The “hot potato effect” explained

Many people continue having trouble with the “hot potato effect,” which to me is the sine qua non of monetary economics.  You either understand it, or you understand NOTHING.  They want explanations they can understand at the individual behavior level.  But that just won’t work in this case.  Part of the problem is that people want to think in terms of a Real —> Nominal transmission mechanism, which seems natural in a world of sticky prices. “Why would Joe buy a new car?”  But that won’t work either.  It’s a dead end.

I’ve tried to explain the hot potato effect using analogies with the apple market, but I think I’ve found a better way: gold.  When compared to apples, gold has properties much more similar to money.  Even better, it used to be money!

I’ll walk you through the HPE effect step by step, and you tell me where you get off the train:

1.  Assume gold sells for $1200.  Interest rates are low, so the expected future price is roughly the same, maybe a bit higher.  Now assume a company discovers a gold hoard so vast that world output soars.  Over a period of years the extra output causes gold prices to fall in half.  But why?  Because before the discovery people were already in equilibrium, they held as much gold as they wanted to hold at existing prices.  The extra gold is a sort of “hot potato” that people try to get rid of.  But obviously not by throwing it away!  They get rid of it by selling it.  But notice that while that works at the individual level, it doesn’t work in aggregate.  Now someone else has the extra gold. (That’s why attempts to understand money at the level of the representative consumer fail.)  The only way for society as a whole to get rid of the extra gold is by driving down the price of gold until people want to hold the new and larger quantity.  Assume the price falls in half.  That also means the value of gold relative to other goods and services falls in half.

2.  Same thing, but assume the company merely announces the big gold discovery, but it is credible.  Now gold prices would plunge on the announcement.

3.  Same thