2020/11/15 by Ufuk Beyaztaş, Han Lin Shang, Beyaztas, Ufuk +1
Economics, Econometrics and Finance · Decision Sciences · Mathematics · #Monetary Policy and Economic Impact #Forecasting Techniques and Applications #Statistical Methods and Inference
paper · pdf · doi:10.48550/arxiv.2011.07664
The bootstrap procedure has emerged as a general framework to construct\nprediction intervals for future observations in autoregressive time series\nmodels. Such models with outlying data points are standard in real data\napplications, especially in the field of econometrics. These outlying data\npoints tend to produce high forecast errors, which reduce the forecasting\nperformances of the existing bootstrap prediction intervals calculated based on\nnon-robust estimators. In the univariate and multivariate autoregressive time\nseries, we propose a robust bootstrap algorithm for constructing prediction\nintervals and forecast regions. The proposed procedure is based on the weighted\nlikelihood estimates and weighted residuals. Its finite sample properties are\nexamined via a series of Monte Carlo studies and two empirical data examples.\n