This paper addresses the inherent procyclicality in widely adopted financial risk measures, such as expected shortfall (ES). We propose an innovative approach utilizing the worst-case higher moment (HM) risk measure, which offers a robust solution to distributional shifts by incorporating adaptive features. Empirical results using historical S&P500 returns indicate that worst-case HM risk measures significantly reduce the underestimation of risk and provide more stable risk assessments throughout the financial cycle compared to traditional ES predictions. These results suggest that worst-case HM risk measures represent a viable alternative to regulatory add-ons for stress testing and procyclicality mitigation in financial risk management.