Interpreting the Monetary Base Under the New Monetary Regime

by on December 1, 2008 at 7:41 am in Economics | Permalink

The monetary regime has changed and, as a result, many people are misinterpreting the recent increase in the monetary base.  Paul Krugman, for example, posts the picture Benbase_2at right.  His interpretation is that the tremendous increase in the base shows that the Fed is trying to expand the money supply like crazy but nothing is happening, i.e. a massive liquidity trap.  (Krugman is not alone in this interpretation, see e.g. this post by Bob Higgs).  Thus, Krugman concludes, Friedman was wrong both about monetary history and monetary theory. 

Krugman’s interpretation, however, neglects the fact that the monetary regime changed when the Fed began to pay interest on reserves.  Previously, holding reserves was costly to banks so they held as few as possible.  Since Oct 9, 2008, however, the Fed has paid interest on reserves so there is no longer an opportunity cost to holding reserves.  The jump in reserves occurred primarily at this time and is entirely under the Fed’s control.  The jump in reserves does not represent a massive attempt to increase the broader money supply.

Here’s a bit more background.  When no interest was paid on reserves banks tried to hold as few as possible.  But during the day the banks needed reserves – of which there were only $40 billion or so – to fund trillions of dollars worth of intraday payments.  As
a result, there was typically a daily shortage of reserves which the Fed made up for by extending hundreds of billions of dollars worth of daylight credit.  Thus, in essence, the banks used to inhale credit during the day – puffing up like a bullfrog – only to exhale at night.  (But note that our stats on the monetary base only measured the bullfrog at night.) 

Today, the banks are no longer in bullfrog mode.  The Fed is paying interest on reserves and they are paying at a rate which is high enough so that the banks have plenty of reserves on hand during the day and they keep those reserves at night.  Thus, all that has really happened – as far as the monetary base statistic is concerned – is that we have replaced daylight credit with excess reserves held around the clock.  The change does not represent a massive injection of liquidity and the increase in reserves should not be interpreted as evidence of a liquidity trap.

Addendum: (For the truly wonkish.)  If you want more, see my earlier post on excess reserves, posts by Jim Hamilton, and David Altig, and especially two very useful Fed articles, Keister, Martin, and McAndrews (n.b. the last section) and Ennis and Weinberg.

1 lzl December 1, 2008 at 8:01 am

On Krugman’s note: I thought that money supply was important, why draw conclusions on monetary base?

2 Yancey Ward December 1, 2008 at 9:58 am

Alex,

You might want to write a second post explaining why the Fed began to pay interest on reserves.

3 Bob Murphy December 1, 2008 at 10:25 am

Alex,

This is an intriguing post. Like Bob Higgs, I too have been very concerned about the growth in the monetary base. Specifically, I am concerned that when the panic subsides and the financial sector returns to “normal” (whatever that means when the government owns half the banks), that Bernanke won’t want to be a party pooper and drain away the reserves he pumped in during the crisis. Hence, massive price inflation (by US standards).

I hate to say it, but I’m wondering if you are so eager to deny Krugman’s point about the liquidity trap, that you’re downplaying the truly massive injections of the Fed. Let me quote you and explain where I think the flaw in your reasoning is:

Today, the banks are no longer in bullfrog mode. The Fed is paying interest on reserves and they are paying at a rate which is high enough so that the banks have plenty of reserves on hand during the day and they keep those reserves at night. Thus, all that has really happened – as far as the monetary base statistic is concerned – is that we have replaced daylight credit with excess reserves held around the clock.

It’s possible that the spike we see in the graph gets the date wrong, but there is no doubt that there has been a massive increase in reserves (and base). At best, I think your bullfrog analogy just shows us that the real Fed interventions started in 2007, with all of the massive emergency loan operations. Now, with the introduction of paying interest on reserves, the Fed can drop the smoke and mirrors and we can see how much extra base it has pumped in.

In other words, I am saying that the spike in the graph is finally showing the true situation.

