« "Presentation of the Liberal position on climate change and the environment" | Main | More reasons to support the PBO »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Nick: Lots of good comments by JKH. Not sure what I can add.

I would like to underline the definition of saving as an income statement non-event, namely, money comes in, but does not go out. Perhaps it may be even more accurately defined as a cash flow event.

My employer credits my bank account with $1000. $100 I leave in the bank, and the other $900 I spend on haircuts, backscratches, antiques, lottery tickets, and shares of Goldman Sachs. I have saved $100, and I have spent, through some combination of investment (both capital and inventory) and consumption, $900. The rest of the economy has had its bank accounts credited by $900, and my bank account has been debited by $900. The $100 that remains there is taken out of circulation. Note I'm talking about financial assets here only, the antiques that I bought are real assets and I now own them, perhaps hoping to sell them for a profit in the future. Nevertheless, it is critical to differentiate actions that trigger income events (which savings does not) from actions that do not trigger income events (which savings does). Savings, is no one else's income. If an economy wants to increase this number, it cannot, for reasons that I believe we both agree on, and this state of affairs is extremely harmful. An economy can have many reasons for wanting to increase this number, one of which may or may not be interest or discount rates. Certainly interest/discount rates are not the only reason, and in certain circumstances, they may be a very unimportant factor, swamped out by other, more important factors.

I tried defining this as "equity" earlier because if you begin with an empty balance sheet and pay in some capital, it resides as cash on the asset side, and equity on the liability side. The terms seemed to be imprecise, and it's a shame Anon did not or could not come up with an accounting entry to capture the concept.

Adding antiques to the haircut economy increases the initial quantity of assets, but I'm guessing they are real assets. If the economy wants more but cannot have more, then there may be shuffling around of them internally, but there will be no net increase. The economy may or may not change its desire for money (at least temporarily) in the hope of somehow translating that into antiques in the future, but if they try, then we're in the same collapse of AD condition as the number of haircuts dwindles.

ADAM P: You are correct, money is not a real asset, but others somehow interpret this to mean that money is not important. I am not accusing you of this, just using your comment to highlight this point. In Nick's model economy, money was a financial asset, but the desire for the economy to have more of it lead to falling AD and lower real asset production (fewer haircuts). Demand for the medium of exchange is demand for something that is nominal (not real), but if it is not met, it has material negative effects on real production.

ALL: Please do note the key macroeconomic, balance sheet analysis by JKH. No matter what happens at a micro level, the quantity of savings (financial assets) in the economy does not change. You just get lower income, and lower real production. It's stuff like this which makes it impossible to scale up micro models to the macro-economy. A household can increase the net financial assets on its balance sheet, but a sector cannot.

NICK: Finally, it would be great if the US stopped issuing bonds. That would be a way to force the issue as to what will or will not happen if the Chinese stop buying them (since if none are issued, none can get bought). I believe Mervyn from the BoE actually suggested this for gilts, although I've heard nothing since so maybe it did not happen. I can tell you what will happen though if the US took this step (which I think it should).

SCOTT/JKH: Are these transactions best captured in the income statement, or in the cash flow statement? Cash flow would make errors like Nick including real assets harder to make. It also keeps things close to the hard debits and credits I like to work through to make sure I'm manipulating my balance sheets correctly. But you also miss many important things in cash flow statements, so I'd love your take on this.

"I would like to underline the definition of saving as an income statement non-event, namely, money comes in, but does not go out."

That's a pretty good way of thinking about it as well, winterspeak. That describes the act of saving, which is a passive one.

From that position of having saved, you can start talking about taking action in terms of changing your portfolio mix of accumulated saving - e.g. switching from cash into bonds, equities, houses, etc.

JKH: let H be flow of haircuts bought, Y be flow of haircuts sold, M be stock of money, and A be stock of antiques, W be stock of assets. Let Xdot be time derivative of X.

Seems to me there are two equivalent ways of writing the individual's budget constraint:

1. Y=H+Wdot W=M+A

2. Y=H+Mdot+Adot

Just differentiate W=M+A with respect to time, substitute it into the income statement Y=H+Wdot, and you get 2 from 1. Maybe 1 would be better than 2 if M and A are jump variables, so Mdot and A dot are not defined (they are infinite). But as you go to discrete time, or allow adjustment costs, it's all the same. You can't say that 1 is right and 2 is wrong. Just two different ways of saying the same thing.

"I tried defining this as "equity" earlier because if you begin with an empty balance sheet and pay in some capital, it resides as cash on the asset side, and equity on the liability side. The terms seemed to be imprecise.."

I don't think that imprecision should be a source of concern, winterspeak. Yes, it's a general classification, but it opens up sub-classifications according to source - e.g. paid in from a new capital issue, or accumulated via retained earnings, or in the case of a household, it's more commonly known as household net worth. The fundamental thing to understand is that this type of equity resides on the right hand side of the balance sheet, below debt. And anything can offset it on the asset side. In the case of newly issued equity, cash is the initial offset. In the case of retained earnings, many accounting events occur along the way that change the composition of assets as retained earnings are being accumulated over time.

Winterspeak, I tackled the bond question at 12:13. Would you agree with my description?

Winterspeak: " Savings, is no one else's income. If an economy wants to increase this number, it cannot..."

this is not correct. An economy can, in aggregate, increase its savings if it invests in real assets. This has absolutely nothing to do with money, money needn't even exist for this to happen.

In a decentralized economy of course financial assets facilitate the direction of resources but that is not essential to what saving actually is at the aggregate leve.

"Are these transactions best captured in the income statement"

Winterspeak, I addressed this at 1:52.

Transactions in current period output are captured in the income statement. Income is revenue. Such purchases of current period output are expenses.

Asset transactions apart from that are captured in the flow of funds statement.

The purchase of antiques is the purchase of assets, not the purchase of current period economic output. It doesn't appear in the income statement because it doesn't generate any current period income to the factors of production (apart from commissions at an auction, or an antique seller's profit margin, but that's minor relative to the capital cost of the antique itself.)

“You can't say that 1 is right and 2 is wrong. Just two different ways of saying the same thing.”


The budget constraint you are constructing cuts across current period income as well as starting assets as reflected in the starting balance sheet. That’s fine as long as you realize what you are doing. You’re certainly not constructing a budget constraint for current period income alone.

Accounting differentiates between time periods. Primarily it differentiates output and income activity for the current time period, versus the cumulative result of all such activity for prior time periods. That cumulative result is measured in terms of a balance sheet. The current period result is measured in terms of an income statement.

The balance sheet includes assets, liabilities, and equity. The income statement includes revenue, expense, and a measure of profit.

“Just differentiate W=M+A with respect to time, substitute it into the income statement”

There you are mixing the balance sheet with the income statement in your approach to budgeting. That’s fine, so long as you understand you’re not budgeting for current period income alone.

In my example, X is saving from current period income. Y has no current period income because X wouldn’t buy a haircut from him. But Y buys a haircut from X. Y does this by dissaving, which manifests itself in Y drawing down money resources from his balance sheet.

Again, I’m not sure what point you’re making, other than the idea that you can construct some sort of budgeting equation that allows not only for current period income, but for the resources potentially available from the starting balance sheet for the period. You can manipulate your approach to “budgeting” as you desire, but that has nothing to do with the fact that the income statement is different than the balance sheet, and that current period saving is a function of the current period income statement, and has nothing to do with the starting balance sheet.

ADAM P: An economy can increase its net holdings of real assets, of course. It can do this without money at all as you say!

I have always always always been talking about financial assets, and an economy cannot increase its net holding of financial assets. Sometimes it wants to and tries to, and the results are very bad for the real economy.

The savings I am talking about is always net financial assets. Not real assets.

Winterspeak, I think we agree. Actually I think we've been saying the same thing the whole time, we're not debating here, just converging on terminology.

