Taking the mystery out of cycles

DETRENDING TAKES THE MYSTERY OUT OF CYCLES

Historical cycles show great potential for low risk trades. But to make cycles tradable it is
necessary to define the tops and bottoms so they can be identified and traded with
mechanical buy and sell signals.

For over 50 years the non-profit Foundation for the Study of Cycles has been
using a simple procedure called centered detrending to identify cycle lengths in all
natural phenomena, including economic activity. The advantage of centered detrending
over Fourier analysis and Spectral analysis is that you can visibly observe the historical
cycle bottoms and tops. This takes the "mystery" out of cycle trading and analysis. When
you can look back over 20 years of price history and actually see the consistency and
tradability of cycles, it is easy to have the confidence to trade them. After all, if you
cannot see cycles in price activity how can you put your money on the line to buy a
bottom or sell a top. We are not looking to trade the big bottoms and tops that occur as
the trend reverses, but the bottoms and tops of the shorter-term cycles occurring in the
direction of TREND.

The process to detrend a cycle is simple:

1. Calculate a moving average the same length as the potential Trading Cycle. For
most markets use a 20-bar moving average of the close. But instead of plotting
the moving average on the day of the last calculation, it is centered (or plotted) in
the middle of the cycle. For example, the 20-day moving average is plotted back
10 days from the most recent close. This is called a centered moving average as
it is plotted in the center of the cycle.

Chart 5 – Centered Detrend -Trading Cycle tops and bottoms are easy to see in the centered detrend. 

2. Subtract the moving average from the high and low of each price bar and plot the
results around a zero line directly below each price bar. The zero line represents
the centered 20-bar moving average.
3. The centered detrend at the bottom of Chart 3 mitigates the effects of trend and
allows the highs and lows of the trading cycle to be easily seen. This process can be easily programmed into many of the popular trading platforms. Alternatively you can use the DPO indicator in StockCharts.com.

Chart 6 - The chart below has been compressed to show more cycles. The highs and
lows of the centered detrend correspond to the highs and lows of the trading cycle. We know from our envelope analysis that 2.5% above or below the MA for 20 TDs contains most of the price movement for the SPX. With the SPX at around 1500 that is 37 points above the zero line is extreme and indicates an imminent sale or 37 points below the zero line is an oversold condition and indicates an imminent buy. But, these are rare occasions and we need tighter ranges (probably1.25% or 18 points above/below the zero line).

　
If you are thinking that this is the indicator for which you have been searching…
there is still one problem. Because the centered detrend lags prices by one-half the length ofthe cycle, or 10 bars for a 20 period cycle, the centered detrend is excellent for showing historical cycles, but cannot be used for real-time trading (same as with envelopes), but it reveals information regarding cycle amplitudes.

 
Chart 6 – Compressed Chart with Centered Detrend – Cycle tops and bottoms occur at the highs and lows of the detrend which shows limits to how far prices can move above and below the centered moving average (which is the zero line in the bottom centered detrend).

　

However, we now have something to look for -- A real-time oscillator that matches the
centered detrend and the cycles.

OSCILLATORS SHOW CYCLE TOPS AND BOTTOMS

Chart 7 - The standard oscillators in Stockcharts.com can be overlaid on the centered detrend (or on top of prices) in the search for an oscillator that tops and bottoms as the trading cycles top and bottom. Not surprisingly, the performance of these oscillators for identifying trading cycle tops and bottoms can be improved by a few simple techniques.

To use an oscillator to identify cycle tops and bottoms look for three characteristics:

1. The oscillator turns when prices turn
2. The oscillator does not "wiggle" much at cycle tops and bottoms
3. The oscillator has amplitude moves that take it to the extremes of an allowable range as the cycles bottom and top.

One such oscillator is the Wilder RSI. If the RSI 3 is used it provides frequent enough oscillation to fit a swing trader's needs. A smoothed RSI with a 3 bar averge is best (RSI3M3). It shows the bottoms and tops of the trading cycle almost as well as the centered detrend and can be used as a mechanical trading signal and cycle identifier… And the RSI3M3 is current to the most recent price bar, turning down in this chart to identify the most recent trading cycle top (not normally identified by the centered detrend (DPO)).

OSCILLATOR/PRICE PATTERNS GENERATED

Chart 7 – The RSI3M3 oscillator is current to the most recent price bar and shows cycle tops and
bottoms as well as the centered detrend.

NOTE: Many use the RSI without understanding its relationship to cycles. By using 20/80 lines we may miss some trades that could be taken if we used 30/70 lines. Still we caught good entries at 1, 3, 5, and 7. We got good exits at 2, 4, 6 and 8. Using 20/80 or 30/70 is an individual decision (20/80 is a more conserative approach).

