Generative modelling technology for financial markets
Skanalytix Pty Ltd is a Melbourne-based company developing data-driven generative modelling technology for financial time series. Its work combines research into financial market behaviour with the development of methods for generating and analysing plausible market scenarios.
Generative modeling for financial time series
The Skanalytix modelling approach is designed to generate plausible financial time series that reproduce important statistical and structural properties of observed markets.
Particular emphasis is placed on evaluating simulated data against historical observations, including the behaviour of return distributions, temporal dependencies and relationships between assets. The objective is not simply to produce realistic-looking price paths, but to reproduce the market characteristics that matter for quantitative analysis.
Data-driven and non-parametric
Rather than assuming a particular distribution for market returns or training a neural generative model, Skanalytix uses a non-parametric, memory-based approach that draws directly on historical market observations.
At each step of a simulation, the model estimates a conditional distribution of returns using the relevance of historical observations to the current simulated situation. This relevance depends on how market situations are represented, how similarity is assessed and how historical observations are weighted.
The approach combines data-driven estimation with explicit modelling choices informed by financial market behaviour. It is designed to remain interpretable while allowing complex distributional and temporal relationships to emerge through sequential simulation.
Andrew Skabar, PhD
Founder, Skanalytix Pty Ltd
Andrew Skabar founded Skanalytix in 2023 following more than two decades working in artificial intelligence, machine learning and decision systems, initially in academia and subsequently through independent research and development. He holds a PhD in artificial intelligence and has a background in physics and mathematics.
His research has centred on methods for learning from complex data and estimating relationships under uncertainty. This work eventually led to the modelling framework underlying Skanalytix and its application to financial time series.
Equity time series and portfolios
Current development focuses on individual equity returns and multi-asset portfolios, with particular attention to reproducing fat tails, volatility clustering, mean reversion and cross-asset dependencies.
Skanalytix can generate synthetic financial time series tailored to specific requirements, supporting applications such as portfolio risk analysis, stress testing and model validation. Organisations interested in exploring these capabilities are welcome to contact Skanalytix to discuss their requirements. Contact Skanalytix.