This project aims to provide a framework for quantifying and analysing the market risk of financial portfolios. Insights into the risk contributions allow for efficient risk capital allocation and increase the financial performance in the long run.
The ultimate goal is a complete valuation and risk assessment framework for most financial instruments, where risk measures and stress test impacts on arbitrary portfolios can be calculated. Hereby, state-of-the-art methods and a light-weight and rapid calculation shall be implemented with free software only.
The code is written in Octave and C++ language and will be useable right out-of-the box with Octave version 4.4.1 (up to Octarisk v0.6.0) and the Financial, IO, Struct, Parallel and Statistics package. Starting from Octarisk v0.7.0 only Octave versions 5.2.0 or higher are supported. The code is licensed under the GNU GPL.
Further information: website
function and method hierarchy (PDF)
Stefan Schloegl (schinzilord@octarisk.com)
IRRer-Zins (IRRer-Zins@t-online.de)
Up to Octarisk v0.6.0: You need Octave in Version 4.4.1 with installed financial, io, struct, parallel and statistics packages. From Octarisk v0.7.0 or higher: You need Octave in Version 5.2.0 or 6.1.0 with installed financial, io, struct, parallel and statistics packages.
Moreover, an efficient linear algebra package is strongly recommended (e.g. OpenBLAS / LaPACK).
Just follow these steps:
- install Octave either from your distro repository (
sudo apt-get install octaveoryum install octave-forge) or directly from the binaries - install Financial, IO, Struct, Parallel and Statistics package. Therefore type at the Octave command line (installation may take some minutes to complete):
pkg install -forge financial
pkg install -forge io
pkg install -forge struct
pkg install -forge parallel
pkg install -forge statistics
- clone the Octarisk repository or download the latest release and unpack it to your desired folder
- add the path of your Octarisk directory to Octave (mind the forward
slashes
/, even under Windows):
addpath ('/path/to/octarisk-latest')
savepath()
- compile required C++ code by typing on the Octave console (C++11 compiler required)
compile_oct_files('/path/to/octarisk-latest')
This takes a couple of minutes and compiles into binaries for even faster pricing (the most performance relevant algorithm have been highly optimized in C++).
- adjust the path to your working_directory (which can be anywhere on the system where you have write access). An example working_folder with pre-defined instruments, risk factors and portfolio is provided in the release.
- run unittests and integrationtests
unittests()
integrationtests ('/path/to/octarisk-latest/testing_folder')
All tests have to be successful (PASS only, no FAILS). Otherwise please make sure all C++ binaries were compiled and financial, io and statistical packages were installed properly. The integrationtests valuate all possible instrument types and calculates portfolio value-at-risk and stress figures (a full end to end test including scenario generation, instrument valuation and portfolio aggregation is performed).
- open script file octarisk_gui.m and set the input path to your installation directory 'path/to/working_folder'. Furthermore adjust the link to the working folder in the second line of the central parameter file 'parameter.csv' or parameter_unix.csv (depending which OS you use)
- start the GUI (which automatically calls octarisk main script) or just call the main script from the console
octarisk_gui ('/path/to/octarisk-latest/working_folder')
octarisk ('/path/to/octarisk-latest/working_folder')
-
the script will run several seconds and gives you two file based reports (you can find the reports under
/path/to/octarisk-latest/working_folder/output/reports), and instrument and portfolio attributes and risk measures on your screen. -
now you can start using the GUI, adjusting the input data (
/path/to/octarisk-latest/working_folder/input/ and /mktdata) or change the risk measurement settings directly in theoctarisk.mscript.
For a more detailed installation instruction please refer to this tutorial.
Octarisk runs on every system with a running Octave environment in Version 4.4.1 just up to v0.6.0. Due to deprecated Octave libraries starting with Octave 5.2.0, the latest releases (starting from v0.7.0) are only supported on Octave 5.2.0 or 6.1.0.
To my knowledge, these Octave binaries are provided for MS Windows and many Linux distributions.
A system with at least 4Gb memory is recommended.
Matlab Compatibility: Due to recent lack of a Matlab programming platform the support for Matlab was entirely skipped. In theory it should be possible to adjust the code, but unknown effort is required to do so. Further code development will only be accomplished with Octave (because it is free and everybody can contribute and run the code.) Write me an Email if you need assistance in porting the code basis to Matlab.
- Version 0.7.0 supporting Octave 5.2.0 and 6.1.0, but dropping support for Octave 4.4 or below.
- Version 0.6.0 introduction of enhanced reporting framework
- Version 0.5.1 introduction of Retail class to allow for valuation of private investors banking and insurance products
- Version 0.5.0 introduction of Solvency II reporting according to Tripartitie v4.0 standard
- Version 0.4.4 bug fixes, stability and performance improvements (dropped support for Octave 4.2)
- Version 0.4.3 fixed compatibility issues with Octave 4.4.0 release, minor bug fixes, more statistical tests
- Version 0.4.2 introduction of CDS and Sobol number generator, minor bug fixes
- Version 0.4.1 introduction of a GUI and a batch mode, minor bug fixes
- Version 0.4.0 massive speed impprovements, enhanced documentation, modified stress test interface
- Version 0.3.5 more instruments, refactored code, enhanced documentation
- Version 0.3.4 various performance improvements, enhanced documentation
- Version 0.3.3 introduction of c++ pricing functions, enhanced volatility framework
- Version 0.3.2 introduction of a session-like instrument valuation interface, new instruments
- Version 0.3.1 enhanced testing features for regression and integration tests
- Version 0.3.0 independent validation of all pricing functions, unittests, bug fixes
- Version 0.3.0-rc1 further enhanced file interface, aggregated curves introduced, bug fixes
- Version 0.2.0 enhanced file interface and bug fixes
- Version 0.1.0 first public release
The following alternative commercial and proprietary solutions exist:
- IBM Algorithmics
- Sungard / FIS Frontarena
- Numerix
- Quaternion (which relies on Quantlib libraries, but is fully proprietary)
- QuantLib is an open source library for instrument pricing. The QuantLib repository could serve as a basis for Octave code adaptions
- Opengamma Strata risk analytics platform is also an open source library for instruments (mainly derivatives) pricing.
Octarisk focusses on being fast, light-weight and easy to set up and maintain. Therefore it contrasts the feature-rich proprietary solutions with massive overheads which are in use by financial institutions.
Nevertheless, Octarisk can be used for portfolios with several thousands instruments and risk factors. Therefore it is best suited to use it as an independent tool to validate the results of proprietary solutions or as a (starting) point for a full in-house market risk measurement solution.
The costs of using Octarisk are low, since the licensing costs both for Octave and Octarisk are zero. The money can be spent on in-house or external support for development of new features and maintenance. As a result you do not face vendor lock-in and have full transparency - forever.