Skip Navigation

A Framework for Projecting Interest Rate Spreads and Volatilities January 2000

By law, interest rates on money borrowed through the federal student-loan program are tied to a reference, or benchmark, rate. For the past several years, the benchmark has been the rate on three-month Treasury bills. In 2003, the formula for setting interest rates in the program will begin using a different reference rate that is also based on Treasury securities. Lenders who participate in the programs have proposed changing the formula further: they advocate using a reference rate from private-sector markets to tie the interest they receive more closely to their costs for making the loans. In December 1999, the Congress enacted the Ticket to Work and Incentives Improvement Act (P.L. 106-170), which changes the benchmark for lender yields on student loans issued between January 2000 and June 2003 to the rate on commercial paper with a three-month maturity. (Commercial paper is defined here as short-term debt issued by financial companies.) The change is a temporary one, covering only the period noted.

In estimating the cost of the various proposals for changing the benchmark, the Congressional Budget Office (CBO) developed a framework, or model, for projecting the proposed alternative rates and their volatilities (in general, the tendency of the rates to change over time). With the change in the benchmark, CBO will now use that framework to project commercial paper rates for its economic outlook and budget baseline. Projections from the model will also be used as the Congress considers permanent changes to the benchmark in the student-loan program. This memorandum discusses CBO's framework and its underpinnings and presents illustrative results for rates on various short-term financial instruments.
 

THE FEDERAL STUDENT-LOAN PROGRAM

The student-loan program covers two types of loans: direct loans from the federal government and loans issued by private lenders on which the government guarantees repayment. All student loans carry variable interest rates that are adjusted annually according to formulas specified in law. However, the rates that borrowers must pay on the loans are limited by statutory caps, and those caps have important implications for the federal budget. In the case of direct loans, the caps limit repayments to the government, thus raising the overall federal cost of the loans. For guaranteed loans, the government makes so-called special-allowance payments to private lenders to cover the difference whenever the formula-based interest rates rise above the caps. Estimates of the cost of the student-loan program must therefore allow for the possibility that interest rates will exceed the caps.

Recent legislation has made several changes in the formulas for interest rates on student loans. For loans issued between July 1, 1998, and December 31, 1999, student borrowers taking new loans pay an annually adjusted rate equal to the rate on three-month Treasury bills plus an additional premium based on the borrower's status: while the borrower is in school or in a grace, or deferment, period, the premium is 170 basis points (a basis point is one-hundredth of a percentage point); otherwise, the add-on is 230 basis points. However, the rate that the student borrower pays is capped at 8.25 percent. (Terms for parent borrowers are somewhat less attractive.) The interest rates paid to lenders are based on the quarterly average of three-month Treasury bill rates plus 280 basis points when the loans are being repaid and 220 basis points at other times.

As of January 1, 2000, the rate on three-month commercial paper will become the reference rate for lender yields. (In general, the yield is the expected return on the loans.) For student loans, the yield will equal the interest rate on three-month commercial paper plus 174 basis points while the borrower is in school or in the grace period and 234 basis points while the borrower is repaying the loan. Lender yields on parent and consolidated loans will equal the rate on three-month commercial paper plus 264 basis points. Under current law, however, the basis for the interest rate formula is slated to change once again in 2003 to Treasury securities whose maturity is comparable with the maturity of student loans. The Treasury and CBO have both interpreted that reference rate as the average for long-term Treasury securities (including 10-year Treasury notes as well as bonds with even longer maturities).
 

CHANGING THE INTEREST RATE FORMULAS

Before enactment of P.L. 106-170 in December 1999, tying lender yields to Treasury benchmarks had raised concerns among lenders and spurred proposals for changing the rate formulas. Private financial institutions making loans through the federal student-loan program wanted Treasury interest rates to be replaced permanently with private-market rates as the new reference. Lenders argued that private-market rates determined their loan costs and that they risked losing money if the interest rates at which they borrowed funds for student loans did not move in tandem with the Treasury's rates.⁽¹⁾ Lenders were also concerned that projected federal surpluses might shrink the market for Treasury bills (which are sold to finance government debt) by enough to make the bills less representative of overall market rates.

Generally speaking, lenders argue that their costs of funds for student loans are closely tied to private-sector markets for short-term securities--particularly commercial paper issued by financial firms and London interbank dollar deposits. (The interest rate on those deposits is known as LIBOR--the London interbank offer rate.) The market for commercial paper is now much larger than the market for Treasury bills and is likely to continue to grow. Similarly, LIBOR plays a role domestically and internationally as a reference rate for many private-market loan instruments, and that role is not likely to diminish.

Using such alternative rates in interest formulas for the student-loan program may reduce the risk that lenders face, but other factors also require consideration. The government has used three-month Treasury bills as a benchmark for the student-loan program because their interest rate is widely regarded as a standard measure of a risk-free rate--that is, free of the risk of default. Another reason for their use is that CBO and the Administration already project future Treasury bill rates as part of the federal budget process. Of primary importance is what happens to federal costs for the program if one of the alternative rates is used. In estimating such costs, CBO must take into account not just the expected differences, or spreads, between Treasury bill rates and the alternatives but also the rates' differing volatilities.⁽²⁾ The alternative rates are generally more volatile than the rates mandated for loans after June 2003. As a result, the probability increases after that date that interest rates will from time to time exceed the caps. Assessing those probabilities and the potential need for special-allowance payments requires analysts to both understand the factors underlying movements of rates in the past and be able to project future rates and their volatilities.

Over the past three decades, interest rates and interest rate spreads have been at times volatile and at times relatively stable. In the 1970s and 1980s, a variety of shocks buffeted the rates, producing relatively large ups and downs. (Such events include the end of the Bretton Woods system of fixed exchange rates and the oil price hikes occurring in the mid- and late 1970s.) Important changes in the U.S. financial system--for example, increased diversity in the intermediation between savers and borrowers and wider dispersion of risk bearing--may help prevent such shocks from battering rates in the future. In the past decade, in fact, markets have generally been calmer, and relatively few big jolts have moved interest rates and spreads.

To estimate the cost of the student-loan program, CBO constructed a general statistical model (that is, a system of interrelated equations) that can accommodate the kind of changes noted above. The two-step model, which projects different benchmark interest rates and their volatilities, takes into account the diminished movement of rates in the 1990s and even finds some systematic reasons for that change in the relatively low inflation and stable monetary policy that the United States has enjoyed since the mid-1980s. (Of course, the projections also allow for the possibility that inflation will climb.) For technical reasons, CBO chose to project the spreads between rates on Treasury bills and on the alternative instruments rather than the level of those rates. In discussing its model and the illustrative projections, CBO uses "rates" and "spreads" interchangeably to refer to the rates' future behavior.

The model's results show that with one exception, spreads between the alternative rates and rates on three-month Treasury bills are narrow compared with past average spreads. (For the most part, the spreads discussed in this paper are measured against the three-month Treasury bill.) The spreads' general narrowing in CBO's estimates stems mainly from the favorable economic conditions that CBO is projecting for the next several years. The exception is the spread for three-month commercial paper. In the past, that spread has been slightly smaller than the spread for one-month commercial paper. CBO's model estimates that it will be slightly wider than the one-month spread over the 1999-2009 projection period.

1. For further discussion of the funding risk, see Congressional Budget Office, "Letter to the Honorable Pete V. Domenici Regarding the Profitability of Federally Guaranteed Student Loans," March 30, 1998 (available at http://www.cbo.gov/Letters.cfm).

2. See the appendix for a more detailed description of how spreads are calculated.

---