Rob Parenteau gets sectoral balances right

Note: This post will only format decently in a browser window opened very wide. The equations will probably be garbled in an RSS reader.

First and foremost, I owe Rob Parenteau a big apology. Parenteau is the originator and first user of the clever term “Austerian”, which I erroneously attributed to Mark Thoma. Thoma never claimed parentage. I first encountered the term on his blog and a quick Google search turned up no antecedents, so I went with that. But Google does not index everything. I apologize for the error, and thank Marshall Auerbach who first pointed it out to me.

Parenteau’s contributions go far beyond a catchy neologism, however. I recommend his most recent post at Naked Capitalism, which is the best use of the “sectoral balances approach” to economic analysis that I have seen in the blogosphere.

The “sectoral balances approach” (frequently attributed to Wynne Godley) decomposes financial stocks and flows by virtue of a tautology. Every financial asset is also some entity’s liability. The sum of all financial positions is by definition zero. So we can write:


Suppose that, quite arbitrarily, we divide the world into a “foreign” and a “domestic” sector. Then we have:


Suppose that, again arbitrarily, we decompose the domestic economy into a public and private sector:


Substituting into our previous expression, we get


We can also write this in terms of changes or flows. Since the sum above must always be zero, it must be true that any changes in one sector are balanced by changes in another:


Two of the flows in the equation above have conventional names, so we can rewrite:




This decomposition has been quite prominent in the blogosphere. I first encountered it in conversation with the always excellent Winterspeak, and associate it with the “Modern Monetary Theorists” or “chartalists”. But it’s been used widely, very recently for example by Martin Wolf.

The usual argument goes something like this: In the aftermath of a terrible credit bubble, in most countries, the private sector is desperate to “delever”, or reduce its indebtedness, which is equivalent to increasing its net financial position. As a matter of pure arithmetic, equation 8 must always be in balance. If the private sector of a country is to force the left-hand term positive, the country must either run a current account surplus (e.g. by exporting more than it imports) or else its government must run a deficit. Some countries may “export their way” to financial health, but not all can, since every current account surplus must be matched by a deficit elsewhere. If we put “beggar thy neighbor” strategies aside and set the current account to zero, any improvement in the financial position of the private sector must be offset by a deficit of the public sector.

This is true by definition. Once the terms have been defined, there is nothing to argue about. If we want the financial position of the private sector to improve (defined as increasing total financial assets less liabilities), and we consider a country whose external account is in balance or deficit, then the public sector must run a deficit.

However, a thing can be true but still misleading. The catch is an assumption, that an increase in the net financial position of the private sector is a good thing, something that we should encourage or at least accommodate. This is where Parenteau is great. He decomposes the domestic private sector into a household and business sector:


(Note that “business” here means any non-household private entity that could have a financial position. It would include, for example, non-profit organizations.)


ΔNET_HOUSEHOLD_FINANCIAL_POSITION is just net household financial income.

NET_BUSINESS_FINANCIAL_POSITION is, by definition, all business financial assets minus all business liabilities (including shareholder equity). On a business’ balance sheet, “all business liabilities (including shareholder equity)” is necessarily the same as “total business assets”. So we can write:


Now use our new definitions to rewrite equation [10]:


Now we can tell what I think is a much more informative story. It is not the “private sector” whose financial position needs to improve. Businesses exist to increase the value of their liabilities to shareholders and creditors. They do not “delever” by reducing the sum of those liabilities. “Leverage” properly refers to the ratio between different sorts of liabilities, debt versus equity, not the total quantity of claims. In a good economy, the financial indebtedness of business entities will be increasing, as the value their real assets grows! Growth in the “net private sector financial position” could come from an increase in household income (yay!) or a decrease in the value of real business assets (yuk!). We certainly shouldn’t make policy decisions based on promoting or accommodating such an ambiguous outcome. Instead, we should craft our policies to be consistent with what we actually want, which is household financial income. (Note that this analysis necessarily excludes nonfinancial income, such as unrealized gains or losses on the value of a home.)

Reviewing equation [17], there are three ways a nation can improve the financial positions of its household sector. It may (i) run a current account surplus, usually by exporting more than it imports; (ii) have the government run a deficit, improving household financial position by having the government run a deficit, or (iii) increase the value of business nonfinancial assets. Approach (i) can’t work for everyone, of course. Assuming external balance, it is obvious (at least to me) that approach (iii) is ideal. Parenteau, I think, agrees:

Remember the global savings glut you keep hearing about from Greenspan, Bernanke, Rajan, and other prominent neoliberals? Turns out it is a corporate savings glut. There is a glut of profits, and these profits are not being reinvested in tangible plant and equipment. Companies, ostensibly under the guise of maximizing shareholder value, would much rather pay their inside looters in management handsome bonuses, or pay out special dividends to their shareholders, or play casino games with all sorts of financial engineering thrown into obfuscate the nature of their financial speculation, than fulfill the traditional roles of capitalist, which is to use profits as both a signal to invest in expanding the productive capital stock, as well as a source of financing the widening and upgrading of productive plant and equipment.

What we have here, in other words, is a failure of capitalists to act as capitalists. Into the breach, fiscal policy must step unless we wish to court the types of debt deflation dynamics we were flirting with between September 2008 and March 2009. So rather than marching to Austeria, we need to kill two birds with one stone, and set fiscal policy more explicitly to the task of incentivizing the reinvestment of profits in tangible capital equipment.

So what is the role of approach (ii), which stimulus proponents and MMT-ers frequently advocate? Note how Parenteau phrases things: because “capitalists [fail] to act as capitalists”, because businesses are not increasing the value of their nonfinancial assets, fiscal policy must be employed to avoid “debt deflation dynamics”. Here we reach the formal limits of the sectoral balance approach. This style of analysis gives us no insight into the dynamics or distribution of financial positions within any of the categories we have carved out.

Nevertheless, consider the following (counterfactual) thought experiment. Imagine that the NET_HOUSEHOLD_FINANCIAL_POSITION is negative, and that people go nuts in a harmful way when they are formally insolvent. Suppose also that the current account cannot be brought to surplus, and that businesses cannot expand the value of their nonfinancial assets in a short time frame. Under these conditions, by running a deficit, government could create financial income for households until their net financial position turns positive and people stop behaving like antisocial lunatics. In this scenario, fiscal policy does nothing to change the real asset position of the economy. But by shifting around financial assets and liabilities, government alters the behavior of agents in the economy in a manner that improves future performance, increasing overall wealth.

In real economies, people may well behave in ways that are harmful to the economy when their financial positions are very tenuous, although their actions are more likely caused by illiquidity than lunacy. But in real economies, some people have strong financial positions while others have weak financial positions, and the sort of intervention described above would be useless if the income created by a stimulus went primarily to households that were not financially stressed. Government funds spent purchasing goods and services from existing firms, or deficits created by income or payroll tax cuts, go first to people who are already employed, or who already have financial claims on businesses, and these may not be the most stressed groups. Designing a “good” stimulus where the object is to alter the character of real behavior by shifting financial variables is well beyond the scope of this post, but it would necessarily involve distributional questions and complex behavioral assumptions. If you target a stimulus to the deeply indebted, you may improve their behavior, but damage the behavior of others who feel aggrieved that prudence went unrewarded. If it was me, I’d make flat transfers unrelated to income or employment status, so that on the one hand the program seems “fair” — the prudent benefit along with the bankrupt — yet on the other hand it is guaranteed to improve the financial position of even the worst-situated households.

What about approach (iii)? What could cause an increase in the value of business nonfinancial assets, improving household financial positions? Fundamentally, there are two ways: Businesses could borrow or use their own cash to purchase real assets from the household and government sectors (holding the public sector deficit constant), or else the value of existing business nonfinancial assets can somehow be made to increase. Parenteau suggests policies that would push businesses to purchase real assets. But note that any sort of increase in the valuation of business nonfinancial assets, including intangible assets, would be sufficient to improve the household-sector financial balance. That would include events as insubstantial as a pure inflation, but also real improvements in business productivity. Again, looking beyond where sectoral balances can take us, distribution matters. If “debt deflation dynamics” occurs primarily through households whose weak financial positions include few claims on businesses, then increasing the value of business nonfinancial assets might not help very much.

p.s. Edward Harrison offered a response to Parenteau’s piece that is very much worth reading. In particular, he focuses on the quality of business investment, a topic about which sectoral balance decomposition can tell us very little. Mechanically, low quality investment should improve the valuation of business nonfinancial assets less than high quality investment, and should therefore exert a drag on household financial balances. Harrison uses an Austrian (though not Austerian!) perspective to suggest that stimulus may reduce the quality of business investing, implying a trade-off between approaches (ii) and (iii) above.

[MMT Note] Agree or disagree, the “MMTers” are among the most interesting and provocative thinkers in the economics blogosphere. In addition to Winterspeak, I’d include Bill Mitchell, Warren Mosler, Scott Fullwiler (who occasionally writes at Economic Perspectives from Kansas City), Marshall Auerbach, and perhaps Parenteau himself in this group. I agree with much but not all of what the MMTers have to say. I have learned profoundly much from disagreeing and squabbling with them. I do hope that Kartik Athreya will someday have the pleasure.

Update 2010-07-01, 6:40 am EDT: For reasons I do not understand (my big fat finger?), this post “disappeared” for a few hours. It reverted from “published” to “draft” in WordPress. The post is back, and the comments seem to be intact, but my apologies to all for the disappearance!


101 Responses to “Rob Parenteau gets sectoral balances right”

  1. MarkS writes:

    Anyone read TPC at Pragmatic Capitalism? I believe he is a fund manager by background so he brings an interesting perspective while mixing a bit of Austrian economics in with his MMT which has made for enlightening reading. Highly recommended.

  2. […] Rob Parenteau gets sectoral balances right Steve Waldman […]

  3. […] Rob Parenteau gets sectoral balances right Steve Waldman […]

  4. vlade writes:

    The main problem I have with MMTs or chartalists is that as far as I can see they take as given that upping the stimulus by any means (a.k.a. “printing”) has no side-effects via human perception of actions.

    The fact is that majority of population does not understand a lot about economy (nor cares to). Thus gov’t going into debt to create stimulus will always be associated with how they view and feel about debt on personal level. If you try to argue that gov’t can print as much money as it wants to, so can never have debt, then it will raise the spectre of inflation (and suspiction of fiat money) – and then even those with the most to gain from inflation (basically just about everyone who has financial liabilities vs. real assets) will argue against it. It’s instinctive, it’s human, it has nothing to do with what is right or wrong. But the resulting actions do have real impact, so you get austerity backlashes at the worst of times, inflation scares in deflationary situations etc.

