Trading Divergence

[sinner]

Re: Reversal Swing Convergence Divergence

Now those conditions are???

tech/a I kindly request you backtest the conditions as presented for mechanical "type 2" trades in Chapter 11 of the current Advanced GET Trading Manual, link posted earlier.

Your right it is raw and really doesnt find divergence.
The problem is you need to get around the Zig Zag indicator often needed to define peaks and troughs.Thats why Im guarded about the code.
I paid for it!

Why on earth ZigZag? Why not Tom DeMark supply/demand points (which are just "fractals" of various duration) if you want something precise, or ATR swings to account for vol?

The "twin peaks" divergence used in the Bill Williams system (a bit different again, uses inflection in AO to enter) is outlined down the very bottom here:
http://www.alpari.co.uk/en/alpari_academy/market_analysis/bill_chaos/ao.html
AO is pretty much identical to the AdvancedGET oscillator, but the "type 2" trade for GET does not mind if the oscillator crosses below 0 between the "peaks" whereas Bill Williams does. I notice frinks divergences above don't cross over the 50 line of RSI which is equivalent to 0 line of the AO or GET oscillator.

Has all your divergence backtesting been using ZigZag?

 

[tech/a]

Re: Reversal Swing Convergence Divergence

tech/a I kindly request you backtest the conditions as presented for mechanical "type 2" trades in Chapter 11 of the current Advanced GET Trading Manual, link posted earlier.



Why on earth ZigZag? Why not Tom DeMark supply/demand points (which are just "fractals" of various duration) if you want something precise, or ATR swings to account for vol?

The "twin peaks" divergence used in the Bill Williams system (a bit different again, uses inflection in AO to enter) is outlined down the very bottom here:
http://www.alpari.co.uk/en/alpari_academy/market_analysis/bill_chaos/ao.html
AO is pretty much identical to the AdvancedGET oscillator, but the "type 2" trade for GET does not mind if the oscillator crosses below 0 between the "peaks" whereas Bill Williams does. I notice frinks divergences above don't cross over the 50 line of RSI which is equivalent to 0 line of the AO or GET oscillator.

Has all your divergence backtesting been using ZigZag?

No but if yor using peaks and troughs of price or oscillator----Zig Zag cannot be used in a formula(Thats how metastock atleast identifies peaks and troughs in price or oscillator---with the Zig Zag formula) If you do you will have a false result in systems testing because the % swing is dynamic those identified in hindsite are now set in stone(Peaks and troughs)---those which are current(You find today) are dynamic and todays peak or trough may disappear in 2 or so days to show a very different peak or trough this is why it is SO difficult to trade in R/T This will become aparent when live traded---you have to code around it.


Type 2 trades from GET arent divergence trades.
They are possible end of wave 5 trades with stochastic as one of the filters.
Ive never coded nor tested the setup and would be interesting to do but it has no bearing on the divergence discussion.

Like all oscillators in your case Sinner (AO) they look fantastic when seen in hindisite particularly when shown on a chart with a well behaving trend its just an M/A.

Realtime they are a completely different animal.

Ill find 3 charts with divergence if anyone else would like to add theirs and lets see how they trade R/T.Use whatever filters you deem appropriate set your own stops if you think necessary and your own exit (Trade management).

 

[glenn_r]

Re: Reversal Swing Convergence Divergence

Nirvana's latest Volume Systems 2 has a swing divergence system that produces some back testing results.

http://www.nirvanasystems.com/oti/pdfs/volumesystems2.pdf

 

[eunza]

Re: Reversal Swing Convergence Divergence

Ok i've updated DUE and SGN - Will update DUE on Monday.

happy to lay my cards on the table here - Tech/A i've done plenty of back testing - lets see from the results as they go as although i'm not stating entry/exit points i'm still showing bullish/bearish on my posts (all bullish so far however will post the bearish when they show up.) I might start at least showing on the chart where i've made the 'call'

still happy to take the criticism but lets just see how it goes

 

[tech/a]

Re: Reversal Swing Convergence Divergence

Ok ii've updated DUE and SGN - Will update DUE on Monday.

happy to lay my cards on the table here - Tech/A i've done plenty of back testing - lets see from the results as they go as although i'm not stating entry/exit points i'm still showing bullish/bearish on my posts (all bullish so far however will post the bearish when they show up.) I might start at least showing on the chart where i've made the 'call'

still happy to take the criticism but lets just see how it goes

Great stuff.

