Saturday, January 23, 2010

True Money Supply versus M2: Why Murray Rothbard Was Wrong

Many money supply watchers have pointed out that although M2 money supply has virtually stopped growing, the money supply measure developed by Murray Rothbard, the True Money Supply, has not stopped growing.

Rothbard developed TMS because he was unhappy with the Fed money measures. Specifically, he was unhappy with components of M2 that were long term in nature and were not immediately available for withdrawal. Rothbard's argument was that in order for something to be considered part of the money supply, it had to be immediately withdrawable into cash, that is, a near perfect cash substitute.

Rothbard designed TMS so that it includes: Currency Component of M1, Total Checkable Deposits, Savings Deposits, U.S. Government Demand Deposits and Note Balances, Demand Deposits Due to Foreign Commercial Banks, and Demand Deposits Due to Foreign Official Institutions.

Components of M2 that are not included in TMS are small denomination time deposits and, curiously, money market funds. I have no argument with excluding small denomination time deposits as part of money supply, however, my contention is that money market funds are indeed part of the money supply, i.e. most who own money market funds consider them as immediately withdrawable funds in the same manner that a demand deposit is. In fact, I would venture to guess that many pay their rent and other expenses with money market funds.

Thus, I consider them, unlike Rothbard, part of the money supply. Until recently this has been nothing more than a theoretical, academic dispute having no real world applications since M2 and TMS generally moved in sync, up or down. Now that has changed, TMS continues to climb while M2 has not.

What money measure an analyst follows, thus, now gives one differing perspective on what is occurring in the economy, namely, are we in a tight money period or loose money period. Watching M2, you are going to view the current period as a tight money period, watching TMS you are going to consider it a loose money period.

Here's what is going on:

Money market fund holdings are declining and being shifted into other M2 components that are part of TMS. For those who watch M2, like I do, this is simply a shift in components. It is a wash as far as money supply is concerned. One component up, the other down.

For TMS followers, since they don't include money market funds as part of the money measure, the shift out of money market funds appears as an increase in money supply rather than simply a shift.

I would argue that this shift out of money markets strongly weighs in favor of my point that money markets should be considered part of the money supply.

The shift started because of the financial crisis, as concerns about money market fund safety continued to escalate, individuals sought out a safer substitute. If they viewed this money as savings, as opposed to needed immediately available funds, they would be moving the funds into T-bills and the like. If in fact, as it appears, they are shifting their funds into safer versions of the M2 components, it signifies that these are what they consider the closest substitutes for money market fund holdings, i.e. another cash substitute. If this is indeed what is occurring, then it is clear that money market funds should be considered part of the money supply.

I do, however, concede that an argument could be made that the decline in money market funds and the simultaneous increase in other components is simply coincidental and that money market funds have been drained and put into other instruments and that the other money supply components climbed by the same amount coincidentally. The only way to resolve this argument would be for someone who had access to complete sets of this data and actually examine where money market funds were shifted.

A further argument could be made that money market funds are simply savings and that the financial crisis caused individuals to decrease their savings and add money to cash and cash substitutes.

Thus, while the empirical evidence does appear to suggest that simply a shift among substitutes has occurred. The empirical data does not make it a completely open and shut case. Ultimately, the role of money markets comes down to how individuals subjectively view them, to the extent they use them to pay bills, it becomes clear they are considered cash substitutes just like demand deposits, and thus, as opposed to Rothbards inclination, they should be considered part of the money supply.


For Federal Reserve money supply data, see here.

For True Money Supply Data, see here.


  1. "The shift started because of the financial crisis, as concerns about money market fund safety continued to escalate, individuals sought out a safer substitute."
    I still think Rothbard had a point. Precisely because people leave money funds in droves shows they don't consider them a liquid instrument in times of crisis. And precisely that is what I would think is what I want from an instrument that pays my bills ... You can NOT pay rent with T-bills.

  2. Lots of money market funds allow you to write checks against them, so they basically look like a bank account. I certainly think of my money that is in money market funds as available on demand, so as cash-equivalent, not investment/savings.

  3. Maven, I disagree. I consider any money I have in a Money Market as liquid, but I no longer consider a MM fund as riskless. I am one of the (I think) many who "sweep" my money market funds in to Treasury Direct T-Bills and C of I instruments. They are completely liquid, because I can shift them back into the MM fund when I want to buy stocks or my checking when I pay a bill. I no longer have to worry about FDIC limits or a MM fund breaking the buck.

  4. Isn't the end result simply that at worst the TMS number is underestimating the money supply, just not as badly as the M2?

  5. The Shostak/Salerno/Rothbard argument against including MM funds in the money supply is that when a person writes a check to a MM fund that amount of cash is withdrawn from their checking account, loaned to the MM fund, then loaned (by purchasing, for example, commercial paper) to a corporation, who deposits in their checking account. Therefore the same dollar can't simultaneously be "money" to both the person who invested in the MM fund and the corporation who deposited the dollars in their checking account. Shostak has a couple of daily articles on that explain the origins of TMS and give the above argument.

  6. I have replied to Peter's comment here:

  7. Yes, holders of money market fund balances would be very surprised to learn that they were not "money." They are certainly checkable deposits. They ARE savings, but in the form of money. Just as much money as are checking account demand deposits. Part of the proof is that they have yielded virtually nothing for 3 years or so, but people and institutions still hold them--because they are money and considered part of their money balances. In a future conversion to gold money, they must be included as deserving of exchangeability into gold, just as should checking deposits and physical currency.