The field of automated/algorithmic trading has been developing rapidly. All companies with trading departments are slowly but steadily moving to automated algorithmic trading. This transformation requires knowledge of previously irrelevant subjects like: market microstructure, communication protocols (FIX, ITCH, OUCH), trading strategies and many more.
Specific area’s of expertise include:
System and architecture design of an automated trading system
Implementation of an automated execution system
Design and implementation of
Risk management specifically for automated trading
Complex trading strategies
A framework for historical and operational backtesting
Emulators for trading related components
Consultancy with the goal to
Explain the basics of trading strategies; stock pricing, options pricing, spread trading, market making, arbitrage and statistical arbitrage
Spread knowledge on the design, development and operational aspects of automated trading systems
Troubleshoot errors in automated trading systems
Analyze logs and other information to identify bugs and other shortcomings
Rather than developing everything from scratch, we consider literature in this field to lay the foundations on which to build.
Mathematical Simulations, Generative Machine Learning
Generative Models achieve the opposite of Calculation Models. A generative model can create complex objects and data from a small dimensional input. Recently, we observe a surge in the development and application of generative models.
Some simple general examples:
Using a starting price and volatility to generate a stock price movement
Generate/predict the movements of crowds, by modelling individuals as particles that moving according to a spatial swarming model
A machine learning (GAN) model which generates ultra-realistic pictures
Mathematical Modelling, Data Science, Machine Learning
Calculation models can be used to solve complex problems. Typically these models use input data for calculation, estimation or classification purposes.
Often a calculation model must be developed for a specific use-case. Depending on the use case we will ensure to apply the most suitable methodology:
Mathematical Modelling - White box model with limited amounts of data
Data Science - White box model with large amounts of data
Machine Learning - Black/Grey box model with large amounts of data
We see software engineering as a way to empower organisations to optimize their core business in a digital world. We can design, develop and maintain software systems.
Even though it’s a challenge, we always aim to deliver software of high quality. We achieve this by proper design, implementation according to recognized industry standards, automated testing and thorough documentation.