Can you whack up a chart on both please.
I think it important to have an entry point otherwise all that's being shown is a possible setup.
Unless of course your saying that your divergences are a precursor to another entry triggered by some other condition?

I also think its important to know when the signal no longer valid---in other words the divergence is no longer applicable as a setup---IE its failed.
On the flip side when you would take profit.
Any stop?

Not hard to post up and then follow.

 

[sinner]

Re: Reversal Swing Convergence Divergence

No but if yor using peaks and troughs of price or oscillator----Zig Zag cannot be used in a formula(Thats how metastock atleast identifies peaks and troughs in price or oscillator---with the Zig Zag formula) If you do you will have a false result in systems testing because the % swing is dynamic those identified in hindsite are now set in stone(Peaks and troughs)---those which are current(You find today) are dynamic and todays peak or trough may disappear in 2 or so days to show a very different peak or trough this is why it is SO difficult to trade in R/T This will become aparent when live traded---you have to code around it.

I have no idea why you are even arguing this issue...not sure why anyone even remotely interested in the concept of a divergence would use ZigZag for anything related to it.

Type 2 trades from GET arent divergence trades.
They are possible end of wave 5 trades with stochastic as one of the filters.
Ive never coded nor tested the setup and would be interesting to do but it has no bearing on the divergence discussion.

Errrr what? Stochastic? The "Elliott Wave Oscillator" is an oscillator which represents the distance between 5SMA and 34SMA, very very similar to the formula for AO. The trading manual specifically states in chapter 7 titled "Elliott Wave Oscillator" that you can use this oscillator to identify wave 5, using divergence. It is included with the software. Specifically, pages titled 7~159 and 7~160 show how to interpret the divergence with waves labelled. The following 4 pages describe how to use the oscillator to interpret waves.

As for your statement saying type 2 trade has no bearing on divergence, please see the first rule of the type 2 trade:

"1. When the Wave 5 makes new highs, make sure the Elliott Oscillator shows divergence between the Wave 3 peak and the Wave 5 peak."

and the appropriate figure in the manual which shows the divergence in question. Bill Williams makes an almost identical observation in one of his books!

No idea what you meant by stochastic?

Like all oscillators in your case Sinner (AO) they look fantastic when seen in hindisite particularly when shown on a chart with a well behaving trend its just an M/A.

Err actually it's showing the difference between two MAs in the case of AO, which is not "my case", I posted it because you kept asking for a set of rules to trade divergences, they are right there. When to enter, when the signal is invalid, etc.

Did you actually read any of the links I posted so far?

Realtime they are a completely different animal.

Have traded plenty of these divergences realtime, with no issues on my part.

The above two quoted statements are frankly pretty condescending.

 

[It's Snake Pliskin]

Re: Reversal Swing Convergence Divergence

actually I thought it did a half decent job for something that got banged together in 20minutes

Every signal it spat out seemed ok. Maybe I just don't understand divergence properly, it's not really something I've looked at in great detail before.

Here are the last couple. Seems fine to me.

View attachment 38269

Thanks for that code though. Off to the races today so will most likely be under the weather tomorrow, and maybe monday, will try and come back to this later.

Hi Professor,

You have labeled the divergence on the price extremes. Are you using the close price or price extreme as the points of price for the divergence?

 

[tech/a]

Re: Reversal Swing Convergence Divergence

Sinner.

I certainly like the idea of the type 2 AGET setup being used.
The rules are very clear and I have GET so can find trades to monitor.
Enrty Exits and Stops are clear.

I presume you have GET as well so can check also how its going.

For those that dont here are the rules as Sinner pointed out and some charts showing a trade.(From Esignal)

I'm sure this exercise will clearly show the difficulties in trading Divergence and Elliott at the same time.
I cannot test this setup as I dont have the code for much of the Elliott component---but we can forward test (Albeit limited---as in sample size.)