To be continued.....

IDENTIFYING CYCLE TOPS AND BOTTOMS USING ENVELOPES

Chart 1 -  This chart shows the ebb and flow of prices, but identification of the Trading
Cycle bottoms and tops requires a little effort to be visible to the untrained eye.

Chart 1 - Do you see the cycles in this chart?

 Chart 2 -Cycles are measured from bottom to bottom. Every time frame of every market has a dominant Trading Cycle averaging from 14 to 25 bars as measured in weeks, days, half days, hours, minutes or ticks. Most Trading Cycles in the stock and  futures markets tend to cluster in the 18 to 22 bar range, averaging 20 bars from bottom to bottom. In Chart 2, SPX exhibits a daily trading cycle of 23 days from bottom to bottom. The trading cycle tops and bottoms are indicated by the arrows.

 Chart 2 - the arrows indicate trading cycle bottoms and tops.

ENVELOPING PRICES

Prices cluter above and below a moving average. If we can identify a potential cycle (bottom to bottom) then we can draw a moving average the length of that cycle. To center that average over the data it averages we need to shift the average half the length of the cycle. Fortunately, many have studied cycles and we know what the likely cycle lengths are. One very common cycle is the lunar cycle which runs about 21-23 trading days (explains the 20 day MA).

We can draw a time shifted MA and draw a line above and below this MA by a given percent. During normal market volatility about 2.3% above/below the MA seems to be a reasonable choice. Below is a chart using a 21 day MA time shifted and an envelope 2.3% above and below that MA.

Chart 3 - Envelope highlight cycle tops/bottoms

ENVELOPES WITHIN ENVELOPES
If we enclose the first envelope within a second envelop of a longer cycle and select an appropriate envelope width then the envelopes should touch or cross at critical points and these points become points to enter (bottom lines cross/touch) or sell (top lines cross/touch). Another common cycle is the 34 TD (Trading Days) cycle.  

 Chart 4 - Envelope in an envelope and channel line signals

As you have already figured out - because the envelopes are offset they lag real time prices and you will have to project the envelope lines. This means some judgment is called for, still it is an improvement over the chart without envelopes.

To be continued......

a guide to Cycles TA

I often get questions about cycles and charts.  I came across a guide to cycles TA.  Hope it helps you in your analysis.

http://www.walterbressert.com/pdf/Manual.PDF

Enjoy the read.

silver update Feb 26 2013

We hit the downside targets I suggested in my Feb 20 update and bottomed as expected.  We have now turned and this short term move up should complete by the end of the week at $30.40,

Here is a visual:

GL  traders....

Gold outlook - Feb 24 2013

Should see more downside, but an upside reaction short term may be needed.  Here is a chart:

GL gold bugs

Mar 2013 outlook

I suspect we are into a pullback of at least 6-7%.  If we get that by mid-March then it will be time to take a second look.  Overall March should be weak.  As usual we will have some short up moves, but the cycles suggest we get more down than up.  By mid April  expect the 20 week Wall cycle to be bottoming (a buy the dip opportunity?).

Here is a visual:

GL traders

Feb 2013 update III

Did today start a correction?  Maybe, one is overdue.  Of course cycles are not always symetrical, often the upleg takes more time than the downleg so it often looks as if the top is late.  But in fact that should be antcipated.

Short term it appears we should get our first down week of the year.  Here is a visual:

I'll try to add a longer term chart later.

Here is a longer term chart:

GL traders

New short term chart:

GL

Silver update feb 20, 2013

WOW - that is a big selloff in silver.  I mentioned last post on silver we would need a big sell down to meet the 28.50 target.  It took longer than projected (4-5 TDs longer) but we got the sell down.  Looks like we might go somewhat lower.  At these levels though if buying silver and you want to buy $10k -  if I were you I'd buy about $3k worth and hope to buy more  at lower prices.

Here is a visual:

GL - Hiho silver away!!!  Who was that masked man?

Gold update February 2013

Gold had quite a move down this past week.  A down move was expected, but this move was earlier and deeper than expected.  Most likely we will see a bounce next week and then continue down into mid to third week of March.  A low in mid 1500s would not be surprising....

Here is a visual:

GL gold bugs.

Hurst cycle lengths

Cycle Lengths as defined by JM Hurst
17.93 years  (Kuznets cycle)
8.96 years  (Jugular cycle)
53.77 months (4.5 years)
17.93 months (1.5 years)
38.97 weeks (9 months)
19.48 weeks (Wall cycle)              
68.2 days (about 10 weeks)
34.1 days (1/4 Wall cycle)
17 days
8.5 days
4.3 days
2.2 days
26.67 hours
5.3 hours
160 minutes
53.3 minutes

David Knox Barker refers to 19.48 weeks as the Wall cycle  (141 calendar days or 20 weeks), so you can see overlap as one would expect and some differences (like the kitchin cycle at 42 months).

