US Debt Duration Grows As T-Bill Share Plunges To Pre-Crisis Levels
One of the saving graces of the burgeoning US Federal debt was, for the majority of the post-crisis days, its composition, split between short-term (Bills), and long-term (Coupons) debt. As Bills have on average yielded well below 0.5% (and recently even less, now that the 2 Year is yielding 0.5%), a sizable portion of the US debt, was for structural reasons, paying essentially no interest. Since in early 2009, over a third of total US debt consisted of Bills, due to massive foreign (especially China) and domestic demand for ultra short-term US paper, nearly a third of US debt was "free" to the US. However, as Morgan Stanley points out recently the Bill portion of total marketable debt (which today closed at a record $8.777 trillion) has plunged to just 22%. The balance, as the debt portfolio has rolled in time, has logically been filled by much higher interest paying coupon debt: from a low of just 55% in early 2009, this has risen to 72% currently, or roughly $6.3 trillion. This is troubling for the Treasury as it means that due to ongoing rolling of Bill debt, there is no longer an easy way to issue "interest-free" debt. In summary, the average maturity of US Treasury debt is now 58 months from just 48 months at the end of 2008, and as Morgan Stanley points out, is directly in line with the average since 1980. The conclusion is simple: very soon what the Fed does on the front end will have increasingly less of an impact on the interest rate the US pays on its borrowings.
More obsevations from Morgan Stanley on the maturity transformation of US debt.
Slower pace seen for further extension of the average maturity of the debt. In the TBAC minutes, Colin Kim – the Direct of the Office of Debt Management – notes that while the front- and intermediate-end auction size cuts will continue, the pace of the extension of the average maturity of the debt will likely occur at a slower pace. According to our economists, these minutes seem to confirm the Treasury’s concern about cutting T-Bill issuance too much and possibly having this impact the liquidity of the front end of the curve. Exhibit 8 – top – demonstrates that the T-Bill share of the total debt portfolio has declined to 22%, or 3% below its 25% 10-year average (or versus a pre-crisis average of 24%), which may be the source of this concern. Therefore, the notes by Colin Kim could mean that the absolute level of T-Bills may not decline further from its current $1.8 trillion level (and will only go up as the size of the debt burden grows).
Additionally, MS' Igor Cashyn muses on the future gross and net issuance of total US debt:
Gradual reduction of coupon sizes in the front to intermediate end continues. Beginning in April 2010, the Treasury steadily cut the issuance sizes of 2s and 3s (down $6B from peak), 5s (down $5B), 7s (down $3B) and 10s (refunding down $1B – not expected to repeat). This represents a cumulative decline $232B of annualized borrowing capacity from April levels. This week’s statement reaffirmed that the Treasury will continue to decrease the coupon auction sizes at a gradual pace, and our US economists are anticipating a $1 billion cut / auction through the end of 2010 for 2s and 3s, and through September for 5s and 7s (Exhibit 5). 10s and 30s are expected to stay unchanged, and putting the issuance in the various tenors into 10y equivalents, we can see that front- to intermediate-end gross issuance will have peaked in F2010 while the back end stays high (Exhibit 6). Treasury officials indicated that once these cuts are complete, the Treasury will likely hold auction sizes constant for a period of time to assess the fiscal outlook.