Understanding Fair-Value Accounting

One of the wonkier discussions that arise inside the Washington beltway from time to time is what accounting method to use for federal credit programs. In a new issue brief, CBO weighs in on the argument for and against using "fair value" accounting versus the current method. We provide a brief overview here, but be sure to check out the full report for more details.

Background on Federal Credit Programs

The federal government supports a number of private sector activities through a combination of direct loans to individuals and businesses and loan guarantees to private lending firms, the largest of which are loans and guarantees for residential properties and postsecondary education. The way that the federal costs, or subsidies, stemming from federal credit programs have been reflected in the budget has changed over time, and may continue to do so.

Prior to 1990, only the cash outlays and receipts for federal loans and guarantees were recorded in the budget. After lawmakers enacted the 1990 Federal Credit Reform Act (FCRA), however, this "cash accounting" system changed to an "accrual accounting" system to reflect the budgetary costs of a new loan or guarantee upfront. Cash accounting hides the true costs of new commitments: a $100 million loan made in a certain year doesn't reflect expected repayments in future years or the estimated total impact on the federal budget (see the table below). The system was changed to an accrual basis in order to reflect the expected future costs of loans, eliminating the incentive to favor loan guarantees (which receive premiums upfront and don't incur costs until later) under cash accounting over direct loans.

The 1990 reforms required credit programs to discount the future expected cash flows to a present value estimate, a process which requires an interest rate. Under those reforms, federal programs would use the interest rate on U.S. Treasuries.

Understanding the Current Debate

In recent years, some organizations and experts (including the Congressional Budget Office and our colleague Jason Delisle at the New America Foundation's Federal Education Budget Project) have argued that the rules under FCRA are not a comprehensive measure of the cost the government incurs when it enters into loans and loan guarantees. They argue that using the Treasury discount rate does not fully account for the market risk the government is taking on and does not accurately portray the costs to policymakers.

A "fair-value" approach to estimating the costs of federal credit programs would assign a "market risk premium" to bridge the gap between Treasury interest rates and the discount rate that private lending institutions would assign when assessing the costs of a loan. Some federal programs already incorporate fair-value accounting, including payments to the IMF and the financial and auto industry assistance under TARP, but it is not widespread practice.

The table below from the CBO report shows how the cost of credit programs would differ under both methods.

Budgetary Treatment of $100 Million Loan under the Three Approaches
  Cash Accounting FCRA Fair-Value
Year 1 $100 million -$1.6 million $1.3 million
10-Year Period -$3 million -$1.6 million $1.3 million

Note: See CBO's detailed version of this example in the report.

But those arguing on the other side point out that these loans are funded through the issuance of Treasury debt, which means there is very little or no market risk, so it is appropriate to use the FCRA rules. They also argue that the additional cost that fair-value accounting records does not actually occur and that using this method would apply inconsistent budgetary treatment to credit programs versus other programs.

CBO dives into the details of what a switch to fair-value accounting would mean in practice, touching on the additional resources and time spent training staff on the new method, the judgment calls about what discount rates to use when there are no similar examples in the private sector, and the need to better communicate program costs under a more complicated estimating approach.

But CBO’s view is best expressed in a section on the cost of market risks to the government and taxpayers, concluding with this passage,

Thus, none of the differences between the federal government and private investors changes the fact that investments with returns that are correlated with the performance of the economy as a whole are risky in a way that other investments are not. Federal credit programs expose taxpayers to that market risk.

The valuation of federal credit programs can be a pretty arcane topic. But at the same time policymakers need to decide what is the best way to reflect the costs, or profits, of the operations of the federal government. Given that the total amount of such loans outstanding in 2011 stands at $2.7 trillion, it's an important debate to be having.

To continue and enlighten the debate, e21 will be holding an event in DC on March 19 about the optimal way to account for the cost of loan programs. To learn more than you ever wanted to know about accounting rules in the federal budget, register for the event here.