Recent Posts

Sunday, October 4, 2015

Net Financial Assets And Equity

An old debate about "Net Financial Assets," a term used in by Modern Monetary Theory (MMT) was reopened by Steve Roth at the article "Where MMT Gets Its Accounting Wrong -- And Right." This generated a lot of comments, and a response by Steve Randy Waldman at interfluidity ("Translating Net Financial Assets"). This also generated discussion at Mike Norman Economics. I largely agree with Steve Waldman's view, but I just want to offer what I think is a more introductory version of what I understand to be the underlying issue. That is, does it make sense to "net out" equity?

A Really Simple Economy

Looking at real world national accounts data and the formal definitions used immediately raises a lot of complications. But we can understand the problem if we look at an imaginary, highly simplified economy,

Let's assume that we are discussing an economy on a small, isolated island that does not trade with the rest of the world (for simplicity). There is a government that issues a fiat currency, and the business sector consists of a single conglomerate -- "Acme Inc,"  -- as well as some family-run firms whose accounting I merge with the operating households.

The government issues paper money and Treasury bills. The amount of these instruments held by everyone else are what MMT refers to as their "Net Financial Assets." It is easily seen that the total (cash) government deficit over a year equals the increase in Net Financial Assets. (This is assuming that the government deficit does not include accrual accounting concepts; the deficit has to correspond to net cash outlays.)

This leads to the common statement within the MMT literature that the government needs to run deficits to accommodate net savings desires of the private sector. If the private sector ("Acme Inc." and households) wants to increase their holdings of government-issued financial instruments, the only way that can be allowed is if the government runs a deficit.

What about private debt?
  • If Acme Inc. issues short-term debt (to households), the increase in household financial assets is matched by the increase in liabilities of Acme Inc.
  • If households issue debt to Acme Inc., the corporation's increase in financial assets is matched by the increase in household liabilities.
In other words, in order for one part of the private sector to increase its holdings of privately issued debt instruments, it requires another part of the private sector to be willing to increase its debt issuance.

The usual state of affairs in modern economies is that most private entities have a rising nominal income, and they correspondingly wish to increase their liquid asset holdings in line with income. (This is known as a stock-flow norm.) In an economy with a single firm, this means that either the firm has to be a net issuer of short-term debt, or the household sector, One could easily see that this could create problems, In a real economy with many firms, there is usually some sector of the economy that is willing to take on short-term debt. However, in a crisis, it is difficult to find credible short-term borrowers, and so the government has to step in.

(At this point, many people would start to raise the issue of the banking system. Although the banking system creates private money in large quantities, it can only do so if there are credible borrowers. Since those borrowers disappear during a crisis, this only has a limited effect on the analysis above. The key advantage of the banking system is "maturity transformation," which reduces the problems created by the mismatch between the large desire to lend at short maturities versus the limited willingness of borrowers to rely on short-term funding,)

What About Equity?

The confusing part of the analysis is the role of equity. In order for private financial assets to net to zero, the shares of Acme Inc. need to be netted out as well. We need to treat the equity as being like a debt. This is implicitly done by the MMT formulation. This is often justified in two ways.
  1. The national accounting treatment essentially treats equity as a form of debt in other contexts.
  2. Businesses do take into account the "cost of equity" in determining their cost of capital. It is treated in practice like any other liability.
Once we realise that equity is supposed to be netted out, the complications raised by Steve Roth regarding valuation changes disappears. An increase in the market value of Acme Inc. shares would increase the value of assets held by households, but it still nets out to zero for the private sector, since there is a corresponding increase in the "liability" for Acme Inc.

However, it appears reasonable to object to this netting out of equity. As someone with a background in the financial markets, I would not normally lump a corporation's equity with its other liabilities when analysing the firm. If we do not allow for netting of equity, the accounting for "net financial assets" would break down. For example, an increase in the value of Acme Inc. shares would increase the value of shares held by households, but there is no longer an increase in the "liabilities" for Acme Inc.

Since the choice of accounting treatment is somewhat arbitrary, you can take either side of MMT's view of net financial assets. Correspondingly, I have no desire to argue about the technical details.

