What we offer
We develop systematic strategies and algorithmic trading systems. Our range of services includes:
Quantitative
Research
Investment
Analysis
Trading Systems Development
Machine Learning and
Data Science
Time Series
modelling
Our Workflow
Idea generation: we review scientific papers, suggest ideas and start investigating the feasibility of candidate strategies with you.
Testing, backtesting: we test our ideas, always questioning hypotheses. It starts from data sourcing and refining, and ends with backtest statistics and a complete profile of the strategy.
Production setup: in this phase, we prepare the trading bot and infrastructure to behave as expected. The exchange API must be translated into functions that allow the algorithm to effectively replicate the strategy in live trading.
Paper trading: normally, a stage where the algorithm runs in a sandbox environment. The target is to replicate as much as possible real case scenarios, to validate the setup.
Small capital runs: at this stage, we run the strategy on test capital.
Live trading, monitoring: we help you in monitoring the runs and ensure the strategy behaves correctly in live trading. The performance is compared with backtest results to identify any implementation failures which could still have passed the other stages.
Assess validity of the strategy: trading strategies are decommissioned when their alpha decays, which can happen due to market structural changes or additional market agents exploiting the same source. This is a continuous work that needs to be done while the strategy runs, and, in case it is not delivering what initially tested and expected, a new idea generation process can start to improve it.