During the previous years, I offered some -very minor- community support on stackoverflow by answering data modeling related questions. It is my intention to restart these efforts in a more serious manner very soon. My focus will most likely remain on R, but I would like to contribute to the Python community as well.

You can find my SO profile on this page.

The PoloniexR package provides a convenient user interface to the Poloniex (Cryptocurrency Trading) REST API.

The package was committed to CRAN and the first version of the corresponding CRAN reference manual can be found on this page. At the time of writing, some documentation fixes are still required.

From a user perspective, more intuitive package installation instructions and tutorials can be found in this blogpost. The source code can be obtained from the following Github page.

View Github for project details and demo.

This project provides a custom implementation of a Texas Hold ’em handevaluation subsystem as described in chapter 5 of the original Loki Poker Bot research paper. For preflop play, fixed relative strength values are provided based upon simplified monte carlo simulations where all players call to the end of the hand. For postflop play, calculations are provided to determine immediate handstrength and future handpotential via complete scenario enumeration (given the opponent hand probability distributions as input).

From this handevaluation system, betting strategies can be derived and opponent models can be added to provide more realistic and adaptive hand probability distributions. The university of Alberta Computer Poker Research Group has performed extensive research on both fronts and has succesfully used Poker as a testbed for AI and reinforcement learning techniques.

View Github for project details and demo.

In the paper and demos on Github we focus our attention on the risk modelling aspects of the portfolio management process. We first offer some motivational examples on why one should stay clear of the normality assumption while assessing portfolio risk. Stylized facts of univariate and multivariate financial market data are illustrated and discussed. We follow up with a few definitions and interpretations of financial risk and focus our attention to the Value at Risk (VaR) and Expected Shortfall (ES) measures. In the remainder of the paper we present alternatives to the normal distribution for modelling and forecasting the latter risk measures:

• We give an exposition on the Generalized Hyperbolic Distribution (GHD).
• We introduce methods and concepts from extreme value theory (EVT) as a means of capturing severe financial losses.
• Conditional risk measurements are presented in the form of GARCH models.
• Copulae are discussed with the purpose of modelling multivariate dependencies between assets.

From a more practical point of view we perform out of sample backtests and run statistical tests to evaluate the risk forecasting performance of the models. Here, we argue that a mixed (E)GARCH-Clayton/Gumbel copula model is especially suitable for asset return modeling and risk forecasting at the portfolio level. We demonstrate how the latter tool can be employed to provide superior risk adjusted returns in a portfolio context by reliably targetting a certain expected VaR (or ES) level during the portfolio rebalancing process.

View Github repository for paper and demos. This project consists of a proof of concept and conceptual demo of a simple stat-arb trading strategy that is based upon cointegration method.

View Github for project details an demos.

This Matlab based Financial Engineering Toolkit contains several smaller subprojects in the domain of Financial Engineering and the Structuring of Financial products. Available subprojects:

• Pricing, Greeks, implied volatility of barrier options under Black-Scholes.
• Pricing, Greeks, implied volatility of Lookback options under Black-Scholes
• Stock path simulation under Black-Scholes and Heston
• Yahoo Option Chain Downloader
• Heston Calibration Toolkit
• Structuring of a Reverse Convertible (Paper and Demo)

View Github repository: This mql4-framework consists of a few dedicated modules that are linked together in order to assist the user with rapid prototyping and testing of simple TA forex strategies. An example expert advisor is also added for demonstration purposes.

View Github repository: This project includes scripts that can be used to set up a basic securities master database. The database structure is fairly generic in the sense that it can contain data for any assetclass type and/or datafrequency. Additionally, the project includes scripts for the downloading, parsing and inserting of timeseries data for the S&P500 stocks, SPDR ETF funds, the BEL20 stocks and its corresponding ETF tracker. Simple scripts for data cleaning, outlier detection and corporate action backadjustment are provided as well.

View Github repository: This project contains a few smaller subprojects that lie in the domain of non-life insurance actuarial statistics. Each of the subprojects contains its own readme file with additional information regarding the added functionality. Currently added subprojects:

• Claim arrival modelling using poisson processes.
• Loss severity modelling using a spliced ‘body and tail’ approach
• Claim frequency modelling using Generalized Linear Models (GLM)