Scott Sumner  

The Fed doesn't have a magic wand

No More Stiglers... Merry Christmas and Peace on E...

Suppose I put a digital sign outside my house, with "Sumner's fed funds rate target" and a number. And suppose I adjusted the target rate from 0.25% to 0.5% last week, right at 2pm on Wednesday. Would I have caused the fed funds rate to increase? Obviously not. So why do people assume the Fed controls market interest rates?

The Fed has tools that I don't have, which allow it to influence market interest rates. They can adjust the size of the monetary base, or they can influence the demand for base money. The latter can be done via changes in interest on reserves or changes in reserve requirements. But prior to October 2008 IOR did not exist, and the Fed found reserve requirements to be too clumsy to use as an effective policy tool. That simplified things, as it means that prior to October 2008, the only way that the Fed could directly influence market interest rates was by adjusting the supply of base money.

One way we know this to be true is that the Fed asked for permission to adopt IOR precisely because in October 2008 they wished to target interest rates at a level that would have required a much smaller monetary base than they preferred. (They had recently injected lots of base money to relieve illiquidity in the banking system.) So obviously the Fed does not have complete control over interest rates---other forces will move them around even if the Fed is not doing anything to the monetary base.

Now here's where people get confused. Because the Fed can influence rates, they assume that any movement in rates is caused by the Fed, as if it had a magic wand. I suppose there is a sense in which that is true, as there is always some counterfactual policy that would have produced a different path of interest rates. But it's not true in the sense that people assume. Here's a recent comment I got:

My brain has a hard time getting past the point that the Fed lowered rates, to try to make more money available for spending, to try to increase prices and NGDP, averting deflation, which sounds like a sensible policy.
This has things backwards. The Fed doesn't lower rates to make more money available, they (sometimes) make more money available to lower rates. But at other times rates fall because of market forces, as there is no increase in the amount of money in circulation. Here are the monetary base and interest rates on August 1, 2007:

MB = $855.960 billion, fed funds rate = 5.25%

And here is the same data on April 9, 2008:

MB = $855.411 billion, fed funds rate = 2.25%

Why did interest rates fall sharply? Perhaps one is entitled to say the Fed caused rates to fall, in the very limited sense that some counterfactual policy would have produced a different path for interest rates. But one is not entitled to also call that an expansionary monetary policy, even though expansionary monetary policies do in fact sometimes cause interest rates to fall.

This last point is something people really have trouble wrapping their minds around. The Fed policy of late 2007 and early 2008 did not inject new money into the economy. By itself that means little, as the monetary base is also not a good indicator of the stance of monetary policy. But when the base is stable and interest rates fall sharply, it is quite likely that monetary policy has become much tighter. If you prefer the Equation of Exchange, then M is not always a good indicator of changes in M*V. But when M is stable, and interest rates fall, then both V and M*V are likely to fall.

And when both the base and the level of interest rates fall sharply, then look out below.

PS. Here's some data from October 1929 to October 1930:

October 1929: MB = $7.345 billion, Discount rate = 6.00%

October 1930: MB = $6.817 billion, Discount rate = 2.50%

Yikes! Falling interest rates combined with a falling monetary base is a recipe for disaster. Less supply of base money and more demand for base money. The value of base money rises sharply. The Great Deflation.

And yet some economists will say "The economy failed to revive on the Fed's rate cuts of 1930." My response:

Screen Shot 2015-12-23 at 12.07.20 PM.png

Merry Christmas.

Comments and Sharing

COMMENTS (9 to date)
Philo writes:

Yet what "some economists say" is literally true (though understated). It's when they add that the Fed's policy was "easy" or "accommodative" (absolutely, not just relatively to some hypothetical even-tighter policy) that they go wrong.

James writes:

Scott: I've noticed that mainstream monetary economists deserve a large share of the blame for the confusion you speak of here.

