Title: Modelling foreign exchange time series with SVJD models with stochastic drift
Authors: Milan Ficura - University of Economics in Prague (Czech Republic) [presenting]
Jiri Witzany - University of Economics in Prague (Czech Republic)
Abstract: Several models with stochastic volatility, jumps, and stochastic drift are proposed and applied to the foreign exchange time series on the daily frequency. Alternative specifications of the drift term are tested, including Markov-Switching, continuous diffusion, pure-jump and jump-diffusion processes. The estimation is performed with Sequential Gibbs Particle Filter algorithm and the proposed models are compared based on their Bayes factors. In addition to the statistical evaluation, investment strategies utilizing the stochastic drift are proposed, and their risk-adjusted profitability is computed. Implications with respect to a possible violation of the weak form Efficient Market Hypothesis are discussed.