Recent Posts

Sunday, January 12, 2020

MMT Critics And Banks

As seen in some comments on my recent articles, critics of Modern Monetary Theory (MMT) often complain about MMT's treatment of banks. I am largely mystified by these criticisms, as they obviously miss the point. Very simply, the existence of banking system, and the fact that bank deposits to be considered to be part of monetary aggregates, is well understood within the MMT literature. The only real debate is about the implications of private banking. To what extent the banking system matters, it is in reference to the business cycle. You would need to read MMT journal articles to judge how well those authors describe the business cycle. One generally notes a lack of reference to said journal articles in criticism. As such, until there are references to said literature, the criticisms should not be taken too seriously.

Adding Banks Does Not Change Much

Banks show up in a number of places in discussion, so I will pick a key example. One quite often sees statements that are equivalent to: tax payments result in the destruction of settlement balances at the central bank. (The term "reserves" would probably be used instead of "settlement balances at the central bank," but the use of "reserves" in this case only makes sense in the case of countries with reserve requirements, like the United States.)

The line of criticism appears to go as this: taxpayer Alice pays off a tax bill drawn on Bob's Bank. Since Alice's bank deposit is destroyed* by writing the cheque, and the deposit is "bank money," and Bob's Bank creates "bank money" independently of central government fiscal policy, this is seen as refuting the MMT view.

Not so fast. The tax obligation will only be considered discharged if the cheque clears. 

There are two cases of interest:
  1. The fiscal arm of the central government banks solely at the central bank.
  2. The fiscal arm of the government holds deposits at private banks.
(Note: my initial draft of this text argued that the first case was the general case, but as was pointed out by an anonymous commenter, that does not apply at least to the United States. I noted the possibility that exceptions might exist, but the original statement by itself was incorrect.)

If the fiscal arm of the government (Treasury/Ministry of Finance) banks solely at the central bank, Bob's Bank has to make a settlement balance transfer. The implication is that there is a destruction of private sector settlement balances (since we are setting aside the fiscal arm).

If the fiscal arm of the central government leaves deposits at private banks, the logic is slightly more convoluted. Firstly, in the modern era, the decision to leave deposits at private banks is voluntary -- but this did not apply in earlier eras where central banks did not exist. The settlement of the tax obligation will just result in a transfer within the private sector banking system. However, the demand deposit held by the fiscal can be transferred at will to the central bank. Such a transfer will destroy settlement balances. In this case, the tax did not directly destroy the settlement balances, rather it created an obligation within the private banking sector that can only be discharged by the destruction of settlement balances.

(Why leave deposits in private banks? In the case of the United States, it makes it easier to run its byzantine payments system. By not destroying reserves, there are less complications for hitting the required settlement balances. Also, the Treasury can aim to increase interest received on these balances. As was pointed out to me on Twitter by George Selgin, the behaviour of the U.S. Treasury has changed over time. Since 2008, the Fed has paid interest on reserves, and so the Treasury has been keeping its deposits with the central bank -- link. In Canada, the Ministry of Finance achieves the same thing by making loans to private banks from its settlement balance.)

If we walk away from completely pointless word-twisting, we see that the underlying principle remains: the existence of a tax forces some entity in the private sector to get their hands on settlement balances -- government money -- to settle those taxes (or increase demand deposits that can only be discharged by transferring a settlement balance). The real threat of the government's coercive power (since non-payment of taxes is illegal) creates a real value for what are ultimately spreadsheet entries. The ability of the private sector to create its own monetary instruments that are layered on top of the monetary base does not change this reality.

Footnote:

* One may note the deliberate use of figurative language.

(c) Brian Romanchuk 2020

26 comments:

  1. Yeah, this is a strange, weird thing that MMT gets criticized for. I mean MMT does not at all ignore the banking system and finance or bank money or anything related to banks. Especially compared to the crap economics I got a degree in, there is a huge emphasis on the actual workings and accountings of money and how it comes to exist and stop existing. And one of the weirdist things is that some critics argue that banks don't really create money- like Paul Krugman in his argument with Steve Keen circa 2013- so MMT is wrong. And some argue that banks do actually create money- but it is in the context of a system whose complexity is such that only the 'anointed' can properly understand it- so MMT is wrong if it attempts to explain it using any generalities that are not specific to each and every country's individual banking laws and regulations and God knows what else.