Only a few minutes . . .very interesting discussion. Just have enough time to add that I think the "equity" measure searched for in the discussion with Anon is probably just the stock of Net Financial Assets. for the pvt sector in a closed economy, that would be equal to the national debt (bonds + currency + reserve balances - cb lending to pvt sector . . .or the stock of vertical money as we MMT'ers like to call it).

"The savings I am talking about is always net financial assets. Not real assets."

If that's the case (which is not surprising), important to make that clear at all times, winterspeak, at risk of repetition, but to avoid confusion.


If you're looking for a name for the equity invested in net financial assets, why not just call it NFA equity, or MMT equity?

JKH . . . that's what I said at 3:32 (regarding NFA).

Regarding NFA again . . . don't know why we didn't call it that . . . brain cramp, I think. We were focusing on a measure on the equity side, and then I had the "duh!" moment a bit later. Glad to see you see it the same way, JKH.


Yes, agree as regards NFA.

The important thing is to be clear when the meaning intended is that of NFA equity.

Versus equity more generally, which means equity accumulated as per national accounts definition of saving.

E.g. it strikes me that Winterspeak (by his own admission) almost always if not always is intending the meaning of NFA equity. No problem there, but others may intend a broader meaning, unless notified the discussion is on different grounds.

"Glad to see you see it the same way"

Typing as you commented.

Absolutely. It's occurred to me before, but the natural terminology seems clear now.


To be clear, if you're constructing a macroeconomic balance sheet for NFA, the highest level entries would be:

Assets: NFA
Equity: NFA Equity

This "NFA balance sheet" would be a subsection of the entire non government balance sheet.

It's the subsection that deals exclusively with the NFA position, which I think is the way you like to isolate it.

(The rest of it deals mostly with non-NFA equity and real investment. Most everything else should consolidate because financial assets cancel against financial liabilities.)

That's a little more than Scott and I just discussed, but I hope he will agree.


I'd then simply refer to the two pieces I noted above as:

non government NFA equity
non government real equity

NFA equity is created by the government deficit

i.e. the government uses its leverage to create NFA equity as an addition to real equity

JKH/Scott: NFA equity is fine. It is what I mean. Unfortunately, unless you already understand all this stuff, you don't understand what NFA equity is! I exactly mean vertical money, but if you don't know how the monetary system works, you will have no idea what that is either!

If I could string together commonly understood balance sheet entries (like paid-in capital and retained earnings + goodwill + etc. etc.) it would be a way to engage with people from somewhere they recognize. It's never worked out, so may be the wrong approach.

ADAM P: I think the point you are making is that an economy can always create stuff (real assets) without needing money (net financial assets, or NFA equity) at all. This is true, it's also trivial in that no one would contest it.

The point I'm trying to make is that an economy, as a whole, cannot increase its NFA equity by itself. Most people would think of this as savings, and would regard the statement "everyone in the united states cannot save more unless the Government runs higher deficits" as being profoundly weird. If I point out that the National debt precisely equals the NFA equity (aka. net savings) of the private sector, that would seem bizarre too. The deficit is rightly thought of as the NFA equity the Govt has "paid into" the private sector to capitalize it, the same way investors "pay into" a start-up to capitalize that. This paid in NFA equity acts as the equity base that all private sector credit extension rests on, as the sector can lever (or de-lever) itself to whatever degree it wishes. It just cannot capitalize itself.

This is extremely non-trivial conceptually, and I can point you to the front page the NYTimes, or the Harvard Econ Department, or any economics textbook to demonstrate just how non-trivial conceptually it is. I can also point you to any recent speech read out by Barak Obama to highlight the danger of not understand how the economy works.

JKH: Your focus on the act of saving as being an income statement event, and then subsequent purchases (or not) of stuff as being restructuring of balance sheet assets is fine. It nicely mirrors the reserve/t-bill interaction in the banking sector. Not sure if people will realize its importance initially though.

Here is the simple haircut version.

Assume no banks and no currency denominated debt, fixed amount of currency, and fixed, flat price of $1 haircut.

If someone saves $1, does that mean the overall number of haircuts falls by 1?

Too Much Fed, yes that's the point. In the absence of investment in real assets any aggregate attempt at savings just reduces todays income. Thus, attempts to use money as a savings medium (the excess demand for money) cause a fall in income.

Money or other financial assets can't be used as aggregate savings. They can only be used to re-direct resources from the production of consumption goods into the production of capital goods.

“This paid in NFA equity acts as the equity base that all private sector credit extension rests on, as the sector can lever (or de-lever) itself to whatever degree it wishes. It just cannot capitalize itself.”

Full disclosure, winterspeak, I’m not with you on this, at least not until I understand what you’re trying to say. I seen this phraseology before, and it’s puzzled me.

NFA equity is a relatively small proportion of total non government equity. Most of the total is reflected in household net worth, which is currently about $ 55 trillion in the US. This number captures the measures of all types of household assets and liabilities, including the value of all financial assets held directly (e.g. bonds, stocks) and indirectly (e.g. though the value of pension and insurance liabilities, mutual funds, etc.), and the value of residential real estate. All of that captures private sector net worth (equity) to which you have to add the net international investment position of the foreign sector, which is probably around $ 3 or 4 trillion (haven’t check lately). Total non government equity in round numbers then approaches $ 60 trillion, down substantially from the corresponding number at its peak of about $ 75 trillion.

NFA equity is a subset of this $ 60 trillion. The public float on the national debt is around $ 7 or 8 trillion I think, and obviously growing. So for the sake of discussion, say NFA equity is $ 8 trillion out of a total of $ 60 trillion in equity. That’s about 13 per cent.

The collapse in total equity reflects the effects of the recession and balance sheet deleveraging by the non government sector. The government is injecting more NFA into the economy as a function of increasing budget deficits and as a means of providing leverage to offset this non government deleveraging. This provides additional income and satisfies excess demand for saving during the recession, attempting to jump start the economy back into gear.

That’s my interpretation of what’s going on.

As surface statements, I certainly don’t agree that “all private sector credit extension rests on” NFA equity. And I certainly don’t agree that the private sector “just cannot capitalize itself”. Perhaps this just reflects your writing style. But it certainly reads like a sort of tautology about the way in which total equity is an absolutely dependent function of NFA equity, which is clearly not the case. Obviously the private sector can extend credit and capitalize itself in the long run. The disproportionate shares of the equity components reflect that (in stock terms).

But if instead you mean that government NFA provision is assisting with the process of credit extension and recapitalization (in flow terms) – that, I can certainly agree with. But that’s not exactly what your words seem to suggest.

Am I just misreading your writing style, or is there a material disagreement here?

What Adam P. said. (unless, of course, the price level falls so that the real money supply expands, or the central bank increases the supply of money, until people no longer want to save more money. And the worry is, in the current context, that that might require a very large money-financed tax cut/transfer).

BTW, I think one of the reasons that economists like me stick to very simple models is that we know we are less likely to screw up the accounting that way.

The haircut example is excellent! Of course, once in equilibrium why would anyone want to save more, as there is nothing to buy but haircuts?

Unless one barber had knowledge of the coming antiques (asymmetric info). Knowing there is a fixed money supply, the only way for saving is to put a barber out of business. So, demand suppression results in lower velocity (="savings"), reduced income, one or more barbers (probably those that believed in equilibrium) go out of business, now we have money available to buy those antiques.

I love neo-classical economics.

Very similar to our banking problem today.

FYI, there is an implied transaction flow in accounting entries - from credit to debit. Examples: pay rent with cash: CR cash > DB rent; purchase inventory with cash: CR cash > DB inventory; borrow cash: CR loan > DB cash; receive cash for sale: CR income > DB cash. The credit is the "source" or transaction origin, the debit is the "resting place" of the transaction.

Also, since accounting has no relevance, lets consider:

C + I + G + X - M = Y

Eliminating government and foreign trade:

Y = C + I

What about accounting?

