Friday, March 30, 2012

Tom McClellan: S&P 500 - First day of the month tendency

I did a study recently, looking at the behavior of the SP500 on the first trading day of a new calendar month.  It has long been said that pension fund flows and automatic mutual fund investments for 401Ks, etc. tend to put a positive bias on the last few days of the month and the first few days of the new month.

What I found is a strong positive bias in recent years for the first day of the month to be an up day.  The last day of the month and the second day of the month do not show that same effect.

The first chart shows an “equity curve” of the first day of the month versus the rest of the month.  It assumes you “buy” the SP500 at the close on the last day of the month, and then sell it at the close on the first day of the month, thereby capturing the price change on just that first day.  The Rest Of The Month line does the opposite.  Since this is just a study, and not a real trading system analysis, no allowance is made for commissions, slippage, or money market interest.  It makes a pretty strong argument for the first day having a decent positive tendency.

Interestingly, this phenomenon is a fairly recent development.  The second chart shows a longer view of the same idea, this time on log scale, and the First Day equity curve is pretty flat through the 70s and 80s.  It really only starts to slope upward around 1997, implying that the first day effect was not really prevalent before then.

When I included the last day of the month into the mix, and also the second day of the month, they did not make the results better and in fact they hurt the performance of the First Day equity curve.  A comparison of the 3 methods’ results is shown in the third chart. 

Does that mean Monday will be an up day?  Not necessarily, of course, but given the recent odds I would not want to bet against it.

Tom McClellan: 9-Month Cycle

Chart In Focus
July 22, 2011

The 9-month cycle in the stock market used to be a very regular and important factor governing stock price movements.  But recent changes in the rules and structures of the markets may have made this cycle go the way of Saturday trading and paper stock certificates.  Or perhaps it has just changed itself into a new form.  Let's take a look.

My lead chart this week highlights what I am talking about.  Before 2007, there were important bottoms about every 185 trading days.  Cycles analysts for years have called this the "9-month cycle", or the "40-week cycle", even though the precise period was a little bit shorter than those numbers.  Big round numbers are easier to say, which is why those names were used.
In addition to the major cycle lows every 185 trading days, there was also a significant mid-cycle low that would appear somewhere in between the major bottoms.  The mid-cycle low was usually not as punctual, and could arrive early or late, even as the major cycle low would tend to be more on time.  This mid-cycle low was a "harmonic" of the frequency of the major cycle low, meaning that they were even multiples of each other.  Harmonic frequencies are a big deal for mechanical engineers dealing with solid structures, but they also show up in other arenas like the stock market.

Starting in 2007, this all changed, as delineated by the red vertical line.  It was hard to understand this change as it was occurring at the time, but easier to see now that we have the luxury of looking back at the historical data.  What appears to have happened beginning in 2007 was that the length of this cycle contracted dramatically, for both the major cycle and the mid-cycle periods.

One of the reasons why it was so difficult to understand this change in period as it was occurring in real time is because of another trait of this cycle, which is known as a "phase shift".  In my historical research, I have identified the 9-month cycle as working on the stock market all the way back into the 1960s, although curiously not so much before then.  One of the more interesting behaviors of this cycle over that time period is that about every 6-8 years, the 9-month cycle would seem to skip a beat, and then start up again on some new schedule.  Here is a great example of this behavior:

Flashback to 2002-06 and the 9-month cycle

In the lower portion of this chart, there is a modified sine wave pattern to help visualize the behavior of the cycle in the SP500's price movements.  The market was following this cycle pattern very nicely up until late 2005, and then it jumped onto a new schedule that just happened to be about a half cycle length off of the original schedule.

So with the knowledge that a phase shift was a possibility with this cycle, it was hard to understand what was happening in early 2008.  And this illustrates one of the big pitfalls with doing any sort of cycle analysis: cycles can change, and so while they may give us nice predictions of what should happen at some point in the future, there is no guarantee that the past behavior will remain in effect in the future.

It just so happens that 2007 was when this cycle changed, and it was also the year that the uptick rule for shorting stocks went away.  It is hard to understand why a rule change like this could make a difference on a market cycle, but I have an explanation that may help.

