An Exploration of Bank Confirmation Process Automation: A Longitudinal Study

This exploratory study examines automation of the bank confirmation process using a longitudinal dataset from the largest third-party U.S. confirmation service provider, supplemented with informal interviews with practitioners. We find a significant increase in electronic confirmation use in the U.S. and internationally. Errors requiring reconfirmation were less than 2 percent of all electronic confirmations. Errors made by auditors were almost five times more likely than errors by bank employees. Most auditor errors involved use of an invalid account number, although invalid client contact, invalid request, and invalid company name errors increased recently. Big 4 auditors made significantly more confirmation errors than did auditors at non-Big 4 national firms. Error rates and error types do not vary between confirmations initiated in the U.S. and those initiated internationally. Three themes emerged for future research: authentication of evidence, global differences in technology use, and technology adoption across firms of different sizes.