    Economy is just and extension of politics – not the other way around.

  5. RueTheDay writes:

    I always end up going back to Keynes. In the General Theory, Keynes identified investment, rather than consumption, as the key unstable variable. He proposed splitting the government budget into an operating budget, which would always be in balance, and a capital budget, which would countercyclically swing into deficit when private investment (ΔBUSINESS_NONFINANCIAL_ASSETS above) falls. Keynes wasn’t far from what is described in your post, and the Hicks-Hansen and Phillips Curve caricatures that became popular had little to do with anything Keynes actually wrote.

    On this: “Businesses exist to increase the value of their liabilities to shareholders and creditors. They do not ‘delever’ by reducing the sum of those liabilities.”

    I’m not so sure. Richard Koo, in his book Balance Sheet Recessions, makes a pretty compelling case that in the aftermath of a burst asset bubble, overindebted businesses fundamentally alter their behavior. Instead of profit-maximizing behavior (more properly, maximizing the NPV of all future profits) they begin engaging in debt-minimizing behavior. Investment is halted, operating expenses are slashed, and all cash flow is diverted to extinguishing debt. While this may not be rational behavior in models that assume firms to be perpetual going concerns, it’s perfectly rational in a world where the future has suddenly become more uncertain and managers see excess debt as a threat to the firm’s very existence.

  6. anon writes:

    The key take away from the Parenteau private sector decomposition is that decreasing the business sector financial surplus requires increasing business net real investment.

    If you put it all simply as an equation of surplus/deficit balances, then:

    H + B + G = CA

    (the sum of household, business, and government balances equals the current account)

    In accordance with this equation, the conclusion from Parenteau (and yourself) is that you can increase the household surplus by decreasing the business surplus (increasing real investment), increasing a government deficit (decreasing a surplus), or decreasing a current account deficit (increasing a surplus).

    The problem is that this is not a menu of independent choices. E.g. you can’t dial up a current account adjustment and expect to be mirrored exclusively in H: B and G may also be affected by such a choice. So this is a menu of non-independent (and in that sense false) “choices” about how to get a predictable effect for the household balance.

    That is what I questioned at Naked Capitalism.

    P.S. you might want to cite your MMT sources in a way that distinguishes between the actual professional originators of the theory and the commentators on it.

  7. Winterspeak writes:

    srw: businesses are pro-cyclical (banks especially so). They will not invest on real assets unless they forecast rising downstream demand. The lever to push is after-tax household income. It is easy to push.

  8. anon writes:


    At Naked Capitalism, Rob Parenteau generously responded to my request for his choice of an antonym for “Austerian”:

    “Ok, I will give it up.

    The opposite of the Austerians is … the Libertinians.

    If you use it, please be smart enough not to go attributing it to someone who did not come up with it first, ok?”

    I promised him I would do that. This is my first proactive good work in that regard. It fits nicely with your apology, I think.

  9. RTD — I think there is a terminological issue that makes these issues hard to talk about.

    Firms may certainly want to reduce indebtedness, meaning alter their capital structure to reduce the financial risk associated with having fixed-value debt against volatile (and in the worst case, insufficient) assets.

    But the phrase “liabilities to shareholders and creditors” is intended to include the claims of common stockholders. Holding firm assets constant, pretty much any firm, at least to the degree that it is guided by financial motivations, would prefer to expand “liabilities to shareholders and creditors”. To do so it is equivalent to increasing the value of firm assets and shareholder wealth (either directly, or the option-value of equity if a firm is insolvent).

    Now to say that firms would like to do this doesn’t mean that a given firm will not work actively to reduce the size of its balance sheet by divesting itself of assets, i.e. by delevering in the typical sense. Since liabilities have costs to shareholders, in terms of interest and financial risk, individual firms may wish to sell assets and eliminate liabilities. If shareholders believe that a firm cannot make value-maximizing use of firm assets, they may reduce balance sheet size via dividends or buybacks, increasing leverage while still reducing balance sheet size. (Note that, at least in theory, the question of managing conventional leverage is orthogonal to the question of managing balance sheet size, or “liabilities to shareholders and creditors”.)

    Regardless of the circumstances of a particular firm, I think it is fair to say that if we treat businesses as a consolidated sector, the economically desirable outcome is for business-sector asset values to expand in real terms. Holding real assets of the business sector constant, this translates to an increase in productivity. (Uh oh, I don’t mean “productivity” in the conventional output-per-labor-hour sense. I should use “efficiency” or “total factor productivity”. But I think you know what I mean.) Even if we are not holding the real assets of the business sector constant, if we are after measurable economic growth (which we might not be, we might be better off seeking nebulous but real household or social utility), we tend to prefer firms purchase and exploit assets real assets owned by households and governments, on the theory that firms are most likely to efficiently exploit real assets to produce measurable economic value. I don’t mean to argue too strongly for that last point — the fact that individual and social benefit derived from household and government assets is hard to measure doesn’t make it unreal, and I am definitely opposed to cartoonishly imagining that business always makes the best use of things so we should privative everything. But even making reasonable allowances for the fact that households and governments hold and should hold lots of real assets, economic growth is almost always accompanied by growth of the asset portfolio of the business sector. Shrinkage of the real asset portfolio of businesses in aggregate is not a socially desirable outcome that we should strive to accommodate.

    But I know that such a shrinkage is not really what you mean to propose with respect to Koo-style deleveraging. (Note: I’ve read some articles by Koo, but not his book.) What you are after there is something like this: (1) Firms are overindebted, in the sense of having fragile and costly capital structures; (2) To manage this, firms individually are inclined to divest themselves of assets and shrink, which composes to the socially undesirable outcome of shrinking the consolidated business sector real asset portfolio; (3) government spending can help prevent this, by generating retained earnings for firms that leave them with more a lower debt:equity ratio and therefore fewer financial fragility costs.

    That is all true, but it’s important to be clear about some things:

    First, the capital structure of firms and of business in aggregate is invisible to sectoral balances analysis. It is an abuse to suggest that the government spending “arithmetically” has any tendency whatsoever to “delever” the business sector. Any increase in the financial asset position of firms occasioned by government spending is precisely matched by an increase in some financial liability (perhaps common equity). Yes, as a side effect of a plausible path for government spending, firm leverage is likely to be reduced. But arithmetically, governments could transfer funds to households who then lend cash to firms, leaving businesses more leveraged rather than less. Again, that’s not so plausible, but don’t tell me it can’t happen as a matter of double-entry bookkeeping. Ultimately what double entry bookkeeping requires is that any deterioration in the financial position of the private sector is matched by financial income to households (domestic or foreign), because households own businesses via financial claims.

    Second, given that government-spending is not necessarily related to deleveraging of the business sector in a capital structure sense, we should consider a variety of alternatives if the business sector does need a more equity-heavy capital structure. Yes, it’s plausible that government spending could create retained earnings that would help delever firms. However, direct government spending to firms has a lot of questionable distributional and legitimacy issues, and arguably reduces the efficiency of firm production of “purchases” become poorly-disciplined excuses for macro-motivated (or corrupt) transfers. We would consider governments just giving money to only some firms to improve their balance sheets entirely illegitimate. We have to be careful that government spending into businesses does not amount to the same thing. Further, there are lots of alternative means of helping delever firms in a capital structure sense. The structure of claims can be renegotiated between debt and equity holders internally. Governments could help achieve this in a deficit-neutral way by finding ways to encourage debt-to-equity conversion without the disruption of a full bankruptcy. Governments could also equalize the tax treatment of interest and dividends, which at the margin would cause firm claimants to negotiate amongst themselves to increase the proportion of equity financing. (Tax-treatment equalization could either increase or diminish public deficits, depending on how it is done.) Again, I’m not arguing that using government spending in hope that it will help firms achieve more robust capital structures is necessarily a bad idea. But it is a hazardous approach, to which there are alternatives, and we should evaluate the trade-offs carefully rather than imagining we are compelled by some sort of arithmetic.

    Re Keynes and government capex — as an artifact of how I have decomposed things, direct government spending on real assets does not alter BUSINESS_NONFINANCIAL_ASSETS. But if we were to decompose the government deficit net capital expenditures and net operating expenses, we would find that increasing either of them would create household financial income, and we might argue that net government capex is “better” as a matter of sustaining the real economy over time (a dynamic which, again, is invisible to sectoral balance analysis). In other words, I like Keynes approach. Maybe it would be useful to do that decomposition, and then we could talk about the trade-offs, both in terms of what is rapidly achievable and in terms of the real-economic efficiency of expenditures, of trying to increase BUSINESS_NONFINANCIAL_ASSETS or GOVT_CAPITAL_ASSETS. That might be a nice way of framing things. In other words, I think I agree with ol’ Maynard and I suspect with you as well on this.

  10. winterspeak — I agree, which is why I am practically often on the same page as MMT-ers, that is I like very broad ongoing transfers to households.

    But it is not sectoral balances arithmetic that gets us there. Rather we are making conjectures about how such transfers will affect household nonfinancial expenditures (whether directly or as a side effect of improved balance sheets, household will spend more), and conjectures about how businesses make real investment decisions. I think we agree on both those conjectures, leading to pretty similar policy preferences.

  11. anon — I agree, and it is important to emphasize that choices are not independent. Accounting identities are useful for pointing out what must be true about the results of any course of action, and for calling into question proposals that purport to leave us in an arithmetically implausible state of the world, however desirable that may be. Some bits of an accounting identity may be plausible “exogenous” — we probably can soup up the government budget deficit if we want (as long as we consider politics exogenous). But we cannot determine where among H, B, and CA in your identity that change will end up without a lot more careful reasoning. We can talk about what we might hope for, however, what sort of balance we might strive to accommodate, although our attempts to tweak things in that direction could always backfire if we get things wrong.

    Re attribution, it’s a hard problem. I have read some academic chartalist work, and should probably attribute, but then I don’t have it remotely handy and I don’t want to have to do a formal lit-review every time I write a blog post. I do my best to attribute where I actually learned stuff from, and I do take some time (and did take some time, both before and after being told of Parenteau’s primacy) to try to track down people to whom credit is due. But I can’t promise that I won’t sometimes fail to give credit where it’s due, or even do very much searching beyond what is easily discoverable on-line, when producing a blog post. I can promise to apologize when I get stuff wrong.