Debit Accounts = Credit Accounts

Assets + Expenses = Liabilities + Equity + Revenue

A + E = L + Eq + R

Lets rearrange:

R = A + E - L - Eq

A little cleaning up:

R = E + (A - L - Eq)

Revenue = Expenses + net(Assets - Liabilities - Equity)

Income = Consumption + Investment

Look similar?

pebird said: "The haircut example is excellent! Of course, once in equilibrium why would anyone want to save more, as there is nothing to buy but haircuts?"

What about someone wanting to retire?

Adam P said: "Too Much Fed, yes that's the point. In the absence of investment in real assets any aggregate attempt at savings just reduces todays income. Thus, attempts to use money as a savings medium (the excess demand for money) cause a fall in income."

What if the rich are using excess corporate profits to save?

Nick's post said: "What Adam P. said. (unless, of course, the price level falls so that the real money supply expands, ..."

So if price deflation is 2% and the nominal fungible money supply is flat, then the real fungible money supply is plus 2%? Correct?

With that scenario, will people use the fungible money supply to save?

Plus, if there are savers, price deflation is to be avoided, and real GDP (quantities) needs to increase then the fungible money supply needs to increase? Correct?

Nick's post said: "... or the central bank increases the supply of money, until people no longer want to save more money. And the worry is, in the current context, that that might require a very large money-financed tax cut/transfer)."

I don't think the term money should be used. It should be either currency or currency denominated debt. With that in mind, what is the difference in these scenarios?

1) price deflation of 2% and the fungible money supply is flat

2) price inflation of 2% and the fungible money supply grows by 4% with currency

3) price inflation of 2% and the fungible money supply grows by 4% with currency denominated debt

"What about someone wanting to retire?"

If you don't have money and the only thing money buys is haircuts, then you let your hair grow or perhaps you don't have hair anymore.

Or, you have a couple of friends that cut your hair for free, if there is sufficient excess barber capacity.

Come on, think through this.

It sure is interesting that the equation of exchange predates national income accounting, and yet in modern terms its treated as being related to GDP! rather than its classical definition as being related to aggregate transactions (actually much more sensible).

I don't recall the history well here. Is there a narrative that does along with this?

Jon: that is interesting. You can sort of think of V in MV=PT as the parameter needed to reconcile a balance sheet with an income statement?

But I think double-entry bookkeeping goes back a long way, and probably pre-dates MV=PT, even though national income accounting is more recent, as you say.

Winterspeak, "everyone in the united states cannot save more unless the Government runs higher deficits" seems weird to me (even with your definition of saving). What about the rest of the world in this?

Also, please explain what "vertical money" is. I clearly do not know how the monetary system works!

JKH: We may have a material difference in opinion. Let's see.

Traditionally, you measure "leverage" by seeing what the non-equity liability to equity liability ratio is. So 10:1 leverage means you have 10 units of non-equity liability, and one unit of equity liability. Balance sheet size is 11 units.

So, look at a combined Govt, non-Govt balance sheet. It nets to zero, of course, but the Govt sector pays NFA equity into the non-Govt sector. So, if you look at just the non-Govt sector, there are these equity liability accounts that do not have an in-sector liability match. The asset they match would be cash, I guess, or at least some sort of bank deposit.

The non-Govt sector cannot create an accounting entry by itself. Any financial asset created in the non-Govt sector must have a corresponding liability within the sector, so it has to net out to zero (within sector). I think we are agreed here.

So, the way I think about it is that the Govt pays in the initial equity (NFA equity) that the non-Govt sector can then expand balance sheets on top of through non-Govt sector credit extension. The non-Govt sector can extend additional credit, if it chooses, or uncreate credit. But it cannot increase its NFA equity. The analogy is of starting a company, paying in some initial equity (which shows up as cash and equity) and then the company can take on debt on top of that to expand its balance sheet. If the non-Govt sector wants to increase its NFA equity, it cannot do that by itself because it cannot create NET financial assets.

Clear? Do we disagree?

I very much agree that the quantity of NFA equity is small in comparison to the total size of the non-Govt sector balance sheet. In Australia, I believe the situation is more extreme, while in Japan the non-Govt sector probably has a much larger % NFA equity component. Also, I will repeat again, that this is just for financial assets, not non-financial assets. If you look at household net worth, and take out real estate, I think that number will drop considerably. My guess is that most households probably do not have significant net worth outside of their houses.

RebelEconomist: "Horizontal money" is financial assets created within the non-Govt sector, so non-Govt sector credit extension. It's called "horizontal money" because it's just expanding balance sheets by creating assets and off-setting non-equity liabilities. There is no net financial asset creation (there cannot be).

"Vertical money" is net financial asset creation where the Govt creates net financial assets in the non-Govt sector by deficit spending. This short up in NFA equity on non-Govt balance sheets.

You can agree whether this model makes sense or not, but that is what Chartalists are talking about when they talk about vertical and horizontal money.


Thanks for the explanation of horizontal and vertical money, which sounds reasonable. And my question about the rest of the world?

I do not disagree with the chartalist view of money at all. In fact, I have been aware of it for years, and think it offers useful insights into the nature and value of fiat money. I just am not convinced that it offers anything as new or powerful as its prophets and their disciples claim.

pebird, I am trying to think this thru so let me rephrase.

What about someone saving so they can retire so they can demand haircuts but do not have to supply any?

Sound better?

RebelEconomist:"And my question about the rest of the world?"

FYI:Scott Fullwiler once wrote about incorporating it here.

Rebel Economist: "Rest of the world" is just another non-Fed Govt saver (or currency user). When I say "Govt" I mean the Fed Gov, as it can issue currency. Everyone else, for that currency, is a currency user, whether they are a household, state Govt, or foreign Govt. As a short hand I refer to them as "private sector".

So does the government debt = the non government savings? Cause the market cap of the tsx > the government debt. That's not including private companies, and bonds. Has someone run the numbers?

That's what I thought, winterspeak. So you have consolidated out the US piece of the global imbalances. Does it not concern you that this makes such analysis rather irrelevant?

edeast said: "So does the government debt = the non government savings?"

How about does currency denominated debt of the gov't plus currency denominated debt of the lower and middle class = the savings of the rich?

Rebeleconomist: I'm not sure what you mean when you say that "consolidating out the US piece of the global imbalances... makes such analysis rather irrelevant". I think it's very relevant for certain questions, and I think other questions (like "what will happen if China stops buying Treasuries!") are fundamentally irrelevant -- which this analysis makes clear!

JKH: Thinking more about an earlier post of yours. You distinguish between reshuffling assets and saving, saying that the act of saving brings assets (and equity liability) onto the balance sheet (and is an income statement activity) while changing the composition of those assets (buying stock, selling real assets etc.) comes after.

Not sure I'm 100% OK with this, as you can change the composition of your assets in every way through private section credit extension, whereas adding to NFA equity is uniquely a savings activity, and impossible at a sector level.

Did I understand you wrong?

The conversation has moved far beyond this, but I still think that it is worth noting:

Nick Rowe: "By common stereotype, accountants are careful people; economists are usually sloppy people."


Nick Rowe: "When I teach Y=C+I+G+X-M I always explain it this way to my students: "try and find a fault in it. You can't. Because whenever you find something wrong with it, I will just re-define my terms to that it stays true."

JKH: "You’re just changing terminology. That’s not helpful, but the accounting requirement is that the equation works."

Nick Rowe: "But I was just trying to explain to my students how accounting works!"



There may be quite a difference in the way we view things at a high level, including a disagreement or disconnect on the nature of non government equity in total. (I can’t tell for sure whether it’s the words you put together that I disagree with, or the analysis itself.) Non government net financial assets is an important strategic component of the full piece, but it’s fairly straightforward from an accounting perspective. I think we mostly agree on that. The larger accounting framework that positions NFA in the context of the totality of economic equity is more challenging.