The more important question is whether the MMT treatment is analytically useful. My opinion is that it is in fact useful. In practice, we see that net equity issuance is negative -- that is, more is spent on equity buybacks than is raised via Initial Public Offerings. In other words, the stock market is a net drain of funds, not a source of funds. Since funding is effectively provided by debt, which should be netted, the MMT treatment is closer to reality. Furthermore, although increases in equity valuation do drive balance sheet decisions, they do not have the same impact as cash income flows. For example, a $100 billion increase in wages is going to have a much more significant effect on behaviour than a $100 billion increase in the market value of equities that are ultimately held by households. (The practice of holding assets via insurance companies or pension funds limits the importance of equity market gains for households, as the gains are not visible or necessarily passed through to them.) Correspondingly, I have little objection to a framework that minimises the effect of equity valuation changes.

In summary, this is a debate that revolves around accounting definitions. Although an argument about definitions is of academic interest, the more significant questions revolve around how the private sector behaves. My feeling is that the MMT literature captures the standard patterns of behaviour well, but one must keep in mind the subtleties behind the analysis.

(c) Brian Romanchuk 2015


  1. Awfully useful exposition of the main point of contention. A crib note that I shall print up and keep in my back pocket.

  2. The strangest of which was when MMT defenders stated, contrary to the accusations by critics, that MMT does not claim money is a public/state monopoly. The critics were baffled by this, and rightly so!

    Yet the inimitable and formidable Randall Wray does that again and again.

    For instance, "...economists and policymakers fail to recognise that money is a public monopoly".

    Mostly Wray writes "money" monopoly but sometimes he writes "currency" monopoly, which is confusing. The state is certainly the currency monopolist. But is it the money monopolist? That didn't seem to be adequately resolved.

    1. The state is a currency monopolist within its own currency area. But it is not a money monopolist because anybody can create money (in particular states in other currency areas).

      The monopoly comes because the state in the area you are resident can force you to use its currency to pay taxes and other state debts. Your freedom to use other moneys is limited by the state's legitimate demands.

      So a state can impose taxes and then set a price for work to be done (build a temple for a $1). To the extent that it has coercive power to jail and execute it can get the price it wants for the real output it desires. That's the monopoly.

    2. John,

      Although I consider myself in the MMT camp, I do not necessarily agree with everything that is considered MMT. (It would silly to announce that I have created a new school of thought within economics, but it actually is 90% MMT.) And there are confusing points within various MMT authors' writings.

      In this case, I think we need to replace "money" with "unit of account," (currency might do it, as you suggest, but it might be hard for many people - including myself - to follow the distinction) in Wray's statements, and things are clearer. Since one of the core roles of "money" is to be a unit of account, blurring the two appears defensible.

      Moreover, the "monetary base" is a state monopoly, and since all private money is just a claim on base money, one could extend the "monopoly" usage.

  3. " For example, an increase in the value of Acme Inc. shares would increase the value of shares held by households, but there is no longer an increase in the "liabilities" for Acme Inc."

    No but the 'holding gain' is an unrealised gain. For that gain to be realised somebody else has to take a loss (buying a share for more than the shareholder equity in Acme Inc.).

    Therefore if there is a 'holding gain' on a financial asset, there is always a 'holding loss' somewhere else in the system. This loss is unrealised and, worse, you have no idea what sector that 'unrealised loss' will get realised in, if at all.

    Hence why it is more appropriate to exclude it.

    The holding gains are really only useful to assess the 'wealth effect' - since they end up being collateral for loans.

    1. Holding gains in real estate (K) and equity share positions are not necessarily offset by any holding loss while prices are rising for the asset class.

      Suppose Acme Inc. is publicly traded with a fixed float of 1000 shares. The prior peak price is $10 per share. If the last sale price occurs for a volume of 10 shares at a price of $11 per share then the other 990 shares are market to market at the new high price of $11. The last owner to sell has a liquid gain of $1100. The newest owner is even. All the old owners have a holding gain of $9900 because the shares are trading at an all time high!

