Questions about Skanalytix
A concise introduction to what the models generate, how the approach differs from resampling, and where realistic market simulations can be useful.
01What does Skanalytix generate?
Skanalytix generates synthetic financial time series representing plausible market scenarios. These are not forecasts of what will happen, but simulated paths that capture important characteristics of observed market behaviour.
02Is Skanalytix intended to predict future market returns?
No. The objective is not to identify a single most likely future path. Financial data provide only one realised history, while many different outcomes could plausibly have occurred. Skanalytix is designed to generate a range of realistic simulations for scenario analysis, risk assessment and related applications.
03What makes the approach non-parametric?
The models do not begin by assuming that returns follow a particular probability distribution or that their relationships have a predetermined functional form. Instead, conditional distributions are estimated directly from relevant historical observations, allowing the shape of those distributions to be determined by the data.
04Is this the same as historical simulation or resampling?
No. Historical observations inform the distributions from which new observations are generated, but complete historical periods or return vectors are not simply replayed. As each simulation develops, the relevance of past observations changes according to the evolving simulated situation. New observations are generated from conditional distributions estimated from relevant historical data, allowing new market paths to emerge.
05What characteristics of financial markets can the simulations capture?
Current work examines whether simulations capture important characteristics of observed financial markets, including fat-tailed return distributions, volatility clustering, mean reversion and cross-asset dependencies. At the portfolio level, this includes changes in correlation and diversification as market conditions evolve.
06What can realistic market simulations be used for?
Potential applications include scenario analysis, stress testing, portfolio risk analysis and model validation. Generating many plausible simulations makes it possible to examine how portfolios or models behave across a much wider range of market outcomes than has actually been observed.
07What markets does Skanalytix currently model?
Current development is focused primarily on equity time series and multi-asset equity portfolios. The underlying framework is not inherently restricted to equities, and its application to other financial and economic time series is also being explored.