Evolutionary Portfolio Optimization During the COVID-19 Crisis: An Empirical Analysis of the Indonesian LQ45 Index Using NSGA-II
TL;DR
Evolutionary algorithms offer a better and more efficient decision support system for institutional investors coping with severe macroeconomic shocks, computationally compared to traditional analytical techniques.
Abstract
Background: The global financial capital markets have faced unprecedented volatility during the COVID-19 pandemic which severely questions the validity and practicality of the classic portfolio optimization. Method: This research studies the capacity to optimise the capital allocation in the Indonesian LQ45’s stock index using Non-dominated Sorting Genetic Algorithm II (NSGA-II) in pandemic era. Method A bi-objective mathematical model specifically with the goal to maximize expected return and decrease portfolio variance was created, not requiring developing a full-scale data-driven software application, in order to be able to generate a non-dominated Pareto front. Results: The results from experiment show that the evolutionary algorithm managed to determine highly efficient portfolios and obtained an exceptional Sharpe Ratio of 1.034 under extreme economic stress. The best asset allocation offered to provide insight into a mode of strategic allocation which actively mixed high-yield equities with low variability defensive stocks - it eliminated the shortcoming of naive diversification. In conclusion, evolutionary algorithms offer a better and more efficient decision support system for institutional investors coping with severe macroeconomic shocks, computationally compared to traditional analytical techniques.