Puzzles about interest rates on debt
The federal government is paying much less interest on its regular debt than do other borrowers. The graph below compares interest rates on Fannie Mae’s bonds to those of Treasury bonds maturing at the same future date, shown on the horizontal axis. The vertical axis is the spread or difference between the rates. The rates are yields from the bond market. They measure what Fannie or the government would need to pay to issue a new bond maturing at a given date, based on the current market prices of bonds issued in the past maturing on that date.
Notice first that Fannie always pays higher rates than does the Treasury. This is a puzzle to begin with because the government has virtually declared that it stands behind Fannie’s bonds. The other puzzle is how much the spread differs across maturities, without any coherent pattern. In particular, investors seem to be allergic to bonds coming due in 2012 through 2014. The reason can’t be that they question Fannie’s ability to repay the bonds, because those maturing much later, in 2029 and 2030, have rates only moderately higher than Treasuries.
What about other government agencies? The Tennessee Valley Authority, a federally sponsored power company, pays rates on its bonds only a little above the rates on Treasury bonds. Similarly, the Federal Home Loan Bank pays only moderate spreads, despite its exposure to mortgages–it loans to banks and savings and loans with mortgages as security.
The government also pays higher interest on some non-standard debt of its own. The graph below shows the spread between the reported yields on Treasury Inflation-Protected Securities (TIPSs) and regular Treasuries of the same maturities. We the reported yield because what is reported is what investors receive and the Treasury pays if there is zero inflation from now until they mature. For each percentage point of annual inflation, the actual yield is a percentage point higher. Inflation has averaged about 2.5 percent per year for the past 25 years, so generally TIPSs’ reported yields are below those of Treasuries and the spread is negative. As the graph shows, that’s true for TIPSs maturing in 2016 and later. But for earlier maturities, spreads are positive! Do investors really believe that inflation is going to be so negative over the next 7 years that TIPSs will be less valuable than regular Treasuries? No responsible forecaster is saying that so much deflation will occur. TIPSs appear to be an unbelievable deal for investors. Economists have been buying them like crazy lately.
Policy should pay attention to the market’s views about the safety and desirability of different kinds of bonds, even when, as now, we can’t make sense out of those views. As we stress in our analysis document, the government should continue its steps to enable Fannie and Freddie to borrow at the Treausry rate. So far, the government has accomplished this by having the two agencies borrow at high rates partly from the Fed, while the Treasury lends the fund to the Fed to pass on to Fannie and Freddie. If the Fed bought all of the agencies’ new debt, the effect of the current scheme would be the same as if the Treasury borrowed on behalf of the agencies directly. If some of the borrowing comes from investors rather than the Treasury, then the government is throwing money away, assuming that it in fact does back the agencies’ bonds fully.
With respect to TIPSs, the message is that the government should not be borrowing in ways that cost it more. If investors have to be rewarded with ridiculously high returns to get them to buy TIPS, then the government should give them what they want, namely regularly Treasury bonds, which the market seems to crave. TIPSs do not account for very much of total Treasury borrowing, however.