3674 replies to this topic

[HDB]

HDB..your system used in context to day-trading IG chart has HUGE potential on scalp side. that cci signal depending on time frame being used could be the coup d'etat for day traders

U reckon I should patent it?

[delta66]

HDB..your system used in context to day-trading IG chart has HUGE potential on scalp side. that cci signal depending on time frame being used could be the coup d'etat for day traders

[HDB]

Hi,

 

I've been using prorealtime simple creation to try and do some back testing, and I just cannot get it to give me good results when using MA's or BB's.

 

Eventually I used trailing stops and limits instead of trying to find a sell signal that worked consistently.

 

The results are pretty good, when using a limit of 125 and a trailing of 250! But I will be the first to admit I might be doing something wrong and need to do more testing.

This also means that if the market dips, you will miss out on signals until it hits the stop or limit, you can change the back test to take every signal and have multiple positions open and then the results are astounding, but your draw down can get pretty big.

 

Something I found by accident was when I tried to test short signals, I entered them in the wrong place so it was looking for short signals and going long instead of short. Surprisingly this ended up giving really good results. So it would seem at first glance that when this system gives you a short signal, you have more chance of making money going long than short!(based on my trail limit method) Also when it gives a short signal it would seem its the end of the down move in an uptrending market.

Actually doing the test correctly got negative results for me.

 

Has anyone found a good consistent exit signal?

 

So far to me it seems the best signal for exits is discretionary based on experience.

 

I shall carry on testing and any advice would be appreciated.

Gringots u got me here..Im actually new to IG and Aksi having only traded ALsi for last 6 months so bear with me :

 

A few questions first

 

1. plz explain limit of 125 and trailing 250

 

2. Whats pro real time? is it part of IG and if so is it 24 hrs trading?

 

3. is your exercise/backtest relating to exit signals only? ie you in a trade and don't know when to get out?

 

Excuse the ignorance andwith  all due respect to your feedback, but if you look at my rules, advise etc, I explained it like I was teaching a 10 year old child ( As per Arshad instructions from old PSG forum). Lots newbies following this thread so plz tell us in laymans language whats happening.

Edited by HDB, 23 May 2014 - 01:32 PM.

[delta66]

wanted to go long on cci/ma crossover(43940+-) but felt reward thin when looking at days top.  maybe i'll grab short again later on..watching for now  

[delta66]

stopped out 44930

[delta66]

moved stop 49965..following manually 18p stop level

[delta66]

stops set at entry

[gringots]

Hi,

 

I've been using prorealtime simple creation to try and do some back testing, and I just cannot get it to give me good results when using MA's or BB's.

 

Eventually I used trailing stops and limits instead of trying to find a sell signal that worked consistently.

 

The results are pretty good, when using a limit of 125 and a trailing of 250! But I will be the first to admit I might be doing something wrong and need to do more testing.

This also means that if the market dips, you will miss out on signals until it hits the stop or limit, you can change the back test to take every signal and have multiple positions open and then the results are astounding, but your draw down can get pretty big.

 

Something I found by accident was when I tried to test short signals, I entered them in the wrong place so it was looking for short signals and going long instead of short. Surprisingly this ended up giving really good results. So it would seem at first glance that when this system gives you a short signal, you have more chance of making money going long than short!(based on my trail limit method) Also when it gives a short signal it would seem its the end of the down move in an uptrending market.

Actually doing the test correctly got negative results for me.

 

Has anyone found a good consistent exit signal?

 

So far to me it seems the best signal for exits is discretionary based on experience.

 

I shall carry on testing and any advice would be appreciated.

 

 

 

 

[HDB]

divergences all over the cci!!  hold on those shorts guys!!

[delta66]

not even a pause through R1..straight through.  lol, as soon as i close it'll turn down. got to be some form of profit taking at these levels to bring it down a little   

[Qbre77]

always sweat when i short this market, current trade no diff.  will look to bail on 50p gains

i am exactly the same, as soon as i go short INSTANT stress!!!!!

[delta66]

always sweat when i short this market, current trade no diff.  will look to bail on 50p gains

[Qbre77]

i also feel it will now head back to pivot

[delta66]

short 44995 contra to CCI but gut feeling

[Qbre77]

closed at R1

[HDB]

CCI divergence...expect a pullback!!

[HDB]

Working on backtesting the system for CFD trades, looking at opening Long positions only at this stage. (Will look into Shorts if I find a workable system for Longs).

 

Using daily prices (O,H,L,C), SMA 3/10 cross and 14 CCI above 0 cross does not produce that many trading signals on all the shares I have tried it against (some 25 different shares), with most signaling less than 12 possible entry points over a 6 year period. This is mainly due to there being a day or two (occasionally more days) difference between the two crosses (SMA vs CCI). By increasing the CCI to a 15 period cycle the possible entry points doubled. I have also run a test using 15 CCI rising where previous CCI is less than 0 (i.e. not just looking for the 0 cross, but only looking at it rising up whilst still negative). This increased the number of potential entries even further. However, most of the shares tested against these variables produced around a 50/50 profit/loss ratio (using a risk/reward ratio of 1:2)

 

Over this weekend I will spend a bit more time working with different variables to see what the possible results would be. Will update the thread in due course.

Ignore the ma's first...Just use the CCI....also for day trading , use cci 25. 30 right up to 50!!

 

Trial and error will get something going.. Once you got the CCI right, then play with ma's to see which ma cross give the signals close to zero!!

 

Appreciate the help and trial test..Well done!!

[delta66]

will wait for opp to get back in, had my entry been closer to pivot(bought at yesterdays action) ..would of kept my longs open

[HDB]

sharks 16  blues 6

 

and im stopped out 50 p loss!!

 

Well done Delta and Qbre..

 

Lovely long candle taking it through 0..

 

Im sure most of your guys caught it!!

[Aragorn]

Aragon appreciate it if you could post your results on the cfd side...Much appreciated!!

Working on backtesting the system for CFD trades, looking at opening Long positions only at this stage. (Will look into Shorts if I find a workable system for Longs).

 

Using daily prices (O,H,L,C), SMA 3/10 cross and 14 CCI above 0 cross does not produce that many trading signals on all the shares I have tried it against (some 25 different shares), with most signaling less than 12 possible entry points over a 6 year period. This is mainly due to there being a day or two (occasionally more days) difference between the two crosses (SMA vs CCI). By increasing the CCI to a 15 period cycle the possible entry points doubled. I have also run a test using 15 CCI rising where previous CCI is less than 0 (i.e. not just looking for the 0 cross, but only looking at it rising up whilst still negative). This increased the number of potential entries even further. However, most of the shares tested against these variables produced around a 50/50 profit/loss ratio (using a risk/reward ratio of 1:2)

 

Over this weekend I will spend a bit more time working with different variables to see what the possible results would be. Will update the thread in due course.