Over the last 10 years we have spent many person years full time writing a software language and tooling which allows banks to collaborate in implementing the complex regulatory calculations that regulators define that all banks must implement. Banks are trying to meet the same regulation, but they all implement it separately (or with a range of different vendors) and then all test it separately. This is very expensive, especially given that there is no competitive advantage in meeting regulation better than another bank.
From many years experience in writing software for regulatory calculations, we know that collaboration is difficult even within one bank, as domain experts, testers and developers use different tools and artefacts. We decided to create a computer language and visual tooling that allowed for easier collaboration, entirely based on Eclipse modelling and Eclipse DSL technologies. About 1 year ago we realised that the European Central Bank has a project with a very similar aim. They have a specialised language (called VTL) for regulatory logic, and an information model (SMCubes) for describing data structures, but no tooling at all (and no plans to make their own tooling). The ECB works with expert groups to create regulatory logic in VTL and data structures in SMCubes and publishes these in an open fashion. We decided to port all our tooling to support VTL (which is very similar to the language we created) and SMCubes, so as to provide open source tooling for the BIRD project.
We are a member of the ECB's BIRD Working Group on testing, meeting the ECB each month in Frankfurt. We provided a demo of the software to the group which was very well received.
The Eclipse Free BIRD Tools project aims to provide visual tooling to support the development, testing, and understanding of the artefacts produced by the European Central Banks BIRD project.
Specifically, this includes:
1.) A runtime for the VTL language which is used by the BIRD project to describe regulatory calculation logic.
2.) A visualization tool to view the different data structures (cubes) used in the BIRD project, these cubes are described using the SMCube Information Model developed by the ECB. https://www.ecb.europa.eu/stats/ecb_statistics/co-operation_and_standar…
3.) A visualization tool to visualize the dependencies between artefacts published in the BIRD project. This includes attribute lineage of how regulatory report cells calculated using VTL logic depend upon columns from input cubes. This is lineage available before VTL programs have been run.
4.) A visualization tool to visualize the dependencies between results of a VTL program and inputs to a VTL program, including intermediate results. This is data lineage of actual results after a VTL program has been run with input data as a set of populated cubes.
5.) Tooling to support the organized testing of the BIRD regulatory logic.
The artefacts produced by the European Central Bank's BIRD project are free and openly available and provided as a public good.
The BIRD project is described at https://www.ecb.europa.eu/stats/ecb_statistics/co-operation_and_standar… , a useful short video overview of the BIRD project is available at https://youtu.be/CUreE2REzvA
The artefacts produced by the BIRD project are a standard input data model and a set of regulatory logic defined in a language called VTL which transforms the input data model into the values required to populate European regulatory reports.
By providing open-source tools for the BIRD project , the Eclipse Free BIRD Tools project hopes to remove barriers to understanding, contribution, trial, and adoption of the BIRD.
We are not aware of any trademark on "Free BIRD Tools" however we would not be surprised if "Free Bird"; is trademarked for another purpose. We do not mind changing the name to something more suitable. We do have a trademark in the U.K. on "BIRDBench", which is a possible alternative name and we are happy to re-assign this trademark to the Eclipse Foundation.
The project is built entirely on top of Eclipse Frameworks, specifically Eclipse Sirius and Eclipse XText.
The core of the solution, (and the majority of the work and innovation) is a set of Ecore files (Eclipse modelling files) which describe a meta-model.
Three of these Ecore files are a representation of packages described in the ECB's SMCubes methodology, which are openly published.
We believe that this project can increase the visibility of the Eclipse Foundation, and its projects, to banks and regulators in the financial industry. We expect this to lead to more interest and feedback in the technical projects of Eclipse, especially the modelling projects which are the base frameworks used in this project. These modelling projects are currently growing in use in high profile projects at financial organizations such as Mizuho and ISDA. In addition, we expect that this project can bring more members to the Eclipse Foundation as a result of financial organisations using open source Eclipse projects more directly.
We believe that the good name of the Eclipse Foundation can give confidence to users of this project that it is run in a vendor neutral fashion, and that it has a very clear description of intellectual property gained from the Eclipse Development Process. We hope that this will increase the use of, and contributions to, the project and its dependant frameworks from the Eclipse Modelling project.
The next twelve to 18 months will be focussed on improving the VTL runtime, improving visual look and feel, improving the ability to automate testing of VTL programs.
We have built an extensive network of banks, regulators, and software companies working with regulatory technology and will be reaching out to these individually.
Our immediate focus is on providing a good video demo to explain the project to catch the interest of more people and to make the code as easy to get started with as possible.
We continue to attend conferences on Regulatory Technology to network with interested parties. We will try to re-engage with the Bank of England, who were interested in a demo we gave them of an earlier release of this project to assist their own improvements in regulatory technology described in their open DRR (Digital Regulatory Reporting) project.
We will continue to discuss with other members of the Eclipse Modelling community who already have an interest in this project, we have been discussing this with them frequently over the last 8 Eclipse conferences that we have attended.
An initial contribution is ready at https://github.com/openregtech/FreeBIRDTools along with a description of how to build the project.
The European Central Bank,
Incquery Labs,
A number of banks, consultancies and software companies.
The project is built entirely on top of Eclipse Frameworks, specifically Eclipse Sirius and Eclipse XText.
The core of the solution (and the majority of the work and innovation) is a set of Ecore files (Eclipse modelling files) which describe a meta-model, and are written by OpenRegtech
Three of these Ecore files are a representation of packages described in the ECB's SMCubes methodology, which are openly published.
The rest of the application is about 20 program files (Java and Xtend) written entirely by OpenRegTech, and about 30 XML files which represent instances of the Metamodel and again are written entirely by OpenRegTech, these XML files are an example usage of the metamodel. Also there are some configuration files, particularly one for configuring Eclipse Sirius layouts, again these are authored by OpenRegTech.
There are three very short simple Scala files which do not import any libraries.
The code is already published at https://github.com/openregtech/FreeBIRDTools along with a description of how to build the project. The published code already has copyright statements following guidelines from Eclipse (using the Eclipse copyright tool) , also in its structure we have tried to follow lessons learnt from being a contributor to Eclipse EMFStore and EMF Client platform.
- Log in to post comments