Diversification of Stationary Data for Optimizing Risk of Portfolio

Diversification is a process of distributing capital that minimizes the exposure of each individual asset. The purpose of diversification is to minimize risk. This paper examines how these two strategies (stationary and non-stationary) can minimize portfolio risk. To achieve this, data on palm oil and copper from 2010 to 2016 are explored, and two methods are used: correlation and mean-variance. Diversification has different degrees, and correlation is used to check the degree of diversification of the two types of data, while the mean-variance method (MV) is used to estimate the risk of the two datasets after the correlation. The analyses in this paper show that the correlation of stationary data leads to moderately weak diversification, while non-stationary data results in very weak diversification. In estimating the risk of the two datasets, stationary data diversifies 93% of the risk, while non-stationary data diversifies 6%. This shows that stationary data minimizes portfolio risk more effectively than non-stationary data.