OpenFinDesk: Algorithmic Trading based on Technical Analysis

An Opensource framework to trade Forex and CFD using technical analysis in Java

This is an opensource Java framework to implement any technical trading strategy in less than a minute by using openFinDesk annotations. OpenFinDesk has many modules. Current module which is based on technical analysis is opensource.

pros

• loose coupling and Java SOLID principles are considered as much as possible.
• can be scaled horizontally and vertically
• using influxdb,csv or any other store you like.
• complete backtesting before running in live mode
• a module to use Hidden Markov Model
• customized log levels
• testing java classes
• separate risk management
• separate money management
• separate position management
• using any approach such as Fibonacci, indicators or a combination of many strategies

cons

• It is only technical analysis.
• It is not based on a reactive architecture like RxJava.
• refactoring is not perfect.

Example1

@OpenFinDeskChanceNode(containerId=5,nodeId=0
        ,openfindeskStrategies={CamarillaStrategy.class, ADXBasedTrendDetection.class}
        ,timeFrames={"PERIOD_D1","PERIOD_H1"})
public class CheckIfCamarillaAndADX extends ChanceNode {
    final static Logger logger= Logger.getLogger(CheckIfCamarillaAndADX.class);

    @Override
    public void executeCurrentChanceNode() {
        this.openFinDeskOrder.setSymbol(featureVector.get(0).getSymbol());
        this.openFinDeskOrder.setPosition("*");
        this.openFinDeskOrder.setOrderType("noOrder");
        this.openFinDeskOrder.setStrategyName("no signal from container 5");
        this.openFinDeskOrder.setAction("doNothingAndWait");
        this.openFinDeskOrder.setVolume("0.01");
        logger.info("featureVector.get(0).getPosition()="+featureVector.get(0).getPosition());
        logger.info("featureVector.get(1).getPosition()="+featureVector.get(1).getPosition());

        if ((featureVector.get(0).getPosition().equals("buy")) && (featureVector.get(1).getPosition().equals("buy"))) {
            this.openFinDeskOrder.setOrderType("openBuy");
            this.openFinDeskOrder.setAction("buyUsingFirstQuantizedLevel");
            this.openFinDeskOrder.setStrategyName("Container5");
        }else if ((featureVector.get(0).getPosition().equals("sell")) && (featureVector.get(1).getPosition().equals("sell"))) {
            this.openFinDeskOrder.setOrderType("openSell");
            this.openFinDeskOrder.setAction("sellUsingFirstQuantizedLevel");
            this.openFinDeskOrder.setStrategyName("Container5");
        }else {
            this.openFinDeskOrder.setOrderType("noOrder");
            this.openFinDeskOrder.setAction("doNothingAndWait");
        }
        decisionIsMade=true;

    }
}
This informs the software to use CamarillaStrategy in daily time frame and ADX in H1 timeframe

Example2

@OpenFinDeskChanceNode(containerId=1,nodeId=0
,openfindeskStrategies={PinBarStrategy.class,TrendStrengthStrategyBasedOnInternalRetracementSequences.class}
,timeFrames={"PERIOD_H1","PERIOD_D1"},disabled=true)
public class CheckIfWeHavePinBarCandlestickChanceNode extends ChanceNode {
    final static Logger logger= Logger.getLogger(CheckIfWeHavePinBarCandlestickChanceNode.class);

    @Override
    public void executeCurrentChanceNode() {
        this.openFinDeskOrder.setSymbol(featureVector.get(0).getSymbol());
        this.openFinDeskOrder.setPosition("*");
        this.openFinDeskOrder.setOrderType("noOrder");
        this.openFinDeskOrder.setStrategyName("no signal from container 1");
        this.openFinDeskOrder.setAction("doNothingAndWait");
        this.openFinDeskOrder.setVolume("0.01");
        logger.info("featureVector.get(0).getPosition()="+featureVector.get(0).getPosition());
        logger.info("featureVector.get(1).getPosition()="+featureVector.get(1).getPosition());
        if ((featureVector.get(0).getPosition().equals("bearishPinBar"))&&(featureVector.get(1).getPosition().equals("sell"))){
            this.openFinDeskOrder.setOrderType("openSell");
            this.openFinDeskOrder.setAction("sellUsingScalping");
            this.openFinDeskOrder.setStrategyName("container1");
        }else if((featureVector.get(0).getPosition().equals("bullishPinBar"))&&(featureVector.get(1).getPosition().equals("buy"))){
            this.openFinDeskOrder.setOrderType("openBuy");
            this.openFinDeskOrder.setAction("buyUsingScalping");
            this.openFinDeskOrder.setStrategyName("container1");
        }
        decisionIsMade=true;

    }
}
This informs the software to use PinBarStrategy in H1 time frame and TrendStrengthStrategyBasedOnInternalRetracementSequences in Daily timeframe

of course you can also use more complex strategies since the infrastructure is based on datastructure of decision tree and you can create any decision tree.

v8.dll

since i created a dll using visual C++, i used a release version since if i put debug version,then i need a Visual C++ Redistributable on each computer. If you dont create a release version for your dll then MQL5 keeps saying: can not open dll even if you have allowed MQL5 to do so.

#mql5 This java infrastructure connects to mql5 via sockets. Mql5 uses your dll file. I created v8.dll based on my C++ codes.

order

separate by comma:

7 parameters: order,symbol,orderType,executionType,tp,sl,volume

• order : "order"
• symbol: "EURUSD"
• orderType:"ORDER_TYPE_BUY"
• executionType:"instant" or "pending"
• tp:"1.11034"
• sl:"1.10226"
• volume:"0.01"