    2. Correction, Obviously the last owner to sell in the simplified scenario above takes out liquidity of $110 by trading 10 shares of stock for $11 dollars of money per share of stock.

    3. "Holding gains in real estate (K) and equity share positions are not necessarily offset by any holding loss while prices are rising for the asset class."

      Yes there is. 'Holding gain' isn't real. If everybody tries to sell it would disappear. That means that there is a notional 'holding loss' somewhere in the system from the person that will pay more than the shareholder equity value for the share.

      You just don't know where it is, or over what time it will appear - if at all.

      That $10 the owner gains from selling at the higher price is a $10 loss to the person buying the share because the actual current value of the company hasn't changed. You just didn't know who to book the loss to until the transaction completed.

    4. Think of it this way: each owner of floating value assets has a basis price recorded at historical transaction cost to acquire ownership rights.

      Mark to market accounting means some owners will be in holding gains relative to their basis cost and other owners will be in holding loss relative to their basis cost. The future is not considered by a unit engaging in mark to market accounting to adjust its net worth up or down on a balance sheet.

      I agree with you that liquidity is scarce so markets tend to average down the price of a floating value asset after a period where the price is consistently averaging up on the turnover at the last sale. If someone bought at a low basis price in the distant past and holds then this owner would still be in a long term holding gain while a recent owner with a high basis cost might be in a holding loss. The quality of ownership in an asset class tends to be a feedback into price movement: more greedy owners holding gains tends to drive prices higher; more fearful owners cutting losses tends to drive prices lower.

    5. "Mark to market accounting means some owners will be in holding gains relative to their basis cost and other owners will be in holding loss relative to their basis cost."

      You're missing the point. It's not current holders I'm talking about. It is future owners who will have to transfer some of their wealth to the current owner to 'pay for' the excess price over shareholder equity value. That is the 'notional holding loss' that balances the 'notional holding gain' to zero in the national accounts.

      The national accounts are valued relative to the shareholder equity of the company not the 'base price' of the individuals.

      And you don't know where in the sectors the 'holding loss' is held because you don't know who will be the future owners of shares that crystallise the 'holding gain'.

    6. This is a paper on challenges in the United States which integrate Flow of Funds (FFA) and NIPA to create Integrated Macroeconomic Accounts (IMA) consistent with standards for developing a System of National Accounts (SNA):

      Integrated Macroeconomic Accounts for the United States SNA-USA (52 pages pdf):

      Table A on page 8 shows Accumulation accounts which include Revaluation sub-account. I have not made an exhaustive study of SNA/IMA accounting, however, I am fairly certain that the change in net worth over a period includes fair value accounting for a holding gain or loss on assets with no consideration of potential future gains or losses appearing in the macroeconomic balance sheets. I don't see any notional future loss being used to offset a holding gain in SNA/IMA.

      During the financial crisis the IMA accounts show a loss in nonfinancial assets of nonbank nonfinancial sectors and deficit spending by the federal government which would offset the loss of net worth under IMA accounting by increasing net financial assets of the sector with the government decreasing its net worth in response to the crisis.

  4. Neil,

    That's a great deal clearer than what Wray wrote. I understand and agree completely what you say about the role of taxation, debts and fines, and presumably to non state actors in a court of law.

    Do you have a specific example that would make this obvious? I've never seen any other money being accepted in the UK other than the British pound. You can make a case that superstore vouchers, points on your loyalty card and air miles are effectively money, but I'd say it's a weak case because these are no different to items being sold at a discount or a sale over a period of time to loyal customers. Or criminal gangs preferring US dollars to the national currency in many parts of the world. Or a shop willing to accept a bar of gold bullion or $1000 for a pint of milk, knowing that it's made a killing in the exchange.

    As far as I can see, the only money that isn't actually "monopoly money" is bank money (which is effectively all money), and that's because the state will redeem it at par, thus making it effectively state "monopoly money". So it would seem that all money is monopoly money just as the MMT critics claim, as Randall Wray has written, though confusingly.

    Is there a definitive example in an industrialised capitalist country in which money circulates that is not the state's monopoly money?

