Munich Personal RePEc Archive

Evaluating Forecast Uncertainty in Econometric Models: The Effect of Alternative Estimators of Maximum Likelihood Covariance Matrix

Calzolari, Giorgio and Panattoni, Lorenzo (1984): Evaluating Forecast Uncertainty in Econometric Models: The Effect of Alternative Estimators of Maximum Likelihood Covariance Matrix. Published in: paper presented at The Fourth International Symposium on Forecasting. London Business School, July 8-11 (8. July 1984): pp. 1-33.

[img]
Preview
PDF
MPRA_paper_28806.pdf

Download (601kB) | Preview

Abstract

Most of the methods proposed in the literature for evaluating forecast uncertainty in econometric models need an estimate of the structural coefficiencs covariance matrix among input data. When estimation is performed with full information maximum likelihood, alternative estimators of such a covariance matrix (Hessian, outer product, generalized least squares type matrix, quasi maximum likelihood type matrix), although asymptotically equ1valent, often produce large differences in practical applications. Experimental results will be given for some econometric models well known in the literature, both with hiscorical data and with data generated by Monte Carlo.

UB_LMU-Logo
MPRA is a RePEc service hosted by
the Munich University Library in Germany.