Skip to content
C# Library for Metatrader 4
Branch: master
Clone or download
Type Name Latest commit message Commit time
Failed to load latest commit information.
MQL4CSharp fix bool cast issue with latest metatrader May 18, 2017
docs fix locking issue. handle exception properly in ontick/oninit/ontimer Feb 13, 2016
.gitattributes add sln file Jan 20, 2016
.gitignore add sln file Jan 20, 2016
MQL4CSharp.sln add sln file Jan 20, 2016 Update Feb 17, 2016


##C# Library for Metatrader 4

For the Java version (using JNI) go here

This library is currently a work in progress. It demonstrates a framework for calling MQL4 commands from within a C# class.

When a command from a C# Strategy is called, it sets a flag which the MQL expert is polling every millisecond within it's onTimer() function. As the C# Strategy code is called in it's own thread, it will not block the MQL code, and the MQL function can execute while the C# function waits for the result. When the MQL expert detects a command waiting, parses it, executes it and sets the result and error code. The C# base strategy polls for the result every millisecond, so as soon as the result is set by MQL, the C# function will then return with the result.

In initial tests, 1 millisecond polling does not show any performance impacts to the system. 1000 bool tests every second for a 4GHz CPU isn't much theoretically. So performance here should be pretty much the same as running the code directly in MQL. This will be configurable.

The basic workflow =>

I mainly started this project as a way to get decent backtesting performance in MT4 whilst using a mainstream language. I picked C# due to simplicity. shows as example strategy using the framework.

To implement a strategy, you just extend the Strategy class and implement the abstract methods.

Methods like:

  • getStopLoss
  • getTakeProfit
  • getEntryPrice
  • getExpiry
  • getLotSize
  • getComment
  • getMagicNumber
  • manageOpenTrades

Also, abstract type should be extended for specific logic to keep things somewhat separate and reusable:

  • BaseSignal
  • BaseStopLoss
  • BaseTakeProfit
  • BaseFilter
  • BaseRiskProfile

You can also just extend the base MQL type which just contains the metatrader functionality, if you don't want to use the Strategy framework above.

I'm using log4net for C# logging, and SmartThreadPool for concurrency.


I've tested and it works nicely. Currently you need to run Metatrader as administrator if you want to attach to an outside interface. Loopback works for unpriveleged user. By default you can connect to http://localhost:1234/ Check it out in action:

While I'm still building, there's not much in the way of Docs, but you can checkout for the full API implementation. It supports the same as the MQLBase class.

There's 2 ways of building REST URIs, either use the global one when you don't care which chart the command will run on, or you can specify the chartid in the URI http://localhost:1234/{chartid}/ (for objectcreate, etc)

I'll also add a websockets implementation possibly using to provide onTick events such as Rates unpdates And also events for Order Open/Modify/Close operations.

Both REST and Websocket implementations will be able to leveraged from Custom Strategies you create. So if you want to tell your frontend that something has happened in the terminal, then use the events api, and if you want to drive commands from your frontend you can make REST requests to the terminal.

Installation Notes:

PM> Install-Package MQL4CSharp

Check here for instructions how to quickly get up and running using nuget:

Some Backtesting notes

Backtesting now works by executing commands in the onTick() method (as well as onTimer()) as onTimer is not supported in Strategy Tester. The onTick method in MQL will now continue to process commands until it detects the IsExecutingOnTick() to be false, so that all commands are run within the same tick. This could probably do with some further optimisation.

Design Notes:

I've put together a rough class diagram here. It shows what I plan to do to provide a REST api into MQL and Custom Strategies. I've found that debugging is impossible with the latest Metatrader (just crashes as soon as you attach), so I've decided that providing a REST interface and ability to JSONify data within the system, should help with debugging purposes. Also, it will be pretty neat to be able to write some JS fountend using this API.

You can’t perform that action at this time.