Saturday, August 11, 2012

Musings in central banking 2: challenge to MMT

I recently wrote down my understanding of central banking, taking the reader from pure banking (PB) to the system we currently have—interbank rate set through open market (OM)—via the simpler set up in which the CB intervenes through the discount window (DW). The purpose was to demystify some aspects surrounding 'printing press' which tends to be used indiscriminately. I took aim at MMT's thesis that the 'government is not revenue constrained'. I am just an amateur, it should be easy to find flaws in my reasoning. MMT followers were not forthcoming in taking up my first challenge. Let's give it another try.

Pure banking vs central banking

In my tutorial, 'printing press, fact vs fiction' (ECB-Watch), I imagined a two banks model, A and B, the clients of which are the grocer's customer and the grocer, and I asked the question what happens when the customer pays the grocer with a credit card? A faces a treasury shortfall, which has to be funded in order for the transaction to clear. In PB, A borrows the required from B.

In DW, the CB takes the role of B: the CB creates a deposit—now called 'reserve'— in A's name and it is immediately transferred to B, leaving A with a loan obligation towards CB by same amount. The DW rate on the loan determines, in turn, the interest on the grocer's customer overdraft.

In OM, banks can lend each other reserves, the supply of which is controlled by the CB. The CB adjusts it—drains or adds liquidity—so as to meet an interbank rate target, itself chosen to meet an inflation target. A particular aspect of this set up is that the money multiplier theory as taught in the average economics class is incorrect.

Criticism of MMT

MMT argues that the government is not revenue constrained. That taxing and spending are merely macroeconomic levers for steering the economy in a particular direction. The only constraint is inflation. These postulates allow the very progressive MMT to promote a government job guarantee program (JG) also know as employer of last resort (ELR).

The PB model is a two level hierarchy: the interbank market—banks supplying credit to each other, and banks supplying credit to the economy.  Ignore, for the sake of argument, the incentives, micro or macro, that have led banks to fuel a credit bubble in the 2000s. The key driver of private credit, under this hypothesis, is demand for loan by credit worthy customers.

I'm going to assume the government receives the same treatment as private sector in the PB model: it must be credit worthy in order to obtain a loan from banks: the government is credit worthy—solvency—constrained. A corporation pays back its debt from net revenue and so does the government; revenue is the form of sales and taxes, respectively (ignore sale of asset).

In transiting from PB to DW/OM, the CB becomes the monopoly supplier of interbank credit—reserves, and thus can set its price—interbank rate. Is the government no longer credit worthy constrained? I don't see how, but such is proposition implied in the MMT claim that the government is only 'inflation constrained'.

To make their case, MMT relies, I think, on the fact that CB can ultimately print as much money as needed to fund the government. How does that guarantee or void the notion 'credit worthy'? That's for them to explain. I never see MMT comparing the PB model to the OM model. MMT should explain why they rely on OM rather than DW for reasoning about macro policy. Remember Einstein's motto: a model should be as simple as possible, but not less.

Perhaps it's time they tackled this challenge.


A response to my tutorial was that MMT does not deny private banks are not revenue constrained, adding 'what would that even mean'? I said: indeed, what does that even mean (for government)? By the same token I should ask them why debt issued by government is called a 'net asset'. This implies the corresponding liability is void. It's absurd: one's asset (financial claim) is another's liability (financial obligation).

The same person added private banks were 'capital constrained'. That's confusing the means (capital ratio) with the end (constraint on private credit growth).

Also see:


  1. The MMT statement is in theory . The CB may fund government debt at 0% over an unlimited timespan . Hence take a current trillion at five years , reduce the rate to zero , and have repayment over a thousand years . As events stand rates are adjusted to suit a needed reality , no one is asking what happens when the government becomes the economy and it turns out to be unproductive or corrupt . The government might then try to pay people to be quiet , but that money would not necessarily stimulate the economy , only inflation. As stands credit inflation (hence monetary) over the last decade has been vast , and distributed . It worked courtesy of ponzi and China . Current attempts at reflation are only adjusting accounting , not working into the economy properly . There is a balance that is natural between inflating returns and depreciating the value of a currency , so a look at trade deficit and real consumer costs gives another aspect to what is going on . The two big fields here might be oil costs and the Chinese peg/debt holdings (change in the peg , use of dollar ) . Europe gives a good comparison to the US - the reality of secured repayment of government debt in the US means it only partly has to be subsidized , investor confidence in repayment remains . In EU that is not the case currently, much as some would like it to be. On the other hand , easy access by government to funding is not necessarily a good deal - the US is not likely to tighten monetary policy because of that , and may end up in its own rut accordingly .