Flattered, but

I was asked if I had a central repositary for my charts etc.  Short answer is I do not and what I have on my PCis not organized.  This blog comes as close as anything to being a central reference. I just got a new PC and did not copy all files from the past couple of years.

I am flattered that it was suggested people would pay for a book (If I had one - which I don't) covering my analysis.  There are lots of books out there on cycles.  David Knox Barker (long term cycles - Kitchin/Wall Cycle expert) has a couple of books.  Google him and read some of his articles to decide if his work would interest you.  Barker also has a paid advisory nservice.  Jim Curry is a fan of JM Hurst and his work on cycles.  Google him and read some of his articles.  He also has a paid advisory service....  Then there is Cliff Droke (fan of Bud Kress and his work).  Google him.  He has a couple of books and offers a paid advisory service.

Of course there is Charles Nenner and a host of others out there willing to take your money.  I am not one of them.  I do this blog because of many years of investing in the market (some times I did well, other times not so well) and extensive reading about different technical approaches to investing/trading in the market.  I am not a writer (Have a BS in math not journalism) so writing a book would be a challenge.  Having a for profit advisory service would be too much like having a job (defeat the purpose of retirement of  not having regular work).  LOL  In short - I don't do this for monetary gain, but to share some insights that may help others.  I also try to offer pieces of information that will help advance the education/understanding of others.

In summary, there are lots of people who have written books on the subject of stock market cycles (check Amazon) and I have mentioned a couple, but read reviews on Amazon to decide as I have not read all their books.  Also, there are many sites offering software for Hurst Cycles (for example).  Software and data feed access can be expensive.  Having not subscribed to or used any of this software I cannot suggest what software may be worth the money,

I use the free access to a couple of charting services and do my charts manually.  Primarily I use Stockcharts.com and manually annotate using the annotation service.  Since I use the free service I cannot save charts on stockcharts.com.

So, you get my insights for free - and pay what they are worth.  Hehehe

February 2013 update II

When you have one cycle topping (34 TD cycle)as another similiar length cycle (23 TD cycle) is bottoming you have a situation where they cancel each other out.  As a result you get a sideways move which may have a slight upward or downward bias. this situation  may exist for several days as appears to have happened over the last few days. 

This situation is about to be resolved.  The Wall cycle should have also topped and is turning  down.  The longer cycles (kitchin and 1/3 kitchin) continue up offsetting to some extent the downside of the Wall cycle.  The 34TD  cycle is turning down, but the 23 day cycle is turning up.  So expect a slow meltdown over the next 2 weeks and then a little more forceful selldown in early March after the 23 TD cycle turns down.

Here is a visual:

GL traders

 

Shorter 8 TD cycle (~ 11 calender days avg) -

FEB 15 update - 8TD cycle:

GL

Big guns warn of war

Kyle Bass, Larry Edelson, Charles Nenner, James Dines, Nouriel Roubini, Jim Rogers, Mrac Faber and Jim Rickards Warn or War

http://www.ritholtz.com/blog/2013/02/forecast-of-war-and-unrest/?utm_source=feedburner&utm_medium=email&utm_campaign=Feed%3A+TheBigPicture+%28The+Big+Picture%29

silver update Feb 11, 2013

If the cycles play out silver ahould put in a bottom by mid week. Silver failed to reach our upper target, but did exceed our lower targer.  We may get a substantial 2-3 day sell down here.

Here is a visual:

GL traders

25 years ago

Oct 19, 1987 we had the biggest one day fall (flash crash) ever in the market.  That wa 25.3 +/- years ago.  And 25.3+ years before that was May, 29 1962 (flash crash of 5.7% for the Dow). Coincident?  Possibly.  We are close to 25.3 years since the Oct 19, 1987 crash.  So with sequestration coming is a crash possible within the next 2 or so weeks?

1912 - crash happened June 1913

1937 - market dropped by half late 1936 - March 1938 (24.8 years after June 1913)

1962 - May 29,1962 market drops 5.7 % (about 25 years)

1987 - Oct 19, 1987 market drops 22.6% (25.3 years after 1962 flash crash)

2013 - Feb 11, 2013 is about 25.3 years (??? crash time ???))

Notice how these incidents (flash crashes) occur 25+ years. Will it happen again?

