Optimized portfolio using a forward-looking expected tail loss

Abstract In this paper, I construct an optimal portfolio by minimizing the expected tail loss derived from the forward-looking natural distribution of the Recovery Theorem. This natural distribution can be used as the criterion function in an expected tail loss portfolio optimization problem. I find that the portfolio constructed using the Recovery Theorem outperforms both an equally-weighted portfolio and a portfolio constructed using historical expected tail loss. The portfolio constructed using the Recovery Theorem has the smallest historical tail loss, smallest maximum drawdown, highest Sortino Ratio, and highest Sharpe Ratio.