  5. Brian: "Moreover, the "monetary base" is a state monopoly, and since all private money is just a claim on base money, one could extend the "monopoly" usage."

    Now why can't Randall Wray and other MMTers be as clear as that? In any case, that does make a great deal of sense because in a crisis you see bank money becoming effectively worthless. Everyone wants the state's monopoly money - the pyramid of money. And that's an MMT concept!

  6. Brian,

    "In other words, in order for one part of the private sector to increase its holdings of privately issued debt instruments, it requires another part of the private sector to be willing to increase its debt issuance."

    I don't think this makes any sense (except in a tautological self referencing way).

    Investors have a portfolio choice. It's not like they are going to die if there's no private debt in markets. It's like I have a choice among fruits: apple, oranges, mangoes but if somehow mango becomes extinct, it's not the end of the world.

    Suppose there's a world with financial assets cash, deposits, private bonds, public sector bonds. Imagine private debt is retired and firms stop issuing debt for some time. Investors will pick among remaining. Or if the debt issuance is reduced, the price mechanism will solve the problem.

    There is of course one way in which this logic works. If you see some circuit literature or some Marc Lavoie's papers ... they argue how if households wish to hold more deposits, firms may be forced to borrow more from banks.

    1. The concern is that entities want to increase their holdings of safe short-term debt, and simultaneously wish to reduce their short-term borrowings. (I did not mention the desire to reduce debt in that sentence, but I believe I mentioned it elsewhere.) Obviously, they could hold government short-term debt, but the supply of such debt is determined by the fiscal position and is essentially fixed in the short term. During a crisis, the supply of "safe" short-term paper is very limited, since a lot of previously "safe" private issuers are no longer viewed as such.

      Outside of a crisis, this is much less of an issue; people are willing to finance practically anything if they think they can beat their benchmarks by 20 basis points. As the circuitist logic points out, a shortfall in demand will either cause firms to borrow more, or the automatic stabilizers will kick in with more government borrowing. This works until the banks refuse to extend further loans to the industrial sector due to lack of confidence that they will get paid back.

    2. My comment was only about private debt not public debt.

      For example, I can make a model where private debt is suddenly retired and nothing goes wrong really.

      The solution Private debt held = 0 will solve.

    3. Hello,

      I would have to think about this, but I would see problems in the real world. In the 1930s, mortgage borrowers could be forced to pay on short notice, and that caused a spiral of bankruptcies. To use Minsky's terminology, not many borrowers are pure "hedge borrowers", there is always an element of needing to borrow to roll over principal.

    4. Brian,

      Your initial point was about forcing of issuance liabilities. In your latest comment, it is about forcing of reduction of liabilities.

      I think the argument doesn't work because in the case of the public sector, one is arguing by consolidating the private sector and can say that if the private sector wants its expenditure less than income, the public sector needs to do the opposite and consequently has increased liabilities. If you argue for the private sector, the other sector consolidated would include the government itself.

    5. I am not directly addressing the "expenditure less than income" argument here. There's a lot of weird possibilities with cash flows that could happen within a model economy with arbitrary cash flow patterns versus what is realistic. I know that there are MMT statements about this; I am not sure whether I agree with them or not. (The MMT statements are based on manipulating national accounts equations, but there is a risk that the interpretation is more complex than suggested. I have not looked at this in enough detail to comment.)

      My concern is that private entities will want to simultaneously raise their short-term assets while at the same time reducing their short-term liabilities. This won't work (unless the government acoomodates it). It's not a particularly deep observation (it's a tautology), but in my view, it shows up in a crisis or "deleveraging" environment. I agree with Mosler's observation that the "demand leakages" from tax-advantaged saving is a driving force behind "secular stagnation," and that can be interpreted as a "mismatch" between the desire to save in the private sector versus the desire to borrow.

  7. In the simplest analysis one considers only the on-balance sheet Treasury liabilities Lt. Then the macroeconomic identity which governs net financial wealth:

    N = K + Ft - Lt

    where net worth N equals the valuation of nonfinancial assets K in money units, plus the stated value of financial assets Ft owned by the Treasury, minus the stated value of liabilities Lt issued by the Treasury. These variables are estimated in the flow of funds and integrated macroeconomic accounts but not in the NIPA accounts.

