I am the editor of the Stock Trader's Almanac & Almanac Investor Newsletter and a Research Consultant at Probabilities Fund Management, LLC. I use historical patterns and market seasonality in conjunction with fundamental and technical analysis...+ FOLLOW THIS TUMBLR
Although the first trading day of September has been up in 13 of the last 21 years, this trend appears to be fading as the S&P 500 has been down six of the last eight first trading days. This trend is reflected in the DJIA while NASDAQ has been up in four of the last eight first trading days. Last year’s nearly 3% across the board shelling was the worst first trading day of September since 2002. This year the first trading day falls on Thursday before the three-day Labor Day holiday weekend and could be influenced by lower than average volume due to the holiday.
For those of you following along in the Stock Trader’s Almanac 2016
you have likely already noticed a Bull symbol on Monday August 29 and a Bear on
Tuesday August 30. For those without an Almanac bullish and bearish days are
based upon the percentage of time the market’s performance has been up on a
specific trading day over the most recent 21-year time period. In order to be
bullish the S&P 500 must have been up more than 60% of the time. If the
S&P 500 fails to rise at least 40% of the time the day is considered
Next week, Monday is clearly a bullish day. Over the last 21 years S&P 500 has been up 71.4% of the time, NASDAQ as well and DJIA 66.7%. S&P 500 has been up thirteen straight times on the third to last day of August with three advances exceeding 2%.
But on the penultimate trading day of August, the track record is nearly exactly the opposite. S&P 500 has advanced just 23.8% of the time and its average loss is 0.48%. DJIA is only slightly better, up 28.6% of the time with an average loss of 0.43%. NASDAQ has been mixed.
Since 1950, September is the worst performing month of the year for DJIA, S&P 500, NASDAQ (since 1971), Russell 1000, and Russell 2000 (since 1979). September was creamed four years straight from 1999-2002 after four solid years from 1995-1998 during the dot.com bubble madness. Bullish election-year forces do little to improve on September’s poor overall performance over the same timeframe. September’s performance does improve slightly in election years, but it is still negative nearly across the board. Only the Russell 1000 and Russell 2000 have been able to escape negative territory and post modest 0.2% and 0.7% average gains respectively in the last nine election year Septembers.
Although the month has opened strong 13 of the last 21 years, once tans begin to fade and the new school year begins, fund managers tend to clean house as the end of the third quarter approaches, causing some nasty selloffs near month-end over the years. Recent substantial declines occurred following the terrorist attacks in 2001 (Dow: -11.1%) and the collapse of Lehman Brothers in 2008 (Dow: -6.0%). Solid September gains in 2010; DJIA’s 7.7%, S&P 500’s 8.8% were the best since 1939, but the month suffered nearly the same magnitude declines in 2011, confirming that September can be a volatile month.
Not a subscriber? Sign up today for a Free 7-Day Trial to Almanac Investor to get a full run down of seasonal tendencies that occur throughout each month of the year in an easy-to-read calendar graphic with important economic release dates highlighted, Daily Market Probability Index bullish and bearish days, market trends around options expiration and holidays. In addition, the Monthly Vital Statistics Table combines stats for the Dow, S&P 500, NASDAQ, Russell 1000 and Russell 2000 and puts them all in a single location available at the click of a mouse.
At the close today, S&P 500 has now gone 33 trading days without exceeding +/– 1.0% on a closing basis. Going back to 1930, a streak of this duration has occurred 56 times including the current run. The longest streak was 166 trading days from March 1, 1963 until October 24, 1963. No such similar streak occurred prior to 1944. With the exception of seven past streaks, S&P 500 was quietly climbing during the extended period of low volatility and the average gain was 3.78%.
However, in the month after the streak ended S&P 500 continued to advance only 50.9% of the time with an average loss of 0.14%. At 3-, 6- and 12-months after the low volatility streak ended S&P 500 was generally back on a positive trajectory with average gains steadily improving as time passed.