If you are reading these comments, I would love to hear your reaction. Like I said in the beginning, I have been pretty dire in my warnings about the huge surge in the base, so if I am making a big mistake I’d like to know about it! But your post here hasn’t persuaded me.

4 Robert Wenzel December 1, 2008 at 11:11 am

Alex,

Helluva theory, but it doesn’t fit the facts.

The monetary base started to climb in August and soared in September. The Fed did not start paying interest on reserves until October, which kinda blows your theory to smithereens.

There were a number of factors that boosted the monetary base, includng an overall fligh to quality, but for the most part it was the Treasury making deposits at the Fed. Here’s the Treasury part of the explanation:

http://tinyurl.com/5r672o

5 Bob Murphy December 1, 2008 at 11:46 am

Alex wrote:

As you can see, the explosion has occurred primarily after the Fed started paying interest on reserves in October.

Well, I agree that the explosion in October was bigger, but there was still an explosion in September. This chart shows about an 8% jump in the bi-weekly figure occurring in September. In this chart, you can see that this was the largest such jump in the history of the bi-weekly series, going back to 1984 (with the possible exception of the 9/11 spike).

And in any event, whether it started before or after the Fed started paying interest, how is this not an increase in the monetary base? If I understood your post, you were saying that this is nothing new, because the Fed was previously using overnight loans and now the banks are just holding on to the reserves more permanently.

So at best, aren’t you saying that the massive injections of liquidity should be traced to the earlier loan programs (which were “unprecedented” at least back to the 1930s)? I don’t see how your post here proves that there is no injection of liquidity going on.

6 Joen December 1, 2008 at 11:52 am

There are two things that matter.

* M+B: The total nominal liabilities of the government.
* M & B: The composition of government liabilities.

Why M+B? Think of the neutrality of money (double M and you just double P). If you double the money supply but keep the stock of nominal debt constant the allocation is likely to be different.

Why M & B? Think of the Modigliani-Miller/Ricardian equivalence theorems. If they holds then whether you have M or B then it is the same and only M+B is going to matter. If the theorems don’t hold then M & B matter. Lets assume they don’t.

The way the Fed usually operates is by buying/selling B and printing/burning M, so M+B is not affected by open market operations. So any effect on the economy has to come from open market operations has to come from the composition of M+B. The interest rate determines how much you want to hold of M and how much you want to hold of B. But what happens if the interest rate is zero? Well then kind of that you don’t care if the extra resources you have are invested in bonds or just sitting around in cash. That is a liquidity trap. How do you get away from it? You have to increase M+B. How? Have the Fed buy other assets with newly printed bills, have the Fed print money and do a helicopter drop (an authentic one), have the treasury issue debt and buy assets from the private sector, have the treasury issue debt and finance and expansionary fiscal policy (although it will be the increase in M+B the one who got you out of the liquidity trap and not the FP). Now lets remember that there is no such thing as a free lunch. Someone will have to pay for this. Who will it be? Most likely foreigners who have tons of US bonds and cash as reserves.

But the Fed has been given the ability to purchase other assets (not necessarily government bods)

7 Mike Sproul December 1, 2008 at 1:42 pm

Alex:

If the fed issues an extra $300B, it also gets another $300B in bonds as assets. There are more dollars, but there’s also more backing for those dollars, so no inflation. The Fed can always use those bonds to buy back the dollars.

8 bjk December 1, 2008 at 3:23 pm

Does it matter that this is only the St. Louis Fed?

9 Bob Murphy December 1, 2008 at 4:36 pm

Mickslam wrote:

The only way the overall money supply can increase is by govt. deficit spending. Every dollar created by a bank is backed by a liability somewhere and the net money must be zero! This is the accounting that E. Fama was talking about. For there to be more liquidity in the system, the U.S. govt needs to spend money without asking for it back in taxes and preferably without borrowing it.

I think this is backwards. Deficit spending by itself doesn’t create money; for every deficit dollar spent, the government borrows it out of the private sector. In this respect, government deficit spending is no more inflationary than GM deficit spending. (Also, I don’t know what you mean about a government spending money without taxing or borrowing it, unless you are talking about printing new dollars.)

