Modeling auditor utility functions

Utility functions for profits and losses are obtained on 14 practicing auditors. Data were elicited through personal interviews with audit partners and seniors. The appropriate utility functions that describe the risk characteristics of the auditors and fit the certainty equivalent responses are found using a general summed exponential function and nonlinear programming to obtain its parameters. The richness of the summed exponential in assuming a variety of utility shapes and its mathematical tractability, for example, in manipulating models makes it attractive for expressing an individual's utility function. Use of the summed exponential mitigates the difficulties in assessing a utility function by simplifying the assessments and circumvents the need for the analyst to decide on the appropriate functional form. The empirical results show, for example, that a firm's seniors are less averse to losses than its partners. With respect to gains, a variety of utility shapes/risk preferences are exhibited with most being risk averse for large sums of money. The implications for an auditing firm are then discussed.