StockFetcher Forums · Filter Exchange · Anyone Do Fractal Analysis Filters?<< 1 2 >>Post Follow-up
yepher
353 posts
msg #31910
Ignore yepher
4/28/2004 4:12:46 PM

Has anyone tried to write a filter that looks for fractal type occurrences? An example of what I mean by this is that the yearly chart shows a double bottom and the monthly chart shows a double bottom and the daily charts show a double bottom (or any other pattern for that matter).

If you have such experience would you mind sharing a filter that demonstrates how to achieve such a result. My point to this question is I am looking for a way to identify when the Long term, Mid Term, and Short terms cycles all peak or trough at the same time which should be a periodic occurrence.




travlr
80 posts
msg #31922
Ignore travlr
4/28/2004 10:35:26 PM

Hey yepher,

Yeah fractals and coinsiding cycles/trends is good stuff, IMO. I do my work with these on more of a weekly,daily,Intraday scale. But for this I use Wealth Lab, do the lack of intra,and etc. on Fetcher. EWZuber mentions alot about using multi time frames and coinciding stochastic cycles.

BTW, I've found it limiting trying to scan for chart patterns. You'd have to use price(percentage) zones, rather than price levels, etc. Though, I believe it is do able to be relatively close to identifing certain patterns, I haven't yet dedicated enough of my time to it. Have fun.

travlr


wildmap
7 posts
msg #31978
Ignore wildmap
5/4/2004 8:17:35 PM

How about EMAs? Long term could be 500 and shorter term 2, and any variant in between any length you can imagine, and the merging of long, medium, and short term should be easy to read.


TheRumpledOne
6,362 posts
msg #38586
Ignore TheRumpledOne
10/27/2005 4:13:33 AM

Fetcher[
/* Fractal Filter */

set{cp8, close * .8}
set{cp88, cp8 * cp8}
set{voltest, cp88 * 1000}

set{HIbTest, count(high 4 days ago below high 2 days ago, 1) * count(high 3 day ago below high 2 days ago, 1)}

set{HIaTest, count(high 2 days ago above high 1 days ago, 1) * count(high 2 days ago above high, 1)}

set{fPeak, HIbTest + HIaTest}

set{fFrac, count(fPeak above 1, 100)}

add column HIbTest
add column HIaTest
add column fPeak

add column fFrac

fFrac above 0


high 10 day high equal high 100 day high

avg volume(22) above voltest


add column cp8
add column cp88

add column avg volume(22)
add column avg voltest

close above 1
/* END FILTER */
]



I got the idea from Scientific Approach to Trading.

MAY ALL YOUR FILLS BE COMPLETE.


alf44
2,025 posts
msg #38589
Ignore alf44
10/27/2005 11:54:13 AM

wildmap has a great idea with the EMAs...

...in fact, Daryl Guppy has created an "indicator" based on Multiple Moving Averages. He uses a group of short-term exponential moving averages AND a group of long-term exponential moving averages as a sort of "early warning" of a possible Trend Reversal.

The short-term EMAs are: 3,5,8,10,12 & 15...

The long-term EMAs are 30,35,40,45,50 & 60...

Guppy says...

"In both time frames - defined by the short-term group of averages and defined by the long-term group of averages - the moving averages converge in the same tight window. The implication is clear. At major turning points, both groups of averages converge and cross almost simultaneously. The multiple moving average indicator does not lag weeks behind the action but generates a consistently tradable signal. We see a Fractal Repetition across different time frames !"

----------------------

When you put these averages on a chart and start looking at them for a while it IS pretty clear "visually" that when all these time frames converge at once...something BIG usually happens.

They are really pretty cool !


Regards,

alf44






alf44
2,025 posts
msg #38590
Ignore alf44
10/27/2005 7:00:12 PM

..another thought...

...when these "fractal repetition" points happen...and these multiple moving averages converge, as the Trend begins to change...the EMAs begin to "rollover."

The sequence goes from: 3,5,8,10,12,15...30,35,40,45,50,60 in a DOWN TREND !

To a sequence of: 60,50,45,40,35,30...15,12,10,8,5,3 in an UP TREND !

As this "rollover" of these EMAs takes place it is visually similar to a pair of ribbons being twisted. In fact, there IS another study called a "Ribbon Study" that looks (at a glance) a lot like this Guppy Multiple Moving Average indicator. They ARE somewhat different however.


