Filtering of an HMM-driven multivariate Ornstein-Uhlenbeck model with application to forecasting market liquidity

This paper investigates the modelling of risk due to market and funding liquidity by capturing the joint dynamics of the Treasury-Eurodollar spread, VIX and a metric derived from S&P 500 time series. We put forward a two-regime mean-reverting model in explaining the behaviour of the liquidity levels in the financial markets. Expectationmaximisation algorithm in conjunction with multivariate filters is employed to construct optimal parameter estimates of the proposed model. The selection of modelling set-up is justified by balancing the best-fit criterion and model complexity. Using market data, the model performance is demonstrated by producing accurate prediction of market illiquidity states.