    In this case, though, my initial sin was exacerbated by being slow to respond. I was informed quickly by Marshall Auerbach of my mistake and should have apologized sooner, but I was still unable to find Parenteau’s earlier use. It is to my discredit that I didn’t do what Barry Ritholtz eventually did, and track down the origin by e-mailing Parenteau, Thoma, and Auerbach.

  12. vlade — I agree entirely. I think the MMT-ers offer some really important insights (especially re how banking and monetary systems might and sometimes do operate), and really useful tools for analysis.

    But a good analytical perspective does not eliminate the complexity of a real-world in which economic activity is social and political. It is an error to use MMT as an excuse to cast that complexity aside, but also an error not to make use of the insights MMT has to offer. It’s easy to get that sort of balance wrong, and a lot of us do.

  13. MarkS — Had encountered the site, but have now added to my reader, which ensures I’ll at least consistently sample the material…

  14. anon writes:

    I was actually referring to your ‘MMT note’, not the Parenteau attribution. No bother.

  15. Indy writes:

    I’d like to see the ceteris-paribus effect on equation 17 of high current and expected inflation / devaluation. It seems to me that some of the variables are “real” and not much susceptible to nominal events, but that others life deficits are more sensitive.

  16. VJK writes:


    I believe MMT offers some interesting insights into money creation process and central bank operations. I am not so sure about their obsession with accounting identities, identities being what they are — mere tautologies — cannot provide any additional insight ‘by definition’ ;).

    Talking about the ‘private sector’ as a uniform blob moving in synchrony, but in the opposite direction, with government deficit/surplus is rather meaningless as such sectoral aggregation loses all the information about the insides of the sector since nominally the private sector consolidated balance sheet nets to zero. It is gratifying, therefore, to see that some MMT’ers are listening to the criticism and attempting to dis-aggregate the blob.

    Re. you posting:

    1. I am not sure about the derivation from (11) onward. ‘Business financial assets’ is a term I did not ever see on business balance sheets. If by BFA you mean ‘government paper’, i.e. cash and its equivalents, then expression(17) ascribes erroneously all government paper ownership real or perceived benefit to the household sector. Since the BFA meaning is not clear, I am not sure about the export/import component contribution either.

    2. Even with the attempted dis-aggregation of the private sector, the picture is still not very useful. In my opinion, further disaggregation into non-financial, financial and foreign sectors, along with income distribution inside each sector should happen in order to try and arrive at any conclusion. The original pub/priv sector aggregation is simply too crude a tool to offer any sensible economic policy recipe.

    3. The MMT concentartion on the accounting side de-empasizes the real-world significance of production, capital growth, income distribution, employment *structural* changes due to globalization, etc.

    4. There are some not quite clearly explained practical and theoretical issues in the money creation dynamics as seen through the MMT lens, e.g.:

    What economic incentive do banks have for paying interest deposits as according to MMT there is no value in banks having deposits.

    With all the emphasis on the accounting, one would expect that MMT can offer an explanation of the monetizing profits conundrum (I posted about it here: ). Such an explanation would have interesting implications for government deficits for example. None was offered so far.

  17. Ramanan writes:

    A bit of history.

    The identity and the usage to reconstruct Macroeconomics is due to Wynne Godley. Here is an excerpt from his master-masterpiece Monetary Economics co-authored with Marc Lavoie.

    In 1970 I moved to Cambridge, where, with Francis Cripps, I founded the Cambridge Economic Policy Group (CEPG). I remember a damascene moment when, in early 1974 (after playing round with concepts devised in conversation with Nicky Kaldor and Robert Neild), I first apprehended the strategic importance of the accounting identity which says that, measured at current prices, the government’s budget deficit less the current account deficit is equal, by definition, to private saving net of investment. Having always thought of the balance of trade as something which could only be analysed in terms of income and price elasticities together with real output movements at home and abroad, it came as a shock to discover that if only one knows what the budget deficit and private net saving are, it follows from that information alone, without any qualification whatever, exactly what the balance of payments must be. Francis Cripps and I set out the significance of this identity as a logical framework both for modelling the economy and for the formulation of policy in the London and Cambridge Economic Bulletin in January 1974(Godley and Cripps 1974).We correctly predicted that the Heath Barber boom would go bust later in the year at a time when the National Institute was in full support of government policy and the London Business School (i.e. Jim Ball and Terry Burns) were conditionally recommending further reflation!We also predicted that inflation could exceed 20% if the unfortunate threshold(wage indexation) scheme really got going interactively. This was important because it was later claimed that inflation (which eventually reached 26%) was the consequence of the previous rise in the ‘money supply’, while others put it down to the rising pressure of demand the previous year.

    According to Marc Lavoie, Wynne Godley thought it (the identity) was a “great revelation”.

    A brief sketch (actually fairly detailed) of how money is created and how various sectors of the economy interact is given in Godley’s text “Macroeconomics”, written with Francis Cripps. The book is out of print – although I managed to get an (unused) copy recently :-)

  18. passing through writes:

    In the future it’ll be obvious that to do “macroec” correctly you need more network theory (and less macroec); you need centrality, power, and all that. Most schools of macroeconomic thought differ in the specific ways they collapse the network structure of economic activity into scalar metrics. The schools mostly talk past each other because they lack a rigorous conceptual vocabulary which which to justify their metrics’ informativeness (or even to elucidate precisely what it signifies without resorting to heuristics).

    MMT-esque accounting-identity-driven analyses are important insofar as they allow one to reconstruct by hand something of the actual network structure of the economy, and then use the structural insight to obtain further qualitative insights.

    Network-theory inspired ideas have slow uptake in the mainstream economics world because most mainstream academics don’t know the requisite math, are resistant to developments that would leave them having to start over again, and tend to think it’s just reinvention of input-output matrices and thus uninteresting.

    Work in this direction keeps getting done in fringes — @ santa fe, by econophysicists, etc. — but remains generally neglected. If you want to develop further insight into the structure of sectors you’ll be best served learning the basics of network theory; it’s a little like being leeuwenhoek.

  19. NKlein1553 writes:


    I don’t think the MMT position is that “there is no value in banks having deposits,” only that banks do not require deposits to initiate a loan. Even under a zero reserve requirement regime such as exists in Canada banks are still required by law to have a positive reserve balance position at the end of a given accounting period. If a bank has a negative reserve position, then it must borrow from the Federal Reserve discount window or from other banks. In such cases banks have to pay a penalty rate; either the discount rate in the case of the former or the federal funds rate in the case of the latter. So deposits do have value for banks, if only to avoid having to pay the aforementioned penalty rates (although there are other reasons as well).

  20. RueTheDay writes:

    Steve – Thanks for taking the time to provide a long and thoughtful reply. It will take me some time to fully digest it.

    You are absolutely correct in pointing out that total liabilities here includes liabilities to common shareholders. This is something I often find myself stumbling over. I recall an article I read awhile back in one of the Post-Keynesian journals or websites (I can’t for the life of me recall the author) where the point was made that – the difference between the return on real capital and the interest rate on financial capital is the root driver of economic activity, indeed the very existence of modern capitalism depends on the former being greater than the latter. This is what Keynes was getting at when he talked about the marginal efficiency of capital, the supply price of capital assets, and the inducement to invest. I sometimes find myself conflating capital with equity, when it fact the capital in this case shows up on the balance sheet under assets. You make a good point about terminology, and in this case we’re combining economics and accounting, which use the same terms to mean different things.

    BTW, I seem to recall you were pursuing a PhD in Finance. How’s that going? I start a part-time MS in Finance next month. I need to get out of consulting as 12+ years of living out of airports and hotels has really taken its toll.

  21. RueTheDay writes:

    NKlein1553 – This is one of the issues I have with the MMT account of banking.

    MMT advocates emphasize the fact that the textbook analysis of the money multiplier is exactly backwards – banks make loans first and then find the the required reserves to support them. This part is correct. They then go on to say that the Fed is thus forced to supply as much in the way of reserves as the banks demand in order to maintain a stable interest rate target. This is strictly true also, at least in the short run, but it hinges on the assumption of the Fed maintaining the interest rate target. In the medium to long run, the Fed Funds rate is anything but fixed, so no, the Fed is not forced to provide as many reserves as the banks demand over that time horizon.

  22. zanon writes:

    NKlein1553: My Gods. You are very quick learner.

    VJK: Banks seek deposit because it lower their cost of capital. It is a business model banks might pick, and if they do pick it, this is why. Banks can and do pick other model as well where they do not seek deposits. The “penalty” associated at IB market, and even more at discount window that NKlein points to is part of this cost structure.

    RTD: Fed announces target FFR because that it was it does. Then it intervenes in ON IB market to try and hit target. Usually it does fine, on occasion it fails. Nevertheless, FFR target is always set by Fed, and as it is a (ST) market rate, Fed either intervenes to hit its announced target (and therefore execute its policy) or it does not.

    It is nonsensical to me your statement “In the medium to long run, the Fed Funds rate is anything but fixed, so no, the Fed is not forced to provide as many reserves as the banks demand over that time horizon.” There is no “long time horizon” FFR, just series of short term — indeed, overnight — actual IB rates that were either close to Fed target (usually) or not.

  23. […] interfluidity » Rob Parenteau gets sectoral balances right […]

  24. RueTheDay writes:

    zanon – Many MMTers (the horizontalists/accommodationists) assert that the money supply is purely a function of the demand for credit, that banks will expand the money supply to whatever extent is required to satisfy that demand, and that the Fed has no choice but to provide the required reserves to support it, if it is to maintain the Fed Funds Rate. My point was that that last “if” is an enormous “if”, and that while the Fed Funds target is essentially fixed by the Fed over the short run, that the Fed absolutely changes it over the medium to long run.

  25. zanon writes:

    RTD: There is no “medium or long run” FFR. There is just ON IB rate, which happens every day anew. Sure, Fed changes target but so what?

    And what happens if Fed does not bother to try and hit own target? It shuts up business and puts up “I go fishing” sign? The whole point of the Fed is to set target and intervene in ON IB market so it hits it. Your point is “well suppose Fed stops doing it’s entire point?!” It might be “enormous “IF”” to you but it still seems pointless pedantry to me.

