U.S. Securities & Exchange Commission
SEC Seal
Home | Previous Page
U.S. Securities and Exchange Commission

Speech by SEC Staff:
Remarks before the 2007 AICPA National Conference on Current SEC and PCAOB Developments


Todd E. Hardiman

Associate Chief Accountant, Division of Corporation Finance
U.S. Securities and Exchange Commission

Washington, D.C.
December 11, 2007

As a matter of policy, the Securities and Exchange Commission disclaims responsibility for any private publication or statement of any SEC employee or Commissioner. This speech expresses the author's views and does not necessarily reflect those of the Commission, the Commissioners, or other members of the SEC staff.

Can a Large Error be Immaterial? RESTATED

Hindsight is 20/20. We've all heard the phrase, but it seems like you only come to appreciate its truth once you realize you've made a mistake. Last year at this conference I spoke about materiality. And in those remarks, I said: A large error can be immaterial. And perhaps I would have been better off leaving it at that. But we're in a world that requires judgment; and because of that folks crave insight into how to make those judgments; and so I forged ahead last year and shared with you the limited amount of thinking a few of us had done at the time on the topic of large errors. And in doing that, I made a mistake - or in accounting parlance an error. So consistent with the model laid out in FASB Statement 154 for reporting the correction of an error, let me disclose my mistake and then restate my remarks.

Last year when I was explaining my view that a large error can be immaterial, I said something to the effect that: 'It's possible for a large error to be immaterial, but the reality is that we don't see those circumstances often.' Shortly after that I followed with something like: 'You might find it instructive that we could only come up with two examples where a large error might be immaterial - break-even years and discontinued operations.' And on the basis of those two caveats some walked away thinking the needle hadn't really moved; that I had just reaffirmed their long-held perception that the staff of the Division of Corporation Finance views materiality through the singular lens of quantitative significance. And yet, that was not my intent; so let me restate.

A large error can be immaterial. I believe that. But since saying it, and even believing it, does not make it so; let me try to frame the issue more clearly.

The Supreme Court has held that a fact is material if there is a substantial likelihood that the fact would have been viewed by the reasonable investor as having significantly altered the "total mix" of information made available. It's a statement of principle; and it doesn't come with a whole lot of implementation guidance. And so as accountants, we've tried to create our own over time. Collectively, we've bifurcated "total mix" into component parts that we call "quantitative" and "qualitative." And individually, we've created rules of thumb. But sometimes in our effort to dissect and apply the Supreme Court view, we lose sight of its underlying principle. That seems particularly true for large errors; and yet with the benefit of hindsight, it's really no surprise. To understand the larger error debate, it may be helpful to understand how we got here.

Accountants by nature bring a quantitative bias to a materiality analysis. As accountants, we find support for that bias in the maxim - all things being equal, a large error is more important than a small error. And for a time that bias seemed to view materiality as a two dimensional world in which size was determinative. Large errors were material and small errors were not. And so the debate focused on what was large and what was small. Ask yourself - were you a 3%, a 5%, or maybe a 10% person? Where was your line between large and small?

But as time went on, it became clear that even small errors can matter to investors. A penny a share may not have seemed like much, but the impact of that penny on investor perception about things like management performance and compensation told us otherwise. Into the small error mindset came Staff Accounting Bulletin No. 99 reminding us of the Supreme Court view to focus on the "total mix." And with the issuance of SAB 99, the accountant's two-dimensional bias seemed to move to three dimensions. Large errors were still material, but small errors warranted more careful evaluation. Think of a small error scorecard with two categories: quantitative and qualitative. On the quantitative side, the insignificant size suggested the error was not material. But the paradigm shift was that you couldn't stop at size, you had to see if any qualitative factors existed that indicated that the error was important to investors despite its insignificant size. So within the population of small errors, qualitative factors caused errors to be material.