Your response suggests this view of total economic equity is something you don’t relate to. I don’t see where you’ve responded to many of the points I’ve made. I think you have a good feel for net financial assets/equity of the non government sector with the government, but I don’t think you have translated that unique piece coherently to a view of the larger picture. The only thing I can do is more or less repeat some of my analysis, because while we probably have a mutual understanding of the nature of the non government NFA position, I have considerable difficulty with some of the statements you’ve made in the larger sense.

The balance sheet of any non government economic unit consists of assets, liabilities, and equity. This works for households, corporations, and the foreign sector. Balance sheet equity is somewhat separable from the idea of whether or not a financial claim in the form of a stock has been issued on that equity position. While equity claims are obviously not issued by households, they have balance sheet equity or net worth positions essentially comparable in basic accounting to those of corporations. And the balance sheet equity of a corporation is conceptually distinct from its specific stock value representation in the sense that alternative measures of the same thing are possible – e.g. book value on the balance sheet as reported, replacement value for the net assets to which the stock corresponds, or breakup value on liquidation. The balance sheet equity position is what the corporation itself has generated. The stock market evaluates that position as a translation of it to a traded financial asset.

It is obviously true that net financial assets contracted entirely within the non government sector sum to zero. But the fact that this accounting mirror imagery is a wash for financial assets between non government counterparties doesn’t mean such financial asset values can’t and don’t factor into the value of non government equity in total. As a simple example, suppose a household holds corporate bonds as part of its net worth or equity. Those bonds are eliminated on asset-liability consolidation when one takes the standard MMT view of the non government sector. And that is obviously correct from a non government sector accounting perspective, insofar as the bonds and other financial assets are concerned. But the fact that a corporate bond is eliminated on accounting consolidation doesn’t mean that the household’s corresponding net worth (the household version of equity) is eliminated. The bond is simply a financial asset representation whereby the future cash flows expected to be generated by the corporation from its real resources in servicing its bond obligations are mirrored in the household equity position that reflects some type of valuation of those future cash flows. The ultimate wealth or equity position lies in the claim of the household on those cash flows, not in the mirror image accounting elimination of financial asset representations on macro consolidation. Indeed, this sort of representation of household equity or net worth is done routinely through the Fed flow of funds Z1 report. Moreover, that measure of household worth, plus foreign sector net worth, is an exhaustive measure of non government equity in total, by construction, because any corporate contribution is reflected in the value of the other two sectors using the value of their financial claims on the corporate sector. Perhaps you are not familiar with the standard sector balance sheet equity or net worth decompositions in that report.

More generally, following directly from the example above, non government equity in total can be represented as the sum of net assets of the household sector and the foreign sector. This refers to net assets in the general sense of the standard accounting mode of assets less liabilities, which of course is equity. This is because, similar to the bond example above, all corporate sector financial obligations (liabilities and equity) are ultimately reflected either directly or indirectly on a consolidated basis as household or foreign sector financial assets. Corporate sector financial liabilities that interface directly with the household and foreign sectors are specifically included in such a representation, because it includes all corporate issued debt and equity or other obligatory value (e.g. pension liabilities) as transmitted via financial claims to the other two non government sectors. Finally, residential real estate and consumer durables are non-financial assets that contribute as well to non government equity. Regarding the actual portfolio composition of the household and foreign sectors, and to repeat the kind of analysis contained in my previous comment, there’s currently about $ 35 trillion in net financial assets of all types held by US households (that’s net with all sectors; not just the government sector). Add to that about $ 18 trillion in residential real estate, some consumer durables, and some foreign sector net financial asset holdings, and you get about $ 60 trillion in total non government equity or net worth. Furthermore, the combination of household and foreign net financial assets in total approaches $ 40 trillion, of which about $ 7 or 8 trillion is the net financial asset position with the government.

With that said, the following statements you’ve made don’t make a lot of sense to me:

“So, if you look at just the non-Govt sector, there are these equity liability accounts that do not have an in-sector liability match. The asset they match would be cash, I guess, or at least some sort of bank deposit.”

“So, the way I think about it is that the Govt pays in the initial equity (NFA equity) that the non-Govt sector can then expand balance sheets on top of through non-Govt sector credit extension.”

“The analogy is of starting a company, paying in some initial equity (which shows up as cash and equity) and then the company can take on debt on top of that to expand its balance sheet. If the non-Govt sector wants to increase its NFA equity, it cannot do that by itself because it cannot create NET financial assets.”

As I explained in the previous comment, it’s misleading to suggest that non government net financial assets with the government are the structural basis for leverage within the non government sector. That seems to ignore nearly $ 50 trillion in non government equity that is separate from the $ 7 or 8 trillion of equity corresponding to its net financial assets with the government. This $ 50 trillion in equity value has nothing to do with and is unaffected by the accounting observation that net financial assets within the non government sector net to zero. The equity value of the non government sector is not affected by this zeroing out accounting consolidation effect.

As noted before, net financial assets provided by government deficit financing are obviously useful as a response to non government demand for net saving, particularly when the latter sector shows weak demand for goods and services. The stimulus provided by government expenditure in doing so obviously helps fill that goods and services demand gap. And in doing these things the government is leveraging its own financial capabilities as an offset to the deleveraging that is going on in non government balance sheets. I believe this is all standard MMT. But to suggest that the “net financial asset equity” or NFA equity, as I coined the term earlier, is some sort of structural equity foundation for the non government sector, when the existing equity base of the non government sector without it exceeds $ 50 trillion, is simply a distortion of the proportionate presence of the two equity components. You periodically make a statement along the lines that you only “care” about net financial assets of non government with government. I think this narrow approach contributes to error in some of the more general statements you make about equity. So it appears we disagree on this subject of equity in a rather fundamental way, unless you’d like to revise the way you’re analyzing it and/or describing it. While I think you have a good feel for net financial assets/equity of the non government sector with the government, I don’t think you have translated that piece coherently or consistently to a view of equity in the larger picture. BTW, I regard absolutely nothing I’ve written here as being inconsistent with MMT. Anybody who wants to suggest otherwise is welcome to explain how they believe that is the case.

And what about the state equity on the balance sheet of the US non-state sector - ie the fact that the people ultimately own, and are responsible for, their own state sector? That would be one reason not to consolidate foreign and US sub-sectors of the non-state sector.

Rebeleconomist: Still not sure what you mean. If the Chinese choose to hold US$, that is their prerogative. The US Govt is ultimately responsible for the US$, just as the Chinese Govt is ultimately responsible for the yuan, no matter who is holding it. But you still need to know how is holding it and be honest in your accounting.

Many countries operate with mixed currency economies, it really isn't that strange. The US does not, although Canada might to a certain degree (it would not surprise me if there are a fair number of US$ circulating up there).

JKH: Give me time to process. I'm glad we've uncovered this area of disagreement.

JKH: Still digesting your comment. let me ask you this.

Suppose you closed out all non-Govt financial positions. So, corporations bought back their stock, households repaid their debt, etc. etc. What financial assets would remain in the private sector?

Of course prices would go crazy if you went through this procedure, so just tell me what line items would remain. Remember, I don't care about real assets.

I mean that, since the people ultimately own their state, the state cannot really create net financial assets by deficit spending on purchases from its own people (assuming that the citizens' equity stake in their own state is a financial asset).

Winterspeak @ 6:00 p.m.

The answer to that is beyond obvious.

Why are you asking that question?

JKH: With respect, your responses are, in general, pretty dense. It's easy to get lost as you work through them. Always worth the effort though, so please do not take this the wrong way! Strangely enough, I understand you more easily when your posts are very short.

As I move through your response, I'd love to get feedback on whether I am on the right path or not. Thus I asked the obvious question.