On the other hand, under a fractional reserve banking system new reserves really can increase the total quantity of money. E.g. if the Fed buys bonds and increases reserves by $10 billion, then banks can ultimately loan out an additional $100 billion. Yes, the banks have offsetting liabilities, but that doesn’t net out the existence of the new money. The public now has an extra $100 billion in its checking accounts to spend on goods and services, while the banks are not forced to restrict their own spending because of the increased liabilities on their books.

So if the private sector could buy government debt without decreasing its own store of money, then government deficit spending would be inflationary. But no, in our system only banks (in the FRB setup) can lend money to people that they don’t really have.

10 indiana jim December 1, 2008 at 7:56 pm

It is LIKE a liquidity trap, but it is not one in the sense of people holding money for the pleasure of holding it. That kind of liquidity trap is silly.

What seems to be happening is that in the face of the hightened uncertainty created by Fanny, Fredie, various bailouts and request for like, and the OBAMA factor (HUGE; his 60 minutes appearance suggests that he thinks FDR was sage for screwing around with the economy willy nilly), the Fed’s injections that firm up US bank balance sheets are not being quickly lent out in the US economy. No kidding? Who wants to invest when the rules are in flux? Only an idiot, of course, or only for a premium return of extraordinary magnitude (the kind that usually can’t be and shouldn’t be believed; proverbial swamp land in Florida “deals”).

Lots of money is waiting to see what the rules are and are going to be over what periods of time. This is not the pleasure of holding cash, which is what the liquidity trap of economics texts, often asserted. Rather it is fear of the shifting sands of the magnitude of the coming socialism that OBAMA-MANIA has ushered in. Again, you can’t blame me, and I told you so.

11 assman December 1, 2008 at 9:01 pm

“Well, I agree that the explosion in October was bigger, but there was still an explosion in September”

I agree. The first big increase in reserves is in September not October.

12 Bill Woolsey December 2, 2008 at 8:43 am

There is no “liquidity trap” at this time. When the Federal Reserve targets a Federal Funds rate, and the effective Federal Funds rate stays stubbornly above target, then we have a liquidity trap. Currently, the Fed has the opposite problem. The effective Federal Funds rate is below target. (It must be that the Fed is lying about its target, I think it is loosely targeting LIBOR or something.)

However, the lower bound of zero (or slightly less than zero, really) on all nominal interest rates is close for some nominal interest rates at this time. While I don’t think there has been a lot of historical episodes of central banks trying to target a negative nominal interest rate and failing, there are sound theoretical reasons why any such effort would be futile.

So, if the Federal Reserve tries to lower interest rates too much futher, then it would hit something very much like Keynes’ liquidity trap. Monetary policy would be unable to make some interest rates fall any lower.

So, rather than there being a liquidity trap, the possibility of something like a liquidity trap is on the horizon.

The notion that an increase in the demand for bank reserves is a liquidity trap or evidence of a liquidty trap is absurd.

It is entirely plausible that lower interest rates would result in an increase in the demand by banks to hold reserves, an incease in reserve ratios, and lower money muitliplier. A negative relationship between curerncy or reserves, and the interest rate on other assets, isn’t a liquidity trap.

It is the job of a central bank to accomodate such changes in reserve demand. That is what it means to be lender of last resort to the banking system.

The demand for bank reserves have risen. The reserve ratio has risen. The money multiplier has fallen. The Federal Reserve is acting as lender of last resort to the banking system by accomodating this increase in reserve demand. There is _no_ reason to believe that it will not respond to a decrease in reserve demand in textbook fashion.

What is peculiar is that the Federal Reserve, for the last year, has been funneling credit to Wall Street. The market is moving it the other direction, and the Federal Reserve is trying to offset that. So, the Federal Reserve has sold off government bonds and made loans.

When it started paying interest on reserves, it said that it was doing this in order to entice banks to keep funds in reserve accounts so that the Fed could lend the money. That is, to Wall Street. (Keep in mind that loans made by all sorts of banks were (and are) bundled, securitized, and then held in the form of asset-backed commercial paper. The Fed is keeping the credit flowing through Wall Street–it doesn’t fall into a black hole there.)

