Harvard economist Benjamin Friedman's proposal to save the Treasury interest costs by reducing the maturity of government debt has the approval of the Democratic Study Group in Congress and its A-men corner in the financial press. It has also thrown the Clinton Administration into its first important internal debate over debt and monetary policy, with the Treasury Department apparently lining up against the idea -- to the point where Secretary Bentsen called, then canceled, a press conference yesterday morning to announce opposition to the move. The White House signaled "time out" when CEA Chairman Andrea Tyson stated at her confirmation hearings that the administration was seriously considering doing just that. The bond market clearly lined up with the Friedman proposal, the 30-year bond tracing its ups and downs yesterday in accordance with the shifting news coming out of Washington. The amounts of money are small. We're talking about $2.5 billion in a sea of debt of four thousand billion. The importance of the debate itself, though, could ultimately be worth trillions, in one direction or the other.
We have in the past been on both sides of this issue. In 1981, we denounced the issue of 15% "Sprinkel Bonds" by the Treasury when Beryl Sprinkel was Undersecretary for Monetary Affairs, precisely on the grounds that the taxpayers would be stuck for 30 years with this outrageous interest rate, only slightly less if the bonds were called in 25. Treasury Secretary Donald Regan, it turned out, had been bamboozled by his old friends on Wall Street into rejecting the shortening of maturities to economize on debt service. Sprinkel, it turned out, had privately agreed with us, but I did not find that out until after I was widely quoted in the financial press -- including the Times of India, for goodness sakes -- saying "The public flogging of Beryl Sprinkel would do wonders for the bond market."
The issue came up, again, two years ago, and we then opposed the shortening of maturities to take advantage of the spreads. It was not clear that much was to be gained given the situation we saw at the time. Lately, with the long bond well below double digits, the Fed has generally opposed the idea of making changes in the patterns of bond issuance on the grounds that the gains would be much smaller than advertised; furthermore, the long bond provides important information about inflation expectations and should not be messed with on an ad hoc basis. As far as I can tell, this is now the position of the Fed, although there is some lukewarm support among the governors for shortening up.
In this case, we would do exactly what Friedman suggests, although we disagree with the reasons he offers for doing so. Supply and demand of government debt does not determine long-term interest rates. If it did, the long bond would boom if the government shortened maturities of new debt to zero, issuing non-interest bearing paper to pay its bills or retiring bonds with cash. A mix of expectations about government behavior determines long-term interest rates, which is why we, at times, experience rising bond yields coincident with falling commodity prices. An argument can be made that, in the current situation, a drastic reduction in supply of 30-year bonds would lower the long-term bond yield slightly, because investors would pay a premium to maintain the preferred maturity structure of portfolios. Buying long bonds rather than short-term bills amounts to a bet that interest rates will fall, and investors would pay slightly more to make such bets. Friedman's claim that shifting a quarter of new debt sales from longer-term bonds to short-term paper would drop rates by half a percentage point, though, depends upon simplistic supply-and-demand calculations which treat bonds as if they were just another commodity. By the same argument, short-term rates would rise if the Treasury dumped more short-term paper on the market.
The reason I would now shorten the maturity of the debt structure is that I believe the price of gold, not the price of the long bond, is the market's purest guess of inflation expectations. The bond market, we have proven to our satisfaction, is happiest when gold is at $350. Fed Chairman Alan Greenspan and fellow governor Wayne Angell are also happier when gold is at $350, although they are clearly not as concerned about the happiness of the bond market as the bond market is, or as we are. If they were, they would not have permitted the gold price to slide below $340, then $330, a clear deflation signal coincident with rising bond yields these past two months. They could have had the Fed do the job of shortening maturities of government debt by buying bonds in the open market with newly created reserves. The Treasury Department cannot create reserves, which only means it can't issue non-interest bearing cash -- the most liquid of all government debt -- to pay its bills. But it can have much the same effect by issuing shorter-term securities, when the spreads are wide, as they are now, and the gold price is low, as it is now. This is probably why the long bond acts so happy when the news of what Treasury will do runs in that direction.
This is a different proposition from Friedman's claim that issuing shorter-maturity paper is a "win-win proposition." I would not be so crude as to suggest a precise shift in volume and assert a precise price effect, as he does. Shifting a quarter of new debt sales to short-term paper might be too much for the market's appetite for more liquid assets, and we would observe the gold price climbing above $350, along with bond yields.
If I were Lloyd Bentsen, I would, though, do exactly what Friedman recommends, and shift a quarter of the new debt issue to shorter paper -- as the bond market has already applauded the idea in advance. If I were Greenspan, and wanted to keep control of monetary policy at the central bank, instead of ceding it to Treasury in this manner, I would get the gold price back up to $350 by shortening maturities through the FOMC! Then, any talk of shortening maturities at Treasury would probably get raspberries from the bond market.
In 1981, the Fed was still totally hung up on monetary aggregates, which led it into the horrendous deflation and recession that followed. At the time, I remember advising White House Chief-of-Staff James Baker III that President Reagan could put virtually complete control of monetary policy into the hands of the Treasury -- by asking Treasury to stabilize the dollar value of our international gold reserves, and to ask the Fed to cooperate. This would, of course, have instantly replaced the M-1 target with a gold target. The deflation and recession would have been avoided, along with the colossal federal deficits that recession brought in its wake. The idea was a bit too avant garde for JBIII.
Leaving aside the exaggerated claims about big interest savings, any proposal to change the maturity structure of government debt amounts to asking the Treasury Secretary to take a view on interest rates. Yes, by issuing long bonds, the Treasury insures itself against future increases in interest rates. But given the Fed's success in containing both inflation and inflation expectations, there is something to be said for the Treasury placing bets on even more such success in the future. Bentsen would surely be doing this if he shortened maturities in the manner described. The effect on bond prices might not produce terribly significant amounts of savings on debt finance, although every billion helps. The process itself would help educate this new Administration on the dynamics of monetary policy. If it really wanted to send a signal to the market that it is betting on lower future interest rates, it could offer a special issue of bonds with principal payments indexed to the price of gold. Instead of picking up a handful of basis points, Treasury would pick up a handful of percentage points. Homestake Mining's gold bonds now yield just 3% on the market; Treasury could probably obtain an even better rate. The idea is not one that Alan Greenspan would have any trouble with. He's been pushing it for at least 15 years.