    I also reject strict horizontalism as being obviously nonsense. But bank credit expansion does NOT mean system needs additional reserves from Fed necessarily. Bank credit expansion is independent of reserve level. Still, it is true that reserve level can fall between requirement for technical reason, and at this point Fed either needs to supply those levels, or reduce requirements, else ON IB market will break down, and therefore the Fed’s ability to actually execute it’s own policy.

  26. VJK writes:


    Banks seek deposit because it lower their cost of capital

    A quick look at the current rates show:

    Sallie Mae Murray, UT offers 1.34% savings account

    FFR is currently at 0.25%
    discount window is at 0.75%

    For Sallie Mae, taking a savings account deposit is a losing proposition since an interbank loan would be cheaper and event the bearing stigma discount window loan would be cheaper as well.

    SO my question remains: what’s the point of offering a checking account in this case ?

    Even with some checking accounts, e.g. Charles Schwab Bank or Evergreen, the rate is about 0.5% which is worse than the overnight settlement at 0.25%.

  27. Zanon writes:

    Vjk: you are looking at wrong number. Cost of capital is marginal cost of bank raising more equity and thus being Able to make more loan. This is not ffr nor is it what bank offers on checking accounts because banks do not loan out deposits.

  28. Steve,

    Thanks for the kind words (we MMTers don’t get many of those! :)
    I’ll deal solely with the issue of corporate surpluses which Rob raised in this excellent piece. He and I (and others, such as Marc Lavoie and Mario Seccareccia) have discussed this at length. Of course, one doesn’t want to punish corporations for engaging in precautionary saving (another reason why proactive fiscal policy directed toward full employment helps so much). But there is another consideration: I would argue that the build-up of corporate cash flow has facilitated the increasing financialisation of our economy. If the government taxes excess corporate savings, it means there are fewer corresponding opportunities for corporate financial engineering, control frauds, etc., and therefore greater financial stability as you have an economy less prone to financialisation. That’s an unalloyed social good.

    In effect, this becomes a tax aimed explicitly at the corporate rentiers who are not reinvesting their super profits in tangible capital equipment, except in tech/telecom bubbles, or in Chinese malinvestment schemes, etc. And it serves an ideological purpose of a) forcing nonfinancial capitalists to, well, be capitalists (as Rob suggested in his piece), not speculators, and b) ties the deficit reduction initiatives, which, as we have argued many times in the past, are insane and suicidal, but are nonetheless being carried out, to making the rentiers pay their “fair share.”

    Now perhaps the best route isn’t taxing corporate saving. I’m open on this point and welcome an exchange of views. In addition to the tax incidence problem of merely shifting it forward to consuming households, there is the obvious fact that it is not because you really want the government to increase taxes nowadays (since it doesn’t need it) but merely to get business to spend. The problem, however, is that investment spending is not very much affected by corporate taxation. It is mostly guided by expected profit/expected demand. So if I were to revise the piece I wrote below, I would probably amend and urge taxation tax the liquid assets that they purchase/hold rather than the flow of corporate saving per se.

  29. NKlein1553 writes:

    Thank you Zanon. It’s only because I’ve had good teachers. Reading the archives of Winterspeak’s, Professor Mitchell’s, and Mr. Mosler’s blogs, the comments that appear on these blogs from the likes of you, JKH, and Ramanan, as well as a few CFEPS working papers these past few months has been far more educational than all the economics courses I took over four years as an undergraduate combined. And I went to a very good University with some big name professors to boot. Basically, Kartik Athreya is an idiot. What Steve writes here on any given day is a thousand times better than anything you’re likely to see coming out of the universities.

  30. VJK writes:

    Zanon @ 27:

    banks do not loan out deposits

    I did not say they do.

    Reserves are part of bank capital though, and you did not offer any numbers to show that the deposits are a cheaper source of capital than other options.

  31. NKlein1553 writes:


    I suspect it has more to do with the stigma attached to borrowing directly from the Fed and other such intangible considerations than anything else. Banks have to be concerned with their reputations, and any bank seen to be relying extensively on borrowing from the Fed and other banks to cover their legally mandated reserve requirements will be seen as a risky place to invest. This could be completely wrong though. I do agree that a more detailed explanation of what kinds of assets count toward banks’ capital requirements and the relative cost of acquiring these assets might be informative. Professor Mitchell had a nice discussion of this coupled with an informative chart listing asset classes and their relative weights here:

  32. NKlein1553 writes:

    The chart is about a quarter of the way into the blog post.

  33. NKlein1553 writes:

    Also, in your statement “Reserves are part of bank capital,” I’m not sure which bank you are referring to. I’m pretty sure reserve balances held by commercial banks at their federal reserve member banks count as assets of the commercial bank and liabilities of the Fed. Not sure what the relationship is between reserves held by commercial banks and the capital adequacy requirements enforced to limit banks’ ability to make loans.

  34. NKlein1553 writes:

    Last comment, I promise. Someone will correct me if I am wrong, but I do not believe the reserves commercial banks hold at their federal reserve member banks count toward fulfilling capital adequacy requirements. Instead, having a large cushion of reserves decreases liquidity risk, which in turn will make the cost of acquiring additional capital go down. So paying interest on savings accounts is a strategic decision by banks to lower their cost of funding. This is one of the possible “business plans,” Zanon referred to above.

  35. […] interfluidity » Rob Parenteau gets sectoral balances right […]

  36. zanon writes:

    VJK: Lol! NKlein1553 is right. You really are out of your depth! Reserves are not part of bank capital. They are held as asset by bank and liability by Fed. Does that sound like bank capital to you?! Bank capital is equity section on liability side. Reserve is asset. Completely different things, and understanding this is essential to having first notion of how banks work.

  37. […] interfluidity » Rob Parenteau gets sectoral balances right […]

  38. JoshK writes:

    You forgot about bankruptcy and default as a way to reduce the debt level.

  39. VJK writes:


    I suspect it has more to do with the stigma attached to borrowing directly from the Fed and other such intangible considerations than anything else.

    It looks like a rather contrived explanation — I do not think the banking cares more about “stigma” than money. Still, interbank loans come cheaper(apparently) and without stigma.


    Also, in your statement “Reserves are part of bank capital,”

    Technically, at a non-loaning bank, a deposit is both an asset (reserve money) and a liability. The asset part (reserve increase) is cancelled by the liabilitry side. If a mortgage loan is deposited at the same bank, the bank loses more (0% risk reserve) than it gains in assets (50% loan risk) in the regulatory capital requirement sense.

  40. VJK writes:

    zanon @ 36:

    I was rather imprecise(see my answer to NKlein1553), sorry about that. However, your comment does not help to explain why banks need deposits — you did not provide any numeric substantiation that it is more profitable for the bank to take deposits rather than use other funding options.


  41. Ramanan writes:


    Imagine a banks loses a lot of deposits because many cheques were written by bank depositors in favour of bank account holders in another bank. Banks settle with their accounts at the central bank. Since the bank which has lost the deposits will have negative balances at the central bank, it has to do something. Simplest is borrow overnight and/or to issue certificates of deposits. For that the bank has to pay an interest slightly above the central bank’s target. The bank doesn’t need the “money”. It just has to bring the balance at the central bank to a small positive number. The interest costs can be lowered if it manages to raise deposits from non-bank customers of other banks. When it gets back the deposits, its balances at the central bank are back to a positive number. Hence banks look for deposits.

    Hope it answers your question. There are more complicated questions you can ask, for example “why time deposits?”

  42. […] where policy makers should be focused rather than on the shorter-term talk about double dips. See Steve Waldman’s commentary on this score. First rate […]

  43. VJK writes:

    Ramanan @ 41:

    Thank you.

    The situation you’ve described is possible but a bit contrived. Why would customers flee the bank in the fist place ? If the service is abysmal, the bank will fail eventually in ts main line of business of loan making anyway. Besides, with the numbers I gave earlier, it makes more sense to rely on the interbank loans rather than “loaning” from the retail customer because the former is cheaper. Of course, with a satisfactorily functioning bank, the average cost of loaning/borrowing over the interbank network should tend to zero in the situation you’ve described.

    Hope it answers your question. There are more complicated questions you can ask, for example “why time deposits?”

    Wy indeed ? ;)

  44. Guy writes:

    I’m leaning so forgive the ignorant question.

    Where does inflation factor in?


    Can’t you have a situation where private domestic financial position improves without c.a. surplus or c.g. deficit because domestic private debt becomes relatively smaller to the value of labor? I’m sure this is just a definitional question.

  45. zanon writes:

    VJK: First, I never say banks need deposits. They do not. Bank can operate just fine with no deposit at all.

    All I say is that if bank decides to be deposit bank, then it has lower CoC. It also has higher operating cost because it needs nice branches etc. So it is business model decision.

    You insist on believing that marginal cost of a loan is bank borrowing cost IB or at FFR. This is simply not true. I have told you what it is, but if you cannot get beyond “banks borrow to make loan” then there is no hope. Neither deposit, nor CD, nor IB, nor OB IB, nor discount window, nor brokered deposit is “funding”. But mix of those, plus many other factors, all play into final you-know-what, which is marginal price of loan.

  46. VJK writes:

    zanon @ 45:

    First, I never say banks need deposits. They do not. Bank can operate just fine with no deposit at all.

    I never said you had said that. Back to my question, though. If the banks “can operate just fine with no deposit at all”, then what specific “business model decision”, from the MMT point of view, makes them choose this deposit taking modus operandi ?

    You insist on believing that marginal cost of a loan is bank borrowing cost IB or at FFR. This is simply not true.

    I do not insist on anything, just waiting for some *numbers* from you to show that there indeed is a sense for the banks to accept deposits, especially timed ones as Ramanan indicated. My numbers may or may be not sufficient to prove or disprove anything, but so far no hard numbers were provided from your side.

    if you cannot get beyond “banks borrow to make loan” then there is no hope That’s a cute try at a straw man , but let’s get back to banks and deposits ;)

  47. csissoko writes:

    VJK: If your basic question is: Why do banks want deposits? The answer is simple — to profit from the interest rate spread. Since traditional deposits pay little or no (see here) interest, they are the best deal for banks.

    Reserves are most definitely not bank capital, because reserve requirements can be met by borrowing from other banks or from the Fed. Note that the fact that banks loans earn a much higher interest rate than the rate banks pay in the federal funds market makes it unlikely that borrowing reserves will be unprofitable. (Though, as RueTheDay argues banks can mismanage their interest rate risk over the medium to long run — in which case their fixed rate loans become unprofitable.)

