B0597
Title: Asymptotic analysis of portfolio diversification
Authors: Fan Yang - University of Waterloo (Canada) [presenting]
Chen Zhou - Erasmus University Rotterdam (Netherlands)
Abstract: Interconnectedness and heavy-tailedness of risks may result in high systematic risk. When the number of risks are fixed, maximizing diversification benefits can lower the portfolio risk. The diversification ratio, also known as the risk concentration, is used to measure the diversification benefit. We aim to construct the most diversified portfolio when risks are heavy-tailed and dependent, modeled by the multivariate regularly variation structure. Since explicit solutions to such optimization problems are generally not available and the computation burden of such solution is very heavy and not stable, asymptotic analysis is conducted as an alternative way to study them. The asymptotic solution is showed to be a good approximation to the true solution and the approximation error is given as well. Our theoretical results are supported by extensive numerical simulations. The estimation of the asymptotic solution is provided and then our strategy is also applied to real market data.
Journal