Regards,

alf44




Railwhore
69 posts
msg #38593
Ignore Railwhore
10/27/2005 9:43:04 PM

This is pretty cool - Not really a filter but gives a graphical picture of what you guys are saying:

Fetcher[draw ema(3) draw ema(5) draw ema(8) draw ema(10) draw ema(12) draw ema(15)
draw ema(30) draw ema(35) draw ema(40) draw ema(45) draw ema(50) draw ema(60)
Average Volume(90) above 100000
Volume above 150000
Close above 12
not otcbb
]



Next will be to set up the ratio calculations


alf44
2,025 posts
msg #38594
Ignore alf44
10/27/2005 10:17:14 PM

THAT'S WHAT I'M TALKING ABOUT !!! lol

"Visually"...these Multiple Moving Averages (MMAs) are pretty cool !

The interaction of the two groups of EMAs really DO illustrate...

...the turning points...

...the pull-backs...

...the areas of consolidation !

Thanks, railwhore for posting that !


Regards,

alf44




Railwhore
69 posts
msg #38595
Ignore Railwhore
10/27/2005 10:52:06 PM

Ok here is a start of a filter for converging emas - if anyone wants to pick up on it go for it. Its basically a bunch of ema ratios and I confined them as you can see and tweak. Set a tighter range for an interesting list. Not sure what to make of this yet.

Fetcher[/* Set up the ratios */
set{r3.60a, ema(3) - ema(60)}
set{r3.60b, r3.60a / ema(60)}
set{r3.60, r3.60b * 100}
set{r5.50a, ema(5) - ema(50)}
set{r5.50b, r5.50a / ema(50)}
set{r5.50, r5.50b * 100}
set{r8.45a, ema(8) - ema(45)}
set{r8.45b, r8.45a / ema(45)}
set{r8.45, r8.45b * 100}
set{r10.40a, ema(10) - ema(40)}
set{r10.40b, r10.40a / ema(40)}
set{r10.40, r10.40b * 100}
set{r12.35a, ema(12) - ema(35)}
set{r12.35b, r12.35a / ema(35)}
set{r12.35, r12.35b * 100}
set{r15.30a, ema(15) - ema(30)}
set{r15.30b, r15.30a / ema(30)}
set{r15.30, r15.30b * 100}

/* Positive numbers mean short-term emas above long term */
/* Lets find smallest ratios to find potential reversals */
r3.60 < 0.60
r3.60 > -0.49
r5.50 < 2.00
r5.50 > -2.00
r8.45 < 2.10
r8.45 > -2.10
r10.40 < 2.10
r10.40 > -2.10
r12.35 < 2.00
r12.35 > -2.00
r15.30 < 1.50
r15.30 > -1.50

draw ema(3) draw ema(5) draw ema(8) draw ema(10) draw ema(12) draw ema(15)
draw ema(30) draw ema(35) draw ema(40) draw ema(45) draw ema(50) draw ema(60)
add column r3.60
add column r5.50
add column r8.45
add column r10.40
add column r12.35
add column r15.30
Average Volume(90) above 100000
Volume above 150000
Close above 12
not otcbb
]




Railwhore
69 posts
msg #38596
Ignore Railwhore
10/27/2005 10:55:36 PM

I suppose you could alter the above filter to find ratios outside the range to find the stocks that have had MAJOR moves up OR down and are ready to start converging again back. Not really a bearish or bullish filter but a nice filter to show stocks that COULD move from the tight ranges to the large ranges and vice-versa.


StockFetcher Forums · Filter Exchange · Anyone Do Fractal Analysis Filters?<< 1 2 >>Post Follow-up

*** Disclaimer *** StockFetcher.com does not endorse or suggest any of the securities which are returned in any of the searches or filters. They are provided purely for informational and research purposes. StockFetcher.com does not recommend particular securities. StockFetcher.com, Vestyl Software, L.L.C. and involved content providers shall not be liable for any errors or delays in the content, or for any actions taken based on the content.


Copyright 2016 - Vestyl Software L.L.C.Terms of Service | License | Questions or comments? Contact Us
EOD Data sources: DDFPlus & CSI Data Quotes delayed during active market hours. Delay times are at least 15 mins for NASDAQ, 20 mins for NYSE and Amex. Delayed intraday data provided by DDFPlus