fdb – Forschungsdatenbank der Europa-Universität Viadrina Frankfurt (Oder)

Projekt

Prof. Dr. Wolfgang Schmid: Surveillance of the global minimum variance portfolio

Projektleitung Prof. Dr. Wolfgang Schmid
Time span 01/2006 - 12/2011
Fakultät Wirtschaftswissenschaftliche Fakultät
Lehrstuhl Lehrstuhl für Quantitative Methoden, insb. Statistik

Zusammenfassung

A financial analyst is interested in a fast on-line detection of changes in the optimal portfolio composition. Although this is a typical sequential problem the majority of papers in financial literature ignores this fact and handles it in a non-sequential way. This project deals with the problem of monitoring the weights of the global minimum variance portfolio (GMVP). It elaborates the tools for the surveillance of the GMVP. Golosnoy and Schmid (2007) introduced exponentially weighted moving average-type (EWMA) control charts for this task based on the processes of the estimated weights as well as of their first differences. Golosnoy et al. (2010) propose new characteristic processes exhibiting better stochastic properties for sequential monitoring purposes. Golosnoy et al. (2009b) deals with cumulative sum (CUSUM) charts applied directly to the GMVP weights and their transformations. The suggested procedures performe better than the standard ones in terms of average run length and the worst case conditional expected delay criteriums.

Abstract

A financial analyst is interested in a fast on-line detection of changes in the optimal portfolio composition. Although this is a typical sequential problem the majority of papers in financial literature ignores this fact and handles it in a non-sequential way. This project deals with the problem of monitoring the weights of the global minimum variance portfolio (GMVP). It elaborates the tools for the surveillance of the GMVP. Golosnoy and Schmid (2007) introduced exponentially weighted moving average-type (EWMA) control charts for this task based on the processes of the estimated weights as well as of their first differences. Golosnoy et al. (2010) propose new characteristic processes exhibiting better stochastic properties for sequential monitoring purposes. Golosnoy et al. (2009b) deals with cumulative sum (CUSUM) charts applied directly to the GMVP weights and their transformations. The suggested procedures performe better than the standard ones in terms of average run length and the worst case conditional expected delay criteriums.

Kooperation

Internationales Projekt Nein
Kooperationspartner Prof. Dr. Vasyl Golosnoy - Universität Kiel