Generating plausible
market scenarios
Skanalytix develops generative models for financial time series, producing simulations that reflect important characteristics of observed market behaviour, including fat tails, volatility clustering, mean reversion and cross-asset dependencies. The aim is to explore a range of plausible market outcomes rather than rely on a single historical record.
One realised path is
not enough
Financial markets provide only one realised path through time. Yet risk analysis, stress testing and model validation require us to reason about a much wider range of market behaviour than that single historical record contains.
Generative modeling provides a way to explore that range through simulations that preserve important characteristics of observed markets while producing outcomes that were not present in the original data.
Generative models for realistic,
scenario-aware simulations
Skanalytix generates synthetic financial time series in which the distribution of future returns evolves in response to the simulated market situation. Rather than drawing outcomes from a fixed distribution or replaying historical sequences, the model generates successive observations conditionally, allowing uncertainty and market behaviour to evolve throughout each simulation.
These simulations can be used to examine portfolio outcomes, estimate conditional risk distributions and explore the consequences of different market scenarios.
Independent quantitative research
Skanalytix Pty Ltd is an independent Melbourne-based company developing generative modelling technology for financial markets. Its work combines quantitative research with the generation of synthetic financial time series for applications in risk analysis, stress testing and model validation. Simulations can be tailored to the requirements of individual organisations.