With this renewed focus, accountants started asking: if qualitative factors can cause small errors to be material, can qualitative factors cause large errors to be not material? With the benefit of hindsight, it's pretty clear that the answer is yes. But the path to that answer wasn't so clear.

Historically, when the staff in our Division received an analysis suggesting that a large error was not material, the primary support most often cited was the absence of qualitative factors that indicate that a small error is material. It often seemed like a bit of a check-list approach. Registrant pulled out SAB 99, read the illustrative list of factors that may indicate a small error despite its insignificant size is material, noted that none of those factors existed and then put their pencil down having concluded the large error was not material. So what was wrong with that approach? It was incomplete. It did not address the significant size of the error. Think of a large error scorecard with two categories: quantitative and qualitative. On the quantitative side, the significant size of the error suggests the error is material. The analysis that needs to be done on the qualitative side is to identify the factors that indicate that the error is not important to investors despite its significant size.

And that brings us to today and what seems to be the more difficult question: what are the qualitative factors that can cause a large error to be immaterial? It's a difficult question. All things being equal, a large error is more important than a small error. But, that isn't terribly helpful in determining which large error is material and which one is not. Couple that with the reality that all things are not equal, the facts and circumstances of one company rarely match neatly with those of another company, and you start to appreciate the difficulty in coming up with a universal list of qualitative factors that indicate that a large error, despite its significant size, is nonetheless not material. So what do we do? In my view, we have to get back to where we should have been all along - the Supreme Court view and its focus on what's important to the reasonable investor.

In the arena of large errors, it's harder to do than it sounds. For one, it seems to require addressing the question: Is there a substantial likelihood that the size of the error would have been viewed by the reasonable investor as having significantly altered the "total mix" of information made available?" Or maybe stated more simply, why doesn't the size of the error matter? For sure, there is not a universal answer to this question; by necessity it will be dependent on a company's unique facts and circumstances. But it's also clear that the answer will depend on the view of the "reasonable investor." And although it has taken considerable effort, I am pleased to announce today that I have managed to track down the phone number of the reasonable investor. It's 202.CALLSEC. I say that in jest to highlight an important point: the "reasonable investor" is not necessarily the person who questions your judgment. At the SEC, ethics rules prohibit the person reviewing your filing from being an investor in your securities. If you have to evaluate whether a large error is material, don't color your analysis by trying to guess what an accountant in the Division may or may not find important. A better proxy would seem to be the folks that are making investment decisions. And as company management, you talk to them on a regular basis. So ask yourself: Why doesn't the size of the error matter to the reasonable investor? What is it about your individual facts and circumstances that supports your conclusion? Or in accounting parlance, what qualitative factors exist that make the size of the error unimportant to the reasonable investor? A high hurdle to climb? Perhaps, but with the right facts and circumstances, a surmountable one.

One last point on materiality: Don't assume how our Division will view your materiality analysis. It's been suggested to us that some companies conclude that an error is not material, but then choose to restate their financial statements any way because they don't think our Division will agree with their conclusion. If that's you, come talk to us first. You can contact our office, the Division's Chief Accountant's office, directly or reach out to us through the Division's Assistant Director group that reviews your filings.

Judgment and Corporation Finance Review

'Well founded judgments are irrelevant to the Division's staff.' It's a criticism that was made to John White, our Division Director, just three months ago; and, in truth, it's one we've heard before. We take that criticism seriously each time it's leveled for one simple reason: reasonable judgment is the foundation of our financial reporting system.

One of the difficulties we have in evaluating that criticism is that it is typically done out of context and after the fact so we don't have the benefit of the facts and circumstances that might support it or an understanding of the type of judgment we're being criticized for. And that last point - not understanding the type of judgment we're being criticized for - is particularly important. Not all judgments are the same. In a speech last year at this conference, Scott Taub explained how judgment is necessary in both rules-based standards, which rely on exceptions and bright lines, and in objectives-oriented standards, which are more reliant on principles and subjective analyses; as well as for transactions for which the literature does not specifically apply. In any dialogue about the Division's willingness to accept reasonable judgment, it will be important to be specific about what types of judgments we're focusing on.

