Saturday, April 05, 2008

Myths About The Monetary Base And Bank Reserves

One of the hottest (if not the hottest) intra-Austrian debates today is between what is sometimes referred to as deflationists and inflationists/stagflationists. This is not a policy debate of course, as all Austrians is anti-inflation, but rather a debate about whether the current recession will be associated with deflation or inflation. Examples of deflationists are Frank Shostak, Mike Shedlock and Gary North while examples of stagflationists include me, Antony Mueller and Peter Schiff. The dispute is largely originated in a dispute over the definition of the money supply. I have already dealt with that issue extensively (see for example here, here and here ) so I will not repeat this here. Instead, I will focus on the appeal made by the deflationists to the development of the monetary base, which have been largely stagnant for the latest year.

The implicit or explicit argument from the deflationists appear to be that 1)The Fed controls the monetary base, so it is a good reflection of how tight its policy is 2) The monetary base determines the money supply, so the stagnant monetary base implies a stagnant money supply. Yet both of these assertions are simply wrong, at least given the current financial structure.

I have actually once answered the monetary base argument before. At that time I pointed out that more than 90% of the monetary base is made up by what in monetary statistics is called currency in circulation, which is to say paper notes and metal coins held by the public. And I also pointed out that the amount of currency in circulation is determined by public preference for making payments using notes and coins versus making electronic transactions. In the U.S. this is also determined by demand in high inflation third world countries for using dollars as means of payments instead of local currencies. Most likely the stagnant amount of currency in circulation reflects the trend towards a cashless society as well as growing repatriation of previously exported dollar notes and coins due to the distrust of the dollar that the decline in its purchasing power has caused.

I also illustrated that point by pointing out that during the inflationary boom of the 1920s, the monetary base was stagnant. By contrast, the monetary base soared during the deflationary depression of the 1930s, as bank collapses caused people to prefer to hold money in the form of cash instead of deposit money.

However, I now realize that this response was unsatisfactory in one aspect. Namely, because my focus on the currency in circulation part of the monetary base seemed to imply that the other part of the monetary base, bank reserves, do in fact have the characteristics that the deflationists claim. That's not what I meant, although now I realize that the post was written in a way which could reasonably be interpreted that way. And as I see Robert Murphy write a whole article focusing on bank reserves as a proxy for monetary conditions it is clear that the issue must be addressed. So I will now clarify: bank reserves are in today's system basically irrelevant too, both as a proxy of Fed policy and of monetary conditions.

The reason is that there really isn’t any demand for bank reserves. To the contrary, banks do everything they can to minimize it because reserves inflict opportunity costs for them in the form of foregone interest income. In the past, banks still felt compelled to keep large reserves because of the risk of bank runs. But with the Fed providing unlimited quantities of liquidity in the case of unexpected increases in withdrawals, this is not an issue anymore. Today, the only thing preventing banks from reducing reserves to zero is formal reserve requirements and the need to have cash available for withdrawals from bank offices and ATMs. But with the banks moving away from deposits with reserve requirements such as demand deposits and instead finance its operations in for example Money Market Mutual Fund accounts (And using so called sweep operations the banks ensure that the level of demand deposits are always minimized even if customers deposit money there) that don't have any reserve requirements the level of required reserves is declining in importance. And with the increasing use of electronic transactions, cash for customer withdrawals is also becoming less important and is at an absolute level very low. Because of this, bank reserves are increasingly disconnected from the level of money supply.

Indeed, if Robert Murphy had looked more closely at figure 1, he would have seen this point. Bank reserves in early 1990 were $60 billion as compared to $42 billion now. If bank reserves really had been a good proxy for the money supply, then that would have implied a cumulative monetary deflation of 30% during the latest 18 years. The Fed under Greenspan would, if bank reserves were really a good proxy of monetary conditions, have been the ultimate hard money institution, providing more deflationary conditions than a gold standard. Nor do the trends show any correlations with the housing bubble, as it started already in 2001 while the monetary base was flat until 2003. And after a brief upswing in 2003-04 it was basically flat after that. In other words, bank reserves have in today's system nearly no correlation with monetary conditions.

But if the Fed performs open market operations, won't that expand bank reserves? Well, no. While it may result in brief spikes, these spikes won't last as the banks will lend out or invest the money the open market operations produce, either as bank loans or investments in securities. The reason why they are unlikely to keep the money more than a few days is the above mentioned fact that reserves represent opportunity costs and that it is more profitable for the banks to lend/invest. Contrary to what Murphy claimed, it is not the Fed that has moved away reserves from the systems, it is the banks themselves. If you doubt that, just check out the statistics for bank credit, which have soared in recent months.

Because the banks have the opportunity to lend/invest the money they get from the Fed and the incentive to do so, the Fed has almost no control over bank reserves. If they started to impose reserve requirements on all deposits, they could have controlled it, but as it is they don't. Nor is it a reflection of credit conditions or monetary conditions as bank credit grows at double digit rates.

To summarize, the stagnant monetary base and bank reserves have absolutely nothing to do with interest rate policy, and is instead a reflection of the trend in the payment system to move away from currency in circulation and deposits with formal reserve requirements to deposits without reserve requirements, combined with the Fed's promise to help all banks with unlimited quantities of liquidity if they need it. The deflationist claim that the Fed is not inflating is not only serious because it implies misleading investment advice, such as staying away from gold and buying treasuries. It is also damaging because it implies that Bernanke is actually mimicking market conditions (something which Murphy actually explicitly wrote in his article), thus effectively destroying all opposition to Bernanke's inflationary policies. Unwittingly, the deflationists are thus helping Bernanke.


