Monday, March 12, 2012

Running on MMT

I’m going to regret this, but here is a short reaction to the Modern Monetary Theory (MMT) uprising, occasioned by reading (after some hesitation) Philip Pilkington’s MMT-inspired attack on IS-LM models over at Naked Capitalism.

1. I agree with the fundamental complaint MMTers have about the LM curve: it assumes a fixed money supply, so that changes in the speculative demand for money (due to i) have to be offset by changes in the transaction demand (due to Y): hence the upward slope.  (The slope is flat during a liquidity trap.)  But the money supply is not fixed; it has a substantial endogenous component.  Note the word “substantial”.

2. But MMT jumps from one corner solution to another.  After rejecting the implausible notion that the money supply is fixed, always at a level determined by the money multiplier times the monetary base, it leaps to the equally implausible notion that the monetary base is completely decoupled from the money supply.  On this view, infusions or withdrawals of liquidity by the central bank influence only interest rates, and the banking system alters its credit creation to meet money demand at the policy-determined price.  Thus the volume of economic activity need have no bearing at all on interest rates; the central bank has a completely free hand.  Do I have this right?

My view is that corner solutions are usually wrong; at least they should be regarded with fierce skepticism.  It would take a lot to convince me that monetary aggregates are completely decoupled from central bank liquidity provision, just as I doubt they can be controlled by monetary policy.  Surely there are limits to credit creation, which we see in a raw form during those episodes in which reserve requirements are reset.  What’s wrong with saying that the money supply is jointly created by the central bank and the private sector as the latter responds to perceived lending (and other asset acquisition) opportunities?

(Note: this post is about just one issue in IS-LM modeling.  There are others, but I am saving them for a day when there is really nothing better to do.)


TheIllusionist said...

The way it works is that banks lend all they find profitable at a given interest rate. They do not take account of reserves until after the fact. So, they lend as much money as they find profitable and then seek reserves after the fact.

If too many banks are seeking reserves at once in the interbank lending market this will put upward pressure on the interest rate (as banks bid higher and higher for reserves). Thus, if the central bank wants to maintain its policy interest rate they must accommodate the banks by injecting more reserves into the system.

This (like most of MMT) isn't theory. It's just how things work. Don't believe me? Go down to your bank and take out a loan. Then ask the loan manager if he checked if the bank has adequate reserves to meet the loan. He won't have a clue what you're talking about. A separate desk deals with reserves after the fact.

In fact, Dean Baker recognised all this (because he's a great economist and knows the banking system well). His disagreement is theoretical.

P.S. You don't have to go to MMT to read about this. The Bank of International Settlements have published papers on this. And most people in central banks who work at an operational level know this well. It becomes particularly obvious when massive inflation takes off in developing countries and bank lending continues apace despite interest rate rises.


Barkley Rosser said...

Here's jumping on you!!! Eeeek!!!

Tom Hickey said...

I don't think you have the MMT position on interest rate right. MMT holds that in a non-convertible floating rate world, the central bank is not operationally constained by maintaining an interest rate relative to reserves, e.g., gold holdings, so it can set the target overnight rate and conduct monetary operations to hit this target independently of the amount of reserves involved. Nothing new here. Everyone knows that the Fed sets the FF rate independently rather than relying on market forces in the overnight market for federal funds.

Under the existing monetary regime the Fed can and does set the target rate in the overnight market for federal funds independently. It does this either by paying interest on reserves at or above its target rate, as now, or else the excess reserves need to be drained by issuance of Treasury securities to offset fiscal deficits that create excess reserves through injection of net financial assets into non-government.

Since a currency sovereign that is the monopoly provider of a non-convertible floating rate currency does not need to fund itself through taxation or finance itself through borrowing in that it issues the currency directly, Treasury issuance is a monetary operation that drains excess reserves so the cb can hit its target rate with minimal OMO. Payment of IOR obviates the need to issue Treasury securities to drain excess reserves. See Warren Mosler's "Soft Currency Economics" and "The Natural Rate of Interest is Zero" at

The central bank can also control the yield curve by setting price and standing ready to adjust quantity necessary to support that price. Conversely in QE, the Fed targeted quantity in nominal terms, e.g., 650B, and let price float.

Since lenders use the overnight rate and the yield curve as benchmarks, adjustment of these influences the rates that lender charge. Adjusting these is how central banks would conduct monetary policy in an MMT world, although MMT prefers using fiscal policy based on the sectoral balance approach and functional finance. The reasoning behind this preference for fiscal over monetary is an integral aspect of MMT as a macro theory designed to achieve full employment and price stability.

Tom Hickey said...

"What’s wrong with saying that the money supply is jointly created by the central bank and the private sector as the latter responds to perceived lending (and other asset acquisition) opportunities?"

This is essentially what MMT says. The interest rate (FF rate in US) is one cost that lenders use in pricing credit. Liquidity is never an issue since the cb stands ready as lender of last resort to provide reserves for settlement to solvent member banks. The availability and amount of reserves is not an issue nor a price determinant. It is the overnight rate and the penalty rate that affect the cost of reserves, hence the spread that lenders charge.

Private lenders make loans based on demand from creditworthy customers willing and able to pay the going rate. It is private lenders that do risk assessment and credit pricing. The central bank influences this pricing through adjustment of the interest rate.

STF said...

