Recent Posts

Tuesday, August 25, 2015

MMT And Net Financial Assets

There have been some recent characterisations of Modern Monetary Theory (MMT) which I believe are incorrect. In my view, the MMT positions are largely common sense positions, albeit phrased in an original fashion. I am luckily outside of academia, so I have little interest in the question of determining publication priority of analysis. It may be that academics within the MMT tradition would offer a more pointed defence of their analytical framework; I am just offering my interpretation of the MMT literature.

U.S. Federal Debt And Depressions

"Asymptosis" wrote the article "Does Reducing the Federal Debt Cause Financial Collapse"? He discusses the observation that U.S. Federal fiscal surpluses generally preceded depressions.

Firstly, I would not place too much weight on that observation, although it is fairly entertaining. Other countries (particularly Australia) have been able to run federal surpluses without the economy collapsing. The article then states:
What all this ignores, of course, is privately-issued debt, something that — confusingly to me — many MMTers don’t talk much about, if at all. I haven’t clarified my thinking on that issue, so I’m going to pass you for the moment to Rodger Mitchell. I haven’t thought through his ideas or data carefully, but I find it useful that he looks at private debt, federal debt, and their relationships over time — something I haven’t found well done elsewhere. 
In my view, this characterisation is misleading. Most MMT analysis discusses government policy choices; since governments have no way of controlling private money creation, private money is necessarily discussed less. However, this lead to the following tweet by Noah Smith.



Noah Smith's characterisation is dead wrong. The correct statement is that MMT holds that government money is of primary importance, while private money is of secondary importance. That is not to say that private money is not significant, but private money can lose its "moneyness" in a financial crisis. Private liabilities are inherently illiquid, and need to be backed with holdings of liquid government liabilities (reserves, bills and bonds). (It should be noted that this was emphasised by Minsky, and presumably others, and was "inherited" by MMT.)

Finally, J.W. Mason followed up with:

I do not have enough information to fully respond to this (and other tweets). I will now attempt to respond to what I believe is the underlying thinking.

Yes, Net Financial Assets ("NFA's") are an accounting construct, but so is practically everything else in macroeconomics. My interpretation of MMT analysis is that it follows from the Stock-Flow Consistent modelling framework, and the demand to hold financial assets is critical for determining the steady-state characteristics of the economy.

Yes, those financial assets include assets issued by the private sector. However, we see that the desire to emit liabilities by the private sector is determined by the state of the economic cycle, including structural components. For example, during the housing and tech bubbles, the private sector was very happy to emit liabilities. Those "animal spirits" in the private sector with respect to net debt issuance is not under the direct control of the government. All it can do is adjust its debt issuance - (under the accounting convention used by MMT, central government liabilities are the "net financial assets" of the non-central government sector) in an attempt to adjust economic activity.

The variability of the willingness of the private sector to emit debt means that there can be no hard and fast rule relating "net financial asset" issuance to economic growth. But at the same time, I have never seen academic MMT writing asserting that such a simple rule exists. Instead, the deficit needs to be adjusted to current cyclical conditions, which is a core view inherited from Functional Finance.

(c) Brian Romanchuk 2015

20 comments:

  1. Strikes me that can all be boiled down to a few simple sentences, something like the following.

    State issued money is a net asset for the private sector. (MMTers call that “Private Sector Net Financial Assets”). Ergo an increased quantity of that money tends to increase demand. In contrast, privately created money nets to nothing. That is, privately created money is debt owed by one private sector entity to another, and it’s a debt which unlike most debts is easily transferrable. That’s what makes it “money”.

    I.e. privately created money IS NOT a net asset for the private sector. Thus it does not have the same demand increasing effects as base money (aka state issued money).

    ReplyDelete
    Replies
    1. The issue is not whether we can restate the idea better, but whether MMT authors express the idea incorrectly (as per Ramanan's comment below). I think the academic MMT writing does cover it in a reasonable fashion, but one could debate whether it was being expressed in a misleading fashion.

      Delete
    2. Privately created debt absolutely increases demand. Steve Keen has covered this extensively. When a loan is created and spent on an item, that is new demand in the economy that didn't exist before.

      Delete
    3. I think Steve Keen revised his thinking, but his earlier writing on this was mistaken. If someone borrows to buy an existing asset (for example, a house), that borrowing would not raise GDP. You need to purchase newly produced goods or services to add to GDP. This messed up the accounting in his models. I believe he corrected his models after this was explained to him.

