Publication at Faculty of Mathematics and Physics |
2015
Abstract
The exponentially weighted moving average (EWMA) model is a particular modelling scheme used by RiskMetrics for forecasting the current level of volatility of financial returns. The aim of this paper is to introduce and study the self-weighted sequential estimation algorithm, which represents a numerically effective alternative to already established calibration approaches.
Firstly, its derivation and theoretical properties are briefly outlined. Secondly, the presented calibration technique is robustified to eliminate destructive influence of eventual additive outliers.
Thirdly, both versions are examined by means of Monte Carlo simulations and real financial data.