    And it is just bullshit. I have been looking for seven+ years for reasonable responses and critiques and have found none. And all I actually have found is that other economists have continually moved their positions closer and closer to the original MMT position. I have not found any convincing refutation of MMT yet and at this point I don't think one exists.

    ReplyDelete
    Replies
    1. Misspelled my own name. Kind of embarrassing. I stand by the rest of the comment though.

      Delete
  2. "The term "reserves" would probably be used instead of "settlement balances at the central bank," but the use of "reserves" in this case only makes sense in the case of countries with reserve requirements, like the United States"

    There is a confusion here. "Bank reserves" is one thing, and "minimum requirements for bank reserves" is another. In some countries there isn't a minimum requirement, but countries always have bank reserves (in modern counties with a central bank).

    In Canada it seems that they use the term "settlement balances" instead of "bank reserves", which is odd, but doesn't change the point. This is just, you know, language and its arbitrary nature.

    And I couldn't understand why you would need to have minimum requirement for reserves in order to make sense of the term "bank reserves". It is as if you told me that the word "speed" doesn't make sense in German highways because there is no speed limit there. I don't see why. The limit is one thing and the speed is another.

    "Adding Banks Does Not Change Much"

    Adding banks changes a lot, and one (of many) evidence is the 2008 crisis. There is no way someone would be able to predict or explain the crisis if banks were removed from the equation. And the GFC seems "much".

    This is core MMT criticism. For example, Bill Mitchell has criticized the mainstream blindness to money and banks many times in his blog.

    What I criticize is the notion that private banks have access to unsecured overdraft facilities at the Central Bank (and hence they are agents of the state), as sometimes put forward by Warren Mosler.

    I believe that there is enough evidence to reject this kind of claim. Banks do not have access to this kind of facility, and they can go broke if they lack enough liquidity.

    The Central Bank will not necessarily back up any bank and avoid its bankruptcy and usually does not have the obligation to do so. It could if the banker is a close friend of the President, or the banker bribed some people in the Central Bank, or promised a high paying job after the mandate and etc, but this is another story.

    What the banks usually do have access to are facilities or operations that allow them to exchange bank reserves with government bonds (and, in some countries, some private bonds). This is completely different from unsecured overdraft facilities.

    ReplyDelete
    Replies
    1. I got the objection to “reserves” from someone else. “Reserves” implies that the balances must be held in reserve at the central bank. If holding them is optional, it doesn’t apply. Settlement balance is a generic term that applies to all countries.

      You would need to list countries that don’t have lender-of-last resort rediscounting facilities. I don’t recall any that didn’t do such operations against things that are not govvies (the Euro area might be an exception, but euro govvies are effectively subsovereigns). The fact that the central bank does not guarantee access to those facilities seems to buttress the point that banks are under the thumb of the government.

      Delete
    2. As for saying that the Crisis of 2008 has anything to do with the arguments about the creation of demand for base money is a complaint that makes zero sense, sorry. You ripped a section heading of context.

      Delete
    3. In the United States the banking sector has federal deposit insurance and Warren Mosler argues that this forces the Fed to provide overdraft facilities in a liquidity crisis. Because the banking sector would not function properly in the long run without government support and regulation it is reasonable to regard banks as a public-private partnership or government-sponsored enterprises.

      Delete
  3. “Since the fiscal arm of the central government generally does not hold deposits at private banks … the fiscal arm of the government is assumed to bank at the central bank”

    What are you talking about?

    The whole purpose of the real world architecture of a dual Treasury banking interface (TGA central/TTLA commercial) is to minimize the potential for unwanted disruptive effects on the central bank excess reserve setting – effects that might otherwise arise from natural mismatch patterns in Treasury’s daily flow of funds. It’s to immunize - as desired – the net effect otherwise of Treasury transactions on that reserve setting. The Treasury cash position (coordinating with the central bank policy) gets netted to a desired reserve effect on the system reserve position (usually more or less neutral) by funds transfers between these account types.

    And yes – those TTLA thingy’s are deposits too - with commercial banks.

    ReplyDelete
    Replies
    1. I have remarkably little interest in the details of the US system. I would have to get a reliable source to see what extent deposits are held by the US Treasury are held at commercial banks, my understanding was that practice was phased out with the creation of the Fed. I’ll note the objection, but if the deposits are held in a private bank for a couple days then transferred at will to an account at the Fed, it just delays when the private sector has to get its hands on base money.

