Paper Title
Portfolio Allocation of Crypto Currency under Linear and Nonlinear Constraints

Abstract
This paper aim to estimate weight portfolio allocation ofCryptocurrency. The traditional portfolio optimization is standard gradient-based optimization which has problem in case where the function is non-differentiable or is difficult to find. This paper proposes to use the mean/CVaR concentration to compare portfolio allocation ofCryptocurrency. The estimated results show that the non-linear constraint optimization with DEoptim provide the best expected return compared with the linear constraint optimization with the gradient-based optimization. Furthermore, we also provide the estimated weight of the non-linear constraint optimization with DEoptim. The result shows that the best balance are invested in ETH is highest which is 27.95 %. followed by LTC (27.79%), BTC(23.57%) and XRP(20.69%). The Return Rebalancing also confirms that optimal weight form the minimum CVaR Concentration portfolio has a better performance compared to equal-weight portfolio. Therefore, this study provides an empirical evidence supporting that non-linear constraint optimization trends to be a more appropriate constraint optimization for allocating the portfolio in Cryptocurrency. Keywords - Cryptocurrency, Portfolio Allocation, DEoptim