When people like James Grant go on TV and criticize the Fed for manipulating the unit of account and causing inflation, the next guest will be either a researcher with the Fed or an Ivy league professor saying that the Fed is just exchanging cash for less liquid assets of equivalent market value. The implication is that the Fed is really just supplying liquidity, not manipulating prices or quantities. Any changes in the prices of goods or assets are attributable to market forces.

Then when some other guest with a more Keynesian outlook complains that the Fed is not doing enough, the same Fed researcher or Ivy league professor will say the the Fed is undertaking a massive program of injecting liquidity, cutting rates, expanding its balance sheet, etc. The implication is that any change in the prices of goods or assets is a consequence of the Fed waving its many magic wands.

Philo: I thought all economists, when using terms like "easy money," use those terms in a relative sense. There is no absolute zero or Kelvin scale for monetary policy. Every monetary policy is tight or loose compared to some other policy.

Scott Sumner writes:

Philo, Yes, perhaps I should have said "despite" rate cuts.

James, Yes, there's lots of poor communication, although in the example you cite it's not clear to me that the Fed is claiming responsibility for all asset price movements.

A better way to handle that issue is to point out that the stock market likes a healthy economy with low inflation, so if the Fed does its job and produces a healthy economy with low inflation, stocks will do well. Indeed if it even moves us in that direct stocks will rally, as they should.

ThomasH writes:

A more enlightening response would be that the '30s Fed (like the 2008-2015 Fed) did not reduce interest rates enough to revive the economy = did not do whatever it would have taken (buy long term government securities or long term private securities or foreign exchange or charge interest on reserves or some other policy instrument they might have invented) to revive the economy, which would have reduced interest rates. How much of that "other stuff" the Fed needed to do would depend on whether governments were following an NPV expenditure rule or (more likely) departing from it.

pmsap writes:

I'm trying not to reason from a price change and yet trying to understand why did rates fall sharply in 07/08 (with the base stable) and what could the Fed have done to avoid (minimize) the recession that followed.

So, let's imagine that some external factor (for instance Lehman crash) increased money demand by a lot. This would decrease both interest rates and velocity [not only because V(I)]. Could this be the "why"?

Now for the 'what could the Fed have done?' part. A neo-fisherian would say something like "leave rates as it is (high) and the public will think all is okay thus making inflation unchanged". A market monetarist would say "just promise to buy the world if it need be but the the base (M) will rise to offset the decrease in V". Does this make any sense?


ThomasH writes:

The Fed can promise what it wants, but unless it is willing to execute and admiral every once in a while (let inflation exceed its ceiling) "pour encourager les autres" it's promise is unlikely to be believed.

Philip George writes:

Nice piece. But wouldn't a better measure of interest rates be some real interest rate: something like the one in the first graph of although I agree that it too is a crude one.

Right now, we seem to be in a situation opposite to that which you have portrayed. The quantity of money is shrinking and interest rates are rising. That seems to be a recipe for a crash in some financial asset market.

Scott Sumner writes:

Thomas, Actually, it's not clear that the Fed needed to reduce interest rates any more. What they needed to do was boost the money supply.

Interest rates fell primarily because of the Depression. If the Fed had prevented a Depression, then rates would not have fallen as sharply.

You said:

"A more enlightening response would be that the '30s Fed (like the 2008-2015 Fed) did not reduce interest rates enough"

This implies the Fed reduced market interest rates in the early 1930s, but they did not do anything to cause interest rates to be lower. We did not need lower interest rates---as lower rates are contractionary, for any given money supply. We needed a bigger money supply.

pmsap, The period I considered was before Lehman failed.

The NeoFisherian view has never been spelled out, as far as I know. Keeping rates constant is not a policy, it's the effect of a policy, so I can't comment on NeoFisherism. When they actually come up with a coherent monetary policy let me know, and I'll comment on it.

Yes, MM would suggest doing whatever it takes, in which case it would take very little.

Thomas. The markets are rational. If the Fed does what it says it will do, it will be believed.

Philip, I'm not sure the quantity of money is falling---which definition are you using?

pmsap writes:

I guess this is one of the most used papers for the Fisher idea:

Comments for this entry have been closed
Return to top