No. 419: The Likelihood of a Continuous-time Vector Autoregressive Model
J. Roderick McCrorie ,
Queen Mary, University of London
October 1, 2000
This paper provides a method that weakens conditions under which the exact likelihood of a continuous-time vector autoregressive model can be derived. In particular, the method does not require the restrictions extant methods impose on discrete data that limit the applicability of continuous-time methods to real economic time series. The method applies generally to higher-order continuous-time systems involving mixed stock and flow data.
J.E.L classification codes: C32
Keywords:Continuous-time, Vector autoregression, Exact likelihood, Time series