In my previous post, I provided my reactions to Hans Hoogervorst's recent letter to the European Securities and Markets Authority (ESMA), in which he expressed concern for the pie-in-the-sky numbers produced by some EU banks on their Greek bonds classified as 'available for sale' (AFS). Now, I would like to pose a hypothetical: what if mandatory auditor rotation had been in effect? Would Mr. Hoogervorst's letter have been unnecessary?
As a preliminary matter, I am not asking these questions with French banks in mind. The banks in France are known to report anything they want, whenever they want. The regulators are powerless to stop them; ergo, there is no incentive for the auditors to do anything that would jeopardize relationships with clients. Vive le Société Générale!
So, setting France aside, let's start with a very simple fact pattern:
Mandatory audit firm rotation is required after seven years, and Bank One's current auditor (Auditor A) is in Year 7 of its term. Despite the fact that market prices for Greek bonds have declined by 51%, the management of Bank One maintains that a 21% haircut on its Greek bond investments is all that IFRS requires.
Auditor A knows that Auditor B is engaged for next year's audit of Bank One. Coincidentally, Auditor A will take over the audit next year of Bank Two, Auditor B's current client.
Of the possible actions/outcomes of the Bank One audit engagement, here are four to ponder:
- Eat or be eaten – Auditor A will insist on a 51% haircut; either because Auditor A actually disagrees with Bank One's management, or it has some ulterior motive. For example, Auditor A's overriding fear could be that the successor auditor (Auditor B) is motivated to clear the decks of all 'gray areas' when it takes over. If such were the case, Auditor B could come in next year and require Bank One to restate Year 7. So, Auditor A could be thinking that it is s a far better thing to administer grief to the soon-to-be-former client than to be grief's recipient – by none less than by another member of the fraternal order of auditors.
- One hand washes the other – Auditor A and Auditor B collude. They both have promises to trade: Auditor B won't mess with Bank One's Year 7 financial statements if Auditor A promises not to mess with Bank Two's Year 7 financial statements.
- Safety in numbers – Auditor A will caucus with its peer group, Auditors B, C, D and E – either informally, or through a mechanism akin to the Center for Audit Quality. Since auditors are happy to advocate for their clients as a group, the quality of the accounting alternatives (from an investor standpoint) will not matter to the caucus as much as it should.
- What goes around, comes around – Auditor A will accede to the client and have confidence that Auditor B won't try to call it to account. Auditor B is likely concerned with being disciplined by the fraternal order of auditors. This is what oligopoly theory would suggest.
You may provide your own subjective probability distribution over these (and perhaps other) outcomes; and draw your own conclusions as to whether mandatory audit firm rotation is worth the costs. But, the analysis shouldn't stop there; my larger point is that mandatory audit firm rotation has to be considered within the context of improving the entire financial reporting regulatory structure.
First, it goes without saying that criminal acts of collusion and fraud must be punished. However, this is not a panacea. If the death penalty doesn't deter murderers, then why should we trust the vows of politicians to enforce the existing securities laws and professional standards with renewed vigor, or to mete out more severe penalties for violations? In the best of circumstances, vigilant bureaucracies eventually fall asleep at the switch; and on the other side of the equation greed is boundless, inventive and infectious.
Second, mandatory audit firm rotation must complement robust audit inspection programs. To see this, simply change the date of my hypothetical from Year 7 to Year 1 of Auditor A's term. Audit firm rotation is so far in the distance in Year 1 that it couldn't possibly have much effect on the auditor's decision making.
The PCAOB's audit inspection programs, while seemingly effective in uncovering irregularities, don't seem to have enough of a deterrent effect. When the PCAOB uncovers deficiencies in an audit, it must publicly identify the auditor, the issuer, the deficiency and the remediation efforts – all in a timely manner. As a matter of principle, the investing public has a right to know which public company (could have) fudged its numbers, and which audit firm is responsible for letting it slip through their fingers. The PCAOB's has an inspection track record of being right far more often than wrong; this record should be used as the justification for undoing the secrecy and procrastination that is inimical to investor protection.
Third, an auditor has a right to be secure in the knowledge that it cannot be dismissed when it takes a stand against management. Mandatory auditor rotation should be accompanied by minimum audit terms, which cannot be lengthened. Once an auditor is engaged, it cannot be dismissed without performance-related cause, and the auditor may not resign simply due to a disagreement.
Finally, it is sad, but true, to state that the only reason why mandatory auditor rotation is being considered is because auditing has drifted too far afield from its original role as envisaged by Congress when it created the first federal securities laws. That role is verification. Today, the auditor's exposure to pressure from clients comes from their role as assessors of the 'reasonableness' of management's judgments. As I have stated repeatedly, and most recently here, it doesn't have to be this way, and I will have more to say on that in a future post.