    BTW: I agree with JoshK. What’s missing here is a discussion of the importance of bankruptcy and default in rectifying the imbalances that have built up.

  48. Ramanan writes:

    VJK @43,

    Deposits move in and out of the banks to other banks all the time because the payee and the payer need not have accounts at the same bank.

  49. VJK writes:

    csissoko @ 47:

    The answer is simple — to profit from the interest rate spread.

    Did you see my message @ 26 re. interest rates ?

  50. VJK writes:

    Deposits move in and out of the banks to other banks all the time because the payee and the payer need not have accounts at the same bank.

    ??? 2+2=4 ;)

    Can you elaborate on your ““why time deposits?”” ?

  51. zanon writes:

    csissoko: No, banks do not profit from interest rate spreads because they do not borrow at low rates to make loans at higher rates. Interest rate may have impact on overall bank profitability through other channels, but it is not as you say.

    VJK: There is thing called liquidity risk which I define as risk of having to borrow at discount window. Discount window borrowing is considered “bad” by bank investor. If a bank has liability composition that is more weighted to deposit than other sources of liability, it will have lower liquidity risk, and thus lower you-know-what (all else equal).

    The number you should look up is you-know-what and you can compare it to banks with different liability compositions.

  52. VJK writes:

    zanon :

    What is “you-know-what” ?

  53. Ramanan writes:


    Time deposits are a bit complicated. First you need to understand why banks need the deposits :)

    Let us say I pay you $100 from my JPM to your BoA account. JPM’s deposits reduce by $100 and BoA’s deposits increase by $100. So yes 2 + 2 = 4.

    Now JPM owes BoA $100 which has to settled at the Fed. JPM can borrow back the $100 from BoA but now its costs have increased. Earlier it was paying less interests (on the deposits) and now ~ Fed Funds rate. So thats bad liability management.

    Of course some banks can keep doing this and not raise deposits. They lend and then issue certificates of deposits. The object of the CDs is to “get back the lost reserves”. Other banks purchase these CDs and raise deposits. Thats another model of a bank.

    A good reference is to get a hang of all this is Marc Lavoie’s “A Primer on Endogenous Credit-Money”.

  54. csissoko writes:

    VJK: You’re confusing a lot of different things here.

    (i) Smaller banks only have intermediated access to the federal funds rate via the bigger banks that actually trade on the market and therefore can’t borrow at the FFR but at some rate higher than the FFR — and borrowing from the Fed is possible, but brings unwanted regulatory attention that makes the real cost significantly higher than the official cost. You seem to have a model in your head that assumes that banks can borrow an unlimited amount at any of these sources and therefore will borrow at the lowest cost source — but you need to also consider what sources any specific bank has access to, what are the non-interest rate based costs of using these sources and how reliable they consider these sources.

    (ii) Deposits have the advantage that they are “sticky”, so it may be worthwhile for a bank to attract new depositors with an advantageous rate that will be offset by fees or lower rates later. For example the Sallie Mae rate you cite is advertised as 5 times the national average which makes it clear that it’s an introductory rate that won’t last. The bank is trying to attract an initial deposit base right now.

    (iii) Comparing demand deposit or overnight rates with savings or term deposit rates is a mistake.

    (iv) I don’t know what evergreen bank you’re talking about, but the one in colorado that I found is paying 0.01% on balances less than $10,000.
    The Schwab Bank rates indicate that you can get 0.5% on a savings account where withdrawals are mostly ATM and ETF (expensive) — and of course easy transfer to other Schwab accounts. The Schwab checking account rate is 0.1%.

    Basically I think that when you see a high rate you need to investigate why it is high (just like AAA CDO investors should have known that above market yields were a very dangerous sign).

  55. csissoko writes:

    Zanon: I think you misunderstood what I meant — or maybe I misphrased it. Banks lend at high rates in order to profit from the fact that deposit (et al.) rates are low. This is profiting from an interest rate spread.

  56. BSG writes:

    Can anyone suggest where/how/if writing off or defaulting on unmanageable debt burdens fits into this analysis?

    Although painful for some (both creditors and debtors), if some defaults are inevitable – and in our current situation many probably are – it seems that the sooner they come, the better. That may be an instant stimulus with the least long term costs, especially if that is an inevitable outcome anyway.

  57. zanon writes:

    csissoko: What banks need to pay for deposit is one cost, and it is cost that factors into marginal cost of bank extending marginal credit, but it is just one of many factors to do so, and that marginal cost of credit IS NOT FFR, or deposit. Banks do not borrow in order to lend, period. Therefore, saying bank borrows at 2% FFR, lends at 5%, and makes 3% spread is not correct.

    VJK: I have told you repeatedly what “you-know-what” is. When you are able to hear and understand what I say many times, you will figure out how this works and answer own question.

    BSG: When there is debt writedown due to default, then what was loan has simply become transfer. Balance sheets contract. When this happen at bank, you get debit on asset side (as receivable shrink) plus debit on equity side (capital). If capital falls below regulatory required level, the by law govt should shut bank down, but they can also choose other things as we have seen how closely Obama admin sticks to law when Goldman Sachs bonus is on line.

    So sure you can write all debt down, banks become undercapitalized/insolvent, therefore do not/cannot lend, and price of assets (ultimately, in extreme circumstance) fall to what unlevered buyer would pay. This is floor. It is path with high real costs, higher than real costs current approach is taking and certainly higher than necessary.

    This is the Obama/SRW problem. If Govt does not recapitalize private sector, then you must support bank capitalization to maintain private sector credit and therefore aggregate demand.

  58. VJK writes:
  59. VJK writes:



    I found an exchange on this blog regarding “why banks need deposits” ( It is interesting but unfortunately somewhat lacking in numeric details — due no fault of the participants. You can take a look at it and make your own conclusions.


  60. csissoko writes:

    Zanon: As I understand things, banks lend in order to get their liabilities into circulation. I still think it is fair to say that they earn a spread (traditionally the lending rate as they didn’t pay interest on balances) on their asset/liabilities. You seem to be conflating the idea that banks earn a spread with the idea that deposits are needed in order to make loans. I believe that these are separable concepts.

    VJK: Re schwab account small print, only valid with linked brokerage account — the profits from which presumably can cover any checking account related losses.

    Based on your link Everbank is clearly an outlier. The fact that a few banks with small deposit bases pay 25 bps over the federal funds rate does not seem to me to be evidence against the idea that banks like deposits (even when they create them by lending) because they are an extremely low cost source of funds. One can usually find a specialized explanation for outliers if one looks hard enough.

  61. NKlein1553 writes:

    Zanon, Ramanan, or anyone else that can answer,

    Given that reserves play a relatively small role in setting the marginal price of a loan and the fact that reserves do not count toward fulfilling capital adequacy requirements, does it follow that granting Goldman Sachs, J.P. Morgan, etc…direct access to reserves through the Federal Reserve Discount Window wasn’t such a big deal? Many commentators (especially progressive ones) make the argument that the current zero interest rate policy serves the double purpose of being a back door way to recapitalize troubled financial institutions (through the interest rate spread subsidy Zanon describes as fallacious in comment #57). Is this argument by many progressive commentators incorrect? It certainly seems so if reserves play only the minor role described by Zanon and other MMT advocates. What about quantitative easing and all the special facilities established by the Federal Reserve Board to facilitate the exchange of “toxic assets,” for reserves? Though I guess you could make the argument that the importance of these facilities was not in the fact that they supplied additional reserves to the big banks, but in their ability to allow the big banks to put off recognizing losses to their balance sheets. TARP was a direct injection of capital so I think I understand how that helped the big banks. However, TARP has largely been paid back. Certainly it doesn’t feel right to say that because TARP has been paid back U.S. government support of the big banks is finished. But to me, the argument that reserves are not significant seems to imply exactly that. I’m probably confused somewhere. Maybe someone can straighten me out =)

  62. csissoko writes:

    BSG: Default on a large scale (e.g. sovereign default or large bank/banking sector default) is clearly one way to resolve built up imbalances. That is, if creditors are foreign it results in an (involuntary) transfer from foreigners to the domestic economy: domestic debt loads are reduced and the expense of the reduction is shunted abroad in the form of losses. If creditors are domestic it results in a reduction in the domestic debt load without any (direct) effect on domestic/foreign balances.

    What I am wondering is whether it is possible to escape the US economy’s current predicament without large scale default. (It’s possible if the economy grows, but can the economy grow given its current debt burden?)

  63. zanon writes:

    csissoko: banks lend to make profit. They lend to good credit because they make money on loan (interest) and lend to each other so during normal times they earn interest on reserves which would otherwise be idle. Today there is OIR so it is different.

    and there is no one on this forum who is clearer about how bank borrowing, whether from deposit, each other, or Fed is necessary to “fund” lending. I am allergic in extreme to even slightest suggestion that that is mechanism, which is why I reject “spread”.

    And yes, it is entirely possible and trivial for US economy to escape current predicament without large scale default.

    NKlein1553: In extreme situation, shadow bank got access to Fed discount window. So if informally they have access when they need it, making it available formally all the time is mere honesty and therefore good for that, if little else.

    As for QE, if Govt makes a loan that it does not think will be paid back, is it still a loan? Or is it a money transfer? There is some of this going on in US, but even more in China where a great deal of monetary policy is actually fiscal policy in disguise.

    My own view is that fend should lend unsecured at discount window. There is no need for liquidity run. Govt should also shut down banks which fall below capital adequacy requirement. The point of banking is for private capital to be put in first loss position infront of public capital, and TARP did opposite of that. TARP being paid back is bogus because banks still have loads of bad asset on book not written down.

  64. zanon writes:

    NKlein1553: I guess in final analysis, whether or not shadow banks got access to discount window is big deal depends on whether you believe insulating them from liquidity risk is big deal.

    Personal, I think liquidity risk is pointless vandalism. The massive scandal, which the non-financial world left and right believes, is that they were insulated from insolvency by Obama.

  65. pebird writes:


    I believe that the reasons for the excess reserve injection were to 1) ensure systemic liquidity (whose system was made liquid is a different question as noted above), 2) provide banks with guaranteed income via paying interest on reserves, 3) drive interest rates to zero. To the extent that the reserve interest income flows to earnings, I guess that is the slow train on recapitalization.

    No one says reserves aren’t important (at least no one worth listening to :) – but they aren’t a necessary precondition for lending – they affect the cost of loans, liquidity and interest rates.

