Asset allocation with multiple analysts’ views: a robust approach

I-Chen Lu, Baibing Li, Kai-Hong Tee

Research output: Contribution to JournalArticle

Abstract

Retail investors often make decisions based on professional analysts’ investment recommendations. Although these recommendations contain up-to-date financial information, they are usually expressed in sophisticated but vague forms. In addition, the quality differs from analyst to analyst and recommendations may even be mutually conflicting. This paper addresses these issues by extending the Black–Litterman (BL) method and developing a multi-analyst portfolio selection method, balanced against any over-optimistic forecasts. Our methods accommodate analysts’ ambiguous investment recommendations and the heterogeneity of data from disparate sources. We prove the validity of our model, using an empirical analysis of around 1000 daily financial newsletters collected from two top 10 Taiwanese brokerage firms over a 2-year period. We conclude that analysts’ views contribute to the investment allocation process and enhance the portfolio performance. We confirm that the degree of investors’ confidence in these views influences the portfolio outcome, thus extending the idea of the BL model and improving the practicality of robust optimisation.
Original languageEnglish
Pages (from-to)215-228
Number of pages14
JournalJournal of Asset Management
Volume20
Issue number3
Early online date3 Apr 2019
DOIs
Publication statusPublished - 1 May 2019

    Fingerprint

Keywords

  • Analysts’ recommendation
  • Black–Litterman model
  • Fuzzy logic
  • Portfolio selection
  • Robust optimisation

Cite this