Will look for prospects both long and short.(Unless there is a problem with this?)

I particularly like the 94% claim will keep an eye on that one.

 

[eunza]

Re: Reversal Swing Convergence Divergence

Great stuff.

Can you whack up a chart on both please.
I think it important to have an entry point otherwise all that's being shown is a possible setup.
Unless of course your saying that your divergences are a precursor to another entry triggered by some other condition?

I also think its important to know when the signal no longer valid---in other words the divergence is no longer applicable as a setup---IE its failed.
On the flip side when you would take profit.
Any stop?

Not hard to post up and then follow.

Shouldn't be an issue for adding approx entry / exit periods on the charts - (ie for my prior posts (in sgn, ogc etc) with charts you could consider the time of each thread posting would be the last traded price I guess. Was thinking probably will mark on a short term (15 min , 1hr or whatever) so at least will be trackable. Don't have a chance today however if time permits on Monday will update existing charts with the point in time that each prior post was made and then in future posts mark points on the latest live charts.

Although I do use price stops (sometimes up to 5%) my stops are generally based more on the levels / momentum of Indicators themselves and do need quite a level of micromanagement. This is all done manually instead of programmed charting S/W.

Also planning on showing failure of support lines / failures when they occur - Exits are usually only enforced on 2 / 4 hr or longer time periods (too much static on shorter periods and othewise will exit prematurely each time)- Although sometimes if the chart is looking very ominous short term i'll exit and look for another entry when conditions look to improve.

 

[BBand]

Re: Reversal Swing Convergence Divergence

Its possible to prove that black is in fact white!

Our esteemed politicians practice this all the time

All they do is be selective in collecting opposition statements, link them together in the appropriate manner - and viola, black becomes white
or - they can just ignore info/statements which does not suit their cause.

"Good" debaters can do this - when faced with this situation - all you can do is walk away or get all hot and bothered .......................

By the way, what was the opinion of Glenn's post #61. The silence has been deafening
Maybe - Oh, they are an OEM and as such you can't trust what they print, they are just trying to sell you something - so that goes into the "ignore basket"
The manufacturers opinion:
"The divergence concept is perhaps one of the most powerful reversal techniques discovered by Technical Analysis"

I like the e signal concept also - they do not trade the actual move created by the divergence, reference Techs last post, they are more concerned as to where the divergence occurs within the overall price action - must have another look at Elliot wave. Or maybe its just simple trendline trading - anyway there looks like two trades there.
One divergence and the other a trendline break

Sinner, I have not read the "notes" but I can see from Techs charts the potential
Also like the "parallel channel" - based on linear regression?, will check it out, looks like an improvement that I could incorporate into my trendline strategy (good filter for false breaks?)

Thanks Glenn and Sinner, I will definately check out your posts

Is there such a thing as being technically "right" ??????????????????

How is your bottom line
Peter










0



0
0

 

[tech/a]

Re: Reversal Swing Convergence Divergence

Ive read Glenns post and had a look at the indicators.
I think we can code them up and test them.

The manufacturers opinion:
"The divergence concept is perhaps one of the most powerful reversal techniques discovered by Technical Analysis"

There are many "claims" like this on analysis in general and few truly professional works to discuss these claims.

One serious researcher seems to be DR Bruce Vanstone.
Google him if interested.
He has some interesting work out there. Here are the prospects found By AGET today for a type 2 setup.

Thought Id post them up and leave it to you to chose those best prospects unless you wish to follow them all.
Should TYPE 2 (AGET) DIVERGENCE TRADES be in a separate thread??
That way the other divergence doesn't get tangled up.(RSCD)

Below also is one of the prospect ALL which has all boxes ticked.
The wave count is rubbish though but its the divergence though isnt it?

It hasnt triggered a buy yet.
(Not outside the Linear Regression Channel).