      In any event, I would not deny that private debt and money is important. However, it depends upon the private sector being liquid, which is to say that they have holdings of government liabilities.

      Delete
    4. This comment has been removed by the author.

      Delete
    5. Interesting.. I know his early models weren't double-entry, which he later switched over to... OK, for credit to raise GDP it has to be spent on newly produced goods... Isn't almost all credit card debt used on newly produced goods or services?

      I'm unsure why spending a newly created demand deposit (created simultaneously with a new loan) on existing goods wouldn't raise GDP. I'm spending money I wouldn't have otherwise spent... The main point Keen makes is that for me to take out credit and purchase something, does not require someone else to have forgone consumption in order to lend me the money, so therefore it's brand new purchasing power (unless Keen revised that in the last year or so, I believe that's still his explanation). Why wouldn't that raise GDP?

      A new loan is new money (with corresponding in-sector liability), it's not the loanable funds model that people like Krugman subscribe to. According to Krugman, purchasing power is simply transferred from saver to borrower, whereas Keen says no saver necessary, it's an increase in purchasing. If I understand correctly, the bank of England and the IMF have both published papers within the last couple years confirming loanable funds is incorrect.

      Delete
    6. Hyman Minksy gives credit to Keynes with respect to his model of the two price economy. So in the market for houses there are existing houses for resale, which were included in past GDP, and newly constructed houses for sale, which are included in current GDP. The price of the newly constructed houses must be comparable to some of the existing houses. When house prices begin to fall this causes recent mortgages to go underwater and investment in the construction of new houses to slow down or stop, so the lack of finance to sell existing houses causes a drop in investment which contributes to GDP. This is not only true for houses, it applies for oil drilling equipment and other assets.

      Delete
  2. Brian,

    While Noah is not right but he seems to see the overkills. For a long time MMTers argued that the private sector cannot save if the government doesn't have a deficit. It took a while for people to emphasize that it is wrong.

    Anyway, I thought JW Mason's tweet was hilarious.

    National accountants do measure and report Net Financial Assets.

    Good point by you that all things are economic constructs.

    ReplyDelete
    Replies
    1. *all economic things are accounting constructs.

      Delete
    2. I recall those debates about savings; luckily my blog was not launched then, so I stayed out of them. I seem to recall that I was sympathetic to the MMT view, but that the way they "branded" their terminology was possibly suspect.

      I understand the desire for MMT to rebrand terminology in order to avoid the negative overtones in existing terms (like government debt and financing), but at the same time, I want to explain fixed income economics using terminology most people will run into.

      Delete
  3. Brian - factor in trade balances, and private debt. For example, running a trade surplus can offset a budget surplus, and visa versa. May be an accounting identity – but one that has some teeth.

    So a trade deficit of 2% GDP and a budget deficit of 2% of GDP kind of cancel each other out, with private debt being the other lever - it would need to increase in the private sector for growth. Looking at the beginning of the GFC in the U.S., the budget deficit was about 1% of GDP and trade deficit around 6% of GDP, once private debt collapsed - ouch.

    This is all Wynne Godley SFB stuff. If one sector is spending less than its income it must be accumulating net claims on other sectors.

    Australia ran a trade surplus for some time, and was able to run budget surpluses. They now have a trade deficit, a budget deficit, and what appears to be growing (ballooning?) household debt.

    https://research.stlouisfed.org/fred2/graph/?g=1GIz

    ReplyDelete
    Replies
    1. I typically lump in the external sector with the private sector; I should really say "the non-central government sector". So yes, the patterns of external trade will also affect the demand for financial assets.

      For example, the Chinese government's accumulation of Treasury assets (and government-guaranteed MBS) skewed the holdings of financial assets of the rest of the "private sector." There was no way that they could buy all those Treasurys from existing holdings, so the system had to bend in a way that resulted in higher Treasury issuance.

      As for Australia, the government deficits coincided with a lot of private sector debt issuance (mortgages in particular). As for the current account, I believe that there are some special factors - a lot of the mining/energy firms are foreign owned, and the current account "flows" are just reinvested profits within those firms. I never covered Australia in too much depth, so I am unsure about the magnitude of those flows.

      But closer to home, I probably should have mentioned the Canadian Federal Government's experiences with surpluses. Once again, it was accommodated by increased household debt.

      Delete
  4. Apologies in advance for venting here... I don't know why there's such an understanding gap around MMT but if history is any guide, your response won't gain any traction. I could load up this comment with links (but I won't... because spam filter) to MMT economists discussing private borrowing/debt/money creation prior to the 2011 asymptosis post. Then another batch following it, starting with their 2012 "MMT 101 A response to critics" that addressed exactly these points, diagramming how MMT reconciles public money creation with the circuit of private money creation.

