中文摘要：

以自利性假设为基础的标准代理理论认为，相对业绩评价是消除管理层薪酬合约中系统性风险、改进激励效果的最优策略。然而，实证研究并未获得相对业绩评价广泛应用的充足证据，由此导致了“相对业绩评价之谜”。通过修正代理人自利性假设，构建基于公平偏好理论的相对业绩评价模型，发现在公平偏好的影响下最优薪酬合约是选择相对业绩评价以实现最优风险分担、提供激励，与选择团队业绩评价以降低不平等贴水之间权衡的结果；当代理人公平偏好较强时，相对业绩评价不再如标准代理理论所预测的是促进代理人激励的最优制度安排。这一研究结论不仅为相对业绩评价在实践中的较少应用提供了行为经济学视角的理论解释，同时对于如何设计有效的管理层薪酬合约具有一定的实践指导意义。

英文摘要：

According to the principal-agent theory based on the assumption of self-interest, relative performance evaluation （RPE） is an optimal strategy of reducing the managerial systematic risk exposurse thus improving their effort. However, there is no sufficient empirical evidence in support of this result, there by forming the so-called ＂ relative performance evaluation puzzle＂. This paper constructed a RPE model based on fairness preference by revising the agent self-interest hypothesis, and found that under the effects of fairness preference the optimal contract is the trade-off between choosing RPE to reduce risk exposures, thus providing incentive and team performance evaluation to reduce income inequity. And RPE will not be the the optimal system arrangement motivating the agents as predicted by the standard agency theory when the agents have strong fairness preference. This conclusion not only provides a theoretical explanation from the perspective of behavioral economics for the empirically observed lack of relative performance evaluation, but also has a certain practical significance for designing an effective payment contract of management.