June 27, 2005


Low Rates Could Be Around for Long Term (EDMUND L. ANDREWS, 6/27/05, NY Times)

Federal Reserve officials, who meet this week, are beginning to suspect that the perplexing decline in long-term interest rates is more than a temporary aberration.

The possibility has major implications for the economy, and it creates new puzzles for Fed officials on how they should respond. [...]

One school of thought holds that low bond yields are a harbinger of slowing economic growth, which would reduce demand for credit in the future. Another school holds that global investors have lower inflation expectations than in the past, which reduces the risk of holding long-term bonds. If either theory is correct, the Federal Reserve would have less need to fend off inflation and could stop raising short-term rates at a much lower level than in the past - perhaps below 4 percent.

But yet another theory holds that long-term interest rates may have been depressed by other factors, including a "savings glut" around the world and efforts by Asian central banks to keep the value of their currencies down by buying United States Treasury securities.

If that is true, the flood of foreign money into the country could be diluting the Fed's effort to prevent inflation. That would imply that the Fed needs to raise rates more than many investors are expecting.


Posted by Orrin Judd at June 27, 2005 5:22 PM

We are much more energy efficient than when OPEC put the squeeze on us in the 70s.

That's a factor, too.

Posted by: Sandy P at June 27, 2005 6:32 PM

Cheap money means low interest rates which means lower returns for the same risk class of investment. How this could possibly contribute to inflation I have no idea.

Any inflation in the marketplace is due to the price of energy, which is the result of a failure in the market to build enough refineries. At some point, the oil industry will respond to this need for increased supply and build more refineries, at which time, the price of oil will drop once again.

Posted by: bart at June 28, 2005 7:02 AM

Remember that inflation is defined as an increase in the money supply relative to the value of the assets in the economy. If you have an economy with a money supply of $100 and only 5 widgets as assets, each widget is worth $20. If another $100 is then printed, increasing the money supply then increases to $200, the price of the widgets increases to $40 with the increase being due solely to inflation. If, on the other hand, someone in the economy comes up with a new use for widgets and is willing to pay up to $40 for a marginal widget, without the money supply having increased, then that is not inflation. In other words, the increase in energy prices due to increasing demand that cannot be satisfied with current refinery capacity (to the extent that's what's happening) is not inflation.

The Fed's problem is two-fold. First, over the last twenty years the Fed's ability to effect the rate of money creation has diminished significantly. So, to use the common metaphor that the Fed's task is akin to steering a supertanker, over the last twenty years the rudder has gotten smaller while the tanker has gotten much bigger. Second, and this is much more recent and not much talked about, the Fed has lost much of its ability to even measure the money supply or to figure out what counts as money. So, bigger ship, smaller rudder, and increasingly worse information about what's going on with the ship.

Posted by: David Cohen at June 28, 2005 9:02 AM

Oh, I should note that part of what makes the whole thing complicated is implicit in my widget example. If the value of widgets increases to $40 without any increase in the $100 money supply, then the economy would be in serious trouble because there would be too little money. So in that case, you want the mint to be printing more dollars -- but if they print too many, you'll have inflation. The mint, or the Fed, because most money is not printed, needs to follow the Goldilocks rule and get it just right. But they don't have the information they need to know what just right is, and they don't have the control to turn of the spigot with anything like the requisite precision.

Posted by: David Cohen at June 28, 2005 9:10 AM

Widgets never cost more than they used to. There are just more, better widgests cheaper.

Posted by: oj at June 28, 2005 9:18 AM

This, too, shall pass.

Posted by: David Cohen at June 28, 2005 9:29 AM

But if the money is being added to the economy on the investment side of the equation, doesn't that increase supply, pushing the curve down and to the right as it were? IOW, if the money gets injected into the economy as a new widget factory doesn't the supply of widgets go up and the price of widgets go down?

Posted by: bart at June 28, 2005 10:05 AM

Bart - Possibly. But the price of things the widget industry buys - materials and labor especially useful in making widgets - will be bid up by the increased demand. If the money supply has increased, the net effect of all this, in the LR, will be that prices in general rise.

But: The amount of output we produce is also relevant, as was implicit in David's remarks. If the money supply doubles but an innovation in the general production technology triples its efficiency, you could see the price level falling. (Twice as much money chasing three times as much output.) But when have we ever seen a tripling of production efficiency at the time scales relevant for monetary stabilization policy?

Posted by: Tom at June 28, 2005 12:03 PM

Or rather, a tripling of productivity while the money supply doubled.

Posted by: Tom at June 28, 2005 2:56 PM


In addition to the unknowns about the ship itself, the radar (sonar?) cannot see the rocks and shoals as well. So the crew argues about what this or that ping means, while continuing forward.

The rocks seem pretty evident to some (the potential for massive claims from Fannie Mae & Freddie Mac, the risk from hedge funds, the pension burden to the government, etc.), but outside of the WSJ, you won't hear about them. Run of the mill inflation doesn't even make the list.

Posted by: jim hamlen at June 28, 2005 4:23 PM

Tom: Alan Greenspan will every once in a while suggest that something very odd happened to productivity in the 80s and 90s but was missed by the Fed because it couldn't be measured. For example, if lawyers keep billing 8 hours per day, how can the Fed capture increases in productivity. Lawyers collectively spent billions on computer systems, Westlaw and Lexis were developed, courts now accept electronic filings, etc., etc., etc. But productivity numbers didn't show any change, because they were measuring billable hours -- the price of which kept increasing. So either lawyers were wasting billions of dollars, or the government was missing a huge boom in productivity.