    The net financial wealth equals (Lt - Ft) because the Treasury guarantees its liabilities Lt and some units owe debt held as financial assets Ft by the Treasury.

    1. Correction, net financial worth:

      N = K + Lt - Ft

      The valuation of nonfinancial assets K has a mark to market gain or loss in a given period which is not captured by the identity. The liabilities Lt and financial assets Ft appear on the Treasury balance sheet which keeps the books for the federal government or monetary Sovereign.

      The off-balance sheet liabilities of the Sovereign government also should be treated like net financial assets (cash flow insurance) by owners of these guarantees so deposits insurance or government sponsored enterprises issue liabilities that are enhanced by the risk free properties of Treasury liabilities.

  8. Brian,

    The link below is to a 28 page paper regarding flow of funds data in Slovenia. I provide the reference because PDF page 2 includes Picture 1 (balance sheet and net items) and Picture 2 (Financial Account Model) which are instructive on the issues net financial assets, but not of equity:

    To my knowledge no one has accurately described the role of equity instruments in national accounts. Regarding my own studies, I have formed concepts that correspond to Picture 1 net financial assets and net worth, but have not considered the definition of own funds, as shown in this reference. I would like to see who owns the equity and net financial assets in the household sector (should be wealthy families and non-profit foundations) and who owes the liabilities to net creditor households (should be working class families and governments acting as financial intermediaries).

    1. I am unsure how accurate the statistics are going to be. We have a handle on what is happening with publicly traded corporations, but what do we really know about private corporations and their valuations? The income tax people hopefully know, but that information is not shared with the national statistical agencies (except with a long lag).

      The big transactions that make a big splash in the national accounts are the various cross-border takeovers. But even then, there is often a lot of debt issued. Since we cannot easily infer who the holders of the takeover target were, it is unclear what the effects are. Also, those takeovers are pro-cyclical - they only occur when times are good. This means that we cannot know much about causality - are times good because of the takeovers, or are the takeovers a lagged response to good times? Based on what I have seen, my bias is that they are just a lagged response to developments elsewhere.

      The other thing to keep in mind with equities is that they are mainly held by either very rich company founders, or via intermediaries (pension funds, insurance funds). A long bull market will result in pension contributions being reduced, which presumably affects consumption behaviour, but that occurs with a very long lag. This lagged impact is very different from debt market behaviour, where borrowing is almost always used to purchase something immediately (and capital gains are relatively small).

    2. I don't regard the statistical accounts as accurate.

      However the structure of the accounts shows how the government may generate net financial assets (per MMT) as the difference between Treasury liabilities issued and Treasury holding of financial assets.

      Those who are concerned about valuations of nonfinancial assets and equity should describe a stock-flow consistent structure, similar to the design of the macroeconomic flow of funds structure. Then everyone can speculate as to how the valuations are determined by feedback between the money and credit markets into the prices of nonfinancial assets and floating valuation financial assets such as equity claims.

  9. A new Table B.1 Derivation of U.S. Net Wealth now appears in the Z.1 flow of funds tables:

    This includes the market value of domestic corporations and the valuation of nonfinancial assets across all domestic sectors, and subtracts net financial claims in the rest of the world.

    Notice the household net worth (Line 34) is generally greater than the national net wealth (Lines 1 = 31).

  10. This comment has been removed by a blog administrator.

  11. This comment has been removed by a blog administrator.

  12. This comment has been removed by a blog administrator.

  13. This comment has been removed by a blog administrator.

  14. This comment has been removed by a blog administrator.

  15. This comment has been removed by a blog administrator.

  16. This comment has been removed by a blog administrator.

  17. This comment has been removed by a blog administrator.

  18. This comment has been removed by a blog administrator.

  19. This comment has been removed by a blog administrator.

  20. This comment has been removed by a blog administrator.

  21. This comment has been removed by a blog administrator.


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.)