Derivatives accounting and risk management expert Thomas Linsmeier joins the Financial Accounting Standards Board on July 1, replacing outgoing member Katherine Schipper. Linsmeier, chairman of the accounting department at the Eli Broad College of Business at Michigan State University, previously served as an academic fellow at the Securities and Exchange Commission.
He sat down with CFO.com editors to discuss FASB’s conceptual framework, fair value, pension and lease accounting, as well as the mismatch problem with derivatives accounting. Here’s what Linsmeier had to say a week before stepping into his new standard-setting role.
CFO.com: When considering whether the conceptual framework should move toward full fair value accounting, is it important that assets and liabilities become the primary focus?
Linsmeier: I think there is a misconception about the conceptual framework [regarding assets and liabilities.] It is necessary to start with assets and liabilities because our accounting model should be representative of economic reality. However, to get revenues, gains, expenses, and losses—which are accounting constructs—measured regularly and well, we need a discipline that looks at asset and liabilities and then captures changes in wealth. But that doesn’t make assets and liabilities the primary focus.
What should be the focus?
Let’s consider your question in the context of the income statement. You start with assets and liabilities, but value them using different measurement attributes. This causes changes in the timing of when you recognize revenues, expenses, gains, and losses. That’s an important issue, not because you started with assets and liabilities, but because you used different attributes. On the asset side, for example, you might have some items at fair value, some at historical cost, some at amortized costs. If you add them up, what do total assets really mean? The issue of net differences in measurement attribute is one that effects both the income statement and balance sheet.
So the issue is not a question of fair value versus historical value, or whether companies should focused on assets and liabilities, but rather how preparers estimate value?
In part, that’s correct. It’s not the asset or liability focus in the conceptual framework that drives this particular issue. It’s the measurement attribute used to recognize the asset that people are raising the question about.
It seems that the debate about the use of fair value versus historical cost would end if users and preparers weren’t so focused on assets and liabilities to begin with. Is that true?
When you talk about historical cost and fair value, those two numbers are identical at an exchange transaction date. Then the issue becomes whether or not you want to re-measure the transaction price at a fair value in the future in the [accounting] model, or take the old transaction price and allocate it over time to the income statement. The real open question when you make that trade-off is, how might investors best be served? Further, which provides more relevant information to investors to help them predict the future economic prospects of a company: When you re-measure to fair value or when you allocate that original cost over time? It is a trade-off between relevance and reliability.
Is one type of accounting more relevant or reliable than another?
You can’t make the assumption — which is currently the subject of a public debate — that fair values are automatically more relevant, but less reliable, and that historical costs are automatically more reliable, and less relevant. For example, if there is an observable fair value in trade or settlement, or a reasonable way to estimate it, you might get more relevant information by doing a fair value. In other instances, [in which] you can’t observe exchange-traded prices, you would have to estimate fair value, which brings in reliability concerns. So, there are relevance and reliability issues present in both [measurement] choices.
If different types of measurements are added to mixed-attribute models to help improve reliability and relevance, won’t those additions increase the complexity of the accounting system, something FASB would like to avoid?
Our accounting model — and the standards in it — have been developed component by component. A weakness in the model is that we have not seriously considered the implications the separate accounting decisions have on aggregating financial reporting across line items. So a mixed-attribute model obviously causes challenges in that aggregation. For example, some income statement items show the effects of changes in key underlying factors in the period the change occurs, while the recognition of other gains or losses are delayed, because of historical cost allocation. It’s my suspicion that the [aggregation problem] is one reason that the CFA Institute is calling for a single-attribute model — to capture the effects of all changes that happens in a period in that period.
Some proponents of fair value say that sufficient disclosure about estimate assumptions usually relieves most concerns about reliability of that assumption. What would you say to them?
There are going to be judgments inherent in either of these two general accounting approaches we are talking about. But we have a tendency to forget that there are arbitrary choices and estimates made within the historical cost allocation model. Those choices may be made to best mimic the economics of the revenue recognition cycle of a company, or for other less relevant related reasons. The point is that [financial statement] users are better served when they understand how and why judgments are made, whether related to fair value estimates or allocations within a historical cost model.