 

[BBand]

Re: Reversal Swing Convergence Divergence

Hi Tech,
That listing is just pure rubbish and demonstrates the typical politician approach

Most of the stocks just would not even be considered due to their lack of liquidity

and I would think that most people who use divergence, would look for them in a TRENDING market, with the view to entering a trade following the warning given by the formation of a price/indicator divergence
NOT IN A RANGE BOUND MARKET!!!!!!!!!!!!!

Most of the stocks in your list meet one of those conditions

This "debate" is meaningless

Bottom line
That listing is rubbish
You qualify as a politician

So I'll just walk away and do something meaningful and not waste my time

Peter

Why do you do this Tech? - you are far above this #%$* !!

 

[tech/a]

Re: Reversal Swing Convergence Divergence

Hi Tech,
That listing is just pure rubbish and demonstrates the typical politician approach

Most of the stocks just would not even be considered due to their lack of liquidity

and I would think that most people who use divergence, would look for them in a TRENDING market, with the view to entering a trade following the warning given by the formation of a price/indicator divergence
NOT IN A RANGE BOUND MARKET!!!!!!!!!!!!!

Most of the stocks in your list meet one of those conditions

This "debate" is meaningless

Bottom line
That listing is rubbish
You qualify as a politician

So I'll just walk away and do something meaningful and not waste my time

Peter

Why do you do this Tech? - you are far above this #%$* !!

What the??

Its a search by Advanced Get for type 2 divergence trades.
The software found them not me.
Have I missed something here?
I thought it was agreed that the Type 2 trades by AGET would be excellent prospects.
Tell me what filters you want----I'm just using those set by the principals E signals Advanced Get software say to use??

I just put the search results up for transparency.
I don't like the setups either that's why I said make a choice---or 2 ALL looks best hence I put it up.
If no body likes anything then wait till next scan then choose.
No smoke and mirrors.

 

[tech/a]

Re: Reversal Swing Convergence Divergence

Thanks to Boggo for pointing out I hadnt updated my charts
ALL has triggered a buy on Friday.
Second chart is easier to see
CLICK to EXPAND

 

[sinner]

Re: Reversal Swing Convergence Divergence

Sinner.

I certainly like the idea of the type 2 AGET setup being used.
The rules are very clear and I have GET so can find trades to monitor.
Enrty Exits and Stops are clear.

I presume you have GET as well so can check also how its going.

For those that dont here are the rules as Sinner pointed out and some charts showing a trade.(From Esignal)

Hi tech, I have no idea why you keep using a stochastic oscillator for your divergence?

Here are some examples, from the past, that I spotted in advance warning of ending of wave 5 impending top/bottom. All of the wave counting was done using the AO. You can check the dates on the posts to see they were called in advance.

View attachment 38313
http://www.forexfactory.com/showpost.php?p=3166879&postcount=154805
View attachment 38314
https://www.aussiestockforums.com/forums/showpost.php?p=560473&postcount=7661
View attachment 38315
http://www.forexfactory.com/showpost.php?p=3669657&postcount=5990

If you were to trade the above divergence signals as the Bill Williams "twin peaks" signal, you would place a sell stop below the low of the bar where momentum goes less than previous bar. I included some crude arrows on the EURUSD 1.5 top. Not to say I took those specific trades, but you can see the divergence can and does indicate market tops followed by at least some reversal (see below) in many cases.

In my experience, a momentum divergence will take the price to test the wave 4 low in a A-B-C type correction followed by a re-test of the high. Euro was an exception so far and led to re-test (and break) of the low.

I believe in the GET trading manual, wave 4 start is the target for a type 2 trade.

In the case of ALL, I would have had a buy stop at 3.5 (1c above the Aug 5 high) with SL at 3.32 if I was trading it as a twin peaks, which I wouldn't because it doesn't really look like a wave 5 low to me using the AO.
View attachment 38316

Sorry to mix up EW in all this.

I particularly like the 94% claim will keep an eye on that one.

I like it too. Like I mentioned already, Bill Williams made identical observations about the AO in his book through his own backtesting. He gives examples which he traded live, including one case where he was on CNBC or Bloomberg or something called a top in the USDCHF against 3 or 4 other analysts and ended up being right. Considering the oscillator formulas are almost identical there is probably something to the idea.

