This study examines the performance of the risk parity approach in optimizing USD-denominated fixed income portfolios, particularly for institutional investors such as central banks. The research addresses limitations in the traditional Mean-Variance Optimization (MVO) method, which is highly sensitive to input data and prone to concentration risk. Risk parity, also known as Equally Weighted Risk Contribution (ERC), distributes risk evenly across assets without relying on expected return estimates, enhancing portfolio diversification. Using empirical data from 2014 to 2023, the study calculates risk contributions, constructs a covariance matrix, optimizes portfolio weights, and evaluates performance through the Sharpe ratio. The empirical findings reveal that although the MVO model achieved a slightly higher overall Sharpe ratio, the Risk Parity approach demonstrated superior stability and resilience across different market conditions, particularly during periods of high volatility such as the COVID-19 pandemic and post-pandemic financial tightening. These results suggest that Risk Parity, by emphasizing risk diversification over return forecasting, enhances portfolio robustness, making it an attractive strategy for conservative institutional investors focused on stability and capital preservation. The study highlights that portfolio optimization performance is context-dependent, with Risk Parity offering better protection in turbulent conditions, thereby supporting its broader adoption for risk-averse institutions in an increasingly uncertain global financial environment.