Adaptive short-term forecasting of selected financial processes

Abstract

A computer based system is proposed for adaptive modeling and forecasting of financial and economic processes, that is constructed with application of system analysis principles. A hierarchical structure of decision making process during forecasts estimation was taken into consideration and the methods were used for describing uncertainties of structural, parametric and statistical nature. To estimate model structure and parameters several mutually supporting estimation techniques were used as well as optimal state estimation procedure for dynamic systems that allowed take into consideration some types of structural and statistical uncertainties. Probabilistic modeling methods make it possible to consider uncertainties of probabilistic type. The problem of short term forecasting for gold price is considered as an example using a set of constructed regression models and Kalman filter for generating optimal estimates of states. The best forecasting results were achieved with optimal filter and autoregression models with trends. Also the models were constructed for conditional variance that provided acceptable quality forecasts for variance (volatility) that could be used for constructing decision making rules in trading operations.