BudFox Posted January 3, 2005 Report Share Posted January 3, 2005 This thread will be used to discuss the proper use and interpretation of technical indicators for time cycle analysis in all manner and duration of trading. We can also exchange theories and personal trading experiences in order to improve our understanding of cycles and the artful interpretation of them. Hopefully, we can get Doc involved too! Link to comment Share on other sites More sharing options...
BudFox Posted January 4, 2005 Author Report Share Posted January 4, 2005 From BTB...... BudFox, what software are you using to do the cycle analysis of individual stocks, in re the charts you posted earlier? and do you know of a site which lists or reviews all packages which are capable of doing cycles, FFT's, etc., on individual issues? tanks! Hey Doz, I use Metastock, the same program Doc uses for his cycle charts. But I've developed my own custom indicator for cycles after many, many, many hours of fiddling around with the indicator programming language. The programming is complex, but the result is very clean and simple. Any charting program that has a rate of change indicator or stochastics or price oscillator can give you a basic idea of cycles, provided you can identify a regular cycle through visual analysis first. Metastock also has a time series price oscillator, which Doc uses regularly. There are several other proprietary cycle indicators and trading systems that can be purchased for ridiculous amounts of $$ for Metastock, Tradestation, and ESignal, etc. hey Bud, Thanks much! Sounds like the endless fiddling has paid off... You could call it "optimizing" instead, and it'd be easier to market... Do you find that your indicator works well on an intraday basis as well? I would like to find something to predict intraday cycles on individual issues; preferably which would also give some indication of the expected -size- of a move, not just the direction. The objective is to have some help deciding -which- 1-2 issues to focus on that day... Bud, is the need to first visually ID a regular cycle so that one can set the parameters of the above indicators to 'tune' it to that cycle? Or is there a different reason? I'm developing an intraday trading strategy which I think will pay quite well, but success depends on achieving decent % rate on calling direction; and only placing bets where the move has good chance of being sizable (i.e. 1-2% or so, e.g. a typical AMZN or SBUX move of 50 or 100 cents over 10-60 minutes). I really appreciate the help, as it might let me get a cycle indicator working that would aid the direction and size call... Yes, the visual ID of a cycle provides the parameter for the indicator. If the settings are correct, the indicator tops and bottoms should match the price. The better the indicator, the less the time lag between the price and the indicator. Also, there are some stocks that show clear cycles and others that do not. Some have several cycles of varying lengths, some have only one. The more the volume, the better the chance for reliable cycles. Unfortunately, no indicator by itself can predict the size of a move. When two cycles of different lengths align in a buy signal, chances are there will be more momentum behind the move. I used to track cycles on the ES future by the minute in Tradestation. That kind of trading can cause heart failure. I only use Metastock end-of-day now.? I hear ya! The choppier moves can certainly produce palpitations...? I'm looking at doing 1-2 moves per day; rather than a continuous all-day thing; so I'm hoping to limit my cardiac risk. QT has what they call a "time series" upper-indicator, but upon trying it, it didn't strike me as useful/usable; and I have not found a description in the help so far, of the algorithm being used; so I did not try to adjust the default parameters. But I should invest some fiddling there. Do you have a suggested starting point? I.e., in terms of "if the cycle is X periods, start by setting parameter Y to the cycle periods times Z", etc... An oscillator type of indicator is always set to half the top-to-top or bottom-to-bottom cycle length. Link to comment Share on other sites More sharing options...
trinharder Posted January 5, 2005 Report Share Posted January 5, 2005 Well kewl, I've been messing around a bit with Fibonacci time and arcs. Will post on it later, I just found this thread this morning. Great idea. Link to comment Share on other sites More sharing options...
DrStool Posted January 6, 2005 Report Share Posted January 6, 2005 I will pop in as often as possible. Very busy with the new site right now, but I will drop by from time to time. Link to comment Share on other sites More sharing options...
trinharder Posted January 6, 2005 Report Share Posted January 6, 2005 It will likely be this weekend before I can add any real comments here. I have, so far, only been looking at Fibonacci time series. I'm not sure you can really call that a cycle per se. The question that comes to mind is: How do you determine which time period to use in a cycle and is it different for each underlying, or a one size fits all? Link to comment Share on other sites More sharing options...
