Starting in 2018, new rules required disclosure of auditor tenure in audit reports. (See this PubCo post.) And, for some companies, those tenures can stretch over a century. For example, KPMG reported that it has audited GE since 1909. (See this PubCo post.) According to this press release from the American Accounting Association, for “the first 21 companies of the Dow 30 to release their reports this year, the average auditor tenure is 66 years.” But long auditor tenure has its critics and its fans. Some argue that long tenure can adversely affect auditor independence and objectivity, while others contend that long tenure avoids the time loss and distraction of having to “onboard” new auditors, provides deep institutional knowledge—leading to higher audit quality—and offers cost savings resulting from that familiarity. However, a couple of recent academic studies call those suggested benefits into question. The press release cited above and this article in CFO.com report on new academic research that concludes that, among the Big 4 at least, the longer the tenure, the greater the fee, notwithstanding the reduction in effort required of the auditor over time. And this press release from the American Accounting Association reports on another academic study that, contrary to popular assumptions, found a positive correlation between relatively short audit tenure and the speed of discovery of financial misreporting. Will these studies renew calls for mandatory auditor rotation?
As discussed in this PubCo post, both ISS and Glass Lewis recommended voting against a proposal to ratify the appointment of GE’s auditor, KPMG, at the 2018 GE annual shareholders meeting, a pretty unusual event—the recommendation, not the meeting. In an even more rare occurrence, about 35% of the shareholders did not vote to retain KPMG (that includes an “against” vote of over 2.2 billion shares). Among the factors cited by ISS for its negative recommendation were audit quality and auditor tenure. Glass Lewis indicated that it usually supports management’s choice of auditor except when GL believes the auditor’s “independence or audit integrity has been compromised.” In its analysis, GL raised the same concerns as ISS, noting in particular the large increase in fees to KPMG in the prior year, as well as its long tenure as GE’s auditor, which has “thrown KPMG’s effectiveness and relationship with the Company into question.”
In the U.S., only audit partner rotation is mandatory. However, mandatory auditor rotation has popped up for decades on everyone’s shopping list of favorite corporate governance reforms: former SEC Chair Richard Breeden imposed mandatory 10-year auditor rotation on WorldCom, when he was the court-appointed monitor for WorldCom following the scandals there, and SOX required the GAO to study the possible effects of mandatory audit firm rotation and report to Congress on its findings. In 2011, the PCAOB floated a mandatory auditor rotation concept release. (See this news brief.) At that time, PCAOB Chair Jim Doty appeared to be a strong advocate of auditor rotation, arguing that any serious discussion of independence, skepticism and objectivity “must take into account the fundamental conflict of the audit client paying the auditor. That leads to consideration of firm rotation as a counterweight to that conflict…” Barely a few months after the concept release was issued (and probably after having his ear bent by more than a few audit firms, CEOs and directors), Mr. Doty walked back his strong position, acknowledging that mandatory auditor rotation would present “significant operational challenges.” (See this news brief.) Finally, in 2014, the PCAOB indicated that it was not pursuing a mandatory auditor rotation project, although the idea remained on its “activity list.” (See this news brief.) But the idea has not since resurfaced.
The first academic study looked at audit-firm tenure and pricing information from 2000 through 2013, totaling “21,855 company-years’ worth of data from continuous relationships between audit firms and corporate clients, with audits divided about equally between Big 4 and non-Big 4 firms.” According to the study, audit fees for the Big 4 increased “on average by 13% between the first year and the second, by about 22% between the first year and the third, then creep up to 28% above the original fee by year 12 and 32% above it by year 14.” This pattern of fee increases was not prevalent among audit firms outside the Big 4; among those firms, fees remained flat or even declined slightly over time. Moreover, the study authors did not attribute the fee increases “to greater workloads that may result as client firms grow bigger or more complex, since the study controls for such factors.” In fact, the study found that average workload declined as auditor tenure increased, “falling by about 6% between years one and three, about 10% by year five, about 14% by year six, and about 16% by year nine.” (Note that the study’s methodology was to use “audit-report lag, the amount of time from the end of firms’ fiscal years to issuance of their annual financial reports,” as a proxy for workload, assuming that the time lag should “become shorter as audit firms are able to complete audit testing more efficiently and provide the desired level of assurance with less audit effort.”) In summary, the study authors conclude that these audit firms “‘charge a sizable tenure-linked fee premium [even as] less audit effort is needed.’” But the authors’ conclusions are not limited to cost. Rather, they suggest, the economics can affect auditor independence: the results indicate “a need to better monitor auditor independence and audit judgments when tenure is long, especially for Big 4 auditors, because economic bonding between the audit firm and client tends to increase over time.”
