The SEC's Fair Value Roundtable: No Fireworks

On July 9, 2008, in Washington, DC, the SEC hosted a roundtable "to facilitate an open discussion of the benefits and potential challenges associated with existing fair value accounting and auditing standards." The roundtable was webcast and lasted about four hours. I admit that I literally fell asleep after listening to the discussion for the better part of three hours, so I missed the end. For all I know, the grand finale was a fireworks display, but I doubt it – this time, both literally and figuratively. When the SEC schedules these roundtable events for 9 a.m. on the east coast, I can't help but wonder what they are trying to tell those of us located in PAC-10 country (Go Sun Devils!). Maybe they would really prefer that no one listens.

Anyway, the topics included, among other things, discussions of the aspects of current standards that could be improved, and the usefulness of fair value accounting to investors. I'm going to address three issues that are so basic and important enough that accounting professors may want to consider using them for a class discussion.

Issue #1: "Held-to-Maturity" Investments

James Tisch of Loews Corp., when talking about his company's insurance subsidiaries, teed up this issue by describing a situation where his company would invest in marketable debt securities whose valuation might be affected by interest rate changes -- even though there would be no changes to the borrower's credit risk. Being an insurance company subject to various regulatory authorities, a rise in interest rates would supposedly force Loews to declare the investments in the held- to-maturity-category of marketable debt securities, the least onerous of three evils (the other two requiring fair value accounting).

Without the held-to-maturity option, the carrying amount of the investment would initially decline as interest rates rose, but could be expected to recover to the amount of the contractual obligation as the maturity date approached. Tisch's view seems to be that either fair value accounting would unreasonably record losses when it is highly probable that the entire investment plus interest will be recovered, or that constraints imposed by regulators trump the accounting that is most appropriate from an investor's viewpoint.

I think that the best way to approach a question like this is to ask yourself a simple question: did Mr. Tisch's company suffer a loss because it chose to invest in fixed-rate, as opposed to variable-rate, debt instruments? Yes it did. While regulators may find that it obscures their own peculiar needs, there must surely more straightforward ways to solve the conflict with investor needs than to muck up the financial statements.

And don't forget that apart from appeasing the needs of regulators, the held-to-maturity category is chicken salad for management: as Mr. Tisch implied, his company would manage its reported financial position by "cherry picking": if interest rates were to decline instead of rise, those same investments are probably classified as trading in order to get the asset and earnings bumps.

In short, FAS 115 on marketable securities could have been a lot simpler if the goals for financial statements could be (and should be) a lot simpler. As another panelist observed, one shouldn't need a legal degree to be capable of reading all the disclosures. I believe the disclosures he was referring to owe their existence to low-quality solutions cobbled together to meet the needs of someone else besides investors. The SEC should be telling other regulators to go and make their own accounting rules if they don't like the ones that are supposed to protect investors.

Issue #2: Fair Value of Liabilities

Joseph Price, the CFO of Bank of America, expressed his opposition to applying fair value measurements to contractual obligations such as litigation (and by the way, one of my more recent posts discusses the misguided way in which the IASB would require fair value measurement for some non-contractual obligations). Mr. Price has no problem with a mixed attribute model of accounting, which is just another way of saying that he has no problem adding apples and oranges.

The larger question, however, is whether any liability should be subject to fair value measurement. In addition to the claims that gain recognition on liabilities from deterioration of credit risk would distort earnings, other speakers pointed out that the character of the gain itself, often incapable of being monetized absent liquidation, creates problems.

The academic, Kathy Petroni, conceded that it can be confusing when an operating loss can be more than offset by gains from writing down the value of one's own debt. However, she is also of the view that the gain on the debt is representationally faithful; in other words, the problem is not with the current valuation of the debt, but with incomplete asset revaluation. This is because not all balance sheet assets are measured at fair value, and not all economic assets are even recognized. Tom Linsmeier, of the FASB and also an academic, made the interesting observation that a write-down to liabilities could be reasonably interpreted by investors as a signal of the asset losses that were not recognized.

As you may already have guessed, I am not sympathetic to stating liabilities at something other than current values. For one thing, we will never get to the point where all of the assets of a business are recognized, so we will never get to the point of measuring all of the components of economic income. Investor's don't expect financial reporting to account for all of the components of economic income. (Actually, that's what changes in stock prices do, but they have the distinct disadvantage of not allowing an analyst to directly identify the drivers of stock price changes.) What investors do expect is that the components of economic income that are measured are measured properly. If the deterioration of a company's credit worthiness creates an opportunity, amidst the other problems it must be experiencing, for it to restructure its debt advantageously, doesn't that opportunity benefit shareholders? Absolutely.