Imagine a wave pool in a laboratory, where scientists create waves to study how they travel through the water.  Now imagine that you remove all of the water, and replace it with 30-weight motor oil.  Because the oil is lighter but more viscous than the water, the behavior of waves in that wave pool would understandably be different.

So thinking of the financial markets, if the regulators were to do something that changes the "viscosity of money", making it flow more or less easily, then we would likely see changes in the way that waves propagate through that medium as well.  Such changes might include restrictions on shorting stocks, the advent of money market funds, the introduction of stock index futures and options, leveraged ETFs, etc.  All of these affect the ease with which money can flow into and through the stock market.

Now, if you look back at the top chart, you can see that the blue numbers are getting bigger again lately.  Those numbers represent the time period between the major lows of this cycle (formerly known as 9-month).  The lowest number was 159 trading days in early 2008, and it has climbed back all the way up to 177 as of the latest major cycle price low.  It may be that after the initial shock, this cycle is working on getting back up to is "natural" frequency.  Or it may be that 159 and 177 are just the widest extremes of a new range of cycle periods that average more like 168 trading days, and that this is the new natural frequency.  We won't know for sure for several more cycles' worth of time, and that's the big problem with this analytical technique.

For what it's worth, and to help your planning, 159 to 177 trading days from the most recent major cycle low equates to a timeframe of Oct. 31 to Nov. 25, 2011.

Advanced GET's Cycle-tool suggests this was the 9 Month Cycle crest. However, the tool is not adjusted according to Tom McClellan's findings.

Wednesday, March 28, 2012

ITD Delta Pattern & Solar Forecast & EAR - MAR Cycle = Low April 3-6

SPX vs EAR - MAR Cycle

23.09.2011  06:18 FR = EAR 0 MAR [H40]
11.10.2011  07:36 TU = EAR 0 MAR [H40]

28.10.2011  14:55 FR = EAR 0 MAR [H40]
14.11.2011  06:49 MO = EAR 0 MAR [H40]
30.11.2011  12:17 WD = EAR 0 MAR [H40]
16.12.2011  09:51 FR = EAR 0 MAR [H40]
01.01.2012  01:45 SU = EAR 0 MAR [H40]
16.01.2012  15:15 MO = EAR 0 MAR [H40]
01.02.2012  04:11 WD = EAR 0 MAR [H40]
16.02.2012  19:30 TH = EAR 0 MAR [H40]
03.03.2012  14:50 SA = EAR 0 MAR [H40]
19.03.2012  17:46 MO = EAR 9 MAR [H40]
05.04.2012  03:48 TH = EAR 18 MAR [H40]

21.04.2012  23:57 SA = EAR 27 MAR [H40]
09.05.2012  07:38 WD = EAR 36 MAR [H40]
27.05.2012  03:51 SU = EAR 45 MAR [H40]
14.06.2012  13:52 TH = EAR 0 MAR [H40]
03.07.2012  13:34 TU = EAR 0 MAR [H40]
23.07.2012  03:07 MO = EAR 0 MAR [H40]
12.08.2012  06:14 SU = EAR 0 MAR [H40]
01.09.2012  21:39 SA = EAR 0 MAR [H40]
23.09.2012  00:34 SU = EAR 0 MAR [H40]
14.10.2012  14:23 SU = EAR 0 MAR [H40]
05.11.2012  13:15 MO = EAR 0 MAR [H40]
27.11.2012  22:48 TU = EAR 0 MAR [H40]
20.12.2012  18:13 TH = EAR 0 MAR [H40]
12.01.2013  23:28 SA = EAR 0 MAR [H40]
05.02.2013  13:11 TU = EAR 0 MAR [H40]
01.03.2013  08:22 FR = EAR 0 MAR [H40]
25.03.2013  05:25 MO = EAR 0 MAR [H40]
17.04.2013  20:38 WD = EAR 0 MAR [H40]
11.05.2013  01:07 SA = EAR 0 MAR [H40]
02.06.2013  13:43 SU = EAR 0 MAR [H40]
24.06.2013  06:48 MO = EAR 0 MAR [H40]

Tuesday, March 27, 2012

SPX vs EAR-VEN Cycle [helio]

