State Space Disaggregation Model with Information Loss Function

Abstract

Different data frequency is a common problem in many research fields; therefore, it should be handled before a particular study is well under way. Many novel ideas including disaggregation techniques, which are the major interest of this study, have been suggested to mitigate the nuisances of mixed-frequency data. In this study, we suggest a generalized framework to disaggregate lower-frequency time series and evaluate the disaggregation performance. The proposed framework combines two models in separate stages: a linear regression model to exploit related independent variables in the first stage and a state space model to disaggregate the residual from the regression in the second stage. For the purpose of providing a set of practical criteria for the disaggregation performance, we measure the information loss that occurs during temporal aggregation while examining what effects take place when aggregating data. To validate the proposed framework, we implement a Monte Carlo simulation and provide an empirical study.

Keywords: Disaggregation; Aggregation effect; State space model; Information loss function; Kalman filter.