    1. U are repeating MMT mantra. Start from pure banking and explain how MMT magic comes into existence by merely creating a monopoly over reserves whose mandate is to set the interbank rate.

  2. Governments hardly ever repay their debt from revenues. Usually they just borrow new money to repay their old debts. So the ability to repay, the creditworthiness, solely depends upon the existence of a future lender for the government. As long as there is a central bank standing ready as a lender of last resort for the government, it's default risk is indeed 0. What happens to the economy in this case doesn't matter for the bond investor. He just wants the promised payment.

    Regarding your problems with "net financial assets", it's about accounting identities. All assets and liabilities resulting from the private sector lending and borrowing from each other net out to 0. The government is in this view an accounting identity standing outside the private sector (by definition). So any government liabilities are net assets for the private sector while any government financial assets would be net liabilities of the private sector. Once you throw private and public sector together, the assets and liabilities net out to 0 again. So basically the talk about "net financial assets" of the private sector is only one aspect of sectoral "Saldenmechanik".

  3. "U are repeating MMT mantra. Start from pure banking and explain how MMT magic comes into existence by merely creating a monopoly over reserves whose mandate is to set the interbank rate."

    Replying here as javascript void .

    Hope am not talking past you :

    Because the central reserve of government debt is 'guaranteed' as good as money (which is due to the central reserve being able/willing to conjure more money in return for more government debt , directly or via favoured bank, at rates and terms that allow the government accounting to always appear viable ) , and because government accounting is seen to be viable due to this 'trick' (e.g. artificially lowering rates), there only exists the need for an excuse to do this , and to maintain inflation levels is that excuse - an original mandate of the CB over the interbank rate . In other words the government enjoys preferred investment status because it is guaranteed by a printing press . After that it is for each person to reason the level of new 'social investment' that they think is due , if any , and to calculate if creating inflation is a distributive effort , simply biased to one part of the population , or an attempt at propping up accounts of business and government (in other words papering accounts and across the board currency devaluation, but via selected inputs of spending to prefered channels, instead of allowing true market adjustments ) . In reality there are revenue constraints however , not quite as we know them maybe - loss of monetary credibility , excessive inflation that demands a tightening on lending so causing a demand for a rise in yields by investors to compensate , which then makes servicing the accumulated debt level impossible without ever increasing taxation and cutbacks in government spending , social unrest and economic decline , etc. MMT ignores that side of events however because it is still temporarily able to boost money supply without major upset.

  4. 'In reality there are revenue constraints'

    That would be a watering down of the MMT claims I read.

    Let's find out exactly what MMT means by 'not revenue constrained' by starting from ground zero.

    A loan officer estimates the applicant's revenue stream to decide whether to accept/reject. If loan officer overestimates revenue capacity, applicant will fall behind payments and ultimately default.

    Does MMT's claim the gov is 'not revenue constrained' void any notion that debt/revenue ratio determines credit worthiness hence ability to raise new debt?


  5. Cont.

    "Does MMT's claim the gov is 'not revenue constrained' void any notion that debt/revenue ratio determines credit worthiness hence ability to raise new debt?"

    No, it doesn't , at least not to my view . True revenue is earned , not borrowed . It is a bit like I print you a piece of paper which you might exchange for goods or services somewhere. You have to return me that piece of paper later - but hey, don't worry, if you can't, I will print and give you another piece of paper, or two even. If we look at the nature of debt expansion , this practice is common in banking (extension of loan terms etc.) , but only to a point. That point is often breached when the lender realizes that if he does not ease the loan terms the asset he has invested in will not repay, and not restructure either ,but may go to bankruptcy and the return will be even less (due to conflicting claims, and valuations at a 'poor moment') - this is really moving into outright speculation by the lender , not contemplated investment, though the adjustment may also have a reasonable 'plan' , and is often obliged by the borrower who threatens to walk from his payment obligations. Whichever, it is a breach of the original understanding calculated by revenue constraint, due to revenue not being adequate in reality .

