Hi, what do you mean by "brokers took their profits?".
Maybe a silly question but I'd like to know.
Profit Protector from CodersGuru
As I told in my previous post, spikes can kill your profit! It's the news or simply the brokers are taking their profits. So maybe this EA needs some kind of "Profit Protector". There's another thread from CodersGuru (great coder!) where he is talking about such a Profit Protector. I'm not a programmer, but maybe HolyGuy or Nich can tell if this is usefull for this EA.
I've once read an article about this, but can't find it anymore (still searching). In short: Brokers (alone or together) know when they have to sell or buy so stoplosses of their clients are hit and they can take their profits. It was a very interesting article! Gives you a different look at forex!! As soon as I found this article I will post it here.
I would read this with pleasure. could u send adress of codersguru profit_protector thread.
We could also stop trading friday at noon. Few hours lost but those bad birds singing "we do not trade against our customers!" would have beak in the water.
Any possibility that the profit protector could be added into this EA.
The code can be found HERE.
The non-farm payroll definately ate into my profits because I was on the wrong side of the trade when the news came out and hit a number of stops.
I don't think it is the fault of the EA just something that happens. We need to keep testing, the EA does it's job it is just that NEWS can be a wildcard in the whole thing.
However, big news does not come out that often so just consider it a cost of trading in this market.
By the way, I tested out for the first time the Weekend Close feature and it works perfectly. Thanks Nich.
Please Read This Post - All Pro-Gen Testers
There are so many man hours going into this EA and we all see it has potential. Yes, we need to work smarter - not harder. Here is what I propose. If Nich is willing, we could have a group of 10 or more of us each send him $10.00 to $20.00 via paypal. Tradestation is about $200.00 for a month. I've used it. You can cancel before your renewal with no obligation. We could give Nich a month with the platform and the code and see what some "real" backtests tell us. Tradestation is the best tool for backtesting. Period. Of course, Nich will have to be interested in doing this. I only pick him because he has been the programmer on the project and doing a great job. We could give a 3 day window for all paypal $$ to be sent to him and at the end of three days, if he has enough for a month plus some $$ for his time then great. If not enough for one month then we'll extend the paypal timeframe.
If you think about all the time we're collectively investing on this and that we have inadequate tools to draw any real conclusions then I think you will agree that $10.00 or so is better then the road we're going down.
After another week of testing, I agree their needs some sort of plan, and with the thousands of $$$ of junk now on my PC, 20 seams very small price to contribute to see if "this is the one"
My question is, will the expert developed on TS give similar result on MT4? and if so will it stay in the open?, I personally would like it to do so.
Just a FYI, I have paid TS programmers in the past to develop bots using TS, with not so good results, looked great in TS backtester, just flopped forward. They just could be lousy coders though.
I do, however, believe its worth a shot. so where do we go from here?
I will be posting the results soon to the new link as requested, you can see the now at the url posted in the past.
Quick question: Were's the link to post results? Found: post 397 https://www.mql5.com/en/forum/173795/page27
Quick Link to Results https://www.mql5.com/en/forum/general
That is the question
Theoretically, if the data is clean and you understand how to optimize the testing to simulate reality; with the right tools it should give accurate results. TS is a great platform. MT4 is a novelty at best. You get what you pay for. It could be in your situation that the data was not the greatest or the coding was bad or maybe just the settings of the backtester were not reflective of actual market conditions. The data quality is imperative. Like programmers say.."GIGO" (garbage in, garbage out). Gaps in the data will make all efforts useless. There has to be a way to streamline the process of seeing whether or not this EA could actually pull a consistent profit. Months of foward tests on demo accounts, I get the impression, may prove fruitless.
Just one more thing
I'm not getting negative about Pro-Gen in general. It makes some great trades and I think there has been some promising showing with the EUR/USD pair. My ideas are about how we could go about optimizing it to actually trade a number of pairs consistently and simultaneously. If it can work on the EUR/USD then it can work on other pairs. Unfortunately, I'm no programmer or I would volunteer instead of attempting to appoint.
Hi all folks!
Wait before paying for TS! My tick backtester will be functionnal and easy to make it run with mysql, Postgre, SQL_server or Access.... (if you know how to code ODBC connnections or I will explain you).
We only need good tick data and I am ok to pay for this and share. My boss is actually paying me doing coding for the community (I will be fired in 3 months so I don't care and him also (I told him)).
At home I have Sql server an tools joined for data analysis and mining.
Once I have reliable results with those tools I will give you free tools to do the same at home by your own way (maybe running linux but as all knows m$ sucks). I will always share my results (I would not have known PG without you).
Just Coming back from a great party, excuse my english but it's hard to hit keyboard right. Sure I won't code this night but you will have a PG backtester soon (I hope monday morning, I only work on it).
Also, I'll post forward statements tomorrow (holyguy7 test list).
Please enable the necessary setting in your browser, otherwise you will not be able to log in.