Advances in Econometrics: Applications of the Kalman filter in econometrics

摘要: Introduction From the point of view econometric modelling, Kalman filter is very little interest. It simply a statistical algorithm that enables certain computations to be carried out for model cast in state space form. The crucial econometrician understand form opens up possibility formulating models are much wider and richer than those normally considered. Furthermore, it often allows setting have more natural interpretation provide useful information on nature underlying economic processes. This second can illustrated clearly at simplest level pure time series model. Indeed, aim this chapter will show how used framework modelling many ways preferable conventional approach based ARIMA proposed links closely with dynamic models, resulting selection methodology akin econometrics. Perhaps clearest indication closeness these starting regression rather theory stationary stochastic unobserved components incorporated into model, provides means estimating them. specification must, some extent, depend priori considerations, since presumably an interpretation, structural one; see Engle (1978). In reduced individual not explicitly available disturbances generate various amalgamated single disturbance term. case linear univariate process.