BudFox Posted January 10, 2005 Author Report Share Posted January 10, 2005 As bullish as a couple of individual biotech stocks are looking right now, the BBH average is clearly topping out. The cycle lengths shown are 180, 90 & 60 day. Link to comment Share on other sites More sharing options...
BudFox Posted January 10, 2005 Author Report Share Posted January 10, 2005 In the above chart, the 180 day is the dominant cycle. Smaller nested sub-cycles are most discernable here as 60 or 90 day. Sub-cycles in sets of two or three are the most common. Though I have seen 5 sub-cycles from time to time. Also, there is often a variation between the number of sub-cycles. It's important to follow the different cycle permutations in order to obtain usable buy and sell signals. These signals are the result of two or more cycles moving in the same direction. Two cycles moving up together create decent buy signals, while three in sync make for great entries. Link to comment Share on other sites More sharing options...
Janitor Posted January 10, 2005 Report Share Posted January 10, 2005 Hi Here is a cycle indicator I use for Metastock. It is very simple. I found it on the net and adjusted it a bit. CY1:= Input("Short Cycle length" ,1 ,100000 ,10 )/2; CY2:= Input("Medium Cycle length" ,1 ,100000 ,80 )/2; CY3:= Input("Long Cycle length" ,1 ,100000 ,160 )/2; M1:= Input("Short Cycle Multiplyer" ,.01 ,1000 ,4); M2:= Input("Medium Cycle Multiplyer" ,.01 ,1000 ,8 ); M3:= Input("Long Cycle Multiplyer" ,.01 ,1000 ,12 ); T1:= Ref(Mov( CLOSE ,CY1 ,S ),CY1/2)+ M1/100*(Mov( CLOSE ,CY1 ,S )); B1:= Ref(Mov( CLOSE ,CY1 ,S ),CY1/2)- M1/100*(Mov( CLOSE ,CY1 ,S )); T2:= Ref(Mov( CLOSE ,CY2 ,S ),CY2/2)+ M2/100*(Mov( CLOSE ,CY2 ,S )); B2:= Ref(Mov( CLOSE ,CY2 ,S ),CY2/2)- M2/100*(Mov( CLOSE ,CY2 ,S )); T3:= Ref(Mov( CLOSE ,CY3 ,S ),CY3/2)+ M3/100*(Mov( CLOSE ,CY3 ,S )); B3:= Ref(Mov( CLOSE ,CY3 ,S ),CY3/2)- M3/100*(Mov( CLOSE ,CY3 ,S )); T1;B1;T2;B2;T3;B3 I adjust the cycles to fit the security (more or less ) and then extend the cycle lines as trend lines (does that make sense? ) same as Doc's lines in the Anals. NB also adjust the cycle multiplier for a tighter fit. I then just use it to determine potential turning points. Where the long, medium and short cycles meet or get very close or where the extended lines cross I identify as potential entry point and then trade on price confirmation (eg price going breaking down thru trendline that is extention of cycle). Worked OK so far Let me know what you guys think! Link to comment Share on other sites More sharing options...
