Stability of Traditional Portfolio Models

Examensarbete för masterexamen

Please use this identifier to cite or link to this item:
Download file(s):
File Description SizeFormat 
213466.pdfFulltext457.52 kBAdobe PDFView/Open
Type: Examensarbete för masterexamen
Master Thesis
Title: Stability of Traditional Portfolio Models
Authors: Sharma, Ruben
Abstract: The idea behind this thesis came from, at that time, a fellow colleague at SEB. They had evaluated the resampling model when doing portfolio optimizations and knew that it was more stable than the original model by Markowitz. What they hadn’t studied, was to what extent and also the instability of the resampling model itself. Therefore the research questions for this thesis were • Research question 1: How sensitive are the two models’ optimal portfolio weights to changes in expected return, risk and correlation. • Research question 2: How does the inherit portfolio characteristic affect the results of research question 1. To study these questions reference portfolios were derived based on historical asset returns of several multi asset portfolios at different risk levels. To investigate the instability of the models the input parameters were stressed in different combinations to see how the portfolios weights changed. In short, the Markowitz model was more unstable then the resampling model, the dispersion increases with portfolio risk level and the expected return parameter is the parameter that has the largest impact on stability for both models. A few exceptions was seen on the upper end of the risk scale. The results in this study confirms the result of previous studies but also challenges other.
Keywords: Grundläggande vetenskaper;Matematik;Basic Sciences;Mathematics
Issue Date: 2015
Publisher: Chalmers tekniska högskola / Institutionen för matematiska vetenskaper
Chalmers University of Technology / Department of Mathematical Sciences
Collection:Examensarbeten för masterexamen // Master Theses

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.