The Least-Squares Criteria of the Random Coefficient Dynamic Regression Model
- 14 May 2012
- journal article
- research article
- Published by Springer Nature in Journal of Statistical Theory and Practice
- Vol. 6 (2) , 315-333
Abstract
The random coefficient dynamic regression (RCDR) model develops from random coefficient autoregressive (RCA) model and autoregressive (AR) model. The RCDR model is considered by adding exogenous variables. In this article, the concept of the least-squares (LS) criterion is used to estimate the parameter on the RCDR model. Simulation results have shown that the proposed coefficient of the AR model provided asymptotically unbiased estimates nearly for most of the six data-generating models. The RCDR model is then applied to a series of daily observations of the exchange rate of Baht/GBP and Baht/EUR to illustrate the methodology. The predictions of LS criteria are used with those obtained on 20 hold-out future values of withheld observations.Keywords
This publication has 12 references indexed in Scilit:
- Forecasting the volatility of stock price indexExpert Systems with Applications, 2007
- Forecasting volatilityStatistics & Probability Letters, 2005
- Parameter estimation for generalized random coefficient autoregressive processesJournal of Statistical Planning and Inference, 1998
- Consistency and Limiting Distribution of the Least Squares Estimator of a Threshold Autoregressive ModelThe Annals of Statistics, 1993
- Conditional Heteroscedastic Time Series ModelsJournal of the American Statistical Association, 1987
- Time Series Regression with a Unit RootEconometrica, 1987
- Generalized autoregressive conditional heteroskedasticityJournal of Econometrics, 1986
- Threshold Models in Non-linear Time Series Analysis1983
- Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom InflationEconometrica, 1982
- Distribution of the Estimators for Autoregressive Time Series with a Unit RootJournal of the American Statistical Association, 1979