Floating-rate Treasury notes

I have had no time to write lately, but I really want to change that, starting today.

What’s that? Something about Europe? Enh, too complicated. I need to ease my way back into blogging with something simple, like explaining why I think two Bloomberg writers and one Duke University Finance Professor are confused.

Around a week ago, there was a brief buzz about the U.S. Treasury contemplating the issuance of floating-rate debt. It seems Treasury first mentioned this on February 1, so I am not sure why it took so long for the financial media to latch on, but here we are.

Anyway, Bloomberg ran a breathless piece on the topic:

Father of Treasury Floaters Says Now Worst Time for Sales

â€œIn an environment with historically low interest rates, the Treasury should avoid floating-rate debt as it introduces risk,â€ Harvey, a finance professor at Duke Universityâ€™s Fuqua School of Business in Durham, North Carolina, said in a telephone interview April 17. â€œIf interest rates go up, it puts the government at risk because they will need to come up with a lot of extra revenue to pay the interest bill.â€

Truth be told, “what the h*ll are they thinking?” was my first reaction, too. I mean, if offered a 3¾ percent 30-year fixed-rate mortgage, would our fine public servants at Treasury really jump at the Option ARM instead? Are they really planning to expose the U.S. taxpayer to interest rate risk when rates can only go up from here?

Short answer: No, not really. For one thing, it is not at all obvious that rates can only go up from here. But never mind that. Let’s try an analogy.

Suppose I knock on your door and ask, “Can I borrow $100 real quick? I will pay you back tomorrow.” You trust me (everybody trusts me), so you say sure, no problem. The next day I return your$100 plus a penny or two of interest, and I ask, “Can I borrow that $100 again? I will pay you back tomorrow.” You say sure, why not, and hand the$100 right back.

We proceed like this every day for an entire year. The interest you demand from me might change a bit from day to day, but every day I repay you, and every day I borrow the $100 again. This is called “rolling my debt”, and it is what the U.S. Treasury does all the time. OK, OK, in the case of Treasury, it is not exactly every day. But it is not exactly$100, either. For example, from January to March of 2012, Treasury borrowed $523 billion and repaid$498 billion in 4-week bills (h/t Bond Girl).

Suppose I next say to you, “Look, obviously I am going to borrow $100 every day. This year, how about if we just sign a contract that I promise to borrow$100 daily and you promise to lend it, and we will agree on some benchmark to use for the interest rate?” And you say sure, why not. And since we have that contract, there is no reason for me to hand you the $100 every day only to have you hand it right back; I will just hang on to it and pay you the daily interest, returning the$100 at the end of the year. Of course, you will probably charge me a premium to enter this contract, since in general you have a preference for perfectly-liquid cash over any debt instrument. (Although with T-bill yields occasionally going negative these days, apparently that preference is not very strong, to put it mildly.)

There are two important points here. First, the only difference in year 2 is that we have a contract compelling us to do what we planned to do anyway. Second, our contract in year 2 is nothing more nor less than a floating-rate loan.

So, no, Treasury is not talking about introducing taxpayer exposure to interest rate risk, because we already have that exposure courtesy the 1 trillion or so being rolled in the short-term debt markets every quarter. The TBAC discussion from February (flip to page 40, h/t Alea) makes it clear that they are presenting 2-year floating-rate notes as a financing choice that sits between rolling short-term bills and issuing 2-year fixed-rate notes. I guess not all public servants are stupid and/or malign. Who would have thought? P.S. The Treasury Borrowing Advisory Committee presentations are really quite good. Worth bookmarking. P.P.S. Did you know there is an RSS feed for Treasury auction results? It is quite possibly the most boring RSS feed in the world. Pray you are not around should that ever change. 5 comments to Floating-rate Treasury notes • Bond Girl FRNs will just get Treasury off of massive bill rolls. More interesting question is which reference index to use. • Nemo I agree the reference index is an interesting question. I hope it’s not LIBOR… Still, it is somewhat odd for Treasury to be worrying about rolling their bills. Just by way of example, on May 3, they sold36 billion in 4-week bills into demand for \$152 billion.

So with interest rates at all-time lows, and with bid-to-cover ratios at all-time highs, I think “why now?” is also an interesting question.

• Greycap

They are just contemplating augmenting existing bill issuance, not replacing it. So probably the index will be the bill auctions themselves.

• HotSam

There is a bit of a difference here. In your loan example, the debt is an overnight loan with a low duration. A contract for a longer term loan has longer duration.

Floaters at longer maturities would have much greater exposure to interest rate risk than short term bills because duration would be greater.

The governments cost of capital must take into account the buyer’s appetite for risk. They don’t just issue debt and the lender holds it to maturity – these are liquid assets which are bought and sold regularly and as HFT must be marked to market.

Floaters would have a lower duration than an equally maturity fixed rate security under the current rate regime and would be sold at a premium to those securities. But they’d have to be sold at a discount to shorter term bonds with a maturity equal to the reset frequency and equal yields?

Or am I wrong, and that’s entirely possible?

BTW, I’d love to hear your opinion (or that of Bond Girl) on the JP Morgan fiasco.

• bob

Isn’t the point of floating rate something to do with reduced processing costs? With floating rate 2 year notes investors presumably would pay less for a two year note, allowing Treasury to sell less 3 & 6 month bills. Probably for a while there they sold more bills thinking they would eventually reduce the deficit and not have to roll half a trillion a month (or whatever). Now it’s clear those short bills will never get paid of in the short term, issue them for 2 years and save some processing costs.