30.08.2010  11:52 MO = Sun   36°  Venus
29.09.2010  00:48 WD = Sun   18°  Venus
28.10.2010  21:13 TH = Sun    0°  Venus
27.11.2010  19:20 SA = Sun   18°  Venus
27.12.2010  14:36 MO = Sun   36°  Venus
25.02.2011  11:29 FR = Sun   72°  Venus
26.04.2011  22:47 TU = Sun  108°  Venus
22.06.2011  22:42 WD = Sun  144°  Venus
16.08.2011  07:42 TU = Sun  180°  Venus
11.10.2011  13:09 TU = Sun  144°  Venus
12.12.2011  13:33 MO = Sun  108°  Venus
12.02.2012  17:18 SU = Sun   72°  Venus
10.04.2012  11:40 TU = Sun   36°  Venus
08.05.2012  12:41 TU = Sun   18°  Venus
05.06.2012  21:11 TU = Sun    0°  Venus
04.07.2012  12:50 WD = Sun   18°  Venus
02.08.2012  03:54 TH = Sun   36°  Venus
28.09.2012  05:09 FR = Sun   72°  Venus
24.11.2012  20:17 SA = Sun  108°  Venus
25.01.2013  09:11 FR = Sun  144°  Venus
28.03.2013  12:34 TH = Sun  180°  Venus
24.05.2013  13:15 FR = Sun  144°  Venus
17.07.2013  11:42 WD = Sun  108°  Venus
12.09.2013  02:28 TH = Sun   72°  Venus
11.11.2013  12:46 MO = Sun   36°  Venus
12.12.2013  02:51 TH = Sun   18°  Venus
11.01.2014  07:24 SA = Sun    0°  Venus
09.02.2014  21:38 SU = Sun   18°  Venus


Prediction of Sunspot Cycle 24-Peak & Long Term Trading Strategy

SIDC: The daily (yellow), monthly (blue) and monthly smoothed (red) sunspot numbers since 1994, together with predictions for 12 months ahead: SM (red dots) : classical prediction method, based on an interpolation of Waldmeier's standard curves; CM (red dashes) : combined method (due to K. Denkmayr), a regression technique coupling a dynamo-based estimator with Waldmeier's idea of standard curves. Peak: January 2013

NASA: The current prediction for Sunspot Cycle 24 gives a smoothed sunspot number maximum of about 59 in early 2013. We are currently over three years into Cycle 24. The current predicted size makes this the smallest sunspot cycle in about 100 years. Peak: January-February 2013

IPS: Peak: December 2012

Last updated 26 Mar 2012 13:03 UT

                         FORECAST SOLAR CYCLE 24
Cycle  Sol. Start  Sol. Max  Max SSN     Length     Rise to Max     Max to End
       Year Mth    Year Mth             Yr   Mth    Years   Mths    Years  Mths
24     2009 Jan    2012 Dec   90.2     11.0 132     3.9    47       7.1    86

IPS will adjust this forecast cycle as the new cycle unfolds. 
The difficulty is ensuring that adjustments are not made for short 
term variation, only for longer term cycle variation. 

NOAA: Given the predicted date of solar minimum and the predicted maximum intensity, solar maximum is now expected to occur in May 2013.

Here is the data supporting the shorter term strategy of buying at solar minimums and selling at the next cycle maximum for an average 70% gain:
Why might stocks consistently outperform in these periods from solar minimum to maximum, and underperform from solar peak down to the next solar minimum, particularly as higher solar activity can cause higher geomagnetism on Earth which affects humans biologically negatively and adversely affects stock market returns?
Well, there is a slight lag in geomagnetic peaks after solar cycle peaks, as shown below, and this fits well with why we have seen an economic recession follow each solar cycle maximum in the last century - it corresponds to the peak in geomagnetism. Historically, this post-solar-peak period has been one of human apathy and peace. Conversely, the period into the solar peak has been one of human excitability, pro-action and economic inflation, which fits well with stock market gains.
Source: Susan Macmillan, British Geological Survey

Solar Cycle 24 began around December 2008 with a solar minimum and it is predicted to peak in July 2013. An average gain of 70% for the Dow over this period would translate as 14500 by mid 2013 (which would mean a new nominal all time high).A recession has closely followed solar peaks for each solar cycle in the last 100 years. The average recession duration is 1 year. The average length of recession-induced stocks bear markets is 1 year 4 months. As the stock market is forward looking, and a leading indicator, we could therefore find the the stock market peaks around the beginning of 2013 and then declines into the solar peak in mid 2013, and then declines through a recession into 2014.

