A new discussion paper released last week by the staff of the International Accounting Standards Board has revived an old, but still fiery fair-value controversy.
At issue: the role of credit risk in measuring the fair value of a liability. According to the paper’s opening statement: the topic has “arguably … generated more comment and controversy than any other aspect of fair value measurement.”
At the heated core of the dispute is the question of why accounting rules allow companies to book a gain when their credit rating actually sinks. The accounting convention, which opponents contend is counterintuitive if not ridiculous, has prompted “a visceral response to an intellectual issue,” says Wayne Upton, the IASB project principal who authored the discussion paper.
For all the hubbub around it, the rule is rather simple: When a company chooses to use the fair value method of accounting, it must mark its liabilities as well as its assets to market. As a company’s credit rating goes down, so does the price of its debt, which therefore must be re-measured by marking the liability to market. The difference between the debt’s carrying value and its so-called fair value is then recorded as a debit to liabilities, and a credit to income.
Consider an oversimplified example to clarify the accounting treatment. A company records a $100 liability for a bond it has issued. Overnight, the company’s credit rating drops from A to BB. That drop causes the price of the bond trading in the market to decrease from $100 to $90. The $10 difference, under current accounting rules, is recorded as a $10 debit to liabilities on the balance sheet and a $10 credit to income on the income statement.
As the company’s credit rating and the price of the bond rise — to, say, $100 again — the accounting is reversed. Income takes a $10 hit, while the liability account is credited.
That accounting oddity has been a lingering problem since 2000, when the Financial Accounting Standards Board introduced Concept Statement 7, which includes a general theory on credit standing and measuring liabilities. The notion was hotly debated again in 2005, when IASB revised IAS 39, its measurement rule for financial instruments and in 2006 when FASB issued FAS 157, its fair-value measurement standard.
Addison Everett, the practice leader for global capital markets at PricewaterhouseCoopers, notes that the debate cooled down over the last 18 months as the liquidity crisis bubbled up. The crisis spotlighted more politically charged fair-value topics such as asset valuation in illiquid markets, classification of financial assets, asset impairment, and financial disclosures, he says.
But the credit risk quandary is back, demanding the attention of investors, regulators, and lawmakers who were carefully watching ailing financial institutions as they posted their first-quarter earnings results. As financial results were disclosed this year, it became clear that IAS 39 and FAS 157 were being used to boost income as banks and insurance companies became less creditworthy. For example, in the first quarter, Citigroup benefited from its credit rating downgrade by posting a $30 million gain on its own bond debt.
A Credit Suisse report looking back to last year, flagged a similar trend. The bank examined the first-quarter 2008 10-Qs of the 380 members of the S&P 500 with either November or December year-end closes, the first big companies to adopt FAS 157. For the 25 companies with the biggest liabilities on their balance sheets measured at fair value, widening credit spreads-an indication of a lack of creditworthiness-spawned first-quarter earnings gains ranging from $11 million to $3.6 billion.
Those keen on keeping the rules intact and allowing companies to book a gain when credit ratings worsen give several reasons for their stance. Most are laid out neatly in the IASB discussion paper. Consistency is one argument. “Accountants accept that the initial measurement of a liability incurred in an exchange for cash includes the effect of the borrower’s credit risk,” according to the paper. There’s “no reason why subsequent current measurements should exclude changes.”
There’s a practical problem with that argument, however. Not all liabilities are financial in nature. Non-financial liabilities, such as those tied to plant closings (asset removal), product warranties, pensions, insurance claims, and obligations linked to sales contracts, are not as easily marked to market as a clear-cut borrowing. Often non-financial liabilities represent a transaction with an individual counterparty that has already placed a price on the chance of not being repaid. For many of those liabilities, “accounting standards differ in their treatment of credit risk,” notes the paper.
One cure is to use a risk-free discount rate for all liabilities in order to apply a consistent measurement approach. But applying a blanket discount rate to the initial measure of debt leaves accountants with the problem of what to do with the debit. That is, for financial liabilities, should the debit be treated as a borrowing penalty and therefore as a charge against earnings? Or should the debit be subtracted from shareholder’s equity and amortized into earnings over the life of the debt? For non-financial debt, should the debit be the recognized warranty or plant-closing expense?
The paper goes on to describe similar thorny issues linked to accounting for the transfer of wealth between creditors and shareholders when a company’s credit rating changes. Since assets equal liabilities plus equity, the value of a company’s assets is always tied to the value of the liabilities. But there’s also a debate about whether a shift in value of a liability – connected with a credit downgrade, for instance – should be dropped into shareholder’s equity. The underlying problem, says the paper, is whether the company’s creditors or owners should absorb the value change.
The IASB staff also shows that excluding credit risk changes can result in an accounting mismatch between asset and liability measurements. If a company’s assets are measured at fair value, then changes in credit spreads on those assets will affect their fair value, according to the paper, as well as the company’s profit, loss, or other comprehensive income, depending on the transaction. So if the measurement of liabilities doesn’t also incorporate credit-spread changes, the accounting mismatch will distort the recorded profit, loss, and OCI, goes the argument.
Observers who want the rules changed point to their other accounting mismatches. One occurs when there’s a change in the value of a company’s liability accompanied by a change in the value of recognized assets that is not reported in the financials. Such unreported assets can include unrecognized intangible assets (patents, trademarks, goodwill) or confidence in a company’s management.
“There are tradeoffs in any debate around accounting standards,” says PwC’s Everett, noting that his firm is currently working on comments to the IASB discussion paper, which are due by Sept. 1. Last fall, the Big Four accounting firm issued its own white paper advocating a limited application of fair-value accounting for liabilities.
One situation that may warrant a fair-value approach is when a liability offsets an asset reported at fair value. In that way, the accounting symmetry will reflect the underlying economics of the transaction. “Unfortunately, this approach decreases comparability across companies, but the financial reporting is more informative than it would otherwise be – an acceptable tradeoff,” adds the PwC paper. “
Between now and the time IASB and FASB issue a standard or guidance that addresses the credit-risk issue, many other fair-value rules will be discussed and put in place, and the boards will have to consider the added effect of those predecessor mandates, according to Everett. Further, “by the time [standard setters] finish debating a new standard, the economic landscape may have changed,” he says.