  66. Tom Hickey writes:

    I had read the Partenau post at NK, but I’m a little late in getting to this one. I am an MMT aficionado, not an economist.

    First, my impression is that many criticizing MMT haven’t read enough of the MMT literature and blogs to know the depth and breadth of MMT regarding both monetary economics and policy options, and specifically those options offered in response to the crisis. For example, some MMT’ers prefer direct stimulus chiefly through government expenditure (Bill Mitchell) and others primarily by reducing taxes (Warrren Mosler proposes a payroll tax holiday). However, both have a range of proposals, so it is not either/or, but a matter of emphasis.

    MMT’ers are not the monolithic bunch who are saying to just open the tap, as many seem to think erroneously. They all emphasize that deficit expenditure must be directed effectively and efficiently, and only used to advance specific public purpose in the amount needed, rather than being dispersed willy-nilly. They criticize reliance on monetary policy as a shotgun approach when it is effective, and often it is just ineffective. They point out that, conversely, fiscal policy can be tightly targeted to where it can do the most good most quickly.

    Secondly, regarding government deficits versus business investment, MMT’ers agree that in a capitalistic economy, the private investment is preferable. Government should act only for public purpose, where private enterprise is either unwilling or unable to do so. MMT’ers are not “socialists.”

    The problem, as Rob Parenteau observes, is the lack of private investment and the substitution of finance capital for productive capital. This led to the third stage of the financial crisis that Minsky called Ponzi finance. The financial meltdown spread to the real economy, drastically reducing nominal aggregate demand as the public sought to improve its financial position by saving and delevering. The problem now is that business is unwilling to invest in an environment characterized by low demand and over-capacity. As the accounting identity shows, either export have to make up the difference or government. It is pretty clear that the US cannot export its way out of this without creating a global problem. Therefore, deficits are the only recourse if deflation is to be avoided. Private investment is pro-cyclical and will pick up automatically with recovery, and deficits will decrease automatically as automatic stabilization reverses direction.

    The question, then, is how to direct the deficit expenditure. This is both and economic question and a political one, since congress is in charge of appropriation. That greatly complicates matters. In order to gain political support, special interests have to be considered as part of the political “equation.” That is the inherent difficulty with relying on fiscal policy.

    Clearly, it is best not to get into this kind of position in the first place. Capitalism is based on incentives. The incentives were wrong and still are wrong. Finance capital remains too strong relative to other factors, and until this is reformed, the situation can at best be meliorated, but not cured. Adopting austerity would be a giant step in the wrong direction.

    Warren Mosler favors a payroll tax holiday, and I believe that Winterspeak does, too. This would put funds in the hands of middle class working consumers, some of whom would save/delever to improve their financial positions, which is necessary before they become consumers again. Others would consumer more than save. That sounds to me like a proposal that could find political support on both sides of the aisle, since it would be popular with the middle class, that is, the majority of voters. Politicians would have a hard time opposing it. Looks to me like a winner. Not that this is all that is needed. The US has to deal with its LT unemployment problem, for example, which is approaching depressions levels. That will eventually lead to social problems and could manifest as social unrest.

  67. Scott Fullwiler writes:

    Well said, Tom. I, too, find the argument repeated in the comments here and on the previous post that MMT’ers think all spending is good or all deficits are equally good simply breathtaking in the lack of understanding of MMT in the first place, for two reasons:

    1. MMT at its core is about understanding the accounting and operational sides of the monetary system (there’s more to it, like Minsky, and so forth, of course). The point here is to teach these to people. Most people don’t understand that a sovereign currency issuer isn’t revenue constrained. Most don’t understand that govt deficits are the equivalent of non-govt net saving (not, Steve and others from the previous blog, that I did NOT say they are the equivalent of non-govt saving). Most don’t understand that banks don’t need reserve balances to create a loan. Most don’t understand that govt bond sales operationally serve to support the cb’s interest rate target, even according to some of the cb’s own literature. And so on. The point here is that it’s very strange to me that someone would expect an MMT discussion on these topics to venture into the discussion of what sort of deficit is the best deficit, as it’s so much work just to get across the issues at hand here without venturing into even more politically charged territory regarding policy priorities (and, in our view, the politically charged rhetoric isn’t relevant to the issues we are addressing here). I might also repeat here what I’ve said in a few other blogs–a world run according to an understanding of MMT doesn’t mean you don’t have to make tough political choices; the point, rather, is that now the choices are made with the proper economic constraint, namely inflation, in mind (or, alternatively stated, the consequences of the deficits are what matters, not the size of the deficits themselves). In practice, instead of CBO’s current practice of estimating future deficits as a result of spending/taxing policies or proposed policies while assuming real GDP, unemployment, and inflation are at some given levels, MMT’ers would prefer to see CBO or similar agency would instead be tasked with the impacts of policy options on real GDP, unemployment, and inflation.

    2. The blogs and literature referred to in (1) notwithstanding, MMT’ers HAVE most definitely put together specific policy proposals. Warren Mosler has a tab at the top of his site for his current proposals, for crying out loud. There are probably dozens of posts on the KC blog critiquing and proposing policy over the past year. Bill Mitchell has made his own policy views quite clear, it seems to me. This is all not to mention about 20 years of research published on the job guarantee and related issues. It’s beyond me how anyone could have read those and come to the conclusion that we take no stand on what sort of deficit is more appropriate. Disagree with the proposals if you like, but don’t claim we haven’t taken a stand. And even saying this much, as Tom noted, there are a range of policy positions one might take given an understanding of (1); not all MMT’ers agree with each other regarding policy.

  68. fresno dan writes:

    I always enjoy your posts, although much of it is beyond me.
    However, what I would question about these acount equaling tautolgies is that the proposition that everyone’s liabilities equals everyone’s assets, however you define and apportion them. Isn’t the crux of the problem that a good number of assets are priced not in accord with reality?
    The house bought for 400K is now worth 200K, and is still on the books at 200K, and the mortgage still costs 400K.

  69. scharfy writes:

    I was under the impression that banks “need deposits” for only one reason.

    It gives them a client who they can later loan money to. Its like having a free cover charge to the nightclub. Or giving away the razors for free, to later sell the blades.

    So they don’t need the money per say, as it tends to be more expensive than the fed funds rate.

    Note, I have CHASE account that they pay me a percent or so on, and they probably hit me with some fees to recapture that, (small change by any metric). But I do have a homeloan through them, and a credit card. That’s why they will pay up to get deposits – its gives them customers.

    Is this incorrect?

  70. VJK writes:


    My attempts to extract MMT’ers explanation as to why commercial banks need core deposits was a bit tongue-in-cheek.

    Since a typical MMT’er’s belief is that”Banks can exist just fine without deposits”, it is not surprising, therefore, that they are unable to provide a simple answer to the simple question (csissoko was the closest).

    Let’s say you get a loan at bank A, with all the capital constraints satisfied, and deposit the check at bank B. Bank A’s reserve account at FRB will be debited for the amount of the loan and the balance will go below the reserve requirements by the same amount(assuming there was no excess before the settlement). Now, in the MMT utopia, the reserve account does not need to be funded before the settlement occurs — bank A will bring its reserve account balance back to normal by ‘freely’ borrowing on the interbank market (possibly back from bank B), or from Fed.

    The problem with this rosy picture is that the banks that try to fund their activity without the core-deposit base and rely on the federal fund/interbank borrowing business ‘model’ usually fail pretty hard as, for example, the Continental Illinois collapse shows to those who care to learn how real-life banking works. Continental Illinois lacked retail banking, branches — saved a bundle of money! — and relied on fed funds — exactly what MMT suggests banks do/should do. And there are others who had the same miserable fate by relying on wholesale funding(interbank loans, brokered deposits, FHLB loans, etc) and neglecting their core-deposit base development.

    So, in short, bank take deposits because the core deposits are the cheapest and most reliable, cet. par., source of funding. And, yes, “banks can exist just fine without deposits”, but not too long as Cont. Ill. demonstrated so well.

  71. Scott Fullwiler writes:

    I don’t think I’ve ever heard an MMT’er say “banks can exist just fine without deposits.” And I know all of them. MMT’ers say that banks don’t need a prior deposit to make a loan. Those are very different statements. A bank is a profit maximizing business like any other. The more deposits it has, the greater the interest spread of assets over liabilities, since deposits are the lowest cost liabilities, and the greater the profit. There are other reasons, but that’s pretty fundamental. Maybe a bank could exist without deposits, but it would be a lost more profitable with them.

  72. rjw writes:

    I have to aree with anon below. Fond as I am of all this use of identities and the Wynne Godley style of reasoning (RIP), there is a logical flaw in the post. You can manipulate an identity to help you understand logical relations, but you not use an identity without losing sight of the fact that the variables are connected by causal relations.

    In your post you reach the conclusion that higher investment will reduce the current account … that should have tripped off an alarm bell. You are incorrectly assuming that you can adjust household balances and the current account mechanically, but that ignores the real causal connections between the variables.

    Higher investment by firms WILL raise household balances (as household save cash dispensed they receive as incomes when firms invest), but of course they will also spend part of that income … and a bit will leak into imports. So houshold balances go up, but so does the current account deficit.

    Similarly, a rise in public spending (higher defcit) will raise household income and raise balances. But housholds can also use some of those balances for consumption, suck in imports, and increase the current account deficit. Overall, household balances go up, and so does the deficit.

    So what do we make of the strong ex post inverse relationship between household balances and the current account deficit that Parenteau had in his post ?

    Well …. such a relationship is quite normal, if we start with a change in consumption behaviour (as opposed to investment, or govt spending) as the driver. Higher autonomous household spending leads to lower overall household balances, but also higher imports, and a higher current account deficit. And vice versa.

  73. VJK writes:


    You wrote: I don’t think I’ve ever heard an MMT’er say “banks can exist just fine without deposits.”

    Here’s what Winterspeak, who judging by his postings is an MMT’er, said ( “Banks can exist just fine without deposits, as we have seen.”

    You wrote: “MMT’ers say that banks don’t need a prior deposit to make a loan”.

    That is a very simplistic point of view. In order to make a settlement through the bank reserve account, money has to come from somewhere. The most robust source of funding are core deposits as some banks learned only too late. The wholesale funding is complementary and may come at a cost that makes profitable bank operation hard or even impossible.

