CCFp weekly signals (set and forget basket trading) | Page 74 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

BASKET TRADE Management: Using multiple trade management EA's on basket trades 11 replies

Basket Knights - A Basket Trading Round Table 1,313 replies

Updated CCFp and CFP indis for relative strength trading 55 replies

TMS basket: Dashboard for a basket of MA-Filtered RSI Signals 135 replies

Set it & Forget it GBP and CHF signals... 6 replies

Attachments: CCFp weekly signals (set and forget basket trading)
Exit Attachments

CCFp weekly signals (set and forget basket trading)

  • #1,461
  • Mar 13, 2017 11:51am Mar 13, 2017 11:51am
Quoting scalprobot
{quote} Thanks for responding, TGS. I have no idea why nobody else responded. People prefer to spend lots of time on doing things by hand? It's very time consuming, I think. The simulation system was actually simulating the whole basket. All automated and every thinkable option would be easy to test (compare that to running backtests by Excel..). Regarding your last point, MT5, yes I'm also surprised why nobody is running the tests in MT5. You can do multipair (basket) backtests with MT5, I think it's pretty much ideal for the trading system described...
Thank you ScalpRobot.

I am so glad that someone was able to convert the EA/Indicator to function on MT5.

I assume you would be able to run "Optimizations" for start time (converted to GMT) and stop time. This would be very helpful, and greatly appreciated if you could. Unfortunately, this is the only type of optimization that could be tested with MT5 and the EA as written.

The only draw back to the EA, and therefore current backtesting, is the fixed lot feature. With that feature, the Risk/Reward of various basket sizes cannot be dynamic in nature. What I mean by that is that it has been proposed, and some of use are utilizing, that the amount of the SL be based upon a set number of pips per pair open. For example if the number of pips per pair is set to 25 when there are 2 pairs open the SL would be 50 pips, when there are 8 pairs open the SL would be 200 pips, and like this week when there are 16 pairs open (which is the max) the SL would be 400 pips. To equalize risk the lot size for 16 pairs might be .1 lots, while the lot size for 8 pairs would be .2 lots, and the lot size for 2 pairs would be .8 lots.

The Excel simulations that are being worked on right now, will be able to optimize for start time, end time, SL and TP based on all the parameters including dynamic lot sizing without having to change the EA. Until that is completed, I would very much like to look at some of the optimization that the MT5 is capable of performing.

GREEN pips to you Good Sir.
  • #1,462
  • Mar 13, 2017 11:57am Mar 13, 2017 11:57am
Quoting Forexpinbar
{quote} I had similar .
In my first analysis I just checked which between the SL and TP got hit first without considering any type of trade management (trailing, BE locking). Now I did the same thing for what the table is concerned, but at the same time for the losing weeks I decided to monitor the results of the basket and see what's the max number of positive pips reached.
The goal is to define the best threshold before locking the BE, considering that this will affect the winning weeks too. That line refers to the max positive pips reached by the basket before going in negative territory.
  • #1,463
  • Mar 13, 2017 12:07pm Mar 13, 2017 12:07pm
Quoting SmoothTrader
{quote} To equalize risk the lot size for 16 pairs might be .1 lots, while the lot size for 8 pairs would be .2 lots, and the lot size for 2 pairs would be .8 lots. The Excel simulations that are being worked on right now, will be able to optimize for start time, end time, SL and TP based on all the parameters including dynamic lot sizing without having to change the EA.
^^^^ very much agree^^^^
  • #1,464
  • Edited 12:38pm Mar 13, 2017 12:26pm | Edited 12:38pm
Scalprobot, how are you able to get high quality tick data into MT5? I always thought this was a significant limitation of MT5, but I heard that metaquotes was looking to fix this. In the US we can now get a free MT5 demo from AMPfutures, but back testing and optimization seems largely worthless without quality ticks. As I understand it, MT5's tester creates test points per bar, like MT4, which severely limits back testing in MT5. Moreover, MT5 does not allow one to import quality ticks like MT4 does (major bummer), but maybe I missed an update.
  • #1,465
  • Mar 13, 2017 1:21pm Mar 13, 2017 1:21pm
Quoting Monty13
{quote} Xmess7. Thank you for your help. the V31, has solved the problem, and now I have all the trades opened, I also have switched the debugging on, for any issues to report to you. Once again, I appreciate your brilliant work.
Monty:
When u think that you are set and have no more problems I recommend switching it off. Otherwise you may end up with fairly big logfiles
l
Our AWESOME IMAGINATION is the true holy grail!
  • #1,466
  • Mar 13, 2017 1:31pm Mar 13, 2017 1:31pm
Quoting xmess7
{quote} Monty: When u think that you are set and have no more problems I recommend switching it off. Otherwise you may end up with fairly big logfiles l

Hi XMESS7 , could it be my main lagging problem with my pc ?? never had lagging issue before and if yes , what is the solution for this , can we erase some files and witch ones??
  • #1,467
  • Mar 13, 2017 1:50pm Mar 13, 2017 1:50pm
Quoting agagukalex
{quote} Hi XMESS7 , could it be my main lagging problem with my pc ?? never had lagging issue before and if yes , what is the solution for this , can we erase some files and witch ones??
The function that Xmess7 is talking about is the <PrintDebugInfo> function on the inputs of the EA. While that is on, it will print out a log file of many of the functions of the EA on a continual basis. Most of us have that set to "false" so those log file will not be printing out.

