My EA results & offering my skills 'free'

New comment
 

Hi,


I've been reading you for a long time and I've finally decided to post the results of my almost Beta version of my EA. So far it's working on EUR/CHF and GBP/JPY. On the other currencies it fails to be profitable but I know why (which is good) but haven't found a way to fix it yet (which is bad) :-DD


I make small profit p/a but will be running on several currency pairs at the time. Anyway, here they go:

-This is EUR/CHF 1D since 2004 (what I got from Alpari 1M files). I can also run it since 2000 i.e. and the pict is just the same but ending higher :-D

This guy does very little orders but I got 100% longs right and 90% shorts right.



-This is GBP/JPY. Same thing here +/- but there are no Alpari files for this pair so modelling quality is worse. Otherwise performance would have been better I would say:



Anyway, thing looks promising to take your girlfriend for a nice diner every now and again but no too impresive. I can improve it a lot. To start with I need to change it so that it works on other time frames and see how it performs there. Some tests show promising results.


The only problem, which is part of the design though, is that I have no SL and drawdowns are big. My SL strategy would be to get 50% out the account on every profitable trade but now I see the API doesn't support that.


Any ideas on how to model a funds withdrawal from the account?


Also, I've been a soft. developer for 12 years and I work for a bank. If you come up with and interesting trading strategy I will code for you 'free' (no cash) but for the right to use it for my own. I have very little free time I usually invest in my EA so only serious proposals and manually tested strategies please.


Thanks in advance.

 
The only problem, which is part of the design though, is that I have no SL and drawdowns are big. My SL strategy would be to get 50% out the account on every profitable trade but now I see the API doesn't support that.



Your sort of where I was a week or so ago I think, I call it the pursuit of accuracy syndrome :). I can get 98+ ish with large stop losses and small take profits but they are so big they kill most of the profits.


In my newest EA I am putting in some more risk management code to help moderate the losses vs gains


Any ideas on how to model a funds withdrawal from the account?



I store the relative balance I want to save in a global variable and subtract that from the account balance which gives me the usable balance. If Usable Balance <= 0 the program doesn't run ( I usually make it 0+ my maximum stoploss so it never goes below 0)

Though this approach requires you to do one of three things:

1, Manually initialize the current saved balance, at the correct level, each time you restart the program

2, Manually withdraw it from the account

3, Have it as a percentage (variable or fixed)

 

TLB

> EUR/CHF and GBP/JPY

Thats a very unusual pair of pairs - same system works on both?

-BB-

 

Hi,


thanks for your answers. I hadn't updated as nobody replied so I thought it wasn't good enough :-DD


I did some good progress and now I can get to around 130k on EUR/CHF H4, plus a bit more in D1 for the 2 pairs which gives me a good return for a period of 4 or 5 years: Around 40 to 50% in total. Very reliable but drawdawn goes from 30 to 50%.


Now I need to put some real MM into it, defend my profits better and get a better exit for trends. As for the last bit I just joined, as the developer, a work group and I already have done some advances in a Taichi EA. I think this Taichi (or Ichimoku which was my inital idea) will solve my headaches. Then the sky is the limit :-DDD if it works...


For MM try to cash in profits by partially closing positions on Fibonacci retraces. Not perfect but it's a good start.


I'll try to update the post, but now I am quite busy with the Taichi.


Cheers

 

Small update on my progress:


2006 to 2009:



2004 to 2009:


 

Nice Job TirantLoBlanc!!!,

Have you tried with live account? Which was your initial balance?

Could you share the code ;)

I'm initiating with EA programming, nut my experience with C and PHP let me work easily, pitifully my statistical knowledge remains in the past, lol...

Anyway, keep it doing the job, finally some day we could obtain good profit and retire...

 
TirantLoBlanc wrote >>

Hi,

thanks for your answers. I hadn't updated as nobody replied so I thought it wasn't good enough :-DD

I did some good progress and now I can get to around 130k on EUR/CHF H4, plus a bit more in D1 for the 2 pairs which gives me a good return for a period of 4 or 5 years: Around 40 to 50% in total. Very reliable but drawdawn goes from 30 to 50%.

Now I need to put some real MM into it, defend my profits better and get a better exit for trends. As for the last bit I just joined, as the developer, a work group and I already have done some advances in a Taichi EA. I think this Taichi (or Ichimoku which was my inital idea) will solve my headaches. Then the sky is the limit :-DDD if it works...

For MM try to cash in profits by partially closing positions on Fibonacci retraces. Not perfect but it's a good start.

I'll try to update the post, but now I am quite busy with the Taichi.

Cheers

It has been a couple of months since you decided to put MM into it. If you have good results, please post on www.mt4live.com, which shows your trades in real time. I'm very interested in this EA.

BTW, is there any good progress with Taichi EA? Any thread about it in this forum?

Thanks.

New comment