Clearly, you and I are not on the same page wrt to non-govt equity. You don't always specific whether you mean *net* non-govt equity, or non-Govt *financial* equity, or net non-govt financial equity. Are those omissions intentional, or just to improve readability? Just wanted to be sure.

If your point is that there are enough real assets in the non-Govt sector that they contribute a meaningful equity base to leverage on top of, and thus are a meaningful addition to NFA equity from deficit spending, that's certainly an interesting point. You would be saying that, although we cannot ascribe an exact price to them, they have SOME value, maybe even a lot. Maybe even enough to swamp the value of NFA equity.

But I don't know if that's what you mean.

Winterspeak @ 6 p.m. # 2

By obvious, I meant the following:

First, you specified in your question the elimination of all non government financial assets. And then you ask what non government financial assets would remain. You answered that in the question – zero.

Then you ask for what remains, but you’re not interested in real assets. I’m sorry, but you can’t understand the world if you’re going to impose such overwhelming constraints on the examination of it. It’s infuriating. Believe or not, MMT is not the world in its entirety. It has a context. That’s much of my point. Given the elimination of all financial assets, is it likely that a sensible answer to the question excludes at least a reference to real assets?

So here’s the answer:

What remains is that the non government sector has real assets offset by balance sheet equity. In addition, the non government sector has net financial assets with the government sector, offset by the same amount of additional balance sheet equity.

I was interested in knowing where that would get you, which is why I asked why you were asking the question.

Now I’ll have a go at your 7:47 p.m. question.

Winterspeak @ 7:47 p.m.

I mean non government equity. I mean what’s on the right hand side of the consolidated non government balance sheet.

The right hand side is a gross equity position.

The left hand side consists of financial assets with the government and real assets, when everything is consolidated, as per the previous question.

The only time I’ve used the phrase net in connection with equity is with respect to what I termed NFA equity, which corresponds to the piece of equity that offsets the net financial asset position with the government.

The reason what I write is “dense” is that I tend to take the time to define and explain things.

I don’t know what those various terms you’ve used above mean. You’d have to explain that to me, not vice versa.

Your other comments suggest you’ve barely absorbed what I’ve written to date, and I’ve written a lot, including many references to the various numerical quantities involved. It appears you taken little note of the relative magnitude of these numbers, since they refer precisely to the magnitudes you’re wondering about.

"The only time I’ve used the phrase net in connection with equity is with respect to what I termed NFA equity"

Correction. I may have used it in the sense of sector positions. E.g. the foreign sector (i.e. foreign sector to the US) has a net financial asset position with the counter party combination of the domestic private sector and the government sector. That position constitutes cumulative saving, marked to market, through cumulative US current account deficits. I may have used the phrase net equity to describe such foreign sector cumulative saving or wealth with respect to its position with the US; I don't recall. But its consistent usage of the term net. It's just that its not net with respect to only the government.

JKH: Yes, real assets were something that I had not paid any attention to.

As far as the magnitudes, I honestly do not know what to make of them. Even if (and I realize this is not the case) NFA equity was the only sliver of the capital base that private sector credit was heaped upon, I honestly do not know what is the "right" amount, and if a given amount of leverage is too much, or not. V has, and can, change pretty dramatically. How much of that is tied to the NFA equity leverage number, I really honestly do not know. The Federal Govt right now is trying to re-ignite private credit extension, while that sector de-leverages after a credit binge. Is that a viable strategy? Again, I do not know.

I do know that the price of real assets depends on the amount of credit available. The cash buyer has a very different willingness to pay than the leveraged buyer. Inflation/deflation price adjustments become really important when real assets come into play, which is one reason why I was setting them aside. You don't need to worry about that kind of thing when everything is nominal.

You really shouldn't get infuriated with me. There are better targets for you to vent your wrath at.


This was a post about the relationship between economics and accounting. The “Modern Monetary Theory” or MMT brand is built on a foundation that includes the accurate presentation of a particular area of accounting focus – that of central bank reserves, deficits, and the banking system. The MMT brand is also associated with a style where its proponents generally rebuke the neoclassical thinking of those economists who simply don’t understand operations and the accounting for the monetary system. This is very understandable. The way I think about it is that it’s very unlikely that an area of economic thinking will hold much interest for me if I happen to know that its premise is riddled with misunderstanding as to how existing modern monetary systems actually work. And I believe MMT is very accurate in its understanding of the accounting issues that are relevant to its focus. But this also sets a higher bar for MMT proponents, of which you are clearly one, in terms of their own understanding of the accounting issues. The particular accounting focus of MMT is in fact a small subset of accounting as it applies to economics more broadly. To the degree that MMT proponents begin to make more sweeping statements about accounting interpretation, it should be incumbent among them to ensure that such statements accord with the facts of accounting outside of the MMT focus. If such statements are inaccurate or misleading, it becomes a case of the pot calling the kettle black. That warrants strong push back.

JKH: I get your point and I accept it. I genuinely don't know how to handle real assets because of the valuation issues I raised in an earlier post. The amount of leverage you extend against a real asset depends on the value of that asset, and the value of that asset depends on the amount of leverage available to be extended against it. It's very circular, as anyone who has lived through the past 10 years can attest.

So, if you close out all private sector credit, you'll be left with a consolidated balance sheet that includes whatever NFA equity paid-out by the Govt, plus real assets. I absolutely do not know how to guess at the $ value of those assets, but I would start by valuing them at what the non-leveraged buyer would pay. And that, at least in part, would be driven by discount rates, so I see a clear role for monetary policy in private sector credit extension (and I have no idea if MMT adherents would agree). It would also be driven, in part, by the quantity of NFA equity in the private sector.

I don't know to what degree accounting can help here, as there are lots of accounting models that deal with precisely this issue, and all have their pros and cons. I am open to suggestions.


Let’s stick with the accounting first.

The Fed flow of funds report Z1 handles corporate real assets by passing them through the lens of financial assets. The non-corporate sectors hold financial claims on corporations, including their real assets. So the direct valuation of corporate real assets is avoided. The value of real assets is reflected through the liability and equity claims side of corporate balance sheets. That’s how non government equity is ultimately valued in such a way that it takes into account the effect of real assets held by corporations. As far as household real assets are concerned, that is captured directly as the value of residential real estate and consumer durables.

So if you assume a closed economy with a net zero foreign sector, non government equity includes:

a) Household real assets such as real estate and consumer durables

b) Net household financial assets such as stocks, bonds, mutual funds, pension and insurance values, less any liabilities such as mortgages and consumer credit.

Note importantly that the definition of net household financial assets as in b) automatically embeds as a subset the net financial asset position of the non government sector with the government sector. E.g. household direct holdings of government bonds are included. E.g. pension fund holdings of government bonds are included by virtue of the indirect reflection of that value through a household financial asset equal to the pension fund obligation to that household

When you add in the foreign sector, note that the foreign sector is itself a net financial asset position, but again in the more general sense like households. And again, any non government net financial asset position with the government is embedded in that more general net financial asset position. E.g. China’s holdings of government bonds are part of that foreign sector net financial asset position.

So that’s the larger, global accounting context for the handling of real assets and everything else as I see it.

Beyond the accounting, you get into judgements about leverage strategies, which I’d prefer to leave as a separate discussion until we’re on steadier ground with respect to the global accounting, in which MMT accounting is embedded. The proper global accounting is as essential a prerequisite to that integrated leverage discussion as is MMT focused accounting to the proper discussion of government deficits.

I guess I'm saying I'd like to see your question put in the context of a more mutual understanding of a global accounting framework that is bigger than MMT (I mean global conceptually rather than geographically, although that too). That said, your question may disappear in a sense, because the original issue of contention was the interpretation of MMT NFA equity versus equity in the larger global context as I've defined it. That larger global equity position includes the reflection of real corporate assets through a financial asset lens. This has essentially been an accounting issue that impedes better discussions about strategies such as leverage. If that obstacle can be removed, the discussions about strategy flow much easier I'm sure.


