Wednesday, February 6, 2013

The Myth Of The Money Multiplier

Something unusal happened yesterday in my weekly department macroeconomics seminar at James Madison University.  Someone had us discuss a paper that changed the minds of pretty much everybody in the room who did not already agree with the paper, which includes a highly diverse group ranging from Austrian, Old Monetarist, New Classical, New Keynesian, Old Keynesian, Post Keynesian, and some generally eclectic pragmatists.  The paper is from the Fed Bd of Govs in 2010, by Seth Carpenter and Selva Demiralp, still unpublished as near as I can tell, and is entitled, "Money, Reserves, and the Transmission of Monetary Policy: Does the Money Multiplier Exist?"  Their answer is not only "no," which I at least have thought was true since 2007 or so, but that the answer has been  "no" probably since the mid-1980s and certainly since the mid-1990s.  It is available at .

The story can be seen pretty much in figures and charts in the back.  Prior to 1984, there was a clear correlation between reserves, loans, and M2.  After then, while loans and M2 continued to go along in a pretty close lockstep, reserves simply have flopped around all over the place.  This seems to have coincided with various things: a definite move to targeting interest rates and apparently following a Taylor rule (although not more recently on the latter), the change in Regulation Q, globalization, a broadening of sources of funds for banks, and particularly more interbank lending and lessening of reliance on reserves and the Fed's reserve base.  Apparently up to the mid-90s, small umdercapitalized banks still were tied to reserves in their lending, but after then that broke down too, with their appearing to be a shift to relying more on large deposits, mostly from other banks and also the shadow banking system. 

That Fed control over the money supply has become a phantom has been quite clear since the Minsky moment in 2008, with the Fed massively expanding its balance sheet without much resulting increase in measured money supply.  This of course has made a hash of all the people ranting about the Fed "printing money," which presumably will lead to hyperinflation any minute (eeek!).  But the deeper story that some of us were unaware of is that apparently this disjuncture happened a long time ago.  Even so, one of our number pointed out that official Fed literature and even many Fed employees still sell the reserve base story tied to a money multiplier to the public, just as one continues to find it in the textbooks,  But apparently most of them know better, and the money multiplier became a myth a long time ago.

Barkley Rosser


Deus-DJ said...

Well if you guys ever have the balls to read something your colleagues or students aren't reading, like the stuff Basil Moore wrote back in the 80s, you would have known this at least since then. I'm sure the Post-Keynesians were rolling their eyes, because endogenous money is a staple of PK economics. Funny how you have to rely on some fed paper to get a clue.

Arijit Banik said...

Deus-DJ is correct so I won't repeat what he has stated.

From a conventional perspective, Piti Disyatat wrote about the money multiplier (and other aspects of monetary policy) in a BIS Working Paper, "Monetary policy implementation: Misconceptions and their consequences" as well:

The 'reserve' issue continues to be a red herring in Canada. Over there, the bank lending channel comes down to the risk appetite of the Tier 1 banks who are constrained by capital and leverage as mandated by OSFI. Their regulatory framework is on top of whatever Basel recommends. No chartered bank in Canada is reserve constrained; the majority of money creation is endogenous so if there is great asymmetry between demand for credit and risk appetite (of the banks) there will not be an increase in money.
The money multiplier is a wonderful fairy tale but textbooks should reflect reality rather than be sources of liturgy.

Jesse said...

When economists 'know' something that the rest of us don't, it is probably a good move to grab your wallet and head for the door. said...

Actually, Deus-DJ, you do not need to get on such a high horse. I am well acquainted with Basil's work and even contributed to the festschrift published some years ago honoring him. What we have here is something a bit different. Even regular neoclassicals know that money is endogenous when central banks targeting interest rates. But endogenous money alone does not necessarily imply a breakdown of the whole sequence of deposits-loans-deposits that this paper shows empirically.

However, as I noted, there were PKs in the room, who were probably the least discombobulated by this paper than some of the others, with probably the most upset and resisting being an eclectic pragmatist who used to work in a bank.

John said...

I'm just an ignorant layman but since I first saws the word "velocity" in the Eighties I realized it was damn hard to measure economic activity. A supply of any volume can move swiftly or sluggishly.

Moreover, all credit inflates the supply as well. Any time I use my credit card I am creating money in the present with the expectation that I will replace it sometime in the future. Meantime, the transaction no longer exists in an old-fashioned "accounts receivable" journal. No. It is immediately logged as a "sale" and the amount becomes part of a bank deposit.

Making the picture more complicated, in the same way that mortgages were (are?) bundled and resold as securities, there is a big industry built on contracting and servicing debts (everything from vacuum cleaners to swimming pools) sold on credit then resold on the aftermarket. So we have credit multiplying itself as well.

Where does anyone get off imagining there can be any reliable metric for this complexity. It's easier to predict volcanic activity in some distant country or weather in my yard six weeks out. said...


The stability of velocity broke down also in the mid-1980s, but it was tied to a different cause, the easing of the ability to move deposits around to different kinds of accounts. Most economists recognized this almost immediately, whereas this matter of the destabilization of the money multiplier has been much slower to sink in, although quite a few central bankers have been aware of it for a long time, such as Charles Goodhart, although most have simply not felt like talking about it publicly. said...


I have received a message from Douglas McMillan, the editor of the Journal of Macroeconomics, informing me that the Carpenter-Demiralp paper has been published in his journal nearly a year ago, March 2012, vol. 34, no. 1, pp. 50-75.

John said...

I didn't know people still believed this. Anyway this is the reason market monetarists use NGDP instead of monetary aggregates.

Shining Raven said...

There is also a paper by Ulrich Bindseil from 2004 that is very closely related to this topic. He traces the history of the idea of controlling reserve positions of banks and compares European, British and American practice.

If you are interested in the question of the "money multiplier", I highly recommend reading it. See below for a quote from the executive summary. Bindseil is Assistant Director General for Financial Market Operations at the ECB, if you don't know him.

Ulrich Bindseil, ECB Working paper series No. 372 (June 2004)


Central quotes from the abstract:

In between, namely between around 1920 and the end of the 1980s, “reserve position doctrine” (RPD) dominated at least in the US, according to which a central bank should, via open market operation, steer some reserve concept, which would impact via the money multiplier on monetary aggregates and ultimate goals.


It appears that with RPD, academic economists developed theories detached from reality, without resenting or even admitting this detachment. Economic variables of very different nature were mixed up and precision in the use of the different concepts (e.g. operational versus intermediate targets, short-term vs. long-term interest rates, reserve market quantities vs. monetary aggregates, reserve market shocks vs. shocks in the money demand, etc.) was often too low to allow obtaining applicable results. The dynamics of academic research and the underlying incentive mechanisms seem to have failed to ensure pressure on academics to ensure that models of central bank operations were sufficiently in line with the reality of these operations.

Joshua Wojnilower said...

[...] Rosser refers to a paper from the Federal Reserve Board’s Finance and Economics Discussion Series that was the source of a recent Quote of the Week [...]

For those interested in the subject also consider reading
Modern Central Bank Operations—The General Principles by Scott Fullwiler (

Principle #3 from that paper: "The money-multiplier view in which the central bank engages in direct targeting of reserve balances or the monetary base is untenable in practice. The only possible direct target is an interest rate target."