This thesis examines the use of multiple measures of performance, both aggregate and disaggregate, in compensation contracts with executives. A multiple action, multiple measure agency model is used to establish a role for additional measures of performance in compensation contracts where a congruent measure of shareholder wealth is also available. Although the shareholder wealth or market measure incorporates information in other publicly available measures considered relevant to the market for valuation of the firm, the additional measures of performance are found to be valuable for contracting. This result suggests that principals look beyond or adjust the market's assessment of firm value in developing optimal compensation schemes with managers.
The model further explicitly characterizes the weights placed on the measures and finds that the signs of the weights on the shareholder wealth measure and the additional measures reflect the direction of the sensitivity of the measures to the underlying action(s) of the manager and that, ceteris paribus, the weights on the measures are increasing in their precision and decreasing in the precision of the shareholder wealth measure. Further analyses find that multiple relative and firm-specific performance measures are valuable together in contracts.
Empirical tests of the relation between multiple measures of bank performance, including a market measure, and CEO compensation contracts are conducted using a sample of commercial banks for the period 1991 through 1994. The analyses provide evidence that accounting information, including changes in aggregate earnings, earnings components and other bank specific measures disclosed in financial and regulatory reports, is associated with changes in CEO annual cash compensation incremental to a shareholder wealth measure. Additional empirical tests provide support for the hypothesis that performance measure weight is increasing in the precision of the measure. Tests of the use of relative and firm-specific performance measures together in contracts suggest that banks rely on information in firm-specific measures rather than in peer measures in determining CEO annual cash compensation. Also, tests incorporating additional, long-term components of compensation suggest that other factors need to be considered to explain the relation between changes in long-term components and changes in bank financial performance.
My Research