BudFox Posted January 10, 2005 Author Report Share Posted January 10, 2005 I added a time series smoother to your cycle envelopes. Let me know if this helps you or not. CY1:= Input("Short Cycle length" ,1 ,100000 ,10 )/2; CY2:= Input("Medium Cycle length" ,1 ,100000 ,80 )/2; CY3:= Input("Long Cycle length" ,1 ,100000 ,160 )/2; M1:= Input("Short Cycle Multiplyer" ,.01 ,1000 ,4); M2:= Input("Medium Cycle Multiplyer" ,.01 ,1000 ,8 ); M3:= Input("Long Cycle Multiplyer" ,.01 ,1000 ,12 ); T1:= tsf(Ref(Mov(CLOSE ,CY1 ,S),CY1/2)+ M1/100*(Mov(CLOSE ,CY1 ,S)),cy1); B1:= tsf(Ref(Mov(CLOSE ,CY1 ,S),CY1/2)- M1/100*(Mov(CLOSE ,CY1 ,S)),cy1); T2:= tsf(Ref(Mov(CLOSE ,CY2 ,S),CY2/2)+ M2/100*(Mov(CLOSE ,CY2 ,S)),cy2); B2:= tsf(Ref(Mov(CLOSE ,CY2 ,S),CY2/2)- M2/100*(Mov(CLOSE ,CY2 ,S)),cy2); T3:= tsf(Ref(Mov(CLOSE ,CY3 ,S),CY3/2)+ M3/100*(Mov(CLOSE ,CY3 ,S)),cy3); B3:= tsf(Ref(Mov(CLOSE ,CY3 ,S),CY3/2)- M3/100*(Mov(CLOSE ,CY3 ,S)),cy3); T1;B1;T2;B2;T3;B3 Link to comment Share on other sites More sharing options...
BudFox Posted January 11, 2005 Author Report Share Posted January 11, 2005 Here's the SPX with three important cycles and accompanying cycle envelopes. So why don't two concurrent down cycles give a proper sell signal in December? Good question! The answer is because there are more important cycles driving the SPX and other major indexes..... the 10-12 month cycle for example. I believe that these multiple cycle influences are the sum total of the innumerable trading strategies at work on the indexes, the futures, and the underlying individual stocks in the average. When there are more than three important cycles at work, as there are on the SPX and other major indexes, identifying buy and sell signals can be very difficult, if not impossible. This is where experience and artistic interpretation takes over. I still defer to Doc in these matters. My future studies will focus primarily on the cycles of individual issues. Link to comment Share on other sites More sharing options...
BudFox Posted January 11, 2005 Author Report Share Posted January 11, 2005 Housing to be taking an extended breather starting this week. TOL Link to comment Share on other sites More sharing options...
Janitor Posted January 11, 2005 Report Share Posted January 11, 2005 I added a time series smoother to your cycle envelopes. Let me know if this helps you or not. <{POST_SNAPBACK}> Hi Bud Yep, helps a lot. Thanks Is it possible to replicate the SPX, TOL charts you posted, in Metastock? What indicator can I use in Metastock to plot those cycles? Link to comment Share on other sites More sharing options...
BudFox Posted January 11, 2005 Author Report Share Posted January 11, 2005 Is it possible to replicate the SPX, TOL charts you posted, in Metastock? What indicator can I use in Metastock to plot those cycles? <{POST_SNAPBACK}> Those ARE Metastock charts. But I'm not ready to hand over my custom indicator just yet. I'll post the same SPX chart with some other indicators soon. Link to comment Share on other sites More sharing options...
BudFox Posted January 12, 2005 Author Report Share Posted January 12, 2005 SPX Five indicators above, ordered from top to bottom in responsiveness to the October 11-week cycle low: 1. Bud's custom Cycle Indicator 2. Projection Oscillator with Time Series Moving Average 3. Stochastic Oscillator with Simple Moving Average 4. Price Rate of Change with manual trendlines 5. Price Oscillator with Time Series Moving Average All of these indicators work well, though the stochastic oscillator gives many false readings. Keep in mind that these signals are meaningless out of the context of other dominant cycles. Link to comment Share on other sites More sharing options...
BudFox Posted January 12, 2005 Author Report Share Posted January 12, 2005 I am offering to create some cycle charts in my spare time for anyone who doesn't have Metastock. Here are the terms: 1. High volume individual stocks only. 2. The stock must be a good long or short prospect based on your own TA. 3. PM me with suggestions and I'll consider it, no promises. 4. Daily charts only, I can't do minute charts. 5. I'll only post charts that I think are useful. They'll be on this thread or in the new image gallery. I don't have time to scan for good setups, so I hope this effort will be mutually beneficial. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.