      And note that I wrote “generally”, are there any other exceptions?

      Delete
  4. "One quite often sees statements that are equivalent to: tax payments result in the destruction of settlement balances at the central bank."

    I might have written that myself earlier in my economic study. Not today.

    Today, I allow central banks to create 'base money'. This allows the CB to control the monetary system. With money in-hand, the CB lends to government.

    Here a citizen paying taxes does not shrink the base money supply. Instead, government does indeed have a bigger balance at the central bank. Only when government repays the CB loan does the base money supply shrink.

    This money creation approach still has an MMT style of monetary freedom. It also has the CB exercising control, which is similar to the situation found in modern monetary systems.

    ReplyDelete
  5. Oh, I got it. I don't believe "settlement balance" is a generic term. "Bank reserves" is much more. I myself took some time to understand what you were talking about - I had to do some research on the Canadian financial system. In the US, UK, Brazil ("reservas bancárias") and even Japan (in their English reports) the term "bank reserves" is used, while "settlement balance" is an alien term.

    As I see it, "reserves" implies that the financial resources are reserved for future use. I particularly would change the term "bank reserves" to "virtual [currency name]", like "virtual dollars", in contrast to dollar notes and dollar coins. Maybe "base dollars" would also do. I just don't think those language debates are much productive. I could make a point that the word "car" is not helpful and should be changed to "quadcycle" to making things clearer, but who am I to change a language?

    Whatever is the case, speed is different from speed limit, as "bank reserves" is different from "minimum requirements for bank reserves" (or "bank reserved requirements" or "reserve requirements"). "Bank reserves" and "reserves requirements" are not the same thing.

    "You would need to list countries that don’t have lender-of-last resort rediscounting facilities"

    According to Fed itself, "All discount window loans must be collateralized to the satisfaction of the lending Reserve Bank". This also applies to many other central banks (but I certainly do not know how all the 100+ central banks around the world work). If the private commercial bank needs, for example, to supply government bonds as collateral (or whatever), it is just swapping one kind of asset to another. If it doesn't have enough collateral to supply, then it will not be able to borrow. This is not an overdraft! This is a secured credit facility. If overdraft facilities were available, no bank would ever go broke, and they would not need to seek deposits from customers, unless those deposits were cheaper.

    This is important. I cannot blame anyone who questions things like "if banks have access to overdraft at the central bank, why would they bankrupt? Why would they seek deposits from customers?". This sort of "agent of the government" claims do raise a lot of questions and not so much answers. I don't think this is a fair representation of how banks work.

    "As for saying that the Crisis of 2008 has anything to do with the arguments about the creation of demand for base money"

    I was not linking the Crisis with the creation of demand for base money. I was linking the Crisis with the "Adding Banks Does Not Change Much" claim. I do believe modelling macroeconomics without banks is a mistake, and adding banks do actually change a lot. At the end of the day, many banks went bankrupt because many customers hadn't paid principal or interest on their loans, and so banks run into liquidity issues - and had no "overdraft facility" to borrow base money from nor could attract enough base money through customer deposits. But maybe I just missed your point.

    ReplyDelete
  6. Selgin has it right with the following references in his twitter response to you:

    “An overlooked aspect of last week's repo turmoil is how it might have been avoiding had the Treasury employed TT&L accounts, as it did before 2008, to avoid "abrupt changes in depository reserves." … but these technical explanations hardly explain why the Treasury and Fed together didn't consider making better use of the TT&L program last week, or why they don't agree to use it more in future, to prevent a repeat of the recent troubles… that option seems far better than having the Fed begin to "unwind its unwind," as I fear it may. A relatively small loss of Treasury revenue seems a price worth paying to avoid making a complete farce of the Fed's long-promised balance-sheet normalization effort... my point is that the Fed would not have had to act at all had the Treasury accumulated funds in its TT&L accounts rather than in the TGA, moving them to the latter only as needed to accommodate spending.”

    The general point is that, notwithstanding the current dormant condition of the TTL accounts, they offer a useful option for future flexibility in the approach to management of Treasury cash balance effects on the system reserve setting. This flexibility is useful insurance in that it seems unlikely that the Fed is 100 per cent certain about the path for future balance sheet management techniques.

