Portfolio evaluation using OWA-heuristic algorithm and data envelopment analysis

Abhay Kumar Singh*, Rajendra Sahu, Shalini Bharadwaj

*Corresponding author for this work

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

Purpose – The purpose of this paper is to evaluate two different asset selection methodologies and further examine these by forming optimal portfolios. Design/methodology/approach – This paper deals with the problem of portfolio formation, broadly in two steps: asset selection and asset allocation by using the two different approaches for the first step and then well-known mean variance portfolio optimization. In addition, the resulting portfolios are compared using Sharpe ratio. Findings – The empirical observations prove the applicability of the methodology adopted in the research design, ordered weighted averaging (OWA)-heuristic algorithm gives us a better portfolio from the sample observations. Also the asset selection procedures adopted in the research proves to be of help when an investor has to narrow down the number of assets to invest in. Practical implications – The analysis provides two different methodologies for portfolio formation – though the asset allocation is based on the mean variance portfolio optimization, the asset selection methods adopted provide a systematic approach to select the efficient securities. Originality/value – This paper shows that OWA can be used to decide the order of inputs for the heuristic algorithm. Also an attempt is made to use data envelopment analysis to find a solution to the problem of portfolio formation.

Original languageEnglish
Pages (from-to)75-88
Number of pages14
JournalJournal of Risk Finance
Volume11
Issue number1
DOIs
Publication statusPublished - 5 Jan 2010
Externally publishedYes

Keywords

  • Algorithmic languages
  • Data analysis
  • Portfolio investment

Fingerprint Dive into the research topics of 'Portfolio evaluation using OWA-heuristic algorithm and data envelopment analysis'. Together they form a unique fingerprint.

Cite this