    You wrote elsewhere regarding central bank operations: “…a bank extends credit and (in most cases) clears a payment on behalf of the borrower does not necessitate that the bank has or otherwise actively seeks out additional reserve balances; instead, at issue for the bank is the price at which it is able to obtain needed reserve balances from other banks or at a penalty from the central bank in order to effect final settlement of the day’s payments.”

    There are several problems with that statement:

    1. Interbank borrowing. Such borrowing is definitely more expensive than core deposits and has to be repaid by the next day (average maturity is about 23 hours in the US). The bank of course can continue borrowing, day after day, that is until it cannot do so any more (Interbank Lending Market `Died With Lehman’ Bankruptcy, In Canada, the interbank market is small and is used for end-of-day minor adjustments, not as a substantial source of funding.

    2. Central bank borrowing. In Canada which is epsilon-close to the MMT paradise, the reserve requirement is zero (obviously not the same as “no reserve”). The bank with an overdraft/negative balance cannot simply borrow “at a penalty from the central bank”. In addition to the penalty rate, the bank can borrow only when it posts sufficient collateral (Tranche 1) or if other settlement network participating banks agree to pay up in the case the borrower defaults (Tranche 2). The borrower bank has to repay the CB loan by 18:00 the next day. If the bank defaults, the CB will seize the borrower’s collateral or the participating banks will have to rescue the miscreant.

    3. “…does not necessitate that the bank has or otherwise actively seeks out additional reserve balances”.

    Due to (1) and (2), it would be extremely foolish for a sane liquidity manager not to “seek out” funds beforehand in order to make the reserve account settlement as smooth as possible, but rather rely on his luck in the interbank market game or CB borrowing to zero the reserve balance.

    In sum, the bank ability to fund its reserve account is as much of a constraint, and occasionally even more, as its capital related ability to absorb losses unless the bank ‘desires’ to expire in pain, barring the TBTF scenario of course.

  74. Jim Bradley writes:

    The equation is misleading. The private sector can delever if the Fed prints the money…

    The Fed doesn’t take on “leverage” like other economic entities. It is the only entity capable of running a deficit without someone else running a surplus.

  75. Scott Fullwiler writes:


    I am well aware of all the points you make. They don’t refute the point I made, which you’ve misunderstood and twisted the meaning of at any rate. And, no, Winterspeak is not an MMT’er in the sense I would use the term, which is individuals that have actually published MMT research, though he is in general agreement with most of the descriptive aspects of MMT.

  76. winterspeak writes:

    VJK: I have not published as I’m not an academic economist (thank God) and therefore not Scott’s definition of MMT. That’s fine.

    Do you consider a shadow bank like Goldman Sachs a bank? It repo’s with other banks and bank-like institutions overnight. It has access to the discount window when it really needs it. It does not have retail deposits. Seems to do OK profit-wise last I checked.

    I agree that there is risk in relying on the ON IB market, and that deposits lower CoC, although they also increase other costs.

  77. Scott Fullwiler writes:

    Hi Winterspeak . . . I meant no disrespect. When someone puts a quote like that and says “MMT says . . . “, I’m assuming they’re quoting something published. I wouldn’t want everything I’ve said as a comment online to be considered the big-T truth about MMT. Published work is another thing, because you spend a lot of time thinking about it before it’s a finished product and it then also goes through a referee process. On the other hand, there are some online comments that would be clearly descriptions of MMT. Some others I’d like to have back.

  78. Scott Fullwiler writes:

    VJK . . . a bit of clarification on my point that you’ve misunderstood my original points:

    You said:

    “That is a very simplistic point of view. In order to make a settlement through the bank reserve account, money has to come from somewhere. The most robust source of funding are core deposits as some banks learned only too late. The wholesale funding is complementary and may come at a cost that makes profitable bank operation hard or even impossible.”

    Of course, banks don’t settle payments with core deposits. More core deposits reduce the funding costs and thus raise profitability, so for that reason the more a bank can have the better. But payment settlement only occurs via reserve balances (you don’t “settle” with your liabilities, you settle with your assets), for which cb intraday overdrafts mostly are used for, and then money markets to cover overnight. Thereafter, the lowest cost option for the bank is to replace the borrowing with more core deposits.

    You said:

    “There are several problems with that statement:”

    “1. Interbank borrowing. Such borrowing is definitely more expensive than core deposits and has to be repaid by the next day (average maturity is about 23 hours in the US).”

    Agree. That’s why I said the most important point is the price at which the bank obtains financing to settle payments. Again, though, that doesn’t stop the bank from finding more core deposits to reduce borrowings. And, another point I made in that quote, is that the reserve outflow from a given withdrawal does not necessarily require overnight borrowings, since what matters is the final settlement balance at the end of the day.

    “The bank of course can continue borrowing, day after day, that is until it cannot do so any more (Interbank Lending Market `Died With Lehman’ Bankruptcy,”

    The reason why the interbank market has dried up is because of the excessively large qty of ER, not because nobody will lend anymore (though immediately following Lehman, there was certainly some counterparty risk–I’ve dealt with that in other published work, and it certainly isn’t inconsistent to the points I’m making. Quite the contrary.)

    “In Canada, the interbank market is small and is used for end-of-day minor adjustments, not as a substantial source of funding.”

    Yes, they only need to settle the end of day negative/positive balances, there are only a few banks, and they all have completely certainty that they can clear negative/positive balances at the end of the day. In the US, theoretically only end of day adjustments need to occur, but interbakn trading is highly decentralized and end of day balances are far more uncertain. Again, I’ve written on all this before, and it’s not inconsistent with my points.

    “2. Central bank borrowing. In Canada which is epsilon-close to the MMT paradise, the reserve requirement is zero (obviously not the same as “no reserve”). The bank with an overdraft/negative balance cannot simply borrow “at a penalty from the central bank”. In addition to the penalty rate, the bank can borrow only when it posts sufficient collateral (Tranche 1) or if other settlement network participating banks agree to pay up in the case the borrower defaults (Tranche 2). The borrower bank has to repay the CB loan by 18:00 the next day. If the bank defaults, the CB will seize the borrower’s collateral or the participating banks will have to rescue the miscreant.”

    Yes, agree completely. And the consequences are if anything more steep in the US with the Fed. Again, no inconsistency with my point that the key is the “price” of borrowing, which includes any non-rate costs of collateral, etc.

    “3. “…does not necessitate that the bank has or otherwise actively seeks out additional reserve balances”.”

    “Due to (1) and (2), it would be extremely foolish for a sane liquidity manager not to “seek out” funds beforehand in order to make the reserve account settlement as smooth as possible, but rather rely on his luck in the interbank market game or CB borrowing to zero the reserve balance.”

    “In sum, the bank ability to fund its reserve account is as much of a constraint, and occasionally even more, as its capital related ability to absorb losses unless the bank ‘desires’ to expire in pain, barring the TBTF scenario of course.”

    Yes, in the US a large % of the fed funds market (prior to Lehman) was via pre-arranged lines of credit. And, again, not inconsistent with my point that the “price” of borrowing in interbank markets is the key variable. Banks can always get financing (if only from the cb at a substantial monetary and/or non-monetary “penalty”)–the issue is the price. And, given the difficulties encountered in interbank markets following Lehman, the US strategy of pushing so much interbank activity onto the “markets” showed its inherent weaknesses in a time of crisis. The Fed stepped in and tried to help, but it took a long time for it to recognize that the issue was the price of refinance, not the qty (and it actually never figured it out completely). We’ve discussed this over and over again, particularly on Mosler’s site as it was all occurring. Again, there’s nothing inconsistent with my main argument.

  79. VJK writes:


    I apologize if my quoting you irreparably damaged your standing with the ‘official’ mmt’ers ;)

    1. Do you consider a shadow bank like Goldman Sachs a bank?

    To the degree I know anything about GS operations, the answer is: no, GS is not a [commercial] bank. According to some information, about 90% of its ’09 revenues came from activities typical for an IB+hedge fund hybrid, such as trading, market bets, corporate debt and equity, etc.

    Now, in ’08 the wolves(IBs) were relabeled as sheep(banks/’bank holdings’) in order to enable, among other things, their access to the discount window as part of saving the wolves from almost inevitable death. Such relabeling, of course, does not change the substance of IB banking.

    2. It has access to the discount window when it really needs it
    The role of such access for profit generation, not as an insurance tool, is most likely greatly exaggerated: apparently, GS accessed the window once, in 2008, during a $10mil borrowing test run.

    3. It repo’s with other banks and bank-like institutions overnight.
    Do you have any numeric info on that ?

    4. Seems to do OK profit-wise last I checked.
    Right, but you need to establish a causal relationship between (3) and (4), otherwise, it sounds as an non sequitur $10bil of taxpayer money chanelled through the AIG rescue package, as payment due on GS’s CDS bets, looks like a more substantial source of revenues.

    In any case, due to (1), I do not think GS is relevant to the core deposit role discussion.

  80. anon writes:

    Goldman does not consider itself to be a bank. That was in their verbal congressional testimony recently. And yes they confirmed they only ever used the discount window once, as a test.

    And there’s been an awful lot of rubbish on the internet over the past year about the banks borrowing from the Fed to invest in Treasuries. It hasn’t happened. It’s not in the published stats.

  81. Scott Fullwiler writes:

    “I apologize if my quoting you irreparably damaged your standing with the ‘official’ mmt’ers ;)”

    Yes, poor choice of words on my part. I was mostly trying to make the distinction between comments and published research. A good example of a comment I’d like to have back and why the distinction is possibly apt (at least in my own case).

    Regarding Winterspeak’s point on GS doing repos and VJK’s request for more info, see charts 2 and 4 here may be of interest:

  82. Scott Fullwiler writes:

    “And there’s been an awful lot of rubbish on the internet over the past year about the banks borrowing from the Fed to invest in Treasuries. It hasn’t happened. It’s not in the published stats.”

    Very true.

  83. Scott Fullwiler writes:

    VJK . . . to restate along the lines of what I should have said the first time:

    My real problem was with the phrase, “a typical MMT’er’s belief is that . . . ” and then when I pushed back you went and found a statement from 1 person in 1 blog that occurred 1.5 years ago. If it’s a “typical belief,” then it should be something rather commonly found.