    The same general case applies to the Canadian system and probably most modern monetary systems. Treasury (or Finance in Canada) has a dual interface in its banking arrangements – an account at the central bank, plus bank accounts and/or other facilities for deploying cash balances away from the central bank balance sheet (TGA, auctioned deposits, whatever). Such arrangements facilitate coordination between fiscal and monetary policy in controlling the effect of short term Treasury cash positions on bank reserves. It’s complementary to longer term bond financing – mitigating undesired effects on the central bank reserve setting.

    ReplyDelete
    Replies
    1. "The general point is that, notwithstanding the current dormant condition of the TTL accounts, they offer a useful option for future flexibility in the approach to management of Treasury cash balance effects on the system reserve setting"

      I believe it is not useful. On the contrary, it brings unnecessary complexity, costs, risks, rent to banks and inefficiencies in general.

      In many countries the Treasury is not allowed to hold bank deposits assets. They must deposit it all at the Central Bank (as if there was only the TGA available for the Treasury). And this poses no issue at all. Brazil is one example, but there are others.

      The issue is when the Central Bank is (for some unreasonable mystical reason) trying hard to reduce its balance sheet and at the same time not making the Repo and/or Reverse Repo facilities available to banks. Yes, then it will not be able to reach the interest rate target, no surprises there.

      If you tie someone's feet and hands and throw him in the sea, you shouldn't be surprised when he drowns.

      Delete
    2. "they must deposit it all at the Central Bank"

      no

      as in:

      "plus bank accounts and/or other facilities for deploying cash balances away from the central bank balance sheet"

      as in:

      " (TTL, auctioned deposits, whatever) "

      as in:

      " It’s complementary to longer term bond financing "

      all monetary systems would have such options

      Delete
    3. Yeah, in Brazil, for example, 100% of the financial resources of the federal government is deposited in the "Conta Única do Tesouro", or "Treasury Single Accounts", which is the TGA.

      This poses no issue to the financial system. The Central Bank never missed its interest rate target.

      By the way, the Brazilian financial system is more modern than of the USA, which is shameful. It's a pity that this is not true for almost everything else.

      Delete
    4. wondering what this is:

      https:

      //riotimesonline.com/brazil-news/brazil/brazilian-government-to-release-over-r100-billion-to-private-banks/

      seems to read as if its a deposit of Brazilian treasury funds with banks - with the intention of increasing reserves

      "to lend out"

      as per mainstream interpretation

      :)

      can't tell for sure

      ?




      Delete
    5. Oh no, this was just a reduction of reserve requirements (the minimum amount of bank reserves that banks need to keep deposited at the Central Bank):

      "RIO DE JANEIRO, BRAZIL - The government is expected to release over R$100 billion (US$26 billion) to private banks in compulsory deposits, which are funds that banks are required to keep on deposit at the Central Bank."

      It was reduced from 33% to 31% of deposit liabilities.

      The mainstream economists believe that banks will lend out the "freed" reserves.

      Delete
    6. André: "Oh, I got it. I don't believe "settlement balance" is a generic term. "Bank reserves" is much more. I myself took some time to understand what you were talking about - I had to do some research on the Canadian financial system. In the US, UK, Brazil ("reservas bancárias") and even Japan (in their English reports) the term "bank reserves" is used, while "settlement balance" is an alien term."

      It is very simple. I read the argument somewhere in an academic text on operations I read years ago. I see no need to dig up the reference.

      It is very simple "settlement balances" is a term that is applicable to all countries, even if the locals use a different one. The concept of "reserves" existed in Canada, but it no longer was applicable when *reserve requirements* were abolished.

      Trying to browbeat me to use "reserves" is just an attempt to impose American cultural imperialism on me. Feel free to consult texts on Canadian history to see how older Canadians view that topic.

      It's my website, my books. I am going to use what I see as the correct term. If you want to translate the term to "reserves," feel free. I am unlikely to quote extensively from anonymous comments on my own website within my books (somewhat sketchy, since readers could assume that I just make stuff up, and since I administer the website). I will cite corrections and observations, but I keep the discussion generic. As such, I do not see any reason to discuss my preferred wording.