Dow-Commodities ratios and consumer price inflation should peak at extremes at the solar peak (as has occurred each time in the last century), suggesting commodities should push on all the way into mid-2013 whilst stocks lag in the last few months.  
In summary, there is a correlation between stock market performance and solar cycles. A profitable strategy over the last century would have been to buy at the solar minimum and sell at the next solar maximum, and repeat for an average 70% gain in each instance.

An even more profitable strategy would have been to buy and hold over 2-3 decades in between 3 specific half solar cycles. This strategy would have produced 10-fold gains each time, and pattern continuation suggests such a repetition from the solar minimum at the end of 2008 looking out to the 2030s, in line with a further secular stocks bull.

Looking shorter term to the solar peak around mid-2013, stocks should track yet higher, and this implies commodities much higher, as an extreme relative pricing of commodities over stocks should be reached around that solar peak, before a secular inversion.
John Hampson, April 2011 @

Monday, March 26, 2012

Tidal CITs @ Willets Point [NYC] March 15 - June 15

2012-03-14 (Wed) = Third Quarter
2012-03-16 (Fri) = Tidal CIT
2012-03-22 (Thu) = New Moon 
2012-03-23 (Fri) = Tidal CIT
2012-03-26 (Mon) = Moon @ Apogee
2012-03-30 (Fri) = First Quarter
2012-04-02 (Mon) = Tidal CIT
2012-04-06 (Fri) = Full Moon = SuperMoon
2012-04-07 (Sat) = Moon @ Perigee
2012-04-08 (Sun) = Tidal CIT
2012-04-13 (Fri) = Third Quarter
2012-04-14 (Sat) = Tidal CIT
2012-04-21 (Sat) = New Moon 
2012-04-22 (Sun) = Tidal CIT
2012-04-22 (Sun) = Moon @ Apogee
2012-04-29 (Sun) = First Quarter
2012-05-01 (Tue) = Tidal CIT

2012-05-05 (Sat) = Moon @ Perigee
2012-05-05 (Sat) = Full Moon = SuperMoon 
2012-05-08 (Tue) = Tidal CIT
2012-05-12 (Sat) = Third Quarter
2012-05-13 (Sun) = Tidal CIT
2012-05-19 (Sat) = Moon @ Apogee
2012-05-20 (Sun) = New Moon = Solar Eclipse
2012-05-23 (Wed) = Tidal CIT
2012-05-28 (Mon) = First Quarter
2012-05-31 (Thu) = Tidal CIT
2012-06-03 (Sun) = Moon @ Perigee
2012-06-04 (Mon) = Full Moon = SuperMoon = Lunar Eclipse

2012-06-06 (Wed) = Tidal CIT
2012-06-11 (Mon) = Third Quarter
2012-06-12 (Tue) = Tidal CIT
2012-06-15 (Fri) = Moon @ Apogee
2012-06-19 (Tue) = New Moon

2012-06-22 (Fri) = Tidal CIT

Sunspots, GDP & the Stock Market

Theodore Modis: Sunspots, GDP & the Stock Market. In: Technological Forecasting & Social Change 74 (2007) p. 1508–1514 []

Solar Activity and Solar Tides caused by the Planets

Ching Cheh Hung (2007): Apparent Relations Between Solar Activity and Solar Tides Caused by the Planets. Glenn Research Center, Cleveland, Ohio. [NASA/TM—2007-214817 @]


In view of the statistics and the issues presented and discussed in this report, one must acknowledge the possibility that some type of tide-solar activity relation may truly exist, despite the widely accepted thought that believes otherwise. Evidence of apparent relations between planet positions and solar activity was observed and presented:

(1) Twenty-five of the thirty-eight largest known solar flares were observed to start when one or more of the tide-producing planets (Mercury, Venus, Earth, and Jupiter) were either nearly above the event positions (<10° longitude) or at the opposing end of the Sun. The probability of this to happen at random is 0.039 percent. This observation supports the hypothesis that the force or momentum balance (between the solar atmospheric pressure, the gravitational field, and the magnetic field) on the plasma in the looping magnetic field lines in solar corona could be disturbed by tides, resulting in magnetic field reconnection, solar flares, and solar storms.