Thanks for engaging,

A few comments:

"it leaps to the equally implausible notion that the monetary base is completely decoupled from the money supply. "

What we say is the causation is backwards. Banks make loans and search for reserves later, assuming they even need them (in Canada, there are no RR, and in the US, they were essentially voluntary due to sweep accounts even before 2008). Alan Holmes at the NY Fed said this 40 years ago--it shouldn't be controversial but it is for those who haven't progressed beyond undergraduate textbook understanding of banking. Claudio Borio at BIS has been writing the same thing on this as MMT, as have increasing numbers of others.

Banks use reserve balances to settle payments and meet reserve requirements. This is well understood in the literature on the federal funds market. They don't use them to create loans--only to settle a withdrawal that might occur when a loan is made, or meet RR if the loan adds to deposits. But if the banking system is insufficient of reserve balances for these purposes, the Fed supplies them to achieve its interest rate target. That's what it means to set an interest rate target.

I see now that I'm nearly done that TheIllusionist has said much the same as I have here. Sorry for the repetition.

"Corner solutions are wrong" only when the theory that makes the "corner solution" a "corner" is the correct theory in the first place. In the real world, if the theory is inconsistent with real world accounting--as the money multiplier or any other view that reserves drive lending is aside from a gold standard monetary system--then it is the theory that is wrong, not the "corner solution."

Scott Fullwiler

STF said...

I apologize if the "undergraduate econ textbook" comment was offensive. I was not intending it that way and should have proofread before hitting "publish." I wasn't intending to imply that Mr. Dorman has not progressed, only that by and large the profession has not. Again, a good deal of recent mainstream literature makes much the same point. But unless you are doing research specifically on this topic, your background is the textbooks by and large, as it is for anyone in a field/topic that is not their specific area of expertise.

Nathan Tankus said...
This comment has been removed by the author.
Nathan Tankus said...

The seminal book providing evidence that the money and banking textbooks have causality reversed is Basil Moore's "horizontalists and verticalists". I hear people are organizing a 25 year anniversery conference next year.

STF said...

"it leaps to the equally implausible notion that the monetary base is completely decoupled from the money supply. "

Since you suggested it's implausible, let's think a bit about what the monetary base is.

first, there is currency, which in normal times is about 95% of the monetary base if vault cash is included, though not that much less if not. Any quick reading of, say, the NY Fed's annual report on open market operations will make it clear that currency (and vault cash) are created/provided by the Fed entirely in response to desired currency holdings of the pvt sector. The Fed doesn't provide these exogenously, and actually can't since the pvt sector always has the option of converting to a deposit, savings account, cd, and so forth, and the bank would simply send the currency back to the Fed in exchange for reserve balances.

second, there is reserve balances. As I noted above in my first comment, banks use these solely for the purpose of settling payments or meeting reserve requirements. Reserve balances do not enable more/less lending. If they did, Japan would have been out of its balance sheet recession 10 years ago.

Furthermore, the quantity of reserve balances is not a discretionary variable for the Fed unless it pays interest on reserve balances at the target rate or runs a ZIRP strategy. Supplying more than banks want to settle payments or meet RR just reduces the target rate until the Fed drains them, and vice versa.

But even with a ZIRP or IOR at the target rate, the additional reserve balances don't enable more lending, as we see right now in the US and, again, previously in Japan.

Any critique of the suggestion that the monetary base does not enable lending should explain how the above description of the parts of the monetary base is incorrect.

Now, it could be that the point that the MB and monetary aggs aren't "decoupled" (and, again, we just say the causation is backwards, not necessarily decoupled, though it could be) isn't necessarily based on lending. For instance, in QE the FEd raises both the qty of reserve balances and deposits simultaneously when it buys securities from non-banks, so obviously there is a direct relationship in that case. That is, I just want to stress that the MMT point (which is the same as the Horizontalist point of Moore, as Nathan pointed out) is about the monetary base not causing lending; there are instances in which monetary aggregates and MB can be directly related and causation can go from MB to the aggregates.

Scott Fullwiler

NeilW said...

There are others, but I am saving them for a day when there is really nothing better to do.)

How about one explaining why IS-LM is still even spoken about when Hicks rejected it decades ago.

"I accordingly conclude that the only way in which IS-M analysis usefully survives - as anything more than a classroom gadget, to be superseded, later on, by something better - is in application to a particular kind of causal analysis, where the use of equilibrium methods, even a drastic use of equilibrium methods, is not inappropriate"

Hicks (1980) 'IS-LM: an explanation' Journal of Post Keynesian Economics, 3(2): 139-54

kevin quinn said...

Oh hell, Peter: Now you've stepped in it! Bonne chance.

kevin quinn said...

Tangentially, Peter, I don't share your scorn for corner solutions. Borrowing constraints, for one example, seem to be crucial to understanding the empirical failure of Ricardian equivalence.

Peter Dorman said...

Kevin, I'll reply to you separately, since I'm posting a longer response to the MMT folks.

One way I could defend myself is to point out that I said that corner solutions are *usually* wrong: I had to hedge in order to avoid a corner solution on corner solutions.

More specifically, though, I think the case of credit constraints can be used to support my position. (1) Such constraints are sometimes binding, not generally always or never. (2) If you introduce time and expectations on the stream of future outcomes, decisions are made today that incorporate the risk of credit constraints, normally not either their certainty or impossibility.

Also, my beef with Ricardian equivalence is more basic than yours, as you know from my previous posts. I think the credit constraint argument plays a greater role in the critique of Modigliani-Miller.