And, by the way, I am not advocating that all liabilities, regardless of the likelihood that a cash outflow will occur, be given recognition. And perhaps, some non-contractual liabilities, due to their nature, should be excluded from recognition. So the problem of incomplete recognition extends to the liability side just as much as to the asset side.

As the old saying goes, "perfection is the enemy of the good". What that means here is that we should not be distorting liability valuation just because some other element is not perfectly taken account of.

Issue #3: Fair Value Accounting for Non-Financial Assets

There was some discussion and support for measuring non-financial assets at fair value, but that support may have been even less enthusiastic than the support for fair value measurement of financial assets.

Logic dictates that whatever approach to fair value for financial assets is taken, that same approach should be applicable to non-financial assets. What's good for the goose is good for the gander; otherwise, we permanently consign ourselves to adding apples and oranges. And speaking of which, I have also pointed out here that some folks who don't care whether they are adding apples and oranges don't even care how assets and liabilities are measured -- just so long as they can control what is reported on the (their) income statement. One of my favorite examples is the historic cost of a tract of land carried on the balance sheet of a foreign subsidiary: when multiplied by today's current exchange rate to translate into dollars, we don't end up with an historic cost in dollars, or a current value in dollars. We end up with what is essentially a random number. How do you test impairment of a random number?

Speaking of impairment, for those of you who have had to apply FAS 144 on the impairment of long-lived assets, or FAS 142 on goodwill impairment, or even inventory impairment, you would know from that unfortunate experience that the impairment model of accounting is perhaps the biggest source of complexity, if not broken altogether. Some would argue that it is a reason, in and of itself, to abandon historic cost accounting and move to some version of current costs.

So, what if we went to fair value accounting for non-financial assets? That might solve the impairment problem, but it would raise another big issue, that being gain recognition before the non-financial assets, usually inventory, were actually sold. In a nutshell, that's why I think proponents of fair value are hesitant to extend the concept to non-financial assets -- more than anything, it exposes the main problem of fair value serving as a core accounting principle.

The appropriate non-financial asset attribute to measure is replacement cost (entry prices), and not fair value (exit prices). To further appreciate this, take for example the issue of transaction costs to acquire inventory. If FAS 157 were applied to purchases of raw materials inventory, transaction costs (perhaps a brokerage fee) would be expensed immediately upon acquisition. We all know that this makes no sense: we immediately have an expense to report before we have any chance at all to generate a return on our investment. (By the way, the same anomaly applies to financial assets, but it has already been established by FAS 157 that the FASB doesn't seem to care much about this.)

A replacement cost approach, on the other hand, would mean that all of the expenditures required to replace the asset should be part of the carrying amount of the asset. If we ultimately sell inventory for an amount greater than it would cost us to replace it, then we have a profit.

Getting back to geese and ganders, if replacement cost is the appropriate attribute to measure for non-financial assets, then it must be the appropriate attribute for financial assets as well.

Oh, Well…

Overall, the roundtable contributed very little to the fair value debate that hasn't already been expressed and considered before. Nonetheless, it reinforces two points that may well conclude that class discussion I've suggested:

First, I would prefer to have a dialogue at the SEC instead of in London at IFRS headquarters. Chairman Cox himself unwittingly pointed this out when he asked one set of panelists whether they believe current accounting rules contributed in some way to the economic issues the financial institutions are now dealing with. What if the answer to his question is "yes"? That, by itself, should settle forever the debate about who should be setting accounting standards for the U. S. capital markets. What if the answer to Cox's question is "we don't know"? QED.

Second, it would be refreshing if for once, an issue were settled by simply asking what it is that investors would want. Why does it seem that policy makers are incapable of doing that?

Comments

[Independent Accountant]

Cox doesn't care what investors want. He wants to protect the interests of the banking industry.

[A Mouse]

"I think that the best way to approach a question like this is to ask yourself a simple question: did Mr. Tisch's company suffer a loss because it chose to invest in fixed-rate, as opposed to variable-rate, debt instruments? Yes it did."

Not if Mr. Tisch's company received a fixed annuity contract that matched the duration and maturity of the bonds they bought. Unlike banks, insurance companies are more match-funded with their asset and liabilities. Investors care about the credit spread between these assets and liabilities. Marking the assets down because interest rates move one direction, but leaving liabilities fixed isn't what this investor wants