The date on which a portfolio is rebalanced can have a tremendous impact on realized performance. We demonstrate that a strategy rebalanced on different dates using the exact same investment process can exhibit return differentials of over 20% across short periods of time. These differences are entirely explained by path dependency, also known as rebalancing luck. In addition, we cannot statistically validate that any particular date is favorable over another. Therefore, our best estimate for the future performance of any individual date is simply the average performance across all dates. Any deviation from that average is an uncompensated source of risk. To mitigate this risk, we use portfolio partitioning which combines the weights of different portfolios using the same investment strategy and rebalancing frequency but are simply rebalanced on different dates. The result is a portfolio that neutralizes path dependency and is much less susceptible to random noise.