Variance and Intuitive Risk Heuristics

Abstract: This thesis attempts to further the understanding of what risk actually constitutes in a financial setting. More specifically, we hone in on the definition of risk used in Markowitz's portfolio selection framework. The existence of the human bias of risk aversion is established, and models purporting to manage that risk abound. Even so, few of those models have taken the time to justify the use of their chosen risk measure. The absence of discussion has led to variance being the oft-chosen unexamined stand-in for how people intuitively evaluate risk, and Markowitz's portfolio selection framework is no exception. This thesis examines the extent to which Markowitz's choice of variance was sound. We also devise a new risk measure that attempts to bring some ideas brought forth by Prospect Theory into the realm of finance, and conjointly test it with variance as a potential replacement.In order to answer the questions posed by this thesis, a survey is conducted where the respondents are asked to rank hypothetical portfolios. The data is analyzed to reveal genuine preferences at group level. The findings suggest that variance leaves much to be desired in terms of being a suitable stand-in for how people intuitively evaluate risk. The risk measure of our own making did fare better, but the tests reveal that it is far from perfect.