At Causality Group we are passionate about implementing new technologies and quantitative methods which allow us to understand better the drivers of asset returns. Automation of investment process does not only simplify the job of quantitative portfolio managers, but increases transparency and allows us to use wide range of techniques borrowed from machine learning, optimal control and statistics.
While our primary aim is to benefit our employees, we also care about our surroundings. By continuously engaging our people in working towards more efficient
mathematical modeling techniques, investing in next generation hardware and advancing open-source software we also help users of information technology and support our counterparties
who choose to trade with us in the global markets. At the same time we allocate capital to companies who we believe to have the talent and commitment to turn it into great business.
We have found that some of the strongest data analysis, deep-learning and big-data ecosystems are implemented under public and open-source license. We also use closed-source applications when we think they perform exceptionally, however we believe that customizability of open-source solutions is impossible to beat.
Open-source developers and data analysts tend to follow a path where their job is most efficient, without much politics involved in what tools to use. This results in strong support for productivity oriented solutions like Python, pandas or R, which we use heavily for data discovery. Unique results call for unique tools. While using commercial applications it is often quicker to get from zero to one, specializing analysis tools and automating workflow is easier to achieve using the source code.
We are combining trading strategies or factors so that the aggregated performance is stronger than any of the individual strategies. Therefore a significant part of our job is to discover the next strategy which will further strengthen our live portfolio.
The mathematical principle of fundamental law of active management, on the other hand, states that increasing number of strategies has decreasing marginal utility with regard to performance of the resulting portfolio. The ideal way therefore to increase our trading profits is to improve every detail in our trading process: starting from our philosophy on factor modeling, through unbiased inversion of covariance matrices, to optimal control of trading strategies in the order book.