Invasive Weed Optimization Algorithm

In financial markets, a primary concern for investors is achieving enhanced returns while effectively managing portfolio risk. Leveraged trading is one potential strategy for increasing returns; however, the fundamental question is to find the optimal degree of leverage that yields the most efficient investment portfolio. This paper investigates the optimal leverage level of cryptocurrency futures within a diversified portfolio comprising digital assets, employing a constrained mean semi-variance model and the Invasive Weed Optimization (IWO) algorithm. The dataset, sourced from Coin Market Cap, consists of daily returns for 10 cryptocurrencies and 5 futures contracts over the period from 2022 to 2025. The proposed model incorporates allocation constraints, wherein each asset in the portfolio is subject to upper and lower bounds on its weight. Due to the imposed constraints, the problem is not solvable via traditional quadratic programming techniques, necessitating the application of the IWO algorithm as the optimization method. Empirical results reveal that incorporating futures into a cryptocurrency portfolio does not inherently enhance its performance. While leverage may increase expected returns, it simultaneously elevates portfolio risk. Consequently, based on the Sortino ratio, the overall risk-adjusted performance of the portfolio does not necessarily improve with the use of leveraged futures.