Earlier today, the U.S. Census Bureau released figures for sales of new single-family houses in July 2016. The headline figure was staggering at a seasonally adjusted annual rate of 654,000. This was a jump of 31.3% compared to July 2015 and 12.4% above June’s revised number of 582,000. These are not hard numbers, rather they are estimates based upon surveys. These eye-popping numbers should not be taken as cold hard fact and are likely to be revised (many times over). Nonetheless, they are positive for homebuilders and for U.S. economic growth at least in the short-term. Another possible explanation for the surge could be the constant teasing by the Fed to raise rates. Perhaps some buyers finally jumped off the fence in July out of fear of higher mortgage rates later on.
In its nearly 46-year history, NASDAQ has recorded just 21 previous 8-week consecutive winning streaks. The current streak which began with the week ending July 1 is still ongoing and is the 22nd such streak. The average NASDAQ gain during the previous streaks of 8 weeks or longer was 20.04% or 2.11% per week during each streak. NASDAQ’s longest streak was 15 weeks straight from December 3, 1971 until March 10, 1972. 13 of 21 (61.9%) past streaks extended to 9 or more weeks.
Two pieces of economic data are currently suggesting the Fed will not be raising rates anytime soon. In the following two charts the 6-Month Exponential Moving Average (EMA) of the 1-Year Change in the Consumer Price Index (CPI) and the Producer Price Index (PPI) have been graphed. Both CPI and PPI are not seasonally adjusted. This combination of unadjusted data, smoothed with a simple EMA produces solid trend lines.
Looking at the long-term history of CPI and PPI there is a clear trend toward price stability. Prior to 1950, there were wide swings in both that ranged from nearly 40% to negative 30% in less than 5 years. Since then the range has narrowed substantially. Zooming in on the time from 1982 to present the current trend of CPI and PPI is visible. PPI’s leadership of CPI is also visible. It is also clear that PPI and CPI peaked in 2011 and have been trending lower since. Last year’s sharp drop in crude oil and other commodity prices sent PPI spiking below negative 5%. CPI also dipped near zero. Although there has been a modest recovery to both in 2016, the overall trend is still lower. Given the Fed’s data dependent nature, it could be years before there is a stable trend toward someplace around 2% (solid black line).
With a little help from the release of the Fed’s minutes from its last meeting yesterday, the market was able to mount a rally that spilled over into today, producing modest gains on both days. The wishy-washy nature of the Fed has once again convinced enough people that any rate hike is most likely to be later, not sooner. Low rates for longer also put downward pressure on the U.S. dollar which in turn has given a boost to crude oil. Even with the ongoing support of the Fed, the market’s current rally appears to be losing some steam. A bit higher in August is still likely, but afterwards a modest pullback in September-October is not out of the question.
In the above three charts of DJIA, S&P 500 and NASDAQ signs of a slowing rally are growing. DJIA is sitting at about the same place it was a month ago in mid-July. S&P 500 is a handful of points higher and only NASDAQ has made any substantial move higher over the past four weeks. Stochastic and relative strength indicators are bouncing around just under overbought. MACD has been negative on DJIA and S&P 500 charts since the end of July while NASDAQ MACD has visibly turned negative over the past three days. All of this does not preclude additional gains, but it does suggest any gain is likely to be unimpressive.
Monitoring market performance on the individual days of the week has been revealing over the years. The insights were so inspiring that our iconoclastic founder and resident consigliere, Yale Hirsch, entitled his 1986 book Don’t Sell Stocks on Monday. In the following charts we have lined up the performance on each day of the week from pages 141 and 142 of the Stock Trader’s Almanac with annual average daily volume deviations for each day.
Not surprising, Monday, or the first trading day of the week shows substantially reduced trading volume on both the NYSE and NASDAQ of 9% below the average. Friday is also below average for NASDAQ. This underscores the recent trend of market gains being concentrated midweek and weakness at the beginning and end of weeks.
Apparently, traders and investors prefer long weekends; or at least not being exposed going into the weekend and being more tentative about taking new positions upon their return. Picking up stocks on Monday weakness and unloading during midweek strength on higher volume would appear to be a prudent strategy for the most part. It also pays to be keen to price and volume action on Fridays and the following Monday for indications of future market direction. Strong volume and price advances tend to be bullish, while back-to-back weakness on normal or elevated volume is frequently bearish.