Time‐scale transformations of discrete time processes

From MaRDI portal





This paper investigates the effects of temporal aggregation when the aggregation frequency is variable and possibly stochastic. In this case the aggregate-time processes have time-varying parameters and non-spherical disturbances even when these characteristics are absent in original-time. These properties are important considerations in terms of forecasting, estimation and testing since they explain in which directions aggregate models are likely to be misspecified and in which way this misspecification can be corrected. The authors derive a representation of the aggregated data-generating process and derive maximum likelihood estimators for its parameters. The resulting models improve the estimation of structural parameters and provide more accurate forecasts as illustrated with an example on inventory decisions.











This page was built for publication: Time‐scale transformations of discrete time processes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4677046)