Abstract
The thesis investigates the existence of herding behaviour in the Johannesburg Stock Exchange. In addition, adopting a descriptive theory of decision-making, the thesis explores the possibility of adding value to investors‟ portfolio by investing solely in stocks driven by human mentality and psychology. Data were obtained from the INET BFA Expert - Iress Database and comprised the universe of listed stocks in the financial industry of the Johannesburg Stock Exchange spanning the period from January 2010 to October 2016.
The thesis is organised in two phases and contributes to the field of financial economics specifically behavioural economics and portfolio management and bridges the gap between the two fields. The thesis offers an intuitive and a psychologically corroborated descriptive investment strategy capable of adding value to investors‟ portfolio. While the first phase of the thesis highlights and describes key investor behaviours which are largely at variance with the rational assumption documented in the behavioural economics literature, the second phase incorporates investors‟ psychology in the stock selection and portfolio optimisation.
The two initial empirical chapters (i.e. Chapter 3 & Chapter 4) were primarily devoted to searching evidence of herding behaviour in the Johannesburg Stock Exchange1. Three advanced methodologies were adopted in testing evidence of herding behaviour. Chapter 5, the last empirical chapter adopts a descriptive decision theory, the Cumulative Prospect Theory and the Mean-Variance portfolio optimisation criterion to optimise and evaluate classified and formulated portfolios based on the Cumulative Prospect Theory.
Following Chapter 2 the literature review, Chapter 3 tested evidence of herding behaviour both at the industry and the sectoral levels adopting the quantile regression model. At the sectoral level, herding behaviour showed asymmetry. While investors in the banking sector exhibited the herding behaviour during the bear market phase, in the real estate sector, investors suffered from the behavioural bias during the bull market phase. However, in the entire financial industry, the results showed evidence of herding behaviour during the bull market phase only.
Likewise, Chapter 4 compared results of two conventional approaches with the Bayesian model in testing evidence of herding behaviour. Apart from the insurance sector, the results showed evidence of herding behaviour in the rest of the sectors during the bear and the bull market phases using the conventional approaches. Similarly, using the conventional approaches and the Bayesian models, investors in the entire financial industry showed evidence of herding behaviour.
Portfolio optimisation results in the last empirical chapter consistently showed that stocks with extremely lower Cumulative Prospect Theory values outperformed stocks with extremely higher Cumulative Prospect Theory values. The results further established the superiority of the Cumulative Prospect Theory as an empirically corroborated theory of decision-making with rich psychological content.
Ph.D. (Economics)