On the Choice of Maturities for New Treasury Issues

Every week the Treasury issues new securities with maturities ranging from a few days to several years. In the first ten days of December, for instance, bills and bonds with aggregate face value in excess of 220 billion dollars were auctioned; these had maturities ranging from four days to thirty years and annualized interest rates ranging from practically zero to 4.4%. About two-thirds of the total was short term borrowing due for repayment within the next six months, and the remainder was raised by issuing three and ten year notes, and thirty year bonds. Since there is no prospect of a budget surplus at any time over the next few years, incoming tax revenues will be fully absorbed by projected spending and the short-term debt will have to be rolled over on an ongoing basis (or converted at some point to long-term debt).

Clearly, the Treasury has a fair amount of discretion regarding the manner in which the deficit is financed. How ought this discretion to be exercised? And how is it exercised in practice? Theories of corporate maturity choice (such as Diamond, 1991) are not terribly useful in answering these questions because the goals and constraints are so different. The Treasury does not have to worry about liquidation by its creditors, for example, and corporations are not generally concerned with the effects of their actions on the term structure of interest rates or the value of the currency.

The issue is really quite important. A couple of weeks ago, Paul Krugman suggested that the Treasury should shift towards shorter-term maturities to finance the deficit if long term interest rates were to start rising. Two days later New York Times reported that the Treasury was in fact doing just the opposite, "exchanging short-term borrowings for long-term bonds." Without a theory (either positive or normative) of maturity choices for new issues, such proposals and actions are hard to evaluate. Furthermore, the costs of making poor judgments can be very high, both in terms of the interest burden on taxpayers, and the effects on private sector choices of changes in the yield curve.

If the future path of interest rates could be anticipated with perfect precision, then the yield curve at any point in time would be fully determined by no-arbitrage conditions. In this case the cost of borrowing would be essentially independent of maturity choice. But there is always uncertainty about future rates, so rates on short term debt, which is a preferred habitat of lenders, typically tend to be lower than long term rates.  In this case the cost of debt service will be smaller if the deficit is financed with short rather than long term obligations.

Aside from the cost of financing the deficit, maturity choices by the Treasury affect the shape of the yield curve and the extent to which private sector institutions engage in maturity transformation. This has implications for the stability of the system and its vulnerability to liquidity crises, as Bill Dudley has recently observed. In light of these considerations, I find it a bit surprising that while the size of the deficit is a topic of endless controversy, there is such little debate about the manner in which the deficit is financed.

---

Update (12/17): In an interesting post, Andy Harless discusses related issues.