A filtering approach to tracking volatility from prices observed at random times (Q862222)

The paper deals with nonlinear filtering of the coefficients in asset price models with stochastic volatility. Following \textit{R. Frey} and \textit{W. Runggaldier} [J. Theor. Appl. Finance 4, 199-210 (2001)], the authors assume that the asset price process \((S_t)_{t\geq 0}\) is given by the stochastic differential equation \[ dS_t=m(\theta_t)S_tdt+v(\theta_t)S_tdB_t, \] where \(v\) is a positive function, \(m\) is bounded, \(B_t\) is a Brownian motion, \(\theta_t\) is a càdlàg strong Markov process. The ``volatility'' process \(\theta_t\) is unobservable, while the asset price \(S_t\) is observed only at random times \(0<\tau_1<\tau_2<\dots\). The problem of estimation of \(\theta_t\) is treated as a special nonlinear filtration problem with measurements generated by a multivariate point process \((\tau_k,\log S_{\tau_k})\). A closed form optimal recursive Bayesian filter for \(\theta_t\) is derived. A number of examples is given.