Comparative analysis between sustainable index and non-sustainable index with genetic algorithms: Application to OECD countries

Abstract

This study analyses the differences in financial portfolio metrics between sustainable index and non-sustainable firms in the market index through the use of the portfolio theory and genetic algorithms from 2007 to 2013. The sample consists in 926 firms of four regions (1) Europe: Germany, Austria, Denmark, Spain, Finland, Italy, Norway, Sweden and United Kingdom, (2) Asia: Japan, (3) America: Canada, United States of America and Mexico and (4) Oceania: Australia. To measure the performance of the portfolio two classical metrics: Jensen’s alpha and Sharpe ratio were considered. We also calculate a conditional metric that measures the number of times the return of a given portfolio exceeds the average market return. The goal is to find a portfolio that maximizes these three metrics using a weighted ratio and compare the results between the sustainable and non-sustainable portfolios. Due to a nonlinear programming problem, we use genetic algorithms to obtain the optimal portfolio. The results show a better performance in sustainable portfolios in eight countries, although the amount of countries increases if only the conditional metric is considered.