    But why bother citing any of that when three years later, Smith can tweet "Why does MMT think money can only be created by government borrowing, and not by private borrowing?" and folks just accept the premise as though the MMT literature was written in invisible ink.

    ReplyDelete
    Replies
    1. Noah Smith's modus operandi on Twitter is to troll people. If it was just his tweet, I would not rise to the bait. I was more interested in the JW Mason response.

      Delete
  5. The claim that Australia ran a surplus and had economic growth must be clarified. Did they run a surplus in US dollars or Australian dollars? Because that's not the same thing at all. If the Australian govt is running a surplus in Aussie dollars then the aussi non-govt has less and less money each year. Maybe they do their accounting funny and aussi dollar deficit is canceled by a US dollar surplus.. Which goes back and bolsters the view that surpluses precede recessions/depressions.

    It should be common sense that a govt surplus reduces private sector balances, and if you do that for too long, there'll be solvency issues in the private sector. A government very obviously can't run surpluses indefinitely in their own currency.

    ReplyDelete
    Replies
    1. The surpluses were in local currency (Australian dollars, or AUD). They did reduce their debt, but I believe the government ended up buying some form of financial assets to keep up the size of the monetary base and the government bond market. I never covered Australia, so I am unsure about the details.

      Delete
  6. This comment has been removed by the author.

    ReplyDelete
  7. Interesting.. I know his early models weren't double-entry, which he later switched over to... OK, for credit to raise GDP it has to be spent on newly produced goods... Isn't almost all credit card debt used on newly produced goods or services?

    I'm unsure why spending a newly created demand deposit (created simultaneously with a new loan) on existing goods wouldn't raise GDP. I'm spending money I wouldn't have otherwise spent... The main point Keen makes is that for me to take out credit and purchase something, does not require someone else to have forgone consumption in order to lend me the money, so therefore it's brand new purchasing power (unless Keen revised that in the last year or so, I believe that's still his explanation). Why wouldn't that raise GDP?

    A new loan is new money (with corresponding in-sector liability), it's not the loanable funds model that people like Krugman subscribe to. According to Krugman, purchasing power is simply transferred from borrower to spender, whereas Keen says no saver necessary, it's an increase in purchasing. If I understand correctly, the bank of England and the IMF have both published papers within the last couple years confirming loanable funds is incorrect.

    ReplyDelete
    Replies
    1. Asset purchases do not count towards GDP. If it were otherwise, every stock purchase would add to GDP. But this does not add to incomes, other than things like brokerage fees (which are income for the broker). With HFT equity trading, if stock purchases were included, GDP would probably be some really crazy number. Ditto for a home purchase; only the various fees end up as part of GDP. (For an existing house; new home construction is investment and counted in GDP.) This is because GDP equals Gross Domestic Income (although there may be measurement errors).

      If you look at it from the point of view of income taxes, anything that would add to your taxable income would add to GDP; transactions that are not taxable (ignoring capital gains) are not. This rule-of-thumb works fairly well for individuals and small businesses, but you would need to look at the more complicated definitions in the national accounts for more complicated situations.

      A lot of smaller loans are to purchase goods and services (car loans, credit cards, etc.). But the big ticket borrowing is often to buy existing assets - homes, margin loans, credit market leverage, businesses buying other businesses or buying back their own shares. Those asset purchases skew the credit aggregates.

      Delete
    2. "Asset purchases do not count towards GDP."

      They don't, but money doesn't stop at its first use. You have to analyse the entire spending chain from loan creation to money destruction or saving.

      If I purchase a house from somebody with borrowed money, they have sold their house *earlier and for a higher price* than they otherwise would have. That means they have money in their hands. Let's assume the recipients are trading down, so they will then spend the money released, which will cause production due to a demand injection.

      What borrowing does is change the demand sequencing in time, which consequently changes the production sequencing in time.

      Remember that Steve Keen is working in the aggregate and is talking about net increases in borrowing - which is extra borrowing after all the repayments are taken into account. So the buying a house from somebody who then just pays off a loan with the money is dealt with.

      With Steve's model it's the time sequencing in the dynamic process structuring that gets altered. You can't see it in the static links.

      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. I get no warning about this, and only go through my "spambox" infrequently. The best bet it to keep comments short, and if you think the spam filter struck, let me know with a short comment.