Cointegration constant estimation under prior uncertainty
Abstract
A new system approach to cointegration constant estimation, based on prior analyses of stochastic components of the problem is proposed. Stochastic components are estimated using parametric identification methods, which allow estimating statistical characteristics of noise under conditions of priori uncertainty. Drawbacks of existing cointegration constant estimation methods are explained and some new methods are proposed.Downloads
Issue
Section
Methods of system analysis and control in conditions of risk and uncertainty