Forecasting using a random coefficient autoregression
Abstract We consider point forecasts for economic time series using a basic random coefficient autoregression. We show using simulations that these point forecasts do not improve much on the point forecasts from fixed coefficient autoregressive models. Various empirical illustrations emphasize the simulations-based evidence.
Authors
- Philip Hans Franses (ORCID: https://orcid.org/0000-0002-2364-7777)
- Dmitriy Knyazhitskiy
Institutions
- Tinbergen Institute (NL)
Publication Details
- Journal
- Empirical Economics
- Published
- 2025-09-24
- DOI
- https://doi.org/10.1007/s00181-025-02824-y
- Primary Topic
- Financial Risk and Volatility Modeling
- Type
- article
- Field-Weighted Citation Impact
- 0.00