As Xmess7 mentioned, he is not looking at this thread often throughout the day. The best way to ask him a question is to PM him with your question.

GREEN pips everyone.
  • #1,468
  • Mar 13, 2017 2:10pm Mar 13, 2017 2:10pm
Quoting scalprobot
{quote} Thanks for responding, TGS. I have no idea why nobody else responded. People prefer to spend lots of time on doing things by hand? It's very time consuming, I think. The simulation system was actually simulating the whole basket. All automated and every thinkable option would be easy to test (compare that to running backtests by Excel..). Regarding your last point, MT5, yes I'm also surprised why nobody is running the tests in MT5. You can do multipair (basket) backtests with MT5, I think it's pretty much ideal for the trading system described...
Thanks again scalprobot for providing these usable results. I think I understand and want to mention something about the lack or response to your post that you mentioned. Obviously you're a capable person and you're able to produce results that helps others in their research whether on this thread or elsewhere so I'm hoping you don't feel too discouraged and continue to share your results because some of us are intently listening. It's not easy for some to see value in something if they don't understand it and this thread does have some newbies to trading in general and, the thread is really active so posts come streaming in quickly, sometimes burying something that doesn't initially appear to be shiny.

Again, I hope you'll continue posting your findings because, some of us are intently listening.

I learned now that MT5 is capable of multiple pair back testing, thanks for letting it be known. I'm well versed in "high-quality" tick data testing on MT4 but I don't know how the data is structured for MT5 and whether it is in tick form or open prices or something else.

1. Do you think MT5 back testing is at least as reliable as MT4 tick data testing?
2. Would you run your original simulator on MT4 to compare consistency between MT4 and MT5 and share those results?
3. Am I correct to gather from your result with those particular settings that the basket was not profitable for that period?

Assuming your FX Pro MT5 is on GMT+0 so the settings for the basket were opening an hour after London open on Monday and closing it 13 hours before market close on Friday.
  • #1,469
  • Mar 13, 2017 3:07pm Mar 13, 2017 3:07pm
Quoting TGS
{quote} Thanks again scalprobot for providing these usable results. I think I understand and want to mention something about the lack or response to your post that you mentioned. Obviously you're a capable person and you're able to produce results that helps others in their research whether on this thread or elsewhere so I'm hoping you don't feel too discouraged and continue to share your results because some of us are intently listening. It's not easy for some to see value in something if they don't understand it and this thread does have some...
Actually when I saw this topic I thought that things could be done much more efficient and that's when I posted my idea of running tests virtually in MT4. So I wrote that algorithm to run those tests, but just a very basic version to give people here an idea and to pick it up from there. Yes, I would say it's reliable. I build in some data tests to be sure that all data is synchronised correctly.

However I do think that also that was rather a poor solution compared to using MT5. To run such tests, MT5 is much more user friendly.

Another thing, signals can differ from broker to broker (or rather from feed to feed).

I'm not planning on using a lot of time for this, I've got my hands full with other things.

About the previous test I posted, yes that turned out to give a negative result. Mind that it was a very basic version of the system: opening basket Monday 8AM, closing basket Friday 8PM. If I'm correct FXPro is GMT +2 (2 hours ahead of GMT).

If this was a project I would spend some serious time on, I would probably look at the way the "strenghts" are being calculated. I think that with a little different approach a lot of improvement can be made.

I've made some minor adjustments to the settings (MA periods and "trade level") and then the results look like this:
Attached Image (click to enlarge)
Click to Enlarge Name: P3.png Size: 79 KB
  • #1,470
  • Mar 13, 2017 3:19pm Mar 13, 2017 3:19pm
Quoting scalprobot
{quote} Actually when I saw this topic I thought that things could be done much more efficient and that's when I posted my idea of running tests virtually in MT4. So I wrote that algorithm to run those tests, but just a very basic version to give people here an idea and to pick it up from there. Yes, I would say it's reliable. I build in some data tests to be sure that all data is synchronised correctly. However I do think that also that was rather a poor solution compared to using MT5. To run such tests, MT5 is much more user friendly. Another...
Excellent, and Thank You again.

Do you have the mq5 or ex5 files for the indicator and EA so that we can use them to begin our own analysis?

And you mentioned small changes to produce these improved results, if I may be so bold, what were your new parameters?

Thank you again in advance for your reply.