Something is certainly happening: lawyers use to spend a day writing a letter, editing it, having it retyped, then they would mail it. A few days later, it would get to another lawyer, who could take a day to draft up a reply (because he couldn't get it in the mail any sooner, so why bother). It would take another couple days to make the return trip. Elapsed time, one week. With email, the same exchange might now take an hour. A lot of white collar jobs have gone through something similar. (It's always amusing to look at the BrothersJudd sitemeter and see that visitors come to the blog during the workweek much more than on the weekends.) My estimate is that the average whitecollar worker does about 2 hours of real work in an 8 hour day.

In other words, it is possible that in the last 20 years productivity has outstripped money supply growth in about the proportion you mentioned. But as we can't really measure productivity in the largest and fastest growing sector of the economy, and we are not sure we can really measure the money supply, how would we know?

Posted by: David Cohen at June 29, 2005 8:17 AM

We could find out what happened to the money supply and GDP. The Fed has figures for the monetary aggregates if you're interested. Maybe I'll zip on over and check.

Regarding the productivity point, if output is actually growing it's ultimately measurable. E.g., in your example there will be more court cases filed (assuming that counts as productive!) Even for quality improvements, which are harder to measure, there will be evidence, e.g., if cars are safer you can't tell by looking at the car, but you'll ultimately see fewer traffic injuries and deaths. Etc.

Posted by: Tom at June 29, 2005 11:29 AM

Alright, I just zipped off to the Fed's Board of Governors web site. In the five minutes that I was willing to take I found only monetary data going back to November 1998. I looked at the M3 monetary aggregate, which grew at an annual rate from then to now of around 6%, about double what GDP did in that period.
But don't take this too seriously unless you also look at M1 and M2, and get more time into the sample.
To anticipate an objection, no one, including Greenspan AFAIK, thinks we missed half of GDP growth in recent years.

Posted by: Tom at June 29, 2005 11:44 AM

Not accurate figures.

Posted by: oj at June 29, 2005 12:04 PM

As I said, that was all very back-of-the-envelope. If you want to supply better figures, by all means, do it.

Posted by: Tom at June 29, 2005 2:46 PM

Tom: My concern is that no one really knows how much "money" is out there. My checking account is part of the money supply, the unused credit remaining on my credit card is not. I can get a home equity loan from my mortgagee in a couple of days, but that's not part of the money supply. Moreover, because "money" increases as the value of the housing stock increases, there's no reason to think that it increases proportionately with M3.

Posted by: David Cohen at June 29, 2005 8:44 PM

David - One certainly could use broader monetary measures - look at M1, M2, MZM, etc. But whatever you use, I doubt it doubled while output tripled. Regarding lines of credit (like credit cards) as money, monetary economists typically exclude them as money because the borrower's ability to borrow is financed by the lender. In other words, when you spend $100 using a credit card the credit card firm then has $100 less because they're paying for it, until you pay them back. So credit cards are oike you following your girlfriend around buying whatever she asks for - in so doing, you're like the CC company; you haven't actually increased the money supply.

Anyway, I'm not denying that deflation could happen if output grew faster than (whatever is the relevant measure of) the money supply.

Posted by: Tom at June 30, 2005 6:29 AM

Tom: Sure, once credit is spent, it's accounted for. But to what extent is unused credit like money. I have "unused" money in my checking account (let's assume) and I have "unused" credit. How does the total of unused credit in the US effect behavior and "act" on the economy as part of the money supply (i.e., people are more willing to spend their money because they're making assumptions about how much "not-quite money" is available to them).

Isn't there a huge credit-based money-supply overhang that simply didn't exist (because people couldn't easily access credit-on-demand) fifty, thirty, twenty and even ten years ago? As I said before, working online I can apply for a home equity line in about 5 minutes and have the money in a couple of days. Even if I never actually apply, doesn't my knowledge that I could apply effect my spending habits and our collective spending habits as if there were a huge, unmeasured ghostly money-supply out there? For example, I might not save for a new roof because I assume that, when a roof is needed, I'll be able to borrow the money. For another example, my parent's generation abhorred debt and considered it somewhat shameful to borrow to buy a car. Now, a car loan is expected. How can that changed expectation, particularly in a low interest economy (to get back to Bart's question), not have affected the economy like an increase in the money-supply?

I suppose another way of putting this is: isn't the important thing about the money-supply, like so much in economics, people's collective expectations about the future money-supply?

Even if I'm right, though, I don't see how the effect could be quantified, so I'm certainly not saying that money-supply has doubled. I am saying that our supertanker is steaming along in uncharted waters and all we have is a very small rudder.

Posted by: David Cohen at June 30, 2005 9:30 AM

"Even if I never actually apply, doesn't my knowledge that I could apply effect my spending habits and our collective spending habits..."


"...as if there were a huge, unmeasured ghostly money supply out there?"

Less obvious. Maybe. There are other factors that aren't like an increase in M. E.g., banks issue a lot of lines of credit. Some development makes everybody want to use their lines of credit at once, giving banks what is delicately termed "liquidity problems." (They actually don't have the funds to lend. This wouldn't happen with a bona fide increase in M.

"I suppose another way of putting this is: isn't the important thing about the money supply, like so much in economics, people's collective expectations about the future money supply?"

It's one of the relevant things, yes. However, there are lots of things that affect current spending. E.g., expectations of future tax policy, expectations of future recession or expansion, etc. The relevant point here is, while the availability of credit may make you spend more, it doesn't lead to deflation. If anything, inflation, precisely because in spending more, people push up the total demand for goods & services. Also, the effect of this to function kind of like an increase in the money supply is limited. E.g., once banks' management of reserves (basically, idle funds sitting around in the bank) has gotten so efficient that they basically have zero reserves (or the legal minimum of reserves), the process stops. We're already very close to this state of affairs, by the way.

Posted by: Tom at June 30, 2005 10:06 AM
« OH, MY (via AWW): | Main | ANOTHER DAY IN THE LAB »