My own hypothesis is that on the whole companies and their management do a good job of making judgments. And that shouldn't be a hard leap to make. Of any group, they're the ones closest to the action and that makes them best-suited to make the delicate inferences and judgments unique to their facts and circumstances. But at times their disclosure documents don't explain those important judgments and sometimes that disclosure actually seems inconsistent with the exercise of appropriate judgment. That's where the staff of our Division enters the mix.

The Division's staff does question accounting judgments. We do it when the basis for an important judgment is counterintuitive or unclear from the disclosure. We do it when the judgments that are disclosed appear to be inconsistent with other judgments or assumptions made by management. We do it when the analogy to accounting literature management cites to support a judgment has been superceded by more recent thinking in other more closely analogous accounting literature that existed at the filing date. In short, we ask questions about an important judgment when we can't understand it or when on its face it seems inappropriate.

But that brings us to a critical point: the mere fact that we ask questions about your judgment does not mean that we've concluded your judgment is wrong. That's true even if the way we ask our questions may suggest otherwise.

In virtually every comment that asks questions about judgment, you should view the comment as an invitation to a dialogue. Admittedly, it's an invitation that may be difficult to decline, but what's important is that you understand that the comment is intended to elicit a dialogue. And it's a dialogue we need you to participate in fully.

We need you to participate fully because there isn't always one right answer; especially when the judgment relates to objectives-oriented standards or transactions where the literature isn't directly on point. In those circumstances, we're trying to answer the question: Is the company's accounting unreasonable? And to do that we need you to help us understand why you made the judgments you did. Sometimes the reason why is self evident from the disclosure document, but if we're raising a question it's probably not.

We need you to participate fully because what we know about your transaction and the judgments you made in accounting for it is limited to what you say within the four corners of your disclosure document. We don't know what you know. We don't have your perspective and therefore we need you to help us appreciate the delicate inferences that underlie your judgment. Again something that ought to be clear from the disclosure document, but if we're asking a question, it's probably not.

And perhaps at its simplest level, we need you to participate fully because accepting a staff conclusion that you don't really agree with stops the dialogue. The staff may view your silence as validation of the comment and you may view it as the staff not accepting reasonable judgment. If you find yourself in that situation, you can request a review of the staff decision by more senior Division accounting staff. Here's how it would work.

Your dialogue with our Division starts with the staff accountant and review accountant identified in your comment letter. But, if you think their conclusion is not appropriate, then continue the dialogue by asking for a review by more senior Division accounting staff. In our Division, the first layer of additional review will come from the Senior Assistant Chief Accountant; there is one for each of the Division's 11 industry groups. If you still don't agree with the staff decision at that level, request that the decision be reviewed by our office, the Division's Office of Chief Accountant. At that level, the decision will be reviewed by one of 7 Associate Chief Accountants, 4 of us are on this panel, and depending on the course of the dialogue with us, we can pull in either the Division's Deputy Chief Accountant or Chief Accountant. And, after working with our Division, you still have an opportunity to have the Division's decision reviewed by the Office of Chief Accountant of the Commission.

The goal of the multiple levels of review is not to create a war of attrition through bureaucracy; it's to ensure that there are appropriate checks and balances within our decision-making process. You're all very familiar with our agency mission of protecting investors, but you may not be as familiar with another aspect of our mission: facilitating capital formation. One way we're able to balance our goals of investor protection and capital formation is to rely on the judgments of our accountants that are most familiar with your filings. But we only achieve that balance if you're willing to request a review of accounting decisions made by our Division's staff that you think are not appropriate. Requesting that more senior review is a way to participate fully in the dialogue and we encourage you to do so.

See also: Slide presentation (PDF)


Modified: 12/26/2007