    In practice raising new debt by government is not revenue constrained (at least as we would normally understand the words) :

    When government spends money it borrows it has an initial ratio to debt or velocity of 1 (plus interest which I will ignore in this example). That debt is created for the government due to gov. being considered (in theory) to be able achieve a similar return or revenue in the future to repay .

    Let us look at what government 'does' - at best it provides a background for the efficient running of the country and stability of society, or something similar. It doesn't, or very rarely , generate wealth directly via an open market - in other words it does not respond or adapt to market prices or wishes (except via elections and similar) . Quite the opposite, it often tries to set the market environment .

    So where does its revenue come from mostly - that is well known , taxation. It legally expropriates wealth from the nation to maintain itself . So far nothing new, and theoretically there would be some kind of beneficial symbiosis going on between the government and true economy.

    You would think that to achieve its balanced revenue , the government would simply tax out of the economy as much as it put in via its debt and according to a certain revenue plan. That would still give it a time and positional advantage of original spender/investor in the economy. Let's go back to the original borrowing at a ratio of 1 though . When that money is spent it is left in the economy for the period of the debt (which until there is a net borrowing reduction means indefinately) . That money will circulate and be taxed at each turn, at each change of hand . Tax is usually by % and so it would be technically impossible to tax every penny or cent back . However this new money is also placed in deposit , and thanks to the wonders of FRL , might be multiplied ten or fifteen times . That newly 'created money' is also taxed in its movement through the economy . Hence , 'revenue constraint' is subject to an understanding of the whole working of the entire national economy , and 'in theory' , the goverment should easily be able to tax back its original input from the FRL expanded monetary base. MMT I suppose also considers that government spending is hence inflationary in monetary terms , and condone gov. borrowing and spending as a means of readjusting the economy to a desired state. However they maybe overlook both the ethics, and the true nature of the real economy, as well as the limits of FRL, and other derivative schemes.

    1. The most straightforward interpretation of 'not revenue constrained' (NRC) is that ability to raise debt independent of capacity to raise revenue–taxes in the case of government. The NRC claim is a cornerstone of MMT. It's meaning should be crystal clear and your comment isn't (I'll give it another try). Why don't you start from pure banking and tell us if gov is NRC then (Y/N).

    2. No. The most straiightforward interpretation of not revenue constrained is (1) government prints check to pay for purchase (2) check deposited at bank A (3) check presented for settlement via CB (4) Bank A credited with reserves by CB.

      Whether (i) CB just writes up reserve based on fiat currency issue ("greenbacks"), (ii) Treasury issues security to CB, CB credits equal reserve amount to Treasury, and transfers reserves or (iii) government pretends to borrow from private market, with payment transferred to its account, and then CB buys bonds to maintain target rate, the final outcome is identical ~ reserves appear in Bank A's account, as a liability of the CB, CB holds effectively non-interest bearing asset, and government has stuff it bought.

  6. By the last meaning government is not revenue constrained (that is for the US, for the ECB the mechanism to allow this is being tested) . The reason I think this is so, is that the there is nothing to stop a central bank from accepting existing government debt as collateral for further refinancing to banks to buy more government debt (carry trade) , or directly accepting government debt on its balance sheet, either from the primary or secondary market purchase . From my point of view , there is little difference in doing this than from simple direct monetization of debt (which is also a further possibility) except that the debt held by the central bank is not written off in the process, but held (indefinately)for eventual repayment (which usually involves it being simply rolled ) . I would be happy to hear there is some kind of technical limit to the ability to do this , but I haven't found one yet , if you look up modern monetary systems at it suggests that the central bank can simply issue new money for new debt (as well as being able to change other parameters , such as interest at DW etc.). In EU the restrictions are political for via the carry trade (closed door ECB vote on collateral ) , and so far prohibited in open by mandate, the Secondary Market Program being the last obvious attempt to openly flex the rules .
    Practical limits are obvious , i.e. how many zeros would fit on a screen etc.
    Government spending would boost GDP and therefore have the ability to lower debt to GDP , but this is an unlevel real economic approach , and as I mentioned previously depends on a functioning true economy - otherwise it is likely to uphold a false or synthetic set of values in the real economy (though MMT followers find many reasons/excuses to justify their approach ) which are not based on natural individual choice , rather collective four yearly limited vote on national policy via elections, and so on.

  7. Will leave my contribution there - maybe as a simple answer via the PB setup you describe , the government is not revenue constrained because the CB is able to literally create money to lend to government , without using existing reserves . The only obstacles are political and sometimes legal restriction, which are often circumvented .

