Spectral risk measures in portfolio selection problems

This thesis examines spectral risk measures. Spectral risk measures, as a subset of coherent risk measures, satisfy all the crucial and reasonable properties that a risk measure should have. A specific characteristic of a spectral risk measure is that it makes it possible for an investor to quantify the risk that arises due to holding a selected group of assets based on his or her personal attitude towards risk. The aim of this bachelor thesis is to discuss the properties of spectral measures of risk and their relations to commonly known measures of risk, but primarily to scrutinize its utilization in the portfolio selection problem. Based on monthly returns of stocks from chosen American stock exchanges we compute the optimal portfolios of stock indices for different risk aversion functions, and consequently we make an analysis of the results. Powered by TCPDF (www.tcpdf.org