Zvi Singer of HEC Montreal and Jing Zhang of the University of Alabama in Huntsville found that auditors discover misstatements far more quickly within the first three years of an audit appointment but, by the time 10 years are up, the quality of their audit is beginning to wane.
And the pair argue that if US regulators are looking to mandate auditor rotation, the maximum period of engagement should be set at less than 10 years.
“Within the first 10 years, a one-year increase in auditor tenure will prolong the misstatement duration by approximately 2.02%,” they say. “This means that, on average, misstatements are 18.18% longer after 10 years of audit tenure than after one year. Beyond 10 years of auditor tenure, the association between auditor tenure and misstatement is insignificant."
This, they suggest, points to a deterioration in audit quality starting early in the auditor-client relationship.
In fact, the research, published in the current edition of The Accounting Review, reveals that a longer audit tenure allows misreporting to accumulate and increase. It found “a positive and significant association between audit firm tenure and misstatement magnitude, suggesting that longer audit tenure not only results in less timely detection of misstatements, but also leads to misstatements of large magnitudes”.
In other words, long audit tenure is actually “detrimental rather than beneficial to audit quality”.
The two academics conducted their research in the light of the decision by the US Congress not to give its backing to a proposal by US audit watchdog, the Public Company Accounting Oversight Board (PCAOB), to introduce mandatory rotation of auditors.
It also follows the EU’s introduction in June 2016 of mandatory tendering every 10 years and mandatory rotation every 20 years.
They based their research on a sample of 3,465 accounting misstatements that took place between 2000 to 2013 and discovered that the number of misstatements peaked in 2003-04, following the collapse of Arthur Andersen in 2002 and implementation of the Sarbanes-Oxley legislation.
Andersen’s demise meant that its clients had no choice but to change auditors. This enabled the academics to establish whether the new auditors detected financial misstatement more quickly than the incumbent auditors since they were able to use as a treatment sample a group of Andersen’s clients which had accounting misstatements starting in 2000 or 2001 and ending after they switched auditors in 2002.
The academics matched the former Andersen clients against a control group of companies that were clients of other Big Four firms and that also had misstatements in their financial statements.
The comparison showed that the former Andersen clients had a “significantly shorter misstatement duration” than the Big Four clients which retained the same audit firm throughout the duration of the misstatement. This, they say, backs up their “argument that long auditor tenure impairs audit quality and highlights the benefit of a fresh look at a company’s financial reports by a new auditor”.
Singer and Zhang also found that while the introduction of Sarbanes-Oxley led to a weakening of the association between auditor tenure and misstatement by around 50%, it did not remove the negative impact of long auditor tenure on the quality of audit.
The research, Auditor Tenure and the Timeliness of Misstatement Discovery, will provide ballast for those – including the PCAOB – who believe that long relationships between auditor and client encourage the development of economic and social ties which in turn undermine auditor independence and audit quality.
As the academics point out, their conclusions provide “direct evidence for the benefit of a fresh look at financial statements by a new auditor”.