They also said that they were doing this to help it target the Federal Funds rate. That means, to keep some banks from lending funds to some other banks at excessively low interest rates.

While the story being told is that “the problem” is that banks won’t lend to other banks, that has been only part of the story. Large money center banks are having trouble borrowing. They aren’t lending to one another. They can’t borrow from anyone else. Well, they are borrowing less and at higher interest rates.

But there has been plenty of interbank lending at very low interest rates. But this is among the smaller banks.

The Fed wants money going to the large money center banks. So, it is paying interest on reserves so that the smaller banks won’t lend to one other, but rather will hold reserves. And the Fed will lend the money to Wall Street.

Again, talk of a liquidity trap is very loose. The element of truth is that if the Federal Reserve wants to see interest rates lower, they can only go a little bit further if they are using open market operations to target the federal funds rate. There is a good reason to believe that that instrument and target can do little more to lower any interest rates.

The rest of the story is an assumption that if monetary policy is to offset dreases in consumption and investment expenditure, interest rates will need to come down more than the remaining decrease possible using open market operations and a federal funds target.

13 John Elder December 2, 2008 at 9:49 am

It is a bit easier to see where the change occured by looking at the data. Below is the percent change in the monetary base (biweekly) from
http://research.stlouisfed.org/fred2/series/BASE

The big % jump started between Sept 10 and Sept 24.

7/2/08 0.0%
7/16/08 1.0%
7/30/08 0.3%
8/13/08 -0.1%
8/27/08 0.2%
9/10/08 -0.2%
9/24/08 8.7%
10/8/08 7.0%
10/22/08 16.3%
11/5/08 7.0%
11/19/08 19.0%

Aside from the recent jump, the next two largest precentage jumps in the history of the series (since 1984) are associated with 9-11 and y2k:

09/19/01 5.2%
12/29/99 1.8%

14 chris December 7, 2008 at 11:38 am

I went hunting the other day… it looked like a duck and quacked like a duck. Then I shot it, walked over to it and picked it up, it was a duck. In any relatively narrow or wide definition it was a duck and it will be a duck in six months too. We are in a liquidity trap for all intensive purposes, I beg anyone to disagree in simple terms. Any takers? I just want an ocaasional “i guess I agree”.

15 Jigar Khatri March 13, 2009 at 8:26 pm

Apparently Krugman is still making the same claims…..

http://krugman.blogs.nytimes.com/2009/03/02/friedman-and-schwartz-were-wrong/

16 Vincent Poncet May 9, 2009 at 4:49 am

The assumption of the article is monetary base is growing because banks prefers to hold money on their account at the fed instead of using it for loans to other banks, firms, individuals or buying assets.

But I do not understand that statement.

When a bank make a loan to another bank, the amount of the loan go from her account at the fed to the account of the other bank at the fed. So, an inter-bank loan does not affect the monetary base aggregate.
When a bank make a loan to a firm or an individual, the amount of the loan go from the account of the bank at the fed to the account of the bank of the firm/individual at the fed. So, a loan does not affect the monetary base aggregate.
When a bank is buying an asset, the amount of the buying go from the account of the bank at the fed to the account of the bank of the counterparty at the fed (could be directly a bank or a non-bank).
So, I do not see case where an activity of a bank could have an impact on the monetary base aggregate.

As I understand, monetary base is only managed by fed activities.

When the fed is doing a loan to a bank, it create the amount of the loan as monetary base and put it on the account of the bank at the fed. The operation from fed balancesheet perspective is : the collateral of the loan as asset of the fed, and the new base money as liability of the fed (banks accounts at the fed are liabilities of the fed).
When the fed is buying an asset, it create the amount of the buying as monetary base.

So, I do not understand all the story about “banks are holding reserves”.

If I’m wrong somewhere, please give me your lights.

17 amanda May 14, 2009 at 12:49 am

It seems we will be in economy ression for quite a long time. We can see the economy crisis effect everywhere.

Comments on this entry are closed.

Previous post:

Next post: