Help needed for a critical decision!

 

Hi everyone,

I started learning forex few months ago. And, I chose MT5 / MQL5 to start developing my first EA for my own trading strategy. I have found out that there is no way to import historical and tick data into MT5 :-) Also there is no way to modify this data! OOP, better performance... etc. will not help me when I see my account is wiped out because my EA has never been tested under more realistic conditions with real tick data. Therefore, I am thinking to switch to MT4, and here are my 2 questiosns:

1. What do you think? switching to MT4 is the right or wrong decision?

2. In MT4 History Center I can import historical data. But where can I get (free or affordable) quality historical data (brokers usually provide poor historical data, I believe).

Thanks in advance!

 
Fx18:


2. In MT4 History Center I can import historical data. But where can I get (free or affordable) quality historical data (brokers usually provide poor historical data, I believe).

There is no central market for Forex . . . where else will you get data other than a Broker ? or a Bank maybe, are you happy to pay for your data ?
 
Fx18:

Hi everyone,

I started learning forex few months ago. And, I chose MT5 / MQL5 to start developing my first EA for my own trading strategy. I have found out that there is no way to import historical and tick data into MT5 :-) Also there is no way to modify this data! OOP, better performance... etc. will not help me when I see my account is wiped out because my EA has never been tested under more realistic conditions with real tick data. Therefore, I am thinking to switch to MT4, and here are my 2 questiosns:

1. What do you think? switching to MT4 is the right or wrong decision?

2. In MT4 History Center I can import historical data. But where can I get (free or affordable) quality historical data (brokers usually provide poor historical data, I believe).

Thanks in advance!



search before you ask, MT4 backtester makes very basic (untrue) assumptions which will affect your results a lot if you really need tick level data

Brokers don't usually provide, tick data. It's the exception when they do.

 
RaptorUK:
There is no central market for Forex . . . where else will you get data other than a Broker ? or a Bank maybe, are you happy to pay for your data ?


Not really happy to pay for historical data :-) What about you? do you rely solely on your broker(s) historical data? complete data with no gaps? would this data be sufficient for testing your EAs?

 
Fx18:


Not really happy to pay for historical data :-) What about you? do you rely solely on your broker(s) historical data? complete data with no gaps? would this data be sufficient for testing your EAs?


Tick data is specific to your broker, and they way they create prices. If your strategy is sensitive to tick data, it might be thrown off by data from another broker.

I'm happy to pay for data if it's what's needed for profit. But so far I've downloaded free and written scripts to check its quality

 
alladir:


Tick data is specific to your broker, and they way they create prices. If your strategy is sensitive to tick data, it might be thrown off by data from another broker.

I'm happy to pay for data if it's what's needed for profit. But so far I've downloaded free and written scripts to check its quality


1. Do you mean it is safer to develop a strategy that uses open/close prices and not every tick?

2. Which free scripts did you use for checking data quality?

 
Fx18:


1. Do you mean it is safer to develop a strategy that uses open/close prices and not every tick?

2. Which free scripts did you use for checking data quality?





1) That's a huge question. Personally, I think anything that only looks at arbitrarily chosen prices (i.e. Open/Close) is flawed to some extent, especially now computing power is so huge. However, the fact that enough people use it means there might be some benefit to giving them extra attention, and for long-term strategies, the flaws quickly diminish to almost nothing.

2) I don't use other people's scripts when it comes to forex. I wrote my own to have confidence in the results. I'd upload them but they're very specific to my needs and wouldn't really work for other people's data

 
alladir:


1) That's a huge question. Personally, I think anything that only looks at arbitrarily chosen prices (i.e. Open/Close) is flawed to some extent, especially now computing power is so huge. However, the fact that enough people use it means there might be some benefit to giving them extra attention, and for long-term strategies, the flaws quickly diminish to almost nothing.

2) I don't use other people's scripts when it comes to forex. I wrote my own to have confidence in the results. I'd upload them but they're very specific to my needs and wouldn't really work for other people's data


Thanks for your feedback!
 
If you have the time read this thread.
 
Fx18:


Not really happy to pay for historical data :-) What about you? do you rely solely on your broker(s) historical data? complete data with no gaps? would this data be sufficient for testing your EAs?

It depends on how you plan to test and evaluate the result of your test and also on the Strategy being tested. I have used M1 data for debugging type testing and for performance testing, I've also used tick data for performance testing. Use google and search for tick data sources you will find some useful info.
 

Hi FX18,


I used to use http://www.forextester.com/data/datasources.html to download free data, but it's only up to date to the last month and once imported i have to convert to all time frames. There is a script that will that automatically for you in code base section.

Recently i've been playing with this http://www.tickstory.com/, which is also free and can produce 99% back tests. Just follow their tutorials.

Good luck.

Reason: