Republicans on Capitol Hill have introduced legislation that would alter the way that federal credit programs are budgeted by moving to what is known as “fair value accounting”.  The choice between accounting treatment may seem arcane, but the implications are very real.  Under the current approach, for example, CBO estimates that FHA will produce a “savings” of $63 billion over the next ten years.  In other words, the program will be a net money maker for the federal government.  However, applying fair value accounting to the same projected net revenue stream would lead to a subsidy of $30 billion.  It doesn’t take a much imagination to predict what might happen to FHA—or, for that matter, the student loan program--should the government adopt a fair value approach.

Since 1990, federal credit programs have been budgeted based on the requirements of the Federal Credit Reform Act (FCRA).  FCRA uses a net present value approach to measure the lifetime costs of a credit commitment in the year a loan is made, where costs are measured by the difference between projected expenditures (excluding administrative costs) and projected revenues.  The idea of the FCRA was to make it possible to compare credit programs with other federal spending items, which are valued on a cash expenditure basis. 


The fair value approach shares many similarities with FCRA accounting.  Both methods use an accrual basis of accounting in that revenues and expenses are recorded when the obligations are incurred independent of when the cash is actually received or paid.  Both use the same projections of future cash flows.  The primary difference between the two approaches rests in the discount rate that is used to calculate present value.  While FCRA uses the risk-free rates on Treasury securities, the fair-value approach attempts to incorporate a measure of market risk by using the discount rate that a private investor would require in a well-functioning credit market.

There is currently no consensus among economists on the discount rate that should be applied in evaluating federal credit programs.  In large part, the answer depends on the question one is trying to address.  If one is primarily interested in estimating the direct future costs to the government, using risk-free Treasury rates is the better approach.  While there is obviously a risk that actual results vary from expected values—either in a negative or positive way—this is also the case with respect to other projections that enter into the federal budget, for example, tax receipts or expenditures on entitlement programs.  In contrast, if one is primarily interested in the value of the subsidy that is being delivered through a credit program relative to the cost of providing the same services through the private sector, a “fair value” approach would be more appropriate since such estimates are based on the return to capital that would be otherwise be required by private investors.

Even beyond the issue of what question is being addressed, there are fundamental problems with fair value accounting that limit its usefulness.  First, fair value estimates can vary widely based on assumptions about the discount rate.  Often the private sector does not provide the same services as the government program and therefore there are many assumptions and conjectures in arriving at the discount rate.  Second, even where there are market prices, a fair value approach would result in a different assessment of a government program day-to-day as market conditions change.

In the end, however, the basic problem with the fair value approach for budgetary purposes may be that it ignores the fundamental role that government credit programs are designed to serve, namely, to provide counter-cyclical stability and/or serve so-called “under-served” segments of the population.  In the recent housing downturn, the private mortgage market all but disappeared.  Even if one could estimate what private investors would have charged in that environment—which would have been difficult at best in the absence of market transactions —evaluating (and presumably pricing) the FHA program to reflect the risk aversion of private investors at that particular point in time would have totally negated the program’s counter-cyclical role and further undermined the housing recovery.    

It is thus not surprising that employing the estimated discount rate that would be demanded by the private sector to make or guarantee a loan typically shows a positive subsidy when the FCRA approach does not.  Indeed, if the fair value approach did not result in a positive subsidy, one could reasonably question whether the government should be involved in the activity at all since the private sector could presumably provide the credit at a comparable or better price.  However, such results do not imply that the government will likely lose money on the program or that the program should be priced at terms that would be required by the private sector.  While one might want to incorporate some measure of risk into the assessment of a program’s costs, applying the discount rates demanded by the private sector for budgetary purposes is not the way to go.

Schnare is a consultant and former senior vice president of Freddie Mac.  She holds a Ph. D. in economics from Harvard University.