GARP Presentation on Interest Rate Risk and the Treasury

Last week, I gave a talk to the Denver Chapter of the Global Association of Risk Professionals (“GARP”) on the topic of “Interest Rate Risk and the U.S. Treasury.”

Click Here to see the slides in a new window (the slide deck was slightly extended and enhanced from the original presentation). The slides have also been posted on the GARP website.

Last month, the U.S. Treasury published a chart (Slide 2) announcing its progress in extending the weighted average maturity of its marketable securities.  According to this chart, the average maturity has increased by 32% since 2008, suggesting that the Treasury has been locking in its funding for longer periods of time, reducing its exposure to rising interest rates.

Several observers have expressed concern that by extending the maturity of its portfolio, the Treasury is interfering with the Federal Reserve’s efforts to remove longer bonds from circulation.  The Fed’s operations are intended to drive down long-term interest rates and stimulate the economy.  Bernanke recently stated that “to the extent that the Treasury actively sought to lengthen the duration of its borrowing it would to some extent offset the benefits” of Fed actions (Slide 20).

I took a close look at how the Treasury has been managing the maturity profile of its debt.  While the outstanding securities have an average maturity of 64 months, the median maturity is only 38 months (Slide 9).  Almost half of the current portfolio will need to be refinanced in the next three years.  The portfolio is heavily front-loaded, with much of it concentrated in short-term T-Bills at near zero rates (Slide 14).   The distribution by interest rate is similar to the distribution by issuance date (Slide 10).

Slide 22 shows that while the Treasury has reduced the share of maturities of one year or less since 2008, the proportion of maturities of up to five years and of up to ten years has flattened out.

I studied the slight increase in average maturity from April 30 to May 31.  I wondered how the Treasury had caused this increase.  To my surprise, I saw that the average maturity of the new securities sold in May was only 27 months (Slide 24).  How could such a short-term basket of securities bring up an overall average maturity that was more than twice as long?

I realized that there were three important drivers of changes in average maturity in May (Slide 25):

1. The securities that remained in the portfolio (about 94% of them) drew one month closer to their respective maturity dates, reducing the average maturity of the portfolio by almost a month.
2. The $585 billion of securities that matured in May increased the average maturity by nearly four months. This seems odd, but these securities were in the portfolio on April 30, and their extremely short remaining maturities weighed down the average maturity on that date.  The disappearance of these securities by May 31 lifted the average maturity up, like a balloon dropping ballast.
3. The front-loaded new securities reduced the average maturity by more than two months.

The net effect was a slight increase in average maturity.  If the Treasury had issued no new securities at all (or, equivalently, had matched the pattern of the existing securities), then the average maturity would have increased by 2.6 months.  This was the “automatic” or “natural” change that had been predetermined by the structure of the portfolio.  The Treasury’s issuance activities in May slowed this growth by 86% to only 0.4 months.

The Treasury’s chart (Slide 2) showed an increase in average maturity from 48.5 months in October 2008 to 64 months in May 2012.  But the average maturity would have increased to 78 months if the Treasury had not slowed its natural growth by issuing mostly short-term securities (Slide 26).

Bernanke’s statement included the condition “to the extent that the Treasury actively sought to lengthen the duration of its borrowing.”  The Treasury is not actively lengthening the average maturity.  It is not flooding the market with long-term bonds that could offset the Fed’s efforts to stimulate the economy.  The portfolio continues to be very front-loaded and very exposed to interest rate risk.

Revised January 21, 2013 with a correction to Slide 9.  The median maturity was 36 months, not 38.

Note: A more recent presentation on the Treasury’s debt can be found here.

Copyright 2012. All Rights Reserved.