Let’s move on to pension accounting and the potential economic consequences. Some critics say that if FASB mandates putting pension liabilities on the balance sheet, then employers will flee from offering defined benefit plans. To what extent should such consequences enter into your decision making on this pension project?
FASB is always concerned about implications on the marketplace. It is the FASB’s public policy mission to ensure that financial accounting and reporting standards capture economic reality objectively. But I want to raise another issue: [The idea of] whether there are significant economic consequences tied to the board’s decisions. Well, every new standard should have an economic effect, otherwise, why are we measuring or reporting things differently? Investors should understand things differently, hopefully better, by changing the accounting model.
Regarding pensions, let me suggest this to you: If companies are making [pension] commitments that they can’t keep, are we serving the investing community and employees better by letting them know now, so that they can take care of it earlier. Or do we want to wait till later?
Lease accounting is not officially on the table at FASB, but we wanted to get your opinion on using bright line tests to determine what constitutes an operating lease and a capital lease.
In some respects, it’s a little early to be saying much of anything. But I think the project well worth considering. At this point in time, it appears that the vast majority of leases, in dollar magnitude, are treated as operating leases, in part because there are bright line tests in Statement 13 that allow companies to structure their leases for desirable accounting results. It is not clear to me that to reconsider the standard means that there should be a change in the bright lines. It might mean it’s worth thinking about the concepts related to operating and capital leases, and then trying to write guidance that causes companies to evaluate those concepts, and report their leases without, perhaps, relying as much on bright lines.
In other words, the guidance would encourage a more principles-based standard for determining the lease type, is that right?
Yes, in a sense. The difference between principle- and non-principle-based standards are not [straightforward.] For instance, you can have very technical standards that are not bright-line based. I believe every standard should come from a core concept or principle, and that in writing standards, [FASB should consider] the degree to which the concept needs to be explained. But, perhaps, not [explained] to the point where bright lines allow people to engineer around [the standard.]
FASB Chairman Robert Herz advocates reducing accounting system complexity. He thinks some kind of national initiative may be required to get all the constituents to drop their own agenda and focus on achieving consensus. Do you agree?
[Reducing complexity] makes a lot of sense to me, and I agree whole-heartedly with Bob Herz that it is not just the standard setters that have to be involved. I know there is pressure to use details to fight litigation or [defend] against SEC questioning, but the complexity problem should be openly discussed. A conversation among multiple constituencies is something that will have to occur if we are going to pull this off.
Circling back to fair value, would adopting fair value help alleviate complexity in something as complicated as, say, derivative accounting (FAS 133)?
You have to understand that FAS 133 started off as an interim — perhaps political — step toward full fair valuing of financial instruments. If we had full fair value of financial instruments, many of the problems related to capturing the economics effects of change on companies — changes in interest rates, commodity prices, equity prices, foreign currency — would go away. But people weren’t ready for, and probably still aren’t ready for, full fair valuing yet. So FAS 133 is dealing with a partial fair valuing of a set of instruments that are being effected by interest rates, currencies, commodity, and equity prices.
So in a sense, FAS 133 was a stop-gap measure in response to the clamor over derivatives problems.
Yes, derivatives were causing huge losses in the mid-1990s, and nobody foreshadowed those losses. So to understand the degree of loss, or gain, occurring at companies that used derivatives, fair value was called for. Derivative accounting would not have been a big issue if full fair value for financial instruments was adopted at that point. But people weren’t ready, so FASB had to build a hugely complex model to make derivatives losses — or gains — transparent, while trying to take out some of the measurement attribute mismatches.
What do you mean by mismatches?
That means only a half loaf gain or loss is represented. Let’s say the company is exposed to interest rates, and it uses derivatives to hedge risk. As interest rates change during a period, the effects on the interest-risk item may not be transparent in that period’s income statement if there is a mismatch in the timing of the recognition of interest-rate changes. The hedge accounting model says, if you can link a derivative to the interest-rate changes in a hedged item, and you document this in advance, you can [book] a change in the accounting so that the income statement effects of the derivative and the hedged item are in the same period, eliminating the mismatch.
And then you mitigate the effects on earnings?
You mitigate the one-sided effect on earnings. A concern of mine is that mitigating any effect on earnings does not reflect what is happening to the company [regarding] interest-rate changes — unless the company is perfectly hedged.