Blogger Ron McK said...

Thanks for this.
I find it ironic that the only economists thinking seriously about the monetary implications of the Fed's policies are Austrians, who would prefer there was no Fed. You would think the Fed believers would be doing the heavy lifting here, but they are relatively silent.

I have a question for you. How much of the growth in total money supply over the period of the housing boom was the result of the Fed's low interst rate policy and how much was the result of the banks and quasi banks spreading increased leverage throughout the banking system. Or were these two just the different sides of the same coin.

Could one have happened without the other. For example, could we have had the increase in leverage without the low interest rates. ie can the banks create a boom just by increasing leverage(without the help of the fed). Would the low interest rates have been less harmful, if the massive inrease in leveraging had not occurred.

Is what we are seeing now a collapse in leverage, despite the lower interst rates

11:19 PM  
Anonymous Anonymous said...

It appears that the problem is that the role of credit money, which is in M2, M3, MZM but not in M1 and M0, has not been analyzed in the Austrian literature. There is no paper on QJAE or RAE talking about securitization, money markets, derivatives from a theoretical standpoint. The most recent treatise on Austria monetary theory, de Soto's, is still only about currency and deposits. After 40 years of fiat money regime I still have to find a real reason for this neglect. May be the result of this is that the theoretical instruments to understand money beyond M1 are missing.


12:50 PM  
Anonymous Anonymous said...

maybe in one of your future articles you could mention what made central banks shift from the friedman monetarist idea of growing money supply steadily (i think volcker still targeted the m's), to the modern-day price-index-led central bank. why did the swiss give up their m targeting? did they also have problems figuring out which m counts?

2:35 PM  
Anonymous Anonymous said...

oh, one other thing. i can't remember any austrian economist ever explaining why it is that since 1946, the fed has always moved the ffr and discount rate after the t-bill market. that is, the market always leads and the fed follows. surely that would obviate the fed's need to enter into repos/reverse repos? a real conundrum.
for more, see

5:49 PM  
Blogger stefankarlsson said...

Newson, the main reason central banks gave up M targeting is that no M -neither M1, M2, M3, MZM, Shostak's M or any other M that I know about- have a good track record of preceding changes in consumer price inflation with any fixed time lag, combined with the increasing popularity of the false idea that inflation targeting is the way to achieve economic stability.

As for the T-bill market, it is probably true that they have generally forecasted fed funds rate changes. But this only reflects that the markets forecast the FOMC movements correct together with the fact that the FOMC is reluctant to disappoint the markets, making their forecasts to some extent self-fulfilling.

I should however note that the T-bill market have moved in a rather erratic way in recent weeks falling from 2% to 0.25% and then move up again to 1.55%, illustrating that they do not always precede FOMC movements.

7:53 PM  
Anonymous Anonymous said...

as you say, if the fed's "open-mouth" policy is so effective, then why the need for repos/reverse repos? i mean in a falling yield climate, the bill yield is below the ffr and the discount rate, and in a rising rate climate the bill yield is above the ffr and discount rate.
i thought that the fed needed to intervene via open-market activities to keep ffr near its target rate? but why, if the bill market gets the primary direction right?

4:03 AM  
Anonymous Anonymous said...

What about the so-called "shadow-banking" system, where credit was created by banks off balance sheet (via SIVs, etc), and that was thought to be "money-good" but no longer is. Is that not deflationary?

9:18 PM  
Anonymous Anonymous said...

Stefan Karlsson says, "Because the banks have the opportunity to lend/invest the money they get from the Fed and the incentive to do so, the Fed has almost no control over bank reserves."

But what happens after the banks lend out the money they get from the Fed? If the bank lends the money to Abe, and Abe buys a house from Bill, doesn't Bill go and put the money back in the bank? Or if he buys stocks from Carl, doesn't Carl put the money in a bank?

How can the banks rid themselves of even one dollar of reserves, other than by issuing cash to customers, and refusing to take it back?

Unless I'm missing something, and please tell me if I am, it seems like the Fed has almost complete control over bank reserves, at least in the aggregate.

6:33 PM  
Blogger Brian Macker said...


You've hit on the problem. The current system is set up to produce the kind of crazy mathematics one expects around a "black hole". With no reserve requirements you get the problem of infinities.

A reserve requirement of 1/2 leads to an increase in the money supply of 2/1. However a reserve requirment of r/n as r tends toward zero leads to a money supply of n/r as r tends towards zero. Well n/0 is infinity.

One of the reasons banks were willing to buy all these trash assets is that there was so much cash floating around the system that they had run out of investment vechicles. There was no real savings going on backed by the money and therefore no true capital to invest and get returns on.

We are currently experiencing the equivalent of a bank run in this environment. The mechanisms discussed in this article, "the Fed providing unlimited quantities of liquidity in the case of unexpected increases in withdrawals", will force the Fed to convert a large part of the monetary inflation caused by the effective lowering of reserve requirements into actual currency. That is highly inflationary.

As far as I can see the current monetary system is completely unanchored, and mostly out of control of the Fed. There are no metrics the Fed can watch to even know what is going on.

They were watching and using the wrong metric since Clinton, and had set interest rates way below market without realizing it. We will now see a wild correction. Fractional reserve deflation parallel with base money inflation. That will be followed by Fractional reserve inflation on top of the new money.

5:51 PM  

Post a Comment

<< Home