    I’m sorry, but misrepresentations of MMT by individuals who appear to have a chip on their shoulder about MMT are a dime a dozen if you’ve paid much attention at all to the blog discussions related to MMT, and I have little patience with that. In your case, you have repeatedly misrepresented MMT (your earlier conversation with Marshall, your suggestion that MMT thinks all deficits are equally good or similar, your use of “typical” here, and your recent misrepresentation of my research above are just the latest examples). Perhaps I am the one misreading and you don’t have a chip on your shoulder about MMT (I hope so), but if that’s the case, asking a few more questions rather than making (to this point, usually erroneous) accusations about MMT would be more helpful.

  84. anon writes:

    I agree there’s an excess of blogosphere cherry picking of interpretations from second hand observers of MMT, rather than more careful consideration of the work of the MMT originals – such as Mitchell, Wray, Fullwiler, and Mosler. This is a sort of disruptional blogosphere virus. The latest, albeit at a more professional level was Yves Smith, who while getting much of MMT right, represented it inappropriately in a particularly fundamental way, when she’s really not in a position to represent it at all. Comment on it, yes. Represent it, no. Those who are truly interested in it should go the sites of the professionals and read the original versions.

  85. VJK writes:

    Scott Fullwiler@78:

    1. you don’t “settle” with your liabilities, you settle with your assets

    Deposits are liabilities on the right side of the balance sheet, but obviously they are, at the same time, assets represented either as cash in vault or as cash balances on the CB reserve account depending on the way they have been deposited. Clearly, you can settle with cash.

    2. for which cb intraday overdrafts mostly are used for[settlement], and then money markets to cover overnight

    I strongly disagree with the statement that overdrafts/money markets are used as the primary source of settlement *after* the loan check was given to the borrower. In Canada, any decent liquidity manager will easily anticipate predictable cash outflow and provide required funds to settle any negative residue using the bank liquid assets. In fact, the situation is no different from everyday liquidity management of routine withdrawals. I am sure US bank liquidity managers act in a similar way.

    Also, it is interesting to note that the interbank market in most cases, consists of excess of cash that other banks obtained as deposits. So, the bank that covers a negative reserve balance with an interbank loan uses deposit cash from another bank, just at a higher price.

  86. VJK writes:


    Re. shadow banking system.

    You may find this interesting:

  87. Scott Fullwiler writes:

    “I strongly disagree with the statement that overdrafts/money markets are used as the primary source of settlement *after* the loan check was given to the borrower.”

    My point is the actual settlement occurs after the loan. As I already acknowledged in 78, there are certainly agreements, such as lines of credit, established before the fact. It is simply a fact in both the US and Canada that a net overdraft position in a reserve account is settled at the end of the day (timing is a tad different in the two, but nevertheless). That was my point, if you had read carefully.

  88. VJK writes:

    Scott Fullwiler@87:

    There is not much disagreement between us re. reserve account settlements, but I am not too happy about how some MMT’ers de-emphasize the core deposits role — leading some smart folks to believe that core deposits are not needed at all. Too much trouble has been caused by FIs trying to conduct their business without such deposits. The shadow bank liquidity crisis, to a large extent, was caused by their reliance on the wholesale market (see the link I provided above).


  89. Scott Fullwiler writes:


    I don’t think you’ll find much de-emphasis of core deposits among myself, Mosler, Wray, Mitchell, Kregel, and others. We’d very much like to see a return to banking based on deposits and basic lending to businesses and households. See Mosler’s policy proposals for banks, for instance. As I started this whole thread out, I think very strongly that your suggestion that MMT de-emphasizes core deposits is completely mistaken. Can you find any other reference besides Winterspeak’s quote buried in comments from 1.5 years ago? I can’t think of any others myself. And I’m still not completely clear on the meaning of Winterspeak’s quote, at any rate.

  90. winterspeak writes:

    My meaning was simply that there is no shortage of bank or bank-when-they-need-to-be entities out there that don’t do much business, or any, with core deposits! This is an observation, not a recommendation.

    Yes, they give up a CoC advantage, but they gain others, and overall, it is a reasonable business strategy. Such entities, for example, banks relying on brokered deposits, may go bust at higher rates, but it still may be worth it for investors to fund such companies. I’m sure investors will create such entities in the coming years.

    The discussion of this in the thread VJK linked to was very good.

    VJK seems to have confused my observation with a recommendation.

  91. Ramanan writes:

    I think VJK’s worries go like this – if bank lending can make deposits so easily, why do you see bankers hunting for deposits ?

    I think he has a point – though he should have put it in a different way – when MMTers say that loans create deposits, they should at the same time emphasize why its consistent with the fact that banks need deposits. Of course its written at various places, probably VJK couldn’t find it or is not satisfied.

  92. […] A much more technical discussion of this issue than I’m capable of writing can be found in the Interfluidity post that my Alexandria co-blogger Steve linked earlier in comments. In this article, Steve Waldman both […]

  93. VJK writes:


    Can you find any other reference besides Winterspeak’s quote buried in comments from 1.5 years ago?

    The loan desk of commercial banks have no interaction with the reserve operations of the monetary system as part of their daily tasks. They just take applications from credit worthy customers who seek loans and assess them accordingly and then approve or reject the loans. In approving a loan they instantly create a deposit (a zero net financial asset transaction).

    The only thing that constrains the bank loan desks from expanding credit is a lack of credit-worthy applicants, which can originate from the supply side if banks adopt pessimistic assessments or the demand side if credit-worthy customers are loathe to seek loans.
    [emphasis in original]

    Apparently, neither capital nor liquidity is a concern for the bank, only the borrower credit-worthiness is. “no interaction with the reserve operations” is simply wrong.

    So, at an individual bank level, banks make loans as they like, and then either borrow the reserves they need at the discount window or overnight market, OR the grow their deposit base to meet their reserve requirements

    Now, there is some reference to liquidity, but the funding source priority is completely reversed(deposit base significance is de-emphasized).

    Loans create deposits. Most people believe you need funds, deposits, or savings to lend. Absolutely not true.

    Bank capital NOT a constraint on lending

    Neither liquidity nor capital are important, one would imagine.


    Loans are created ex nihilo, at the stroke of a pen, or by punching a key on the computer, as long as the borrower is credit-worthy. The only limit to this process is given by the amounts of loans which can be granted to credit-worthy

    The same simplistic picture.

    As any bank manager knows, loans are not created ex nihilo, both bank capital AND liquidity, ensured primarily through the bank’s deposit base, are needed for such creation. How well the shadow banking system, as well as some non-shadow banks, fared by neglecting either is well known by now.

  94. VJK writes:

    I am not sure how to edit a message — there is no preview option unfortunately. Here’s the last paragraph with proper italics:

    I wrote:

    As any bank manager knows, loans are not created ex nihilo, both bank capital AND liquidity, ensured primarily through the bank’s deposit base, are needed for such creation. How well the shadow banking system, as well as some non-shadow banks, fared by neglecting either is well known by now.

  95. Ramanan writes:


    I think you should worry less about the shadow banking system right now. The crisis makes it difficult to explain such things. Think of a country which doesn’t have a shadow banking system, such as India for example.

    Firstly, as a matter of accounting, loans make deposits. First think of the banking system as a whole. Closed economy, no government spending or taxes or government bond sales for a time period, say a day. Deposits do not change – they just shuffle around. When a loan is made, the banking system’s deposits increase. Do you agree till this point ?

    Then start thinking what happens.

  96. Scott Fullwiler writes:


    You’re confusing literature that is intended to explain why banks are not operationally constrained by reserve balances or deposits in creating loans (Marshall’s quote, my quote, etc.—since that’s what the vast majority of people think, so that’s what we’ve been combating) with the suggestion that banks don’t need liquidity, capital, or core deposits. The former is true. The latter is not. We’ve never suggested anything different.

    Context matters. Please consider it in the future, or again, simply ask rather than accusing.

  97. VJK writes:


    When a loan is made, the banking system’s deposits increase. Do you agree till this point

    Of course ;)

  98. Scott Fullwiler writes:

    OK, VJK, let’s go through these a bit:

    Marshall–the point here, which we’ve made MANY TIMES (if you want a “typical” MMT statement, this is it, since it seems to be so difficult for you to find them on your own), is that the loan officer doesn’t go to the liquidity manager to see if the bank has enough “money” to make a loan. Whether or not to expand the bank’s balance sheet is set at a strategic level (and the loan may be sold, anyway). Certainly, as individual loans are closing, the liquidity manager must incorporate this information into the managing of net changes to the reserve account, but that’s a separate point BECAUSE that’s not the point that 99% of the economists are worried about–they’re all thinking that “more reserve balances=more ability to create loans.” That’s what Marshall’s dealing with here. If your preferred detailed model of a bank’s operations aren’t covered in his post, you need to consider this context.

    Also, regarding “the funding source” is completely reversed again misses the context. For sure, an individual bank starting up and thereafter needs to build capital and a deposit base to appease regulators and to be profitable. Consider the example Ramanan gives you in 95.

    Mosler–you’ve misunderstood (yet again!) the point he was making. First, he said that bank capital is endogenous–this is the basic sustainable growth framework that explains how fast assets can grow in order to keep capital ratios constant, given ROA and retention ratio. This is standard and completely valid. Second, he did say “in the very short run” capital is a constraint, because the capital from profits hasn’t arrived yet. Third, again, overall he’s combating here the lack of understanding of basic accounting–loans create deposits, as Ramanan again referred to.

    Lavoie–Yet again, a counter to the traditional deposit multiplier. Marc has a very thorough understanding of how banks function.

  99. Scott Fullwiler writes:


    Regarding 97, since you agree, then I can’t see anything about MMT that you disagree with regarding banking. No MMT’er would have strong disagreements with the micro-level details you’ve been describing for commercial banks; as I said, our literature has primarily been targeted toward explaining a different part of the process.

  100. winterspeak writes:

    Ah yes, the old “why do banks seek deposit” question.

    I think JKH said it was vestigal in the last thread on this. That must be right then!

    Either way, it certainly isn’t required (even though it may be recommended) and its benefits are clear. What is less clear but just as true is that a bank (or bank-like-entity) can and do just fine as a bank (or bank-like-entity) without taking on any deposits.

    This has been clear in the context of all the discussion on this.

  101. VJK writes:

    winterspeak @ 100:

    I think JKH said it was vestigal in the last thread on this. That must be right then!

    There is no arguing with faith, I’d guess.

    just as true is that a bank (or bank-like-entity) can and do just fine as a bank (or bank-like-entity) without taking on any deposits.

    Wow, just wow…