The long and the short of it

Right now there is essentially universal agreement that interest rates are going up. And this prediction holds for long-term interest rates as well as for the typically far-more-predictable short-term interest rates.

Do you know why short-term rates are typically more predictable than long-term rates?

It's because those are the rates over which the The Fed, and more specifically the Federal Open Market Committee and its Chairperson, currently Alan Greenspan, has the most direct control. That control comes from some fairly complex nuts 'n bolts sorts of things, but in essence comes down to The Fed's regulatory authority over most of the banking system, and it's ability to control the interest rates that banks charge each other for overnight loans.

Do you know why banks make overnight loans to one another?

It's mostly because banks are required by various regulatory authorities to keep a minimum amount of their deposits on hand in cold, hard cash (or, in this day and age, in intangible, not-there-at-all digital information).

These cash funds on hand are know as reserves.

Having reserves, the thinking goes, makes the banks more stable and less likely to fold. Lack of reserves, people say, had a lot to do with bank failures during The Depression because, without reserves, the banks had s no cushion against failing borrowers.

* * *

Here's how this reserves regimen works.

Say that Bank A makes a loan. In doing so, it has, in essence, exchanged some of its cash for the borrower's IOU and maybe the borrower's collateral agreement in favor of the bank, such as a mortgage. Now that IOU does not count towards Bank A's reserves requirement, so, if Bank A was just barely meeting its reserve requirement before the loan, it might well fall below the reserve requirement after the loan, because in doing so it has exchanged its cold hard cash for an IOU.

So Bank A needs to add to its reserves before it closes up shop for the night. It needs to get some cold hard cash, and it needs to get it today.

So let's say Bank A calls up Bank B, to see if Bank B has more than ample reserves, in which case it might be willing to exchange some of Bank B's cold hard cash for Bank A's IOU.

And let's say that Bank B is flush, and is happy to loan Bank A some of its cash, so they do the loan. Bank A gets some of Bank B's cash, bringing Bank A's cash holdings up to at least its reserve requirement, and Bank B gets Bank A's IOU.

Now the loan that Bank A gets from Bank B is usually a short-term loan because Bank A, if it is to stay in business, needs to have other cash coming in from depositors and from borrowers making payments back to the bank on the bank's loans, so it is OK with not having the right to keep the borrowed cash long term. In fact, it would much rather replace Bank B's cash with a customer's cash deposited into a new savings account, let's say, paying 0.25% (25 beeps).

And that's how banks make a good deal of their money: they borrow short-term money (e.g. Bank A borrows short-term money from Bank B or from Bank A's savings account customers) and lend out long-term money (e.g., as a 30-year mortgage to all of us).

As of this week, the short-term rate Bank A will pay to Bank B is 2%, and the 30-year mortgage rate Bank A might make carries an interest rate of about 6%.

Sounds like a nice business, doesn't it? Bank A borrows $100k from Bank B at 2% and lends that same $100k out at 6%, which means that it makes 4% -- often called a 4% spread -- on the whole combined deal. And a 4% business is a pretty nice business, isn't it, particularly if your 6% borrower gives you a mortgage to secure that 6% IOU?

* * *

But long-term rates are a whole different story. Long-term rates are set by the huge supply and demand pushes and pulls the world over. The Fed can try to move those rates, but those supply and demand pushes and pulls the world over are just absolutely gargantuan, so if The Fed steps into that market and tries to move interest rates by buying or selling long-term bonds from its portfolio, it is by no means assured of being successful in moving those rates.

For example, and thinking back to the interest rates up, bond prices down and vice versa rule, it the Fed wants to see long-term interest rates move lower, it necessarily wants to see increases in the price of long-term bonds (because the rule says, bond prices up means interest rates down).

So to get those prices higher it might start buying a lot of long-term bonds, which would, all other things remaining constant, increase the demand for long-term bonds, which in turn would tend to increase the price of long-term bonds and, voila, there you have it: long-term interest rates down.

And let's take this a step further by assuming that those long-term bonds the Fed is buying happen to be U.S. Treasurys (yes, that's how it's spelled), which are long-term bonds issued by the federal government. Then, in that case, the net result of all this is that U.S. debt is being taken out of circulation.

So there's less U.S. debt out there and lower interest rates. Sounds like the late 1990s, doesn't ?

If, instead, the U.S. government is putting more of its debt out there, then it is increasing the overall supply of long-term bonds, which would tend to depress the prices of long-term bonds (over-supply drives prices down), which would mean higher interest rates.

Sounds like the 2000s, doesn't it?

* * *

But the real wild card here is all the other suppliers and demanders of long-term money out there. And right now the biggest question mark is the demand of foreign governments and their central banks (their Fed-equivalents), and foreign investors en masse. How much do they want to hold long-term U.S. debt?

Well, that's a hard call. Probably they want to hold less of it, especially as the dollar weakens relative to other currencies, because a weak dollar means that all dollar-denominated assets decrease in value to the extent the holder of those dollar-denominated assets envisions needing to convert those assets into assets denominated in other currencies. So that drives foreign demand down.

With less demand and more supply, prices should really fall, meaning interest rates should really go up.

* * *

But you know what? This soft of analysis has led people for the better part of this Millennium to say that long-term interest rates were going up, and going up right away. And you know what? They didn't. They are pretty much where they were last year, right after they ticked up a bit after they swooned during the spring 2003 deflation scare.

So when there is pretty universal agreement that short-term interest rates are going up, they probably are. But when those same folks start telling you that long-term interest rates are going up, they are much more apt to be wrong.