GREEN pips to you Good Sir.
  • #1,471
  • Mar 13, 2017 3:46pm Mar 13, 2017 3:46pm
Quoting nonlinear
{quote} I should have been clearer. If mt4 shuts down and is restarted (and the expert is restarted too, via auto boot or a similar app - all of my terminals are on vps), will the expert continue to manage the previously opened basket positions? Thanks!
Yes it will, that is the reason for the global variables
When the EA is restarted it looks at those global variables (which are still there) and records them so that it knows exactly in what state it was before it wnet off line. Hec u can even transport the global varialbes to another machine using thr same account and the EA would pick up from where it left off on tha tmachine too. The only drawback isthe u have to transfer the global variables manually.
Our AWESOME IMAGINATION is the true holy grail!
  • #1,472
  • Mar 13, 2017 4:08pm Mar 13, 2017 4:08pm
SmoothT, did a little digging and found this, which looks like what scalprobot was using for the indy. He seems like a veteran. Wish I was a better coder.
Attached File(s)
File Type: txt ccfp_w_rules.mq5.txt   42 KB | 603 downloads
  • #1,473
  • Mar 13, 2017 4:20pm Mar 13, 2017 4:20pm
Quoting nonlinear
SmoothT, did a little digging and found this, which looks like what scalprobot was using for the indy. He seems like a veteran. Wish I was a better coder. {file}
Unfortunately, I know nothing about the mq5 language, and only enough about the mq4 to get me into trouble.

With that being said, I am putting it out there for anyone that knows how to convert this indicator and EA to run on the MT5 platform to help us out. This will give us the ability to do some much needed backtesting that we can't do on the MT4 platform.

GREEN pips everyone.
  • #1,474
  • Mar 13, 2017 4:46pm Mar 13, 2017 4:46pm
Quoting xmess7
{quote} Monty: When u think that you are set and have no more problems I recommend switching it off. Otherwise you may end up with fairly big logfiles l
Thanks Xmess7.
  • #1,475
  • Mar 13, 2017 5:01pm Mar 13, 2017 5:01pm
============I`m not sure anybody needs this anymore============
Hello
these are M1 charts for 2016.
-all are based on 1 minute open prices so we can see the lowest DD and highest point of the basket more clearly.
-all data is from Dukascopy.com.
-you already have all signals for 2016 in a zip file which I had uploaded earlier. also you know how I have collected them
-I have seen some differences between these M1 charts and H1 charts which I had sent them 2 weeks ago...and I couldn`t find
out why !!? so please check them maybe you can help on this and let me know to fix the chart. also if you found any other
mistakes inform me please to correct them.
-in some charts you can see a kind of jump in pips...and I think something is missed or wrong in there.
Attached File(s)
File Type: zip 2016 CHARTS.zip   6.5 MB | 521 downloads
  • #1,476
  • Mar 13, 2017 5:13pm Mar 13, 2017 5:13pm
Quoting Tiivaa
============I`m not sure anybody needs this anymore============ Hello these are M1 charts for 2016. -all are based on 1 minute open prices so we can see the lowest DD and highest point of the basket more clearly. -all data is from Dukascopy.com. -you already have all signals for 2016 in a zip file which I had uploaded earlier. also you know how I have collected them -I have seen some differences between these M1 charts and H1 charts which I had sent them 2 weeks ago...and I couldn`t find out why !!? so please check them maybe you can help on this...
Thank you Tiivaa for all of your work.

These charts contain a great deal of valuable information, and will take some time to visually go through them to unlock some of their secrets (SL, TP, ProfitLock, Trail Levels, etc). I look forward to analyzing it.

GREEN pips come to those that study.
  • #1,477
  • Mar 13, 2017 5:30pm Mar 13, 2017 5:30pm
Interesting numbers Tiivaa. Thx
  • #1,478
  • Mar 13, 2017 7:24pm Mar 13, 2017 7:24pm
Thank you Tiivaa
  • #1,479
  • Mar 13, 2017 8:59pm Mar 13, 2017 8:59pm
Could anyone clarify how the profit lock and trailing functions work? If one sets proft lock at 500, with a 200 trail, what happens at 500 pips in profit? If profit hits 520 pips and then retraces to to 500, what happens? Does it close out? Or is 320 actually locked due to the 200 trail? Thanks!
  • #1,480
  • Mar 13, 2017 9:05pm Mar 13, 2017 9:05pm
Quoting nonlinear
Could anyone clarify how the profit lock and trailing functions work? If one sets proft lock at 500, with a 200 trail, what happens at 500 pips in profit? If profit hits 520 pips and then retraces to to 500, what happens? Does it close out? Or is 320 actually locked due to the 200 trail? Thanks!
Profit lock is when the first lock sets in (at what pips). So if you set profit lock at 500. Once basket hits 500 pips the TP lock will be 250 pips. If you set the trailing to 200 then at 700 pips profit for basket the TP lock will be 350. The lock will always be half of the profit lock. the trailing lock is how many more pips for the next lock.
49 traders viewing now