Performance-Based Incentives in a Dynamic Principal-Agent Model