Back to search

Πανεπιστήμιο Μακεδονίας

Asset pricing and prediction with time-varying betas

Abstract

dc:description

The theory of asset pricing relies heavily on the principles of present value calculations and the hypothesis of efficient capital markets. The former implies that the price of an asset, not necessarily stock, is a function of the expected future yields discounted to the current data. However, in the field of investments, there seems to be an important but unanswered question. This particular question is concerned with why some assets earn substantially higher average returns. Financial economists have developed different models to address this question in the context of theoretically or empirically motivated asset pricing models. The CAPM is considered to be the most widely used asset pricing model. However, in view of its empirical shortcomings, researchers have made many attempts to refine the theoretical foundations and improve its empirical performance. One of the main assumptions of the CAPM is the stability of beta coefficients. Because this assumption does not hold in reality, in this thesis we develop a novel approach for capturing the time variation of betas by treating the pattern as a function of market return. To do so, we construct a new two-factor model (TFM) which incorporates variables targeting to absorb the information conveyed by betas’ instability. The model is free from subjective bias problems related to the selection of a critical threshold. In addition, the important implications of predicting betas motivated us to examine empirically the accuracy of our TFM’s betas. Our analysis uses stocks traded on S&P 500 and covers a long period of time and different kinds of portfolios. The tests for finding out the model’s performance along with other well-known models proposed in the literature indicate that this novel approach is very promising. In addition, in this thesis we examine the existence of herding behaviour in different financial markets. Herding behaviour can have significantly consequences on the market efficiency, since it might aggravate volatility of returns destabilizing financial markets. Furthermore, nonlinearities of stock prices have been attributed to the existence of herding. Hence, in order to obtain a better understanding of herding, we examine its effects on market volatility. Furthermore, we seek for the existence of herding in factors other than the market risk such as co-skewness and co-kurtosis. Finally, we explore the existence of contagion effects of herding as well as if the measure of herding is affected by various unexpected macroeconomic shocks. We believe that the empirical findings reported in this thesis will help market participants and investors in many ways.

Degree

thesis:*
Grantor dc:publisher
Πανεπιστήμιο Μακεδονίας
Year dc:date
2015

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Messis, Petros
  • Μεσσής, Πέτρος

Subjects

dc:subject × 7

Rights

dc:rights
Statement dc:rights
  • BY_NC_ND
Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/35361

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Messis, Petros; Μεσσής, Πέτρος. Asset pricing and prediction with time-varying betas. Πανεπιστήμιο Μακεδονίας, 2015. http://hdl.handle.net/10442/hedi/35361