Search for Auditors; Don’t Rotate

In March, the Public Company Accounting Oversight Board held hearings about whether to require public companies to change — or “rotate” — their external auditor periodically. Meanwhile, the European Union is proposing to require mandatory rotation every six or 12 years, and the lower house of the Dutch Parliament recently voted to require auditor rotation every eight years.

At the PCAOB hearings, various investor advocates and pension funds argued in favor of mandatory rotation. They found fault with the lengthy relationships between many auditors and the companies they audit — the auditors of almost 36% of all companies in the Russell 1000 have held that position for 21 years or more. According to the supporters of auditor rotation, this coziness creates a potential conflict of interest: an auditor’s desire to maintain a good relationship with its client could conflict with its duty to rigorously question the client’s financial statements.

Mandatory auditor rotation could reduce this conflict. Since auditors would know that their engagement would come to an end after a fixed period, they would have less incentive to curry favor with management. At the same time, mandatory rotation could encourage existing auditors to perform more thorough audits, because the firm would fear that a new auditor would expose any previous errors or omissions.

On the other hand, public companies vigorously protested that the benefits of mandatory rotation would be outweighed by its costs. Because multinational corporations are very complex, an auditor must develop extensive company-specific knowledge to understand the company’s finances. Studies have shown that audit quality is relatively poor in the initial years of an auditor’s engagement, largely because the auditor is unfamiliar with how the business works. A policy of mandatory rotation would increase the frequency of these “initial years.”

Furthermore, newly appointed audit firms need to put forth a significant upfront investment in order to gain the necessary institutional knowledge — an investment that probably will be passed on to clients in the form of higher audit fees. In a 2003 Government Accountability Office survey, audit firms estimated that first-year audit fees could rise on overage by more than 20% under a policy of mandatory rotation. (The survey can be found at www.gao.gov/new.items/d04216.pdf.)

Perhaps most important, mandatory rotation undermines the role of the audit committee in overseeing the audit process, as expanded by the Sarbanes-Oxley Act. That act shifted the reporting relationship of the auditors to the company’s independent audit committee from its executives; the audit committee now has the power to appoint and terminate the auditor. To bolster the role of audit committees, the PCAOB should give them more discretion in choosing an auditor, rather than impose a one-size-fits-all policy of mandatory rotation.

Given these competing arguments, I favor a compromise proposal. Specifically: The PCAOB should mandate that the audit committee of every public company issue a request for proposal every 15 years for the audit engagement, but they should allow the existing auditor to bid on this RFP.

My proposal would help keep existing auditors on their toes. The existing auditor would be concerned that it could be replaced by a new auditor, which would catch any mistakes that have been made. Yet this proposal would not automatically impose the large costs associated with mandatory rotation. The independent directors on the audit committee would only decide to hire a new auditor if such a choice met a cost/ benefit test in that particular circumstance.

While an incumbent auditor would have the inside rail in this RFP process, I believe these RFPs would typically lead to legitimate bidding contests. As shown by the bidding process for many large defense contracts, the benefits of competitive bidding can be achieved with two serious bidders. The prospect of 15 years’ worth of audit fees could certainly encourage at least one other large audit firm to seriously respond to an RFP. Audit fees for 15 years might even persuade one or two midsize audit firms to develop the capability of auditing multinational companies.

One potential bidder might be a firm that provides non-audit services to the company in question. Such a firm would have an advantage in the bidding process because it already has some familiarity with the company. Of course, an audit firm that performs non-audit services for a company cannot legally be that company’s auditor. However, the regulators could allow any qualified firm to respond to an RFP on the condition that it cease providing non-audit services in the event it wins the RFP.

Regardless of the number of bidders, an RFP process would reinforce the critical role of the audit committee. The prior system encouraged auditors to ingratiate themselves to management — potentially by neglecting to ask tough questions. By contrast, an RFP process would give auditors an incentive to make great efforts to serve the needs of whoever controlled that RFP process — the independent directors on the audit committee. To serve the needs of those directors, the auditor would try to be more forthright about debates with management and red flags in financial statements.

In short, advocates of auditor rotation identify a real problem: a potential conflict of interest that could dissuade auditors from vigilantly exercising professional skepticism. Although auditor rotation could mitigate this conflict, a mandatory approach would be very expensive and could undermine the role of the audit committee.

Instead, the PCAOB should require the audit committee to issue an RFP for the auditor engagement every 15 years, but allow the existing auditor to participate in the bidding process. Without imposing a rigid system of mandatory rotation, this process would enhance the auditor’s willingness to make tough calls and reinforce its primary allegiance to the independent audit committee.