This is quite good.


You're on his favorite blogs list, so I find the topic a little more than coincidental, including a haircut example.

JKH and winterspeak, IMO people need to look at the situation like this. There is a domestic lower and middle class, a domestic gov't, and domestic rich. There is also a foreign lower and middle class, a foreign gov't, and foreign rich. Now add in retirement and its economics.

Here is one question. How should retirement be accounted for?

Ah yes, I was about to post the link to Andy Harless. It seems pretty likely he must have been reading this conversation or at least some of Nick's previous posts (and the comments).

Fascinating discussion, though I'll need to set aside some time to read all of it.

In the meantime a note on Krugman: he does indeed state that the deficit enables private saving rather than vice versa. It's implied in:
and in:
and stated fairly explicitly (with a link to Brad Setser) in:

I believe Krugman is a closet or un-selfdiscovered MMT'er, and a potential valuable ally to their cause.

But Warren Mosler, who is as out front an MMT'er and leader of their cause as there is, doesn't:


Leigh Caldwell, I had a look at the 3 posts.

I would be interested to know what the graphs would look like in the second link ( http://krugman.blogs.nytimes.com/2009/07/15/deficits-saved-the-world/ ) if the private sector "surplus" was broken down into the rich and the lower and middle class.

JKH: Yes, I had read Andy's post. But this is well-trodden ground in macro. Nothing I said about S=I is original with me. I stick close to the mainstream on this topic, except I depart a bit when I start talking about money.

Paul Krugman is writing about S-I=G-T within the context of an economy in a recession with the rate of interest stuck exogenously. He would almost certainly say very different things about causation in a different (normal) context. He might sound a bit "MMT-ish" in that context, but there is no way he would say the same things otherwise.

Semi 0ff-topic. I just realised where I had heard the name "Warren Mosler" before. He's the guy that makes really neat cars! I had read about him on car blogs, "PistonHeads" I think. I wonder how he's getting on with GM (not General Motors)?

"Paul Krugman is writing about S-I=G-T within the context of an economy in a recession with the rate of interest stuck exogenously."

How about is this correct?

(S-I) of the rich = (G-T) minus (S-I) of the lower and middle class

Too much Fed: Yes, that's correct. So is:

(S-I) of men + (S-I) of women = G-T

(S-I) of right-handed people + (S-I) of left-handed people = G-T

Which illustrates my point: the accounting identities that are useful to you depend on your theoretical perspective.


Apart from making good cars, Warren Mosler also writes well. You may want to look at "Seven Deadly Innocent Frauds" http://mosler2012.com/wp-content/uploads/2009/03/7deadly.pdf

Ha! As I pasted this link, I got reminded.. he is also running for the President for 2012.


Re Krugman, I was agreeing with Leigh’s comment.

Also, the following:



Thanks for hosting this discussion. I hate to "put you under the microscope", as it were, but I find your inability to "get it" fascinating (and somewhat disheartening). I mean, you're not as arrogant as someone like Krugman or Delong, and you seem genuinely to want to engage with new ideas. But I am forced to conclude that an early exposure to neoclassical thinking leads to economic brain damage (it's like learning BASIC as your first computer language...)

Thanks Jim! I can't afford to be as arrogant as PK or BDL; I'm not as good as them!

As you get older and more out-of-date, it's actually easier to be less hostile to new ideas. It's the young bright guys, fresh out of grad skool, who have most to lose if a new idea makes worthless all their highly specialised theory-specific human capital. All we old guys have left is our general human capital, because all our specialised knowledge got blown out of the water decades ago. (I got that idea, which I think is brilliant, from an even older guy - John Chant.)

I DID learn BASIC as my first (and only) computer language. How did you guess? Funnily enough though, my earliest exposure to economics was very Keynesian.

Given the topic of this post, I can’t resist commenting on the following. It’s a staggering example of a very well known economist who hasn’t a clue as to the basics of financial accounting. Here’s Mark Thoma on excess reserves:

“the Fed has injected liquidity into the system by increasing bank reserves substantially — massively is a better description — and the result is that a very large quantity of excess reserves has accumulated within the banking system. Reserves sitting idle within the banking system, as they are now, are not much of a problem and they provide insurance against unexpected losses in other areas of a bank’s balance sheet.”

In this case, we have one of the most widely read economists in the blogosphere not understanding that profit and loss flow through the income statement and the capital account – not central bank reserves, for goodness sakes. This is a truly mind boggling error.

The rest of his diagnosis of excess reserves is wrong as well, in a way that extends beyond the accounting alone. It’s the usual stuff. I don’t have the patience to go into it.

Here’s the complete mess:


JKH . . . all I can say is, wow!

Frightening, Scott.

OK, I confess. I don't get what's wrong with what Mark Thoma said either, in that paragraph you quote.


It's because "excess reserves" just change around the content of the banking system's balance sheet, they don't provide "insurance" against losses. Let's say you owe $100000, have $50000 in a savings account, and $60000 in a checking account. If I transfered $10000 from the savings to the checking, would you say you were "better insured against losses"?

The other problem in Thoma's presentation is that he implicitly clings to this fractional reserve nonsense, implying somehow that these reserves affect lending decisions by a bank. A bank will make a loan (and create a deposit) if it finds a good credit risk. That means the borrower is "creditworthy". The whole top down premise that somehow you shove enough money at the banks and build up their reserve positions and force them to lend is based on a completely wrongheaded paradigm.

Trying to get anyone to understand that Treasury "borrowing" is just re-arranging assets has been near impossible, in my experience.

And that's a very simple transaction compared to loans creating deposits, for example! Or deficit spending funding NFA equity.

Thoma has no clue, and is in excellent company.

My post said: "(S-I) of the rich = (G-T) minus (S-I) of the lower and middle class"

Nick's post said: "Too much Fed: Yes, that's correct. So is:

(S-I) of men + (S-I) of women = G-T

(S-I) of right-handed people + (S-I) of left-handed people = G-T

Which illustrates my point: the accounting identities that are useful to you depend on your theoretical perspective."

I am trying to approach this from a "what's happening in the real world" perspective.

Can you see that when the lower and middle class stopped going into currency denominated debt to the rich/politically connected domestic/foreigners (and some even tried to default) that the rich/politically connected domestic/foreigners got the gov't to bailout them out of some of the bad currency denominated debt and got the gov't to go further into currency denominated debt for the lower and middle class so the rich/politically connected domestic/foreigners could maintain their excess savings?

I am going to post this and see what everyone thinks.

Jim Baird said: "It's because "excess reserves" just change around the content of the banking system's balance sheet, they don't provide "insurance" against losses."

It seems to me that bank reserves are debt instruments (let's say fed debt) and that the U.S. gov't won't allow the fed to default/fail. If so, does swapping bank reserves for bad debt just mean that the fed is trying to "hide" the bad debt until more can be created to offset the losses?

Plus, if the gov't won't allow the fed to default/fail, are bank reserves actually gov't debt (possbily future gov't debt) in disguise and is "insurance" because the gov't debt "socializes" the losses on taxpayers?

Marshall Auerback, so do you think Steve Keen is right? and from:


"In the real world, banks extend credit, creating deposits in the process, and look for reserves later."

And, "Thus loans come first—simultaneously creating deposits—and at a later stage the reserves are found. The main mechanism behind this are the “lines of credit” that major corporations have arranged with banks that enable them to expand their loans from whatever they are now up to a specified limit.

If a firm accesses its line of credit to, for example, buy a new piece of machinery, then its debt to the bank rises by the price of the machine, and the deposit account of the machine’s manufacturer rises by the same amount. If the bank that issued the line of credit was already at its own limit in terms of its reserve requirements, then it will borrow that amount, either from the Federal Reserve or from other sources.