(2) From the daily planet positions during the period from 1840 to 2000, an 11-year cycle of the alignment of Venus, Earth, and Jupiter is observed. This cycle approximately matches the sunspot cycle. When the two cycles were least matched, the sunspot numbers were low (1875 to 1930). When best matched, the sunspot numbers were high (late 1950s). This supports the hypothesis of resonance and beat between the cycle of small tides caused by the planet alignment and the cycle of independent, large, and irregular nontidal solar activity. Mercury produces significant tides, but is not considered because it does not have an 11-year-cycle resonance with the nontidal solar activities, either by itself or by aligning with other planets. The observed relation between the predictable planet position and unpredictable solar flare suggests a way to forecast the times of the largest solar flares (X9.0 and larger, both near- and far-side from Earth) shortly after giant sunspots appear: They are most likely to start when these sunspots rotate into a region where at least one of the four tide-producing planets is either overhead or underfoot (within 10° longitude). They are least likely to occur when these sunspots are at 36° longitude or further away from the overhead or underfoot points of all these four planets.

The theory of resonance between the cycle of nontidal solar activity and the cycle of the alignment of Venus, Earth, and Jupiter forecast a low sunspot number in the coming solar cycle number 24 unless the current solar minimum would last for a few more years to reduce the current mismatch between these two cycles.

Edward R. Dewey: Economic and Sociological Phenomena related to Solar Activity and Influences

Saturday, March 24, 2012

Tom McClellan: SP500 Undervalued vs M2

The 40-year cycle for the stock market is not due to bottom until around 2022, and so between now and then it gives us the expectation of a sideways decade for the stock market, much like the 1970s or the 1930s. At the same time, the leading indication for the stock market that is given by oil prices says we should see a strong upward trend for stock prices during the 2010s.  These two expectations are in conflict with each other, but I think I can resolve that conflict with this week's chart. It looks at the SP500 Index level divided by the value for the monetary aggregate measure known as M2.  For the uninitiated, M2 consists of all of the currency in circulation and in bank vaults, plus travelers checks, demand deposits, other checkable deposits, savings accounts, and money market funds.  It is a commonly used measure of the total amount of "money". When it comes to analyzing the overall stock market, I like to say that there are only two fundamental factors which matter.  Forget earnings, dividends, book value, revenue, etc.  When it comes to analyzing the overall stock market, the only two factors which matter are (1) How much money is there? and (2) How badly does that money want to be invested?

That's where this week's chart comes in.  The Fed has been inflating the money supply in order to resuscitate the US economy.  Over the past 12 months, M2 is up 9.8%, and it has doubled since 2000.  So in terms of fundamental factor #1, the Fed is doing all it can to help lift the stock market. But that money has not been feeling inclined to go to work pushing up stock prices, and the result is that this ratio of the SP500 to M2 is now back down to levels it saw during the big bull market of the 1980s.  And if the Fed keeps on inflating the money supply, as Fed officials have stated is the goal, then that will push this ratio down even further. So that's how we can get a big uptrend for stock prices during the 2010s like the oil price leading indication suggests we should see.  But that is not necessarily a good thing, because along with all of that money supply growth comes inflation in consumer prices.  So far, the Fed keeps saying that inflation has remained low, but that is due in large part to the fact that housing makes up a 41% weighting in the Consumer Price Index (CPI).  If you factor out the depressing effects of the weak housing market, the price inflation in all of the other CPI components has been as high as 5% recently. 

One way to have stock prices go up is to have the companies backing those stocks become more valuable.  But another way is to have the dollars needed to buy those stocks become less valuable, so that you need more dollars to buy your favorite stock.  That appears to be the path that the stock market is now on, which explains how we can have a rise in stock prices in this decade, but not in real stock values.