This paper formulates a number of new portfolio optimization models by adopting the sample median instead of the sample mean as the efficiency measure. The reasoning behind this is that the median is a robust statistic, which is less affected by outliers than the mean. In portfolio models this is particularly relevant as data is often characterized by attributes such as skewness, fat tails and jumps that are incompatible with the normality assumption. Here, we demonstrate that median portfolio models have a greater level of diversification than mean portfolios, and that, when tested on real financial data, they give better results in terms of risk calculation and concrete profits.
|Title||Format (Language, Size, Publication date)|
|Paper||pdf (it, 381 KB, 12/10/12)|
via Cantarane, 24
VAT number 01541040232
Italian Fiscal Code 93009870234
© 2020 | Verona University | Credits