Forecasting With Difference-Stationary And Trend-Stationary Models

Preprint
Publisher Website
Google Scholar
    • preprint
      • Published in RePEc
Abstract
The behavior of difference-stationary and trend-stationary processes has been the subject of considerable analysis in the literature. Nevertheless, there do not seem to be any direct comparisons of properties of each for both being potential data-generation processes. We look at the consequences of incorrect choice between these models for forecasting when parameters are known, and when parameters have to be estimated. The outcomes are surprisingly different from established results. (This abstract was borrowed from another version of this item.)
All Related Versions

This publication has 0 references indexed in Scilit: