Are CFOs and Auditors Too Cozy?

Familiarity may breed contempt, as the old adage claims, but not in corporate accounting, according to some new research.

In accounting, the term "familiarity threat" refers to the threat to auditor independence that arises when a CFO or other top executive of a company being audited was formerly employed by the accounting firm conducting the audit. In the years leading up to the notorious corporate accounting scandals at the turn of the century, about one third of the largest U.S. corporations, including such disgraced firms as Enron, Global Crossings, and Waste Management, had top executives who were alumni of their companies' external auditors.

It is hardly surprising, then, that the principal legislation in response to those scandals, the Sarbanes-Oxley act (SOX) of 2002, took aim at this alumni effect, section 206 decreeing it unlawful for accounting firms to perform audits if a top financial or accounting executive of the client was employed by the auditor during the preceding year. Unsurprisingly too, Canada, the E.U., and the U.K. have imposed similar prohibitions, Canada for one year and the E.U. and U.K. for two.

Are such restrictions an effective answer to familiarity threat? A paper in the March issue of the journal Accounting Horizons, published by the American Accounting Association, suggests not.

The new research, the first experimental study post-SOX of the alumni effect among North American auditors, tests the willingness of Big-4 managers to adopt a client’s position on a conjectural accounting matter. It finds that 76% do so if the client’s CFO is a former colleague at their Big-4 audit firm, while only 44% do so if the CFO is not. And the alumni effect occurs even if it has been two years since the CFO left the audit firm, double the minimum span required in the U.S. and Canada and the same as the minimum mandated in the U.K. and E.U.

"Obviously, a one-year or two-year cooling-off period is not enough to avoid the alumni effect, particularly if it requires overcoming social bonds that colleagues often develop," comments Michael Favere-Marchesi of Simon Fraser University's Beedie School of Business, who carried out the study with Beedie colleague Craig E. N. Emby. "It may be that five or ten years would be enough. Alternatively, it may be that audits of companies where a CFO or other higher-up is a former engagement partner should be banned entirely, as some research on auditor independence has suggested."

He adds: "How the accounting profession or corporate world would react to proposals for such restrictions or prohibitions is, of course, another matter.”

The study’s findings are based on an experiment conducted via the Internet with 140 managers of Big-4 accounting firms in offices throughout Canada and the U.S. The managers, who averaged about seven years of auditing experience, all received the same background information about a corporate client and its industry as well as a draft of the current year’s financial statement.

In addition, participants were randomly assigned to one of three experimental conditions, as follows:

■   One group was told that until two years ago the client company’s CFO was a partner in their accounting firm, a colleague with whom they worked on engagements involving this very client and others;

■   A second group was told that the CFO formerly attained partnership at another Big-4 accounting firm;

■   A third group received no information about the CFO’s prior employment history.

Participants were asked to assume the role of a continuing audit manager on the account. The key issue in the experiment was the valuation of goodwill, an asset on corporate balance sheets that arises when a firm purchases a company for more than the fair value of its net assets, goodwill representing the difference between that identifiable fair value and the purchase price. The client being audited was a medium-sized company in the biotechnology field, where purchased goodwill is often an accounting issue. The valuation of goodwill can change from year to year, depending on a variety of circumstances, and is widely regarded by accountants as a fairly subjective, even speculative, matter.

In the case at hand, the CFO maintained that the value of goodwill should be unchanged from the level of the previous year. Evidence to that effect, however, was mixed. Participants received information that, in the words of the study, was “sufficiently negative to suggest that goodwill impairment might be a definite possibility but not so overwhelmingly negative that subjects would automatically conclude impairment.”

In the first group of participants, who were told the CFO was a former colleague and engagement partner, 35 of 46, or 76%, agreed with the CFO that goodwill was not impaired and should be set at its prior level. In the second group, who were told the CFO was formerly with another Big-4 accounting firm, 23 of 48, or 48%, agreed to no impairment, and in the third group, who received neither of those indications, only 39% agreed.

In sum, being told the CFO had formerly been a Big-4 partner inclined participants to agreement on goodwill impairment but not nearly as much as the alumni effect did. And least likely of all to be swayed were participants whose CFO had neither alumnus status nor Big-4 imprimatur.

The professors also tested the degree of confidence in their decisions among auditors who agreed with the CFO that goodwill should not be impaired. Those in the first and second groups reported, on average, about equal levels of confidence, with both being significantly higher than the average confidence in the third group.

The results lead the authors to observe that “the fact that the current auditors are more likely to concur with the client’s position when the client is a former audit partner of their firm can be interpreted as indicating reduced professional skepticism, but the results of the [confidence] test…suggest that source credibility also influences the auditors’ confidence in a CFO’s position.”

In conclusion, they urge regulators “to push for a more robust cooling-off period covering a wider range of management positions,” noting that the possibility of a longer period has been raised by the SEC’s Public Company Accounting Oversight Board and adding that even an outright ban on auditors’ taking jobs with clients has been proposed in some academic research.

The study, entitled "The Alumni Effect and Professional Skepticism: An Experimental Investigation," is in the March issue of Accounting Horizons, published quarterly by the American Accounting Association, a worldwide organization devoted to excellence in accounting education, research, and practice. Other journals published by the AAA and its specialty sections include The Accounting Review, Auditing: A Journal of Practice and Theory, Issues in Accounting Education, Behavioral Research in Accounting, Journal of Management Accounting Research, Journal of Information Systems, Journal of Financial Reporting, The Journal of the American Taxation Association, and Journal of Forensic Accounting Research.