    Two articles from websites that cover monetary policy and are worth following , and which give more insight into the whole process:

    1. There is no CB in the PB I describe, so I'll have to disregard your answer. But thanks for the effort. PB stands for pure banking. The description given in this post is a summary of that given in 'Musings in central banking 1'.

  8. Effective sovereign governments of whatever form, democratic or autocratic, have power. More specifically, they have the power to impose tax obligations on those they govern, obligations that can only be discharged by surrendering back to the government a required quantity of the currency "liability" that the government itself has issued. They also have the power, through their systems of legislation, courts and law enforcement, to see to it that the currency they issue must be accepted universally within their jurisdiction for the payment of all private debts. These two powers enable the government to sustain high demand for their currency, even in circumstances in which the inertia of the currency as a customary medium of exchange flags.

    That infrastructure of power, the apparatus and institutions that enable governments to make and uphold rules - i.e. to govern - is what makes it the case that governments are able to secure general acceptability of the currency they issue. The only "credit-worthiness" that a government needs to maintain is some general confidence that the government will be able to maintain its power for the foreseeable future, and thus preserve the general acceptability of its currency through the techniques mentioned above.

    1. 'inertia of the currency as a customary medium of exchange flags' is perplexing, not clarifying.

    2. Let me clarify: sometimes a particular form of currency becomes established within a particular geographical region as a customary medium of exchange. Since human customs and conventions have a certain tendency to persist, and since a society's possession of a monetary exchange system is an enormous convenience, then once such a currency custom has been established it can remain in place for some time. That's the inertia.

      But human beings are also fickle, and if the system of exchange that relies on a particular currency whose exchange value is purely conventional enters into crisis, the currency convention might collapse. In the modern world, then, our monetary systems do not rely on convention alone. They are heavily governed and regulated. The state issues the fundamental form of currency, and enforces a variety of laws that are aimed at assuring that the currency will be sought and valued, and accepted in exchange.

      In any currency system, the entity that issues the currency has seigniorage power: it can create additional purchasing power for itself (or others) by creating additional units of the currency for itself (or others), rather than by acquiring units of the currency from others. So long as it doesn't create so many additional units as to radically inflate prices or destabilize the currency, that can be effective.

    3. This is removed from the terms of the debate I set forth in this post.

    4. Well, whatever. If the terms of the debate require accepting your erroneous premises based on a fictional PB model of a monetary world that doesn't exist, then your erroneous conclusions will naturally follow. But if you are interested in exploring the nature of the constraints that do and do not apply to actual governments that actually exist in the actual world, then it is not a good idea to limit the discussion in advance to fictional models describing the banking system in some imaginary world.

    5. PB is an abstraction of the system without CB. It's internally coherent. MMT's claim that 'gov is revenue constrained' OTOH, is, as things stand in this debate, a fiction: it requires a leap. Have a look at the next post titled 'Musings in central banking'.

    6. Therefore it is a system with direct government minted money and unregulated currency-reserve banking?

      Your PB has nothing to do with the MMT claim, since the MMT claim is a claim regarding how things work under existing money institutions, and your PB appears to be a fictitious just so story with no coherent institutional foundations, similar to the fairy tales told in lieu of economic history by the monetarists.

  9. I reject the whole premise of your argument on the grounds that you're engaging in fallacious reasoning. You set up a hypothetical model of "pure banking" vs. "central banking" which has no bearing on the world in which we actually live. You then issue a challenge to the MMT community to defend its operational description of the modern banking and public finance system, but you refuse to allow any realism in the discussion. Instead, you force them back into this nonsensical pure banking world, which allows no room for real world interaction between the state, the CB and the financial system and ask them again to explain the 'no financial constraint' argument.

    I won't play this game. I'm open to a legitimate discussion concerning these issues, but you need to first fully specify your model with warranted assumptions. Here's a hint: you assume the government faces the same underwriting criteria that a household or firm does. This is clearly an unwarranted assumption, as governments do not go to banks for loans. You further assume that CBs control the quantities of reserves. The Volker experiment of the late 70s early 80s has demonstrated this to be an unwarranted assumption, and as such central bankers have given up on quantity targets altogether, and now target rates, which they have demonstrated control over.

    So, try this again, but do so with good assumptions, a realistic model, and then perhaps we can talk.