Last week, the PCAOB convened its Standing Advisory Group to discuss the topic of auditing accounting estimates and fair values. I was invited to be on a panel providing investors’ perspectives.*
This post mainly consists of an unofficial transcript (with some minor modifications) of my introductory remarks, followed by some brief thoughts regarding the discussion that ensued. The webcast is available here, and you can download my slides here.
* * * * * * * *
Good morning, and thank you, Marty (Baumann), for inviting me to be on the panel for this very important discussion. I’d like to begin by sharing my perspective on investors’ perspectives.
A major challenge for regulators is dealing with differences between what investors say they want; and what others think that investors should want. Both perspectives are important, and I have decided I should assign myself the role of discussing what investors should want.
Fortunately, though, I won’t need to say much about how I know what investors should want. This is because the question we are dealing with today is very specific; and in my opinion, there is little controversy about the answer.
The question is: when a judgment is required to arrive at a number in a financial statement, how should investors want that judgment to be made? And the answer is: an investor should want the judgment to be made in an unbiased manner.
In the 1930s, when verification was the driver of audit quality — and attesting to the reasonableness of estimates was less of a factor — the SEC concluded from its investigation of the McKesson & Robbins fraud that auditors needed to be explicitly told something that today is second nature: that it is not OK to issue an audit report without having examined inventory and receivables. These were the beginnings of some of the fundamental rules of audit engagements.
Today, the balance between verification and attesting to the reasonableness of estimates has shifted dramatically. Therefore, I want to ask whether the evolution of the fundamental rules of audit engagements have been responsive to that shift.
Basically, AU § 342.03 states that management is responsible for the judgmental components of financial statement numbers; and what management chooses to consider when forming its judgments, is a matter of management judgment itself.
This longstanding foundational rule — which to the best of my knowledge has no direct basis in the securities laws — may have worked well enough in the past, but perhaps it needs to be re-examined. Does it promote the unbiased judgments that investors should want, or does it hinder them?
Let me ask the question a different way. Imagine that Accounting Professor X permitted students to grade their own exams. In determining one’s grade, a student may take into account the intention to learn the material better during the coming months while studying for the CPA examination. Professor X understands that she must rein in unreasonably high grades, but that’s not as easy as it sounds. All of the students are giving themselves the ‘benefit of the doubt,’ so to speak. Under these rules of engagement, Professor X certainly can’t, and does not wish, to confront every student and remove the bias from every grade.
Despite its obvious flaws, Professor X must like her system. We know this, because she is the one who wrote the rules into the course syllabus. Whatever the costs of the flawed system and whomever should bear those costs, we also know that Professor X has fewer confrontations with students over grades than any other professor; and the students must think that she’s really cool.
So, here’s my question — As a future employer of Professor X’s students – who will rely on grades to identify her best students – are you being well-served by the rules of engagement for her class? What if the entire university system permitted students to grade their own exams?
My point is that AU § 342.03, however it came into existence, from an investor perspective looks like a standard created by auditors to benefit auditors. And management, like Professor X’s students, is happy to play along. But, the system does a disservice to investors, because it deprives them of unbiased judgments — and ever more so as accounting standards increase in complexity.
To summarize, AU § 342.03 is a foundational rule of engagement, and it is not conducive to unbiased judgments. Even the most highly-qualified and intentioned auditors can be put between a rock and a hard place.
Consequently, under AU 342.03, the best that an auditor can do is to subjectively evaluate for itself whether management has an “appropriate” – or some would say “reasonable” – basis for its estimate. When the present doesn’t look much like the past, this can be a big problem. Personally, I find it most concerning that these rules of engagement enable inappropriate wealth transfers from investors to managers. Investors should not be content with a system under which management is essentially permitted to grade its own exam.
The teller is Walter Schuetze, one of the original members of the FASB, a long-time KPMG partner and a former SEC Chief Accountant, and one of the most plain-spoken individuals you will ever meet. His story is the most straightforward way I can think of to explain why we are discussing audits of estimates today; and why we have come to the point where I believe a fundamental shift in approach is needed.
To this point, I hope I have persuaded you, if you already didn’t know, that AU § 342.03 has some fundamental limitations. For decades, policy makers have acted as if it could not be changed, but that presumption now needs to be challenged, and that’s what I would like to encourage the PCAOB to do.
On this slide, I have barely outlined the start of an iterative process to gradually change how estimates are built into financial statements. Initially, we would scope-in only financial instruments for which Level II or III fair values are already being reported by large financial institutions. These financial institutions would engage independent appraisers to estimate the fair value of those financial instruments.
The auditor would still have a key role, but it would be engaged for this purpose only to verify certain facts. With respect to the work of the appraiser, auditors would verify that factual information provided by management to the appraiser is accurate; that the appraiser met specific independence standards; that the appraiser performed the work in accordance with GAAP and in accordance with their engagement letter with the issuer; and that the appraiser’s calculations were accurately made.
If only this first iteration were to be implemented, that would be substantial progress, indeed. But, I also want to look ahead to the logical endpoint: to purge financial statements of all judgment bias — most likely by replacing management’s judgments with market-based drivers of value, to be estimated by independent experts. Let’s see where that would lead us.
First, both auditing and U.S. GAAP would be much less complex, and much less fraught with risks of restatement and litigation. Second, it would take auditing back to its roots, but it would also create new opportunities for audit firms. Since auditors would no longer have to second-guess management in order to have a reasonable basis for its opinion, it should be possible to reconsider the degree to which non-audit services for clients are constrained. Third, other costly restrictions and regulations that have been discussed recently, and which auditors have opposed, would become moot.
I want to acknowledge, that a 2003 speech given by Walter Schuetze to the New York State Society of CPAs touches on many of the topics I have discussed. For additional background and perspective, I encourage you to read it.
Finally, the caveat. In my brief time, I have provided you with only the barest outline of a new path forward. We will not be able to resolve even a few of the questions that we all have regarding implementation and practicability, but that doesn’t mean there aren’t solutions.
I can’t think of any good reasons why practical solutions would not exist, and why financial reporting regulators would not want to look for them.
* * * * * * * *
Indeed, lots of good questions were asked in the ensuing discussion regarding the implementation challenges I had alluded to in my final caveat; and if those are the most challenging, I came out of the conference feeling even better that substantial progress could be made toward removing bias from financial statement judgments, if the PCAOB and SEC could overcome the inevitable political squabbles.
During that discussion, I also stated that I had no problem with an auditing firm also providing independent valuation services to issuers that were not also audit clients. But, at lunch, Bob Herz suggested that there should be nothing wrong with auditors taking the facts provided by management and performing the measurements themselves, without further input on the measurements from management. I think that is an excellent idea that is also worth pursuing.
And, by the way, I have just finished reading Bob’s recently published memoirs, Accounting Changes: Chronicles of Convergence, Crisis, and Complexity in Financial Reporting. I highly recommend it.
*I am a member of the Standing Advisory Group of the PCAOB. The views expressed herein are my own and do not necessarily reflect the views of the Board or its staff.