I trade on financial markets as well as invest in energy storage technologies, cyber-security, high performance computing, technology reach entertainment. I am a passive investor in a number of businesses.
I am a self-taught programmer. I started programming when I was 7 years old. I started ...
more
I trade on financial markets as well as invest in energy storage technologies, cyber-security, high performance computing, technology reach entertainment. I am a passive investor in a number of businesses.
I am a self-taught programmer. I started programming when I was 7 years old. I started trading on financial markets in 1992. Since 2005 I've been developing custom algorithms for modelling and predicting of non-linear systems behaviour that find application in portfolio construction, stress testing, harvesting alpha. I use inputs ranging from traditional to alternative data, with large focus on social sentiment analytics to deliver performance with high Sharpe ratios. I don't use any conventional portfolio methodologies. Since 2007 I have been developing algorithms for non-stationary time series prediction using a broad range of Machine Learning concepts (mainly SVM, HMM). Just for the purpose of a quick verification please use the link to an independent trade analytics website where the performance of my algorithms’ live trading has been captured during a series of tests: https://www.myfxbook.com/members/PiotrPietrzak
The link shows the performance of a custom made algorithm that predicts directional price moves using pattern recognition for signal extracted from one dimensional data (ticks). The robustness is provided by components that target momentum (to handle black swan events or abrupt changes in market sentiment), mean reversals (to handle typical market fluctuations) and trend (to handle fundamental divergences).
More recently I have focused on electrical energy trading concepts, specifically modelling revenue streams for energy storage.
I have at my disposal 40.000 computing cores and custom made modelling software (C#, C++, CUDA, MPI). I am backed up by a group of 20 strong contributors (programmers, mathematicians, engineers). My system is optimized to process billions of data points to find new patterns in data. Outcome is a non linear, low hyper parameter model that prevents over fitting. Dynamic libraries containing the updated model are exported to my trading servers around the World. I am capable to supply my technology to any professional setup with low latency delivery via custom APIs or FIX. I have several live-proven algorithms for FX predictions. As regards to stocks I can deliver solution that meet specific indexing, overlays needs. I am very interested in collaboration related to energy and emissions markets predictions.
I AM NOT OFFERING ANY FINANCIAL SERVICES TO GENERAL PUBLIC.
less