FASB Could Finally Get Loan Accounting Right – Well, Less Wrong

One of the most positive financial reporting policy developments in a long time is the FASB's announcement, about a month ago, of its intent to require fair value accounting for nearly all financial instruments. An exposure draft is supposed to be coming out pretty soon, but that will mark only the onset of a protracted battle royal with issuers of all stripes. After all is said and done, it's going to make the IFRS shootout look like a tea party.

In addition to throwing their collective weight around, I predict that opponents will provide nothing new, but will instead be calling their old war horses back into service.  Among the first should be the argument that fair value is not relevant when management intends to hold their investments in debt instruments until maturity; consequently it will inject meaningless volatility in reported earnings in the vast majority of instances as loans are repaid in full. 

I can think of two responses to the argument. The obvious one is that much has changed since 1993, when the FASB voted 5-2 to adopt FAS 115, and acceded to the held-to-maturity camp, to allow issuers to blissfully disregard readily available market values. Granted, in some respects the standard was a small step forward in that some financial assets came to be measured at fair value, but it (along with a previous standard on loan impairment, FAS 114) enshrined the pernicious view that managers' assessments of the future consequences of their own mistakes are superior to readily determinable market values and market-based interest rates.

The FASB's current actions can only be seen as a tacit admission that amortized cost accounting for loans was a serious mistake, to put it mildly. I completely disagree with the Financial Crisis Advisory Group (whose accounting bona fides and independence should be subject to serious scrutiny), when it stated that "…it seems [weasel word] clear that accounting standards were not a root cause of the financial crisis." Nothing is more clear that accounting shenanigans fully licensed by GAAP and IFRS played a huge role in the financial crisis.  They delayed early warning signs to bank regulators regarding capital adequacy, and far too many decisions by managers were driven by the accounting result that could be obtained. 

My second response, however, is more subtle, but much more important because it enunciates a principle that the FASB would do well to consider in its deliberations.

A Simple Example

The following statement of facts and resulting analysis can also be downloaded from an Excel spreadsheet, here:

• On December 31, 20x0, Lender Company invested $10,000 in a bond issued on that date with the same face amount $10,000. To keep things really simple for now (and to defer discussing differences between replacement cost and fair value), there are no transaction costs.

• The terms of the bond provide for two payments: $1,000 on December 31, 20x1, and $11,000 on December 31, 20x2. Both payments were made in full.

• Lender Company had only one other asset on December 31, 20x0: cash in the amount of $1,000. It had incurred no liabilities, and engaged in no transactions, except those related to the bond, through December 31, 20x2.

• As a rudimentary, yet sufficient, substitute for real-world measurements of inflation, we will blissfully imagine that there is only one consumption good in the world: beer. As of December 31, 20x0, a keg of beer cost $100. Immediately after the two payments on the bond, the price per keg rose to $110 and $121, respectively.

If 'wealth' is defined as command over goods and services (in this case, beer) Lender's economic income can be straightforwardly calculated for each year, and in total, as follows:

 

By measuring income in monetary terms, we have a standard against which to measure any system of financial reporting Lender Company might adopt. That standard is correspondence of reported earnings to economic income.

The first set of columns in the table below comprises Lender's financial statements under the current held-to-maturity model. The second set is a sufficient approximation of the change that the FASB is proposing. Although income over the two years is equal to reported interest revenue under both systems, reported net income differs from year to year. Some would argue that the difference is a needless distraction since the loan was paid in full. The FASB will argue that there is information content to the volatility, especially in economic times when an alarmingly high proportion of loans are not actually paid in full; and especially in longer-term and more complex lending arrangements.

  

My argument is that the volatility is a natural consequence of inflation, and has its own economic consequences, even when all loans are paid in full. This can be seen from a third set of financial statements, below. In addition to changing the measurement attribute for the investment (as the FASB proposes), I am also changing the unit of measure—from nominal dollars, whose purchasing power declines over time due to inflation, to 'constant units of purchasing power':

Reported income now matches economic income exactly. So, by this standard for earnings quality, even the FASB's ambitious proposal constitutes only a partial answer.  It's a weak one at that because the FASB will still permit lenders to overstate profitability on loans with fixed interest rates and fixed maturity amounts.   Should banks that benefit from government-provided deposit insurance be making loans that fully transfer inflation risk onto its stakeholders? Whatever your answer, you should know that if the accounting makes fixed rate loans look more profitable than variable rate loans, then fixed rate loans will be on the first page of the playbook. 

It is also important to mention that fair value and replacement cost accounting are the same when dealing with financial instruments, until you add transaction costs to the picture.  Then, replacement cost will be the only system that yields economic income. That's because the transaction costs are properly seen as part of the investment as opposed to the inconsistent and rules-based treatments for them under both GAAP and IFRS.

In summary, if the principle goal of financial reporting is to generate a net income number that reflects periodic economic earnings, the FASB's proposal heads in the right direction, but two elements are missing: (1) adjustments for changes in the purchasing power of the unit of measure, and (2) proper treatment of transaction costs.  The main point is that volatility of earnings should matter, even when loans are paid back in full.

So, congratulations to the FASB for announcing its intention to require fair value accounting for all financial instruments, including loans and debt instruments. Although I know it won't happen, their opponents should console themselves with the fact that their earnings will still be overstated—even if they will be less manageable and more volatile.

Comments

[Independent Accountant]

Good show. I'm so old I even remember SFAS 33. It amazes me how many bankers do not understand changes in fair value of long-term bonds as a result of changes in nominal interest rates. These fools still do not understand what caused the 1979-86 S&L crisis.

[Raza]

Hi Tom - what an excellent post you have here on the subject. I could not agree with you more on this. Excellent post.Raza