Stock Trader’s Almanac). Actual stats on the most observed Hebrew holidays have been compiled in the table here.
We present the data back to 1971 and when the holiday falls on a weekend the prior market close is used. It’s no coincidence that Rosh Hashanah and Yom Kippur fall in September and/or October, two dangerous and sometimes opportune months. We then took it a step further and calculated the return from Yom Kippur to Passover, which conveniently occurs in March or April, right near the end of our “Best Six Months” Tactical Switching strategy.
Perhaps it’s Talmudic wisdom but, selling stocks before the eight-day span of the high holidays has avoided many declines, especially during uncertain times. While being long Yom Kippur to Passover has produced more than twice as many advances, averaging gains of 7.2%. It often pays to be a contrarian when old bromides are tossed around, buying instead of selling Yom Kippur – and selling Passover.
Although not an
official market holiday, Rosh Hashanah is observed by many New York area
schools and many Jewish colleagues will also spend time observing the holiday
with family and friends. Their absence can dampen trading volumes as positions
are squared ahead of the holiday. This year Rosh Hashanah begins at sunset on
Wednesday, September 20. This also happens to fall in the middle of one of the worst
weeks of the year, the week after September options expiration.
Over the last 21 years the day before Rosh Hashanah has been up more than 50% of the time, but there have been some large declines that result in average losses across the board while Rosh Hashanah (or the next trading day) has recorded average gains. More recently DJIA, S&P 500 and NASDAQ have all been up 7 times in the last 8 years on the day before (and five straight) but those gains have been short-lived as the next day has been weaker, down 4 of the last five years.
Next week is the week after September options expiration week and it has a dreadful history of declines particularly since 1990. This week has been a nearly constant source of pain with only a few meaningful exceptions over the past 27 years (shaded in grey). Substantial and across the board gains have occurred just four times: 1998, 2002, 2010 and 2016 while many more weeks were hit with sizable losses.
Since 1990, average weekly losses are even worse; DJIA –1.07%, S&P 500 –1.00%, NASDAQ –0.98% and a stout –1.50% for Russell 2000. End-of-Q3 portfolio restructuring is the most likely explanation for this trend as managers trim summer losers and reposition portfolios for the upcoming fourth quarter.
Let’s step out on another limb today and just go ahead and declare the Los Angeles Dodgers this year’s World Series Champs. Never mind completing the balance of regular season games or playoffs, they have the best record in all of MLB today. Congratulations LA Dodgers, 2017 World Champs!!! This is exactly what is happening today with the market. New all-time highs in September (and August) apparently automatically mean that “Selling in May” failed this year and the market can only go one direction from here. Let’s not forget its September 14, 2017 and the “Worst Four/Six Months” don’t actually end until November 1. That’s over six weeks from now and those six weeks are riddled with some nasty sell offs throughout recent history.
In the above chart, we can see DJIA, S&P 500, NASDAQ and Russell 2000 rallying off their respective mid-August lows in the top pane and solid support by Advance/Decline lines in the lower panes. But, note how quickly things have changed in the past. A late-July peak in A/D lines and subsequent pullback was accompanied by a similar move in the major indexes. And from their respective May 2017 closes through yesterday’s close DJIA is up 5.5%, S&P 500 3.6%, NASDAQ 4.2% and Russell 2000 is up 4.1%. It would only take a few 1 percent daily losses to wipe out those gains. The next six plus weeks have a history of that and often much more.
First off, next week is the week after September options expiration week and it has a dreadful history of declines particularly since 1990. S&P 500 has been down 22 times in the last 27 years with an average weekly loss of 1%. Then even before the trading week concludes, the eight-day span between Rosh Hashanah and Yom Kippur commences. DJIA has averaged a 0.6% loss during the period going back to 1971.
And finally, October has a frightful history of market crashes such as in 1929, 1987, the 554-point drop on October 27, 1997, back-to-back massacres in 1978 and 1979, Friday the 13th in 1989 and the 733-point drop on October 15, 2008. During the week ending October 10, 2008, Dow lost 1,874.19 points (18.2%), the worst weekly decline in our database going back to 1901, in point and percentage terms. Absent a bear market in progress, October’s historical tendency for bear market bottoms is a moot point. October’s record in years ending in seven is of concern. October 2007 was a bull market high, in October 1997 DJIA plunged 12.4% from October 7 to October 27 and during the crash of October 1987 DJIA tanked 23% in two days.
Keeping all of this mind and recognizing that more than six weeks remain before the “Best Six/Eight Months” of the year actually begin on November 1, it is far too early to sound the “all clear.” We will continue to maintain a defensive posture in the Almanac Investor Portfolios. We will also look forward to enjoying the remainder of the baseball season and playoffs.
Last month we presented updated 1-Year Seasonal Pattern charts for DJIA, S&P 500 and NASDAQ with historical patterns beginning at the start of the second half of 2017. At that time, volatility was on the rise and the market was slipping modestly lower and since then the market recovered and DJIA, S&P 500 and NASDAQ all closed at new all-time highs again today. At today’s close DJIA, S&P 500 and NASDAQ are all performing modestly better than their respective best case scenario seasonal patterns for this point in the second half of the year. The prospects for a tepid second half of September still remain and even if September does buck historical patterns this year with solid gains, October could still present an issue as two of four (the most optimistic) seasonal patterns tracked in the charts above and below, turn negative.
Folks have been questioning recently the validity and efficacy of seasonal market patterns as the equity markets have shown resiliency over the worst six months and the major indices log new highs, except for the Russell 2000. I have been hearing a good bit of frustration with our worst six months seasonality and out rather cautious stance. But it is at times like this when traders and investors throw in the towel on these evidence-based patterns that they often come home to roost.
We have cited recently many of the fundamental, technical, psychological and geopolitical risks and reasons the market is ripe for a correction in addition to the seasonals. Our Best Six and Eight Month Switching Strategies have served us and our clients and subscribers well over the long and short terms.
So while the market has held up pretty well since “Sell in May” it has not done much since our June 9 NASDAQ MACD Sell Signal when we went more fully risk off. At that time we held our big winners, sold underperformers, tightened up stops, limited new longs and implemented some defensive positions. After logging double-digit gains from our October 24 MACD Buy Signal, since June 9 DJIA is up 4.0%, S&P 500 2.7%, NASDAQ 4.0% and the Russell 2000 0.1%.
In addition, September may look extremely robust on first blush, but look at the update chart below of a typical September with 2017 so far overlaid. It still looks like we are on track for a late month selloff, or at least set up for some month-end softness.
September’s option expiration week is up 54.3% of the time for DJIA and slightly better for S&P 500 and NASDAQ since 1982, but is has suffered several sizable losses. The worst loss followed the September 11 terrorist attacks in 2001. In the last fourteen years, NASDAQ has the best record during September’s option expiration week, up twelve times. Triple-Witching Friday had been firm with all three indices advancing every year 2004 to 2011, but S&P 500 has been down five straight since and DJIA and NASDAQ four of five.