Another academic study, reported earlier this year, concluded that mandatory rotation of audit partners—as opposed to firm rotation—required by SOX, “did not eliminate the negative effect of long auditor tenure on audit quality.” This study looked at 3,465 corporate misstatements by U.S. companies during a 14-year period, focusing exclusively on serious accounting errors that occurred and were corrected during the tenure of the same auditor. The question examined was how the duration of auditor tenure affected the auditors’ speed in addressing misstatements. The study found that, in about 35% of these cases, the error occurred only in one quarterly financial statement, but not in the subsequent annual financial statements; in the other cases, the misstatements were repeated in one or more annual financial statements on which the auditors opined. The study authors found a positive correlation between auditor tenure and misstatement duration: “‘In other words, auditors with shorter tenures are faster to discover financial misreporting.’ For example, when auditor tenure was three years or less, the average misstatement duration was a little less than a year, whereas when it was 11 or more years, average duration was about a year and a half, more than 50% greater.”
However, the authors also found that this disparity tailed off after tenure of about 10 years: “Up until 10 years, they find, ‘a one-year increase in auditor tenure increases the misstatement duration by approximately 2.02 percent [so that] on average, misstatements are 18.18 percent longer after 10 years of auditor tenure than after one year. Beyond 10 years of auditor tenure, the association between auditor tenure and misstatement duration is insignificant. In conclusion, the benefits of a fresh look exist only in the first 10 years of the auditor-client relationship.’” (The 10-year timeframe has particular significance in light of the EU mandate that certain entities engage in a “tender process,” i.e., put the audit out for bid, after auditor tenure of 10 years, although they may then decide to extend the tenure period for another 10 years.)
The study authors also looked at the effect of the change in auditors compelled by the collapse of the audit firm Arthur Andersen and reached a similar conclusion. The authors compared the duration of accounting misstatements that began when Andersen was auditor and ended after a switch to a new auditor against the duration of misstatements for companies that retained a single Big-4 auditor over the same time period. The authors found that the duration of misstatements for the companies that retained the same auditor was on average 15% longer, a difference they concluded was statistically significant. The authors contended that their study “provides telling evidence in favor of renewed consideration of mandatory auditor rotations. ‘Overall,…these results support the PCAOB concern that a long auditor-client relationship may compromise audit quality.’”
Advocating for the other side is this study published in The Accounting Review, reported in CFO.com. This study reaches the conclusion that auditor rotation actually impairs professional skepticism. Skepticism is “a perspective universally viewed as essential to effective auditing. The primary reason traditionally advanced to require rotation is that it encourages skepticism.” However, this study turns that assumption on its head: “’Professional skepticism requirements are intended to elevate auditors’ skepticism of their clients and, ultimately, audit quality,’ the paper says. ‘This benefit disappears and even reverses when auditors rotate. That is, rotation and a skeptical mindset interact to the detriment of audit effort and financial reporting quality.’” The authors suggest that rotation appears to increase the incidence of aggressive financial reporting together with “low-cost, low-effort” audits, boosting the likelihood of audit failures, most significantly during the first two years following rotation. Not that these factors are intentional, the authors contend, but rather reflect a “subtle psychological effect about which [audit] decision-makers rarely have accurate insight.” Analyzing their evidence, the study authors speculate that “‘[r]otating auditors, aware that they will not be in a long-term relationship, will … likely perceive themselves to be less competent in evaluating the honesty or dishonesty of the [corporate] manager relative to auditors who do not rotate.’ As a result, ‘rotating auditors would find it difficult to garner psychological support for the probability of manager dishonesty, leading them to be less likely to choose high levels of audit effort than non-rotating auditors.’” The study authors caution that new auditors “should be extra vigilant in fraud planning and procedures.” (See this PubCo post.)