Abstract In 1978 the Cohen Commission concluded that the required financial statement disclosure of information pertaining to accounting principles made auditor reporting on consistency unnecessary. Accordingly, the Commission recommended that the audit report be revised to omit any reference to consistency, and the Auditing Standards Board followed the Commission's view in the proposed statement on audit reports. However, the final version of the statement--Statemenf on Auditing Standards No. 58--did not eliminate the consistency modification in situations where clients make a change in principles. Both Board members and staff of the Chief Accountant of the Securities and Exchange Commission voiced concern at the paucity of research regarding the usefulness (or lack thereof) of the consistency modification. This study assesses whether the consistency modification is useful by examining stock returns around the time that the audit report was released for 293 firms that early-adopted Statement of Financial Accounting Standard No. 87--Employers' Accounting for Pensions. One hundred and forty-three firms received consistency modifications and 150 firms did not receive consistency modifications. Two types of tests are conducted. The first test compares the stock returns for firms receiving and not receiving consistency modifications. The second test uses multiple regression to isolate the effects of the consistency modification on share prices. Separate analyses are conducted for subsets of firms for which highlighting or signalling arguments are most plausible. The findings in this study do not indicate that stock prices are affected by the receipt of a consistency modification. Based on this result, elimination of consistency modifications would not result in loss of information to the equity markets. Comparing this result to conclusions of studies examining "subject to" qualifications suggests that the information contained in consistency modifications may be of less significance than information that was communicated in "subject to" reports.
Mittelstaedt et al. (Sun,) studied this question.