/FASLR

Free Actuarial System for Loss Reserving

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

FASLR: Free Actuarial System for Loss Reserving

codecov tests docs Join the chat at https://app.gitter.im/#/room/#faslr:gitter.im

FASLR (pronounced: fazlur) is a Qt-based frontend for open-source loss reserving packages. Current plans are to support the Chainladder package.

The actuarial methods and example data used in this project are derived from publicly available papers and data sources. The GUI is developed in Python using the open-source PyQt6 package.

Project Governance

FASLR will assist in the proper governance of periodic reserve reviews. In addition to being a GUI in which actuarial analyses can be done, it will also serve as a portal through which current and past analyses can be managed. Each reserve analysis will have metadata that indicates its status (in progress, review needed, signed-off), and by storing past analyses, FASLR will make it easy to compare quarter-by-quarter reviews without having to awkwardly navigate company shared folders.

Basic Interface

FASLR will have a simple layout. The collapsible side pane will contain a navigable tree consisting of past reserve analyses, organized by jurisdiction, LOB, and unit of time.

The analysis pane is where the user will be able to conduct a reserve analysis. This will be the area where the actuary can view triangles, select and perform actuarial methods, and view graphs. The user will also be able to toggle between an analysis view (doing the work) and project metadata (project status, peer review, sign off).

basic-interface-filled basic-interface-filled basic-interface

Department Integration

FASLR will support departments with multiple users, who will have the ability to check out and lock analyses as they are being used. Data will be stored on a database, although the ones that will be supported have yet to be determined. Lightweight deployments for single-users or personal use will use a SQLite data store.

client-server-model

Contributing

Please refer to the document CONTRIBUTING.md for the contributing guidelines.

You can evaluate the working version of FASLR by executing the main.py program in the faslr directory:

cd FASLR
python -m faslr

Alternatively, if you have a virtual environment set up:

cd FASLR
source venv/bin/activate
python -m faslr

To learn more about the current structure of the repository, read ARCHITECTURE.md.

Supported Operating Systems

FASLR is developed on Linux, specifically Ubuntu 22.04, which is probably what you want to use if you want to capture the same look and feel that you see in the documentation and in the development blog.

However, Python and Qt are cross-platform, which means that you may try it on other systems, but there are no guarantees on my end that it'll work as well as it does on Linux.