1912 market heads down

1912 DJIA declined 23.5%. Duration of bear market: 26 months. Crash came

in June 1913, but bear had started in November 1912. Bear low came in

December 1914. Since the stock market was closed for 4 months in 1914

by the war, a true picture of the decline here is impossible. One source

says it was 36.3%. But I’ll take the more conservative published DJIA

figures as my base.

2013 Looks a Lot Like 1937 in Four Fearsome Ways
Will 2013 be 1937? This is the question many analysts are posing as the stock market has dropped after the U.S. election. On Nov. 16, they noted that industrial production, a crucial figure, dropped as well.

In this case, "1937" means a market drop similar to the one after the re-election of another Democratic president, FDR, in 1936.

The drop wasn’t immediate in that case; it came in the first full year after the election. Industrial production plummeted by 34.5 percent. The DJIA dropped by half, from almost 200 in early 1937 to less than 100 at the end of March 1938.

　

The Flash Crash 1962
"The stock market careened downward yesterday," reported The Wall Street Journal on May 29, 1962, "leaving traders shaken and exhausted." The Dow Jones Industrial Average fell 5.7% that day, down 34.95, the second-largest point decline then on record.

"The drop took place on volume so heavy," added the Journal, that the "ticker wasn’t able to finish reporting floor transactions until 5:59 p.m., two hours and 29 minutes after the market closed."

Oct. 19, 1987 | Stock Market Crashes on ‘Black Monday’
On Oct. 19, 1987, a day that became known as "Black Monday," the stock market crashed as the Dow Jones Industrial Average plunged 508 points, or 22.6 percent in value, its largest percentage drop ever.

Feb. 11, 2013 Market Crashes - DJIA down 2015 points

Black Monday II" happened Feb 11, 2013

NOT a prediction, just an observation....

February 2013 update I

I hesitate to declare a top is in, but it appears to be the case.  Given that I expect a pullback from the 2 shorter cycles (23TDs and 34TDs) into the February 22 and 27 time frame as 23TD cycles bottoms first around Feb 22 and the 34TD cycle bottoms later around Feb 27.  So the absolute bottom should be in that window.

Here is a visual:

 

GL traders

Feb 9 update:  New highs on S&P and NAZ but DJIA lagging slightly.  May be a top, but higher higs Monday not totally unexpected.  Here is an updated chart:

GL

Gann Square of 9

http://www.stockmaniacs.net/gann.html

The important parts of the Square are the center, the Cardinal Cross composed of the vertical and horizontal rows that intersect at the middle of the square, and the lines extending at 45 degrees, 135 degrees, 225 degrees, and 315 degrees that constitute the Fixed Cross. The Cardinal Cross and Fixed Cross are used to determine likely points of support and resistance

gold outlook - February 2013

When it comes to gold I have no special knowledge or insights, but I was asked for an opinion.  So I have attempted to take a look. The commodity gold may or may not track the PM miners.

Looking at the charts it appears a top (around 1700) should occur the second week of February (Feb 11?).  Then it should pull back 60-70 points (1630-1640) before the middle of March (Mar 11-13?).  So take profits in paper gold (ie ETN GLD).  I assume if you hold physical gold it is not for short term trading.

Here is a visual:

GL gold bugs - I like acres where I can grow food (you can't eat gold).

Reading the VIX - February 2013

One way to look at the VIX is to compare it to the SandP 500 index.  We can do this by graphing a ratio of $SPX and $VIX ($SPX:$VIX).  If VIX is 12 and SPX is 1200 then the ratio would be 100.  Here is the ratio charted:

When the ratio is below 80 we should consider the market oversold and look to buy (see green arrows). When the ratio is over 105 we should consider the market over bought and consider taking profits (see red arrows). Adjustments of buy/sell levels may require adjustments over time as the ratios adjust to different channel levels.

February 2013 outlook

January is over and played out much as anticipated. Seems I was off a couple of days on the top (assuming we set a top early this last week of January).  Maybe the 2 days the market was closed for Sandy affects my counts   (not sure how something like this that closes markets affects the cycle day counts)?

I see a lot of people talking about a severe pull back, and I am sure at some point we will get a 40-50% correction like 2008-2009 or in the early 2000s.  Seems to me though the market holds up longer or pulls back less for longer than we expect.  Now I expect February to show weakness as the Wall cycle turns down .  The 33-34 TD cycle also should have topped and turned down for the first 3 weeks of January.  Offsetting is the Kitchin Cycle (up), the 1/3 Kitchin Cycle (up) and 1/4 Wall (22-25 TD cycle up).  Overall I believe the 33-34 TD and Wall Cycle will win, but there is no clear cut case for a huge drop.  So The market should frustrate the bears (a 30-40 point pullback in the S&P 500) and provide little comfort to the bulls.

Here is a visual:

GL traders