Monday, January 08, 2007

Grading the WSJ's understanding of yield curves...

Today's Wall Street Journal article titled "Grading Bonds on Inverted Curve" by Michael Hudson has me feeling frustrated. I must vent.

The yield curve inverts for exactly one reason. Bond traders believe that short rates are likely to fall in the near future and therefore are willing to lock in longer-term investment rates even though those rates are lower than short-term rates.

Think of it this way. You can buy a 3-month bond or a 10-year bond. Right now the 3-month bond is yielding 4.92% and the 10-year 4.65%. If you thought rates were going to remain stable, you'd buy the 3-month bond. However, if you thought that 3-month rates would be lower than 4.65% for the next 10-years, you'd rather lock in the 10-year rate now.

So there I've explained in simple English exactly why the yield curve inverts. There is nothing more magical to it.

Here is why I'm frustrated. Because Mr. Hudson and a myriad of other commentators are obsessed with linking the inverted yield curve with a possible recession. Do you know why the yield curve often inverts ahead of a recession? Because during a recession, the Fed is usually cutting rates. That's right folks. The inversion happens because the Fed is cutting short-term rates. The relationship with any impending recession is indirect.

Now let's talk about today's economy. The Fed seems likely to cut rates once or twice in 2007, with or without a recession. So that puts the Fed Funds rate at 4.75%. Short-term Treasuries normally trade through Fed Funds, an average of 8bps. So if Funds are at 4.75% and there is normally 8bps between Funds and T-Bills, then 3-month bills are at 4.67%, damn close to where the 10-year is today. Throw in that there is some chance that the economy will get even weaker and the Fed will cut more aggressively, well then it seems like the 10-year is priced about right.

The thing is, you don't need new-fangled ideas about how the world has changed to explain today's yield curve in light of the consensus that there will be no recession. All you need are basic bond fundamentals.


Unknown said...

feeling your frustration and would just like to add/throw into the mix concept of some corporate malfeasance (not sure if that's the right word) of sorts. pension under-funding! leading into the tech bubble, and the fall-out years afterwards, folks managing really big portfolios of securities got comfy with some assumptions that, while seeming right at the time, just didn't stand up. assuming stocks would return mid-teens ad infinitim just NOT what the folks over at ibbotsons tells us. be that as it may, all sorts of folks got overweight 'riskier assets' than they were supposed to. i'd argue some of this conundrum and inversion has come as folks get their allocations back into line (evidence today's Strips data released today suggesting continued demand for long principal strips by pension funds). there's a wall of money out there that needs relative safety of Treasuries as well as long duration to offset liabilities.

that whole entire long-winded mindless giberish and rant has NO mention of recession and more than likely doesn't need one. there's more to this bond mkt stuff than some of those dudes over at the wsj would lead us to believe.

what are your thoughts about pension fund demand as supporting role in yld curve inversion?

MM said...

Sounds grea, but what about this:
Just coincidence?

Thanks for Accrued Interest, btw; finally a good bonds blog!

Accrued Interest said...


Pension spending has to have an effect on the yield curve, no doubt. I think companies increasing pension spending is helping spreads too. A lot of people thought that the lack of new 30-year auctions were the cause of the inversion, but we've had two auctions now, and guess what? Still inverted. I think its worth noting that between 3-10 and 3-30 there is still a positive slope.

The point of my frustration is that you can argue the yield curve inversion <> recession without making the infamous "its different this time" claim. I wasn't trying to discount other flattener influences such as pension fund spending.


I'll follow up in my next column.

Anonymous said...

You can buy a 3-month bond or a 10-year bond.

Actually, as I am sure you know, the 3 month debt obligation is a treasury bill, and the 10 year obligation is a treasury note . . .

From the federal reserve:

Bills—securities having a maturity of one year or less . . . Notes—securities having an initial maturity of one to ten years—and bonds—securities having an initial maturity of more than ten years . . .

Please keep your terminology straight. Thank you.

Accrued Interest said...

Glad you cleared that up for me.

Anonymous said...

Sir, i wanna ask very stupid question:

why is there a 8bps spread between tresuries and Fed Funds Rate?


bond rookie

Accrued Interest said...

Rookie: The 8bps isn't constant, and in fact is MUCH wider right now. But Fed Funds is technically a credit-risky security, in that it is actually nothing more than an overnight loan from one bank to another. The Fed does manipulate the rate, but its still a loan to a bank. Therefore it ought to have a spread over tsy.

Anonymous said...

thank you very much!