This research paper tests the effectiveness of Power-Log optimization for managing the downside risk of investment portfolios. It uses Power-Log utility functions, which are based on tenets of behavioral finance, to give investors the ability to build downside protection directly into a portfolio. Comparing optimal Power-Log portfolios with matched mean-variance efficient portfolios, we find that the optimal Power-Log portfolios have lower downside risk, while delivering higher geometric average return. They also provide much better downside protection against unanticipated market shocks, such as the one in 2008, in contrast to the disastrous performance for matched mean-variance efficient portfolios. Power-Log optimization succeeds in managing downside risk effectively, while mean variance analysis fails to protect investors from such risk.