      Delete
    7. " this was just a reduction of reserve requirements "

      ok

      thought that might have been the case - sentence structure was odd

      Delete
  7. " (TGA, auctioned deposits, whatever) "

    … meant (TTL, auctioned deposits, whatever)

    ReplyDelete
  8. “If we walk away from completely pointless word-twisting, we see that the underlying principle remains: the existence of a tax forces some entity in the private sector to get their hands on settlement balances -- government money -- to settle those taxes (or increase demand deposits that can only be discharged by transferring a settlement balance). The real threat of the government's coercive power (since non-payment of taxes is illegal) creates a real value for what are ultimately spreadsheet entries. The ability of the private sector to create its own monetary instruments that are layered on top of the monetary base does not change this reality.”

    If settlement balances are presumed to exist by argument, then the full scope of private sector banking obviously exists as well. There is no other reason for settlement balances to exist. Therefore, “the ability of the private sector to create its own monetary instruments that are layered on top of the monetary base” is a given. In that context, the private sector “layer” of commercial banking is the logical foundation, and the fiscal authority’s sourcing of taxes from settlement balances is more the add-on to that architecture. Settlement balances don’t exist for purposes of state tax collection without first acknowledging that the commercial banking system requires these balances as a priority in its broader function of accommodating private sector bank competition. In the context of that broad functional perspective, the interface with fiscal authority taxation is a relatively minor presence.

    So the primary purpose of settlement balances is to enable private sector bank competition – the adjustment of balance sheet management strategies. Those strategies and their effect on the distribution of settlement balances incorporate adjustments for both assets and liabilities. And the commercial bank liability perspective clearly shows that taxes must be paid from pre-existing commercial bank deposit balances.

    This temporal/causal order for commercial bank money as it pertains to taxation is opposite of that for the MMT meme for settlement balances (i.e. along the lines that “state spending precedes taxation”). While settlement balances are necessary for the collection of taxes, they are not sufficient. Taxes are paid first from bank money – bank money that is created by banks, not the state. Therefore, a “paradigm” that is expressed in terms of settlement balances alone therefore cannot be logically complete or accurate. The implicit structural assumption of bank competition for assets and liabilities – including deposit liabilities - contradicts the usefulness of such an exclusive focus on settlement balances.

    In such a context, the MMT meme that “state spending precedes taxation” doesn’t cut it.

    Nor does the attempted fix: “state spending or lending precedes taxation”


    ReplyDelete
    Replies
    1. (1) Since a modern currency is defined by being equivalent to a deposit at the central bank, if a brand new currency came into existence, state spending has to come first. Once a currency is in existence, creation/destruction of reserves is a continuous cycle. Does winter come before summer or after it?

      (2) Banks cannot pawn off the transfers to the central bank with private deposits. They have to get their hands on government deposits. They are not the "foundation."

      Why are you wasting your life making rhetorical points about primers that you are unable to even provide references for?

      Delete
    2. " And the commercial bank liability perspective clearly shows that taxes must be paid from pre-existing commercial bank deposit balances."

      What? There is nothing in your comment/argument that supports that conclusion. Nothing I can see at least.

      The government decides what it will accept as payment of taxes. It has the army, the police, 'the law', and hopefully, the consent of the people. If you start your argument there, I can't see how you end up where you do.

      Delete
    3. "Does winter come before summer or after it?"

      Now we are talking about the creation of money--right?

      Here is my solution:
      People work before they are paid. Hence, payment for that work is recognition of work completed. Hence, money is acknowledgement for work performed. Newly created money is acknowledgement for work performed.

      I claim that work comes before money; not really a winter/summer question.

      This idea makes money seem like a micro-economic coupon. Enhanced, of course!

      Delete
    4. "Hence, money is acknowledgement for work performed"

      But how do people determine how many units of money to pay for one hour of work? Is it worth 1, 10, 100, 1000 or 10000 units of money? Why?

      Someone somehow has to determine the value of one unit of money, even if for just one commodity that will serve as benchmark.

      If that someone is the market, how it does that? If it is the government, how does it work?

      The "money as units of labor" does not seem able to answer those questions. You do have to resort to some other kinds of theories like the “mass delusion theory” that assumes that, for some strange reason, people somehow start believing and agreeing to each other that one arbitrary unit of money is worth some defined amount of economic value.

      Delete
    5. This is discussed in my latest article.

      Delete

Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts.

Note: if you want to post comments from Apple devices (iPhone, iPad), you apparently need to turn off "prevent cross-site tracking" in Safari privacy settings. (The reason presumably is that another URL handles comments, and so the user session needs to be preserved when redirected to that site. I don't like this, but this is not enough to make me switch my hosting service.)