Αξιολόγηση επενδυτικών στρατηγικών για την διαμόρφωση δυναμικών χαρτοφυλακίων Hedge Funds

Abstract

The present thesis is organized around 5 chapters. Chapters 1 and 2 provide qualitative hedge fund research: they introduce the concept of hedge fund investments and their characteristics, they examine the commonest investment strategies as well as the size of the hedge fund industry and the recent developments in it. Chapters 3 and 4 focus on quantitative aspects: they deal with new methods of hedge fund performance evaluation and portfolio allocation that are necessary stepping stones towards understanding how to select appropriately and how to and monitor hedge funds. Chapter 5 summarizes and discusses the main findings of the dissertation and provides directions for future research. For the purpose of the present thesis, the Hedge Fund Research classification scheme was followed; its published indices were adopted for our empirical analysis. The reason for doing so lies neither in the uniqueness of this classification, nor in the fact that these indices were bias-free; It is the fact that according to empirical studies, the HFR disposes of the most comprehensive database (see Agarwal et al., 2009a) that accounts for our choice.Chapter 1 covers the introduction of most of the basic definitions, history and general characteristics of hedge fund investments. An understanding of some of the concepts introduced there, enables the reader to follow the rest of the thesis to a considerable degree, however most chapters are principally self-contained. In chapter one, we give the definition of hedge funds together with a brief historical review of the hedge fund origins. The characteristics typically associated with hedge fund investments are described in detail and compared to the traditional investment ones. The growth of the industry over time, current trends, the geographical distribution of assets under management and the key groups of hedge fund investors are also discussed in chapter 1. In chapter 2, we describe the key hedge fund strategies. Although the hedge fund industry is considered a distinct and rather uniform asset class, it is incredibly heterogeneous and diverse. The arduous task of the hedge funds’ classification has been undertaken by data providers which collect, process and report the hedge funds’ performance along with further useful information such as the company’s name, strategy, assets under management, their management and incentives fees, etc. Consequently, the second chapter examines the most popular hedge fund index data providers, relevant indices, and the most serious bias of hedge fund data. Next, we present an overview of the various strategies pursued by hedge fund managers, their trades and corresponding risk factors. Finally, the historical performance and the statistical properties of hedge fund indices feature in separate sections of this chapter. Given the lack of well-established asset pricing models in the specific hedge fund context, a factor-based model, termed ‘the hedge fund multifactor model’ (HFMM) is introduced and assessed in chapter 3. This model adapts and extends benchmark asset pricing models by incorporating a number of ‘location factors’ in them. ‘Location factors’ is a term covering proxies for different buy-and-hold strategies on equity, bond and commodity markets as well as ‘option-based factors’ to control for nonlinear risk exposures of hedge fund positions (Fung and Hsied, 2002a; Agarwal and Naik, 2004). To evaluate the risk-adjusted performance of the suggested hedge fund model, the latter is compared with and contrasted against two rival and well established benchmark asset pricing models, namely the Fama-French (FF) three-factor model (Fama and French, 1993) and the Carhart (1997) four-factor model. The issue of time-varying variances and co-variances of hedge fund returns is also addressed along with its implications for the construction of optimal hedge fund portfolios. The empirical methods proposed in this chapter, which are employed to forecast hedge fund expected returns and variances-covariances and to construct optimal hedge fund portfolios in the context of mean-CVaR optimization, have been also evaluated as to their performance and robustness.Chapter 4 deals entirely with asymmetric dependence in the joint distributions of hedge funds returns and stocks (S&P 500 index), bonds (10-year Treasury bond) and commodities (S&P Goldman Sachs commodities index) and the importance of the former in asset allocation. We seek to model the dependence of assets’ returns and concentrate our efforts on defining the asymmetric dependence between them in the context of copulas. Formal definitions of the copula method, the most popular parametric copula functions, their properties and other related concepts are illustrated in the same chapter. In it, we also focus on a special case of the vines, the canonical vine copula model (Bedford and Cooke, 2002; Aas et al., 2009). We generalize the dynamic canonical vine copula model of Heinen and Valdesogo (2009) proposing an alternative Markov-switching dynamic canonical vine (MS-DCV) approach. Also among the aims of this chapter is to investigate the economic significance of asymmetric dependence of risky assets in portfolio decisions. In this perspective, we examine the portfolio choice problem of an investor with constant relative risk aversion preferences, who is willing to allocate his/her wealth to a set of risky assets including hedge funds, stocks, bonds and commodities.

Τα τελευταία χρόνια αποτελεί κοινή πεποίθηση ότι οι παραδοσιακές επενδυτικές επιλογές όπως οι μετοχές, ομόλογα και τα αμοιβαία κεφάλαια δεν μπορούν να καλύψουν τις διαφοροποιημένες, σε σχέση με το παρελθόν, ανάγκες των επενδυτών. Η αντιστάθμιση των κινδύνων, ο βαθμός διαφοροποίησης του χαρτοφυλακίου, η επίτευξη θετικών αποδόσεων ανεξάρτητα της πορείας των αγορών, είναι πλέον κάποιοι από τους στόχους που πρέπει να ικανοποιεί η σύνθεση ενός χαρτοφυλακίου. Η ικανοποίηση των παραπάνω απαιτήσεων έρχεται μέσα από ένα ταχύτατα αναπτυσσόμενο τομέα επενδυτικών υπηρεσιών όπως είναι τα hedge funds. Τα hedge funds αποτελούν ιδιωτικές επενδυτικές κοινοπραξίες οι οποίες οργανώνονται και διαχειρίζονται από επαγγελματία διαχειριστή επενδύσεων που αμείβεται με βάση την απόδοση των επενδύσεων.Σκοπός της διδακτορικής διατριβής είναι να συμβάλει στον εμπλουτισμό της επιστημονικής γνώσης, αφενός, με την κατανόηση και ανάλυση των χρηματοοικονομικών καινοτομιών που εισάγουν τα hedge funds στην επενδυτική κοινότητα. Και, αφετέρου, με την εφαρμογή ποσοτικών μεθόδων και την παρουσίαση προτάσεων για την αξιοποίηση του κλάδου των hedge funds ως μια σύγχρονη επενδυτική επιλογή η οποία μπορεί να προσφέρει σημαντικά πλεονεκτήματα όταν συνδυαστεί με τις παραδοσιακές επενδύσεις. Σε ένα πρώτο επίπεδο ανάλυσης η μελέτη εστιάζει στα ποιοτικά χαρακτηριστικά του κλάδου, τα σημεία διαφοροποίησης απο τις παραδοσιακές συλλογικές επενδύσεις, το μέγεθος των επενδυμένων κεφαλαίων, τις στρατηγικές και τη σχέση μεταξύ κινδύνου και απόδοσης που προσφέρει κάθε μία από τις βασικές κατηγορίες των hedge funds. Μετά την διαδικασία αξιολόγησης των χαρακτηριστικών κάθε στρατηγικής, σε ένα δεύτερο επίπεδο ανάλυσης, μας δίνεται η δυνατότητα να αξιοποιηθεί αυτή η γνώση για την κατασκευή δυναμικών χαρτοφυλακίων. Εφαρμόζοντας καινοτόμα οικονομετρικά υποδείγματα στα πλαίσια των θεωριών χαρτοφυλακίου μελετάμε και προτείνουμε την σύνθεση χαρτοφυλακίων με ιδιαίτερα χαρακτηριστικά όπως σημαντικές αποδόσεις με περιορισμό του κινδύνου, προστασία των επενδυτών σε καθοδικές αγορές.