首页 | 本学科首页   官方微博 | 高级检索  
     检索      


The empirical similarity approach for volatility prediction
Institution:1. Department of Statistics and Econometrics, Ruhr-Universität Bochum, Universitaetstr. 150, D-44801 Bochum, Germany;2. Department of Statistics, University of Augsburg, Universitaetstr. 16, D-86159 Augsburg, Germany;1. Department of Otolaryngology, Head and Neck Surgery, Technical University of Munich, Munich, Germany;2. Department of Otolaryngology, Head and Neck Surgery, University of Ulm, Ulm, Germany;3. Division of Neuropathology, Institute of Pathology, Technical University of Munich, Munich, Germany;1. Centrum Wiskunde & Informatica (CWI), P.O. Box 94079, 1090 GB Amsterdam, The Netherlands;2. VU University Amsterdam, The Netherlands;3. Uppsala Universitet, Department of Mathematics, P.O. Box 480, 751 06 Uppsala, Sweden;4. Mathematisch Instituut, Universiteit Leiden, P.O. Box 9512, 2300 RA Leiden, The Netherlands;1. Department of Industrial Management, Tampere University of Technology, P.O. Box 541, FI-33101 Tampere, Finland;2. Techila Technologies Ltd, Itsenäisyydenkatu 2, FI-33100 TAMPERE, Finland
Abstract:In this paper we adapt the empirical similarity (ES) concept for the purpose of combining volatility forecasts originating from different models. Our ES approach is suitable for situations where a decision maker refrains from evaluating success probabilities of forecasting models but prefers to think by analogy. It allows to determine weights of the forecasting combination by quantifying distances between model predictions and corresponding realizations of the process of interest as they are perceived by decision makers. The proposed ES approach is applied for combining models in order to forecast daily volatility of the major stock market indices.
Keywords:Case based decisions  Empirical similarity  Forecasting combinations  Volatility forecasts
本文献已被 ScienceDirect 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号