Skip to content

Latest commit

 

History

History
88 lines (61 loc) · 7.21 KB

README.md

File metadata and controls

88 lines (61 loc) · 7.21 KB
Raw

IBKR Facade

Interactive Brokers API wrapper and trading toolbox

This project is built around the Interactive Broker's TWS library for Java. If you have a TWS or IB Gateway running, you can use this as a market data source for time series analysis, REST gateway to the market, or even as a trading bot framework.

The IBKR facade based on Spring Boot so can leverage the capabilities provided by the Spring Framework as well which can be really useful for implementing automated trading strategies.

In this project you'll find:

  • The basic functions of the TWS API exposed to a Rest API
    • Searching in Interactive Brokers' instrument master data
    • Fetching the option chain
    • Subscribe to market data for certain instruments, and store the time series data in Redis
    • Create and manage orders
    • Retrieve open positions
  • Market data streamed into a Redis server from where you can do it whatever you want (eg.: save if for further analytics; generate OHLC data automatically with built-in compaction policy; using as a data source by creating a pub-sub message queue; build historical datasets, etc.)
  • Built-in sample trading strategy (not a real working strategy, just an example of course), which is based on periodical time series analysis looking for trading signal and placing orders.

Prerequisites

As step zero, you need an account at Interactive Brokers.

TWS or IB gateway

Trading Workstation (TWS) or IB Gateway has to be installed in order to provide access to Interactive Brokers' infrastructure. If you are using TWS, make sure that the API connection is enabled, and if you want to placing orders through the library, the connection is not restricted to "read-only".

See more: TWS API initial setup

Market data subscription

In order to access market data you'll need an account at Interactive Brokers with subscription to those markets and instruments you need. If you want to use a paper trading account first (which I highly recommend) instead of your live account, you need to share your market subscriptions between your accounts.

See: Sharing market data subscription

TWS API

You have to download the TWS API from https://interactivebrokers.github.io for Java. It contains a TwsApi.jar file which you need to have on your classpath. (You can find a working version of it in the /lib folder of this repository.) This is a legacy JAR maintained by Interactive Brokers and you cannot find it in the central Maven repository, so if you want to build this project with Maven, you need to manually add it to your local artifact repository (see pom.xml dependencies section), or make it available on the classpath some other way.

Redis server

Redis is a powerful and extremely fast tool for storing market data. If you have a Redis server with RedisTimeSeries extension up and running, you can configure its access through the application.properties file, then the library will send the market data to the Redis for the subscribed instruments.

If you don't want to install Redis, you can try Redis Cloud for free.

How to use

  1. Checkout the project
  2. Add the lib/TwsApi.jar to the classpath from your IDE or alternatively publish it to a maven repository. Don't forget to modify pom.xml if needed!
  3. Check your application.properties for:
    • TWS or IB gateway host and port number
    • URL settings for REST API
    • Redis server connection
  4. Build the project with maven
  5. If you have your setup up and running, IBKR Facade should connect to Interactive Brokers automatically.

You will be able efficiently use the software only if you are familiar with the terminology and concepts used by Interactive Brokers, especially Contracts and conid as the unique IB contract identifier.

See: Search Interactive Brokers Contract Database

Use cases

REST API

Every basic function needed for using the system is exposed through a REST API, such as:

  • Contract search
  • Subscription to the market feed for certain contracts
  • Getting price information
  • Placing orders
  • Getting positions

You can change the port number (8082 by default) and context-path (default is /ibkr) from the application.properties file.

The endpoints are documented using Swagger OpenAPI annotations. After you started the application you should find the documentation under the following URL: http://localhost:8082/ibkr/swagger-ui/

Market data analysis (Redis TimeSeries)

If you have your Redis ready you can subscribe to market data feed of any instrument available on Interactive Brokers through your brokerage account.

From Java, you can use the ContractManagerService.subscribe() method or via HTTP you can use the /subscribe endpoint. If everything works as expected, the Contract itself should be saved to Redis under the conid of the Contract as key, and two time series should be created for storing the price information.

If you want to change or extend the functionality of this, you need to extend the TimeSeriesHandler class. You can utilize the full power of Redis from calculating OHLC data automatically, or using it as a pub/sub service, it's all up to you.

Once you have subscribed to an instrument, the price stream (bid/ask changes) will be written into the Redis by the following pattern:

  • A ContractHolder will be created and saved by it's conid used as a key
  • The ContractHolder contains the Contract descriptor provided by Interactive Brokers
  • The ContractHolder has a streamRequestId which is used for identifying the communication "stream" with the TWS itself. When you send a request to Interactive Brokers through TWS with a unique number (the IB documentation refers to this identifier as streamId or tickId), this number will be used in the response method calls. We save this number into the ContractHolder as streamRequestId.
  • The time series will be available under the following key format: stream:[streamRequestId]:[BID/ASK] so basically two series are stored per subscribed instrument: one for the bids and one for the asks.

See: RedisTimeSeries

Trading strategy automation

Since you have hands-on market data, with the methods of OrderManagerService, ContractManagerService and PositionManagerService there is no predetermined way how to implement a trading strategy, the only limit is your imagination :)

You can find an example implementation in the strategy package, which periodically checks the prices of Apple stock from Redis looking for a trading signal. Once the trade performed (you have an open position) it checks the price movements for a possible exit.

Important: This is not a real strategy. Don't even think about using it real conditions!