 

[tech/a]

Re: Reversal Swing Convergence Divergence

I have no idea why you keep using a stochastic oscillator for your divergence?

Pretty easy really.
Its the one used by AGET in their type 2 trades.
Happy to use whatever you like.
But thought you'd suggested the type 2 trade.

Must be missing something once again?

Please feel free to suggest any chart in future be it a Stock or Forex or Futures chart so we can observe the out perfomance of this trading method.
My selections just dont seem to cut the mustard.---I'll refrain then.
Look forward to watching things unfold in Realtime rather than hindsite.

So type 2 trades are out then?

 

[sinner]

Re: Reversal Swing Convergence Divergence

Pretty easy really.
Its the one used by AGET in their type 2 trades.
Happy to use whatever you like.
But thought you'd suggested the type 2 trade.

Must be missing something once again?

On the website, current trading manual uses the Elliott Wave Oscillator. If you go to the website and look at chapter 7 and chapter 11 you will see them using this oscillator. I have no idea how you can say GET uses sto for their type 2 trades when the example screenshot you posted is using the Elliott Wave Oscillator! The rules that you reposted, state plainly "make sure the Elliott Wave Oscillator shows divergence with its wave 3 peak..."

Here is some information on that oscillator:

http://www.tradingfives.com/articles/elliott_oscillator.htm

My selections just dont seem to cut the mustard.---I'll refrain then.
Look forward to watching things unfold in Realtime rather than hindsite.

What? I posted those screenshots to show how I personally interpret the divergence, that is all! Because you seem to have trouble understanding either the "twin peaks" or type 2 divergence I wanted to show some examples from my own experience.

The type 2 trade rules plainly state there must be a pullback divergence with the wave 3 peak of the Elliott Wave Oscillator, you use a STO to show a divergence that barely has a pullback at all on the proper oscillator which is supposed to be used and now you're going to take your ball and go home because I pointed out it doesn't fit the trade parameters of either the divergence I like to use or the one you are supposed to be backtesting?

 

[It's Snake Pliskin]

Re: Reversal Swing Convergence Divergence

Pretty easy really.
My selections just dont seem to cut the mustard.---I'll refrain then.
Look forward to watching things unfold in Realtime rather than hindsite.

With respect to the hindsight aspect, any description or informing of a strategy may need a hindsight chart to help with the transfer of information. With respect to a live trade the foresight comes from the knowledge we have gained. Trading knowledge aren't we? (more practical than theoretical)

Considering you don't trade divergence setups why not let yourself learn something from others who are keen to discuss and give some realtime examples?

 

[tech/a]

Re: Reversal Swing Convergence Divergence

Snake
I dont trade divergence as I have tested it over 1000s of portfolios and it doesnt deliver an edge. I'm attempting to show others how and why divergence to trade is a waste of time in the long term.

This meets a whole heap of flack as "everyone Knows" trading divergence is one of the most powerful tools in technical analysis.Peoples beliefes are being challenged and they dont like it!

People are getting all tied up in th TYPE of ocsillator used.
Doesnt matter----
ALL oscillators are derived from OPEN/CLOSE/HIGH/LOW/RANGE/VOLUME or OPEN INTEREST.

Sinner seems to think altering the oscillator will make a vast difference.
You can see the difference in the chart attached.(Well maybe you can but I cant!) ---perhaps a few are learning something??

So we have one demo in progress from Eunza and nothing else as my example was evidently doomed to failure.(Admit the return to zero condition is yet to be met)

Happy for Sinner to pop up some AO divergence setups to forward trade them.

Sinner how'd the races go?

 

[eunza]

Re: Reversal Swing Convergence Divergence

Ok - So i've placed a few posts on the RSCD analysis on SGN, OGC and DUE for the last week or so.

As a way of approximately tracking when and where the hypothetical entries / holds or exits are i've added vertical lines for BUY (green), HOLD (Blue) and will use red for any possible exits. (none as yet)

Any discrepencies on the charts with times etc let me know.

Again these entry and exit Posts are PURELY HYPOTHETICIAL for the purpose of the RSCD discussion.