If the entire banking system is at its reserve requirement limit, then the Federal Reserve has three choices:

refuse to issue new reserves and cause a credit crunch;
create new reserves; or
relax the reserve ratio.
Since the main role of the Federal Reserve is to try to ensure the smooth functioning of the credit system, option one is out—so it either adds Base Money to the system, or relaxes the reserve requirements, or both.

Thus causation in money creation runs in the opposite direction to that of the money multiplier model: the credit money dog wags the fiat money tail."

Marshall Auerback said: "A bank will make a loan (and create a deposit) if it finds a good credit risk. That means the borrower is "creditworthy". The whole top down premise that somehow you shove enough money at the banks and build up their reserve positions and force them to lend is based on a completely wrongheaded paradigm."

It seems to me we are headed to a two tier credit system. If a bank is politically connected and does what the fed wants, the gov't/the fed will bail you out of bad debt. If a bank is not one of the two, it is on its own.


I agree with the comments of Marshall Auerbach, Jim Baird, and Winterspeak. They’ve emphasized the general theme of reserves, also noted many times in comments throughout your Chartalism and Accounting posts. Jim Baird also hits on the salient accounting point. Thoma’s errors are not that surprising; it’s a recurring pattern with him.

I quoted the part with the insurance reference because it was so apocalyptically egregious from the perspective of the theme of your post, which as I saw it opened up the question of appropriate integration of economic and accounting analysis.

To the degree that that Thoma’s insurance analogy is invoked, its correct application is very different in the case of each of central bank reserves and bank capital.

The insurance concept with respect to central bank reserves has to do with liquidity risk. From a systemic point of view, the central bank supplies a sufficient amount of reserves in order for the overnight rate to trade at or near the policy target rate. In times of crisis, this may require more reserves, because banks may be hoarding reserves. The reason they’re hoarding is due to an abundance of caution with respect to their own ability to attract sufficient funds in order to maintain their reserve accounts at required levels. This was a big factor in the early days of the crisis. To the degree that banks were cautious in such hoarding, they could be said to be “self-insuring” their own liquidity risk. To the degree that the central bank responded, it could be said to be the “insurer” of that risk for the system as a whole. As the crisis progressed, and as additional central bank initiatives took effect, that liquidity insurance motive took on less importance.

As the crisis evolved, the Fed increased excess reserves as a natural consequence of its initiatives on the asset side of its balance sheet. It created new money through its own credit expansion. That money came back to its liability side as reserves. The Fed chose to leave those reserves in place and pay interest on them. Thus the excess reserve strategy was eventually associated more the balance sheet requirements of central bank asset management, than those of commercial bank asset management.

On the other hand, the insurance concept with respect to bank capital has to do with risks other than liquidity. The purpose of capital, literally, is to absorb unexpected financial losses. These losses have to do directly with risks other than liquidity – credit risk, interest rate risk, foreign exchange risk, etc. Insurance against losses is equivalent to a put option, with a financial payoff in response to a certain event. The event in this case is a financial loss due to risks such as those noted. The payoff is the charge to capital (allowing for the effect on capital position of any current period net income).

Thus, there are two broad categories of banking risk – liquidity risk, and a group of risks that have to do directly with financial loss. That second group is sometimes rolled up into the terminology of capital or solvency risk. Liquidity and capital risk can display interactive dynamics, particularly in financial crises, but they are distinct risks. To say that this is a huge topic on its own would be an understatement.

Capital and liquidity are distinct also from an accounting perspective. Accounting statements are in a sense the ex post representation of realized risk. You can’t think properly about risk in the future, unless you can project conceptually the ex post accounting representation of various possible outcomes of that risk.

Capital risk, and the risks that underlie it, are represented ex post in the income statement and balance sheet. Any financial loss will be entered as a charge to income and/or capital.

Liquidity risk is represented ex post in the balance sheet and the sources and uses of funds statement. E.g. a bank that has accumulated excess reserves through deliberate hoarding, or through relatively passive participation in a systemic excess reserve environment will show its share of reserves as a balance sheet asset. The change in that “stock” or asset will be reflected as a “flow” in the sources and uses of funds statement for the relevant accounting period. The principal amount of such flows will never be recorded on the income statement. The only related aspect that touches the income statement is the interest income paid on reserves.

That brings us back to Thoma, who said that “reserves sitting idle ... provide insurance against unexpected losses in other areas of a bank’s balance sheet.” In addition to the fundamental background errors noted by the commenters above, and following from Jim Baird’s more specific comment, this statement by Thoma is an almost inconceivably incorrect conflation of entirely distinct accounting measures. To use his insurance analogy, the insurance functions pertaining to each of central bank reserves (i.e. liquidity) and capital should be entirely different, as described above. Central bank reserves are not some repository of net worth to be used as a cushion against losses. Thoma’s confusion on how to read a balance sheet correlates strongly with his misinterpretation of excess reserves. That seems quite relevant to the topic of your post.

The following should be of interest regarding the Thoma piece:


And, Too Much Fed, that quote from Keen is correct.

Regarding Winterspeak's point, I just did a post this week on that precise issue with the accounting operations (they're quite simple, but most don't know how to do them) and it was posted at various places around the web. A number of commenters claimed it was "dangerous" or "irresponsible" for me to explain this.

Saw your post, Scott, which I thought was very good. Some interesting subtleties there about bank versus non-bank purchases of bonds.

That's an odd reaction from those commenters.

Thanks for noting Bill Mitchell's piece. Hadn't seen it yet.

Scott: Somehow I missed all your posts. Can you point me to one? Thanks!

Happy Thanksgiving everyone

The original post from this week is here:


That was my first one since July, actually.

With a thumbs up from "Flow5".

Now that's a compliment!

Some here will be interested in the "discussion" going on at:


Thoma appears more defensive than normal. I'm collecting quotes from big name economists who have said things that are quite embarrassing if an understanding of reserve accounting ever becomes widespread. Thoma should have his own page on my list after the past few days.

Commercial bank reserves at the central bank are the most liquid form of asset they can hold. And also the safest (except for the inflation risk, but since their liabilities are subject to that same inflation risk, that's probably a good thing). So an individual bank is safer (with respect to both liquidity shocks and shocks to the value of its risky assets) if it has a higher proportion of its assets held as reserves than if it lent them out in some sort of risky and illiquid loan. Isn't that what Mark Thoma meant?

OK, to say they are "insurance" against other risks isn't strictly accurate. If I have a safe asset, equal in value to my house, that isn't the same as having insurance against my house burning down. But I will accept the word "insurance" as a metaphor here.

Where I might disagree with Mark Thoma is that this might be true for an individual bank, but I am less sure if it is true for the commercial banks as a whole. (If they all expanded loans, would this make economic recovery stronger and reduce risk?) But that's a different argument.

“So an individual bank is safer (with respect to both liquidity shocks and shocks to the value of its risky assets) if it has a higher proportion of its assets held as reserves than if it lent them out in some sort of risky and illiquid loan. Isn't that what Mark Thoma meant?”

On an individual bank basis, that may have been what he meant, but it’s wrong relative to the non-liquidity risks that involve financial losses, as I discussed above. Reserves are zero risk weighting, and require no capital underpinning on a risk weighted basis. They are neutral in the sense of both risk taking and associated capital requirements.

The same bank with risky assets instead of reserves would require additional capital. So the choice between the two asset alternatives as presented is a false one, because one of them requires additional capital.

The correct comparison is between two different banks (effectively) with two different risk asset portfolios and two different capital requirements. The fact that they have two different reserve positions is irrelevant to the assessment of non-liquidity financial risks, which is the relationship to which Thoma was referring.

Let’s call a spade a spade. To say reserves are insurance against other assets is strictly inaccurate. The insurance metaphor is wasted if used nonsensically.

OK. I've now read the whole of Mark Thoma's post. It looks OK to me. (I might have a little less faith than him in fiscal policy, and more in monetary policy, but that's a separate argument).

But on the argument in the comments, with R. Winslow, I think they are talking past each other. At root of the misunderstanding is the accounting/economics distinction between quantity sold and quantity supplied. In this case the quantity in question is reserves. R. Winslow (who must be one of you guys, under a different nom de plume?) is talking about reserves *sold* by the banking system; Mark is talking about reserves *supplied* by the banking system. And YES, they are different. If (like the French) we used the word "offer" instead of "supply", (and if economists were more careful with their language in making this distinction), the misunderstanding would disappear. And it's compounded in this case by the distinction between an individual bank, which can lose reserves, and the banking system as a whole, which cannot (unless there's a cash drain or an OMO, or something).

Oh god, I expect I should do a post on this, to let you accountants and MMT'ers have another go!

Scott Fullwiler, I thought that "In the real world, banks extend credit, creating deposits in the process, and look for reserves later." was correct.

I scanned your post "What If the Government Just Prints Money?"


First, I don't believe anyone should say the phrase "prints money". It is not specific about what is happening. I believe it should be what if the gov't prints currency, what if the gov't attempts to sell gov't debt denominated in currency, and what if the gov't allows the fed to "print" bank reserves to attempt to produce private debt denominated in currency.

Since the fed can't print its own currency and IMO the gov't won't allow the fed to default/fail, bank reserves are actually gov't debt in disguise.

Also, "As such, this post considers whether a given deficit resulting in more reserves in circulation and fewer bonds held by the non-government sector raises the likelihood of spiraling inflation, ..."

Do more bank reserves (IMO fed debt, which is gov't debt in disguise) and fewer bonds mean the fed itself is becoming "riskier"?

Mark Thoma said in his post: "When someone starts from perspectives such as that (and with notions such as reserves don't matter for loans and there is no money multiplier), there is nothing for people to learn from shining a light on these beliefs."

How about in certain situations there can be no money multiplier (as in the money multiplier is zero)? Or, is it more like there CAN BE no debt multiplier?

"More nonsense. I usually let Winslow's silliness go, but not this time. On the last point, in a liquidity trap, the interest rate can't fall to clear the market (so quantity is demand determined). How hard is that to understand?"

How about there is a problem in a different market besides the "interest rate" market?

St. Louis Fed Series: MULT, M1 Money Multiplier



I’m not sure what point you are making in responding to the MMT related criticism of his Thoma’s piece, either above or on his blog. But he likes it.

I can only comment additionally about his 5.1 per cent example, which you referenced. It is wrong as an illustration of what it purports to illustrate. I don’t know whether you or he actually understand that the Fed sets the floor for the policy rate using the interest rate on reserves (in today’s excess reserve environment). I can’t decipher the exchange there. I think you probably do. But in either case, that’s not where he’s really wrong in that example.

Again, he doesn’t acknowledge issues related to capital or risk, which are fundamental to understanding the limited role of reserves. And as a result, he presents false choices about bank lending.

The choice is not between leaving reserves at the Fed at 5 or 5.1 per cent versus lending money to a business or consumer at, or below that rate, which is the example he portrays.

The only legitimate choice using an example with those kinds of rates is between leaving reserves at the Fed or lending them overnight in the interbank market – i.e. the choice is between receiving interest on reserves or receiving the fed funds rate. That’s the choice for comparable risk, and it’s the only choice where a bank has alternatives for transacting directly in reserve funds. Banks don’t lend reserves to non-bank customers. That’s a factual observation from MMT and one of the aspects of the monetary system that Thoma ignores or doesn’t know about.

So the choice he does present is completely false. Banks don’t lend to businesses and consumers at rates that are comparable to either the interest rate paid on reserves or the fed funds rate. They lend at rates that reflect risk – at rates that reflect credit risk premiums, maturity risk premiums, and liquidity risk premiums. They lend at a spread.

Mostly importantly, banks lend or acquire risk assets based on the availability of capital to support that risk. That’s why they charge the interest rate premiums they do.

And that’s why banks don’t undertake risk lending as a function of reserves. Banks are capital constrained but not reserve constrained in taking on risky assets.

I.e. banks don’t make decisions on risk lending by looking for arbitrage opportunities between interest on reserves and the Fed funds rate, as Thoma portrays it. That’s an absurd argument.

Again, for the nth time unfortunately, MMT detractors don’t acknowledge the difference between central bank reserves and bank capital (i.e. risk lending according to credit standards). In short, they don’t acknowledge or understand the topic of risk. The relative uselessness of reserves in the function of lending to acquire assets with credit risk (and other risks) is a fundamental point made by MMT.

Nick, I’m not seeing a connection between your comments here (or at Thoma’s) and the MMT related points on Thoma’s piece. And I’m not sure declaring victory over Winslow resolves it. I’d be interested if Scott F. or some of the earlier commenters here can add to this.


" I don’t know whether you or he actually understand that the Fed sets the floor for the policy rate using the interest rate on reserves (in today’s excess reserve environment). I can’t decipher the exchange there. I think you probably do."

Yes. I understand that. The Fed's (now) just like the BoC.

Sure, there's a spread. Mark ignored that spread for simplicity. Economists often talk like that. Add in a spread, and it doesn't affect what he said, in terms of the directions of change.

Suppose banks are capital constrained, and that's an absolutely hard constraint (as opposed to a trade-off). They can't lend a single $1 more without more capital. OK, if the profit opportunities are there, they get more capital. That's like saying an increase in the price of apples won't cause orchards to want to sell more apples, because they don't have the trucks to take the extra apples to market. Buy/rent more trucks. If the price of apples is high enough, it's profitable to do that.

Must go to work.


If I may, I'll throw in my 2 pence.

I think the issue that is hanging everyone up is that Nick and Thoma have in the back of their minds a sort of gold standard analogy. Under the gold standard exchange was facillitated by notes with a promise of gold convertability but only fractional backing. Nick and Thoma are trying to make an analogy where reserves play the role of gold and credit plays the role of the notes.

I think that the reason they keep gravitating to this conceptual framework is that everyone wants the government to have a nominal budget constraint, for some reason people don't like the idea that governments can have real constraints but no nominal constraint (which is how I like to phrase the issue). It's their insistence on this conceptual framework that leads them to continue to insist that the quantity theory makes any sense.

This brings up two points:

1) Their is NO direct analogy between a fiat money system and a commodity backed system. (With a commodity backed numeraire the government's real constraint shows up in nominal terms, this doesn't happen in the fiat system no matter what numeraire you actually use to express things).

2) Keeping point 1 firmly in the front of our minds, if you put a gun to my head and force me to make the best gold standard analogy I can then it's capital that plays the role of gold. The reason for this is that capital is the real side quantity, it is in positive and limited net supply. Reserves are zero net supply, that's why they don't constrain anything.

I'm trying here to translate JKH and friends' point into a language more natural for economists (these MMTers annoy me to no end by insisting on putting everything in accounting terms), I'm quite certain they'll (all) tell me if they disagree.

Adam P.,

That’s an amazingly insightful comment, IMO. You’ve put all the pieces together in a unique way.

I’ve not yet seen that analogy between the gold standard and capital. I’d have to think about it a bit more, but my initial take is that it’s brilliant, IMO. I’ll come back on that.

On the accounting issue, it pains me in a way. I’m not an accountant, but logical accounting is a critical input for the MMT story. The reason it gets in the way so much is that MMT critics (proactive or de facto) inevitably don’t understand the accounting that’s necessary to understand MMT. And they don’t understand the degree to which their failure to understand the accounting for the monetary system has contaminated their paradigm for economics. One of the reflexive defences is to dismiss the importance of the accounting by resorting to the use of the term “accountant” in the pejorative sense. That’s extremely unhelpful and ill advised. You’re not doing that, but I can understand your frustration with the recurrence.

Accounting is the mechanical under body of MMT. The really important ideas are those that you’ve summarized so well.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad