TraderX is a Sample Trading Application, designed to be a distributed reference application in the financial services domain which can serve as a starting point for experimentation with various techniques and other open source projects. It is designed to be simple and accessible to developers of all backgrounds, with minimal pre-assumptions, and it can serve as a starting point for educational and experimentation purposes.
It is designed to be runnable from any developer workstation with minimal assumptions other than Node, Java and Python runtimes. The libraries and toolkits it uses are meant to be as vanilla as possible, to preserve its approachability by developers of all levels.
It contains Java, NodeJS, Python, .NET components that communicate over REST APIs and messaging systems and are able to showcase a wide range of technical challenges to solve.
More detailed information about this project can be found in the website which is generated
from the code under the docs
directory of this project.
Learn more about the project - including a brief demo, in the Keynote Demo session that was presented at the Open Source in Finance Forum 2023
The application has been modified so that it starts with a seed of only Buy
trades so that we can open stock positions. The naive rationale is that you should not be able to sell securities that you don't have. Therefore if you do not have an open position with given security - you cannot sell it - and if you do have the security - you can only sell as much as you currently have. After all the securities have been sold - a position is closed.
Trades and Positions are additionally saved to XTDB so that we can show a history of changes and consecutive trades. We introduced Prices (as discussed above - randomly generated rather than taken from some Exchange feed - for demonstration running locally they are a good enough approximation of 'live market').
The project consists of multiple moving parts, and you can see how things hang together by reviewing the architecture and sequence diagrams located in the docs directory.
Component | Tech Stack | Description |
---|---|---|
docs | markdown | Architecture and Flow Diagrams are here! |
database | java/h2 | A simple self-contained SQL database |
reference-service | clojure | REST service (off a flat file) for querying ticker symbols, providing stock prices and also storing trades and positions in XTDB in order to provide history for Reports |
trade-feed | node/socketio | Message bus used for trade flows, as well as streaming to the GUI |
people-service | .Net core | Service for looking up users, for account mangement |
account-service | java/spring | Service for querying and validating accounts |
position-service | java/spring | Position service for looking up positions and trades by the blotter |
trade-service | java/spring | Service for submitting trade/order requests for further processing |
trade-processor | java/spring | Trade Feed consumer which processes trade/orders |
web-front-end | html/angular or react | Interactive UI for executing trades and viewing blotter. Note: the AngularJS GUI was an initial contribution and contains account management capabilities. The React GUI was contributed during a hack day and may not work for managing accounts, but it does work for executing trades and viewing the blotter |
xtdb | bitemporal database | open-source immutable database with comprehensive time-travel. XTDB has been built to simplify application development and address complex data compliance requirements. XTDB can be used via SQL and XTQL |
This is installed locally through normal git clone operations.
In order to get things working together, it is recommended to select a range of ports to provide all running processes with, so that the pieces can interconnect as needed. To run this all up 'by hand' here are default ports which are used, and you can easily export these variables to your favorite shell.
export DATABASE_TCP_PORT=18082
export DATABASE_PG_PORT=18083
export DATABASE_WEB_PORT=18084
export REFERENCE_SERVICE_PORT=18085
export TRADE_FEED_PORT=18086
export ACCOUNT_SERVICE_PORT=18088
export PEOPLE_SERVICE_PORT=18089
export POSITION_SERVICE_PORT=18090
export TRADE_PROCESSOR_SERVICE_PORT=18091
export TRADING_SERVICE_PORT=18092
export WEB_SERVICE_ANGULAR_PORT=18093 #Angular
export WEB_SERVICE_REACT_PORT=18094 #React
The recommended starting sequence to let everything find what it needs is:
database
xtdb
reference-service
trade-feed
people-service
account-service
position-service
trade-processor
trade-service
web-front-end
The easiest way to run up the entire system is using Docker Compose. This should work on your local computer using Docker Desktop / Docker Compose (tested on Mac Silicon) and also in Github Codespaces.
If using Github Codespaces it is recommended you select an 8-core type machine with 32GB RAM to ensure all the components have the required resources to start.
To do this
- Select the Green Code menu at the top of this page
- Select the Codespace tab then click the three dots '...' and select 'New with options...'.
- Change the machine type to '8-core' and click 'Create codespace'
As of writing, personal Github accounts receive 120 free core hours per month for using Codespaces, see the most recent details here
Once you have cloned the repository locally or once your Codespace has started, from the root traderX directory run
docker compose up
On first run this will build all of the containers from the project specific Dockerfile's and then start them in the correct sequence.
The Docker containers are configured via Docker Compose to connect to a shred virtual network enabling them to communciate whether running on your local computer or via a Codespace.
Once everything has started the WebUI will be accessible at http://localhost:8080 (even if using a codespace, the localhost URL will be mapped through from your local browser to the Codespace).
When building locally in your company, if you are using a corporate artifact repository, you might need to override certain settings such as mavenCentral() in gradle, for the Java projects.
In order to do this, we have designated a .gitignore
'd folder where you can leave company-specific build scripts. This folder is not managed by git and can be modified locally.
Create a .corp
directory and in there you can create a settings.gradle
file which will allow you to build all gradle projects
# in the traderX main directory
mkdir .corp
touch settings.gradle
The settings.gradle
file should contain any overrides on your repositories
and plugins
block but should also contain these contents:
rootProject.name = 'finos-traderX'
includeFlat 'database'
includeFlat 'account-service'
includeFlat 'position-service'
includeFlat 'trade-service'
includeFlat 'trade-processor'
This will include projects in directories at the same level as the .corp directory.
You can also store a separate gradle wrapper here, if you need the distributionUrl
in your gradle.properties
to differ from the public internet one.
To build and run these projects, you can do the following:
###### From traderX root #####
# Note: gradle or ./gradlew can be used, depending on your path
gradle --settings-file .corp/settings.gradle build
# Build specific project
gradle --settings-file .corp/settings.gradle database:build
# Run specific project
gradle --settings-file .corp/settings.gradle account-service:bootRun
##### From inside the .corp directory ####
cd .corp
./gradlew build
./gradlew account-service:bootRun
A great way to interact with the TraderX community is to attend the bi-weekly Friday TraderX meetings. Email [email protected] to be added to the meeting invite directly, or find the meeting in the FINOS Community Calendar.
- Fork it (https://github.com/finos/traderx/fork)
- Create your feature branch (
git checkout -b feature/fooBar
) - Read our contribution guidelines and Community Code of Conduct
- Commit your changes (
git commit -am 'Add some fooBar'
) - Push to the branch (
git push origin feature/fooBar
) - Create a new Pull Request
NOTE: Commits and pull requests to FINOS repositories will only be accepted from those contributors with an active, executed Individual Contributor License Agreement (ICLA) with FINOS OR who are covered under an existing and active Corporate Contribution License Agreement (CCLA) executed with FINOS. Commits from individuals not covered under an ICLA or CCLA will be flagged and blocked by the FINOS Clabot tool. Please note that some CCLAs require individuals/employees to be explicitly named on the CCLA.
Need an ICLA? Unsure if you are covered under an existing CCLA? Email [email protected]
Copyright 2023 UBS, FINOS, Morgan Stanley
Distributed under the Apache License, Version 2.0.
SPDX-License-Identifier: Apache-2.0