Audits of internal control over financial reporting (ICFR) are typically “integrated” with the audit of the financial statements (FSs)—both audits are conducted by the same audit firm, which designs procedures to satisfy the objectives of both audits simultaneously. A common assumption is that integrating the two audits is more effective and efficient than performing them separately. However, this assumption has not been tested empirically. Using a sample of Chinese companies that employ different audit firms for their FS and ICFR audits (i.e., nonintegrated auditors), we find evidence that challenges this assumption. Specifically, we find ICFR audit quality is higher for nonintegrated audits compared to integrated audits. Moreover, total audit fees are lower for nonintegrated audits, despite higher ICFR audit fees.

Using two natural field experiments, we examine how warning individual retail borrowers that their loan performance will be reported to a public credit registry before and after the loan take-up affects their borrowing behavior. We show that credit warnings reduce default rates by 3.7 to 7 percentage points and increase loan take-up rates by 4.1 percentage points, which suggests that credit warnings benefit both lenders and borrowers. The main drivers appear to be borrowers’ anticipation of a reduction in lenders’ informational rents and improved repayment incentives. Moreover, the reduction in default rates is comparable for borrowers who receive the credit warning before and after the loan take-up. As credit warnings received before but not after a loan take-up can affect the borrower pool, and thus the overall credit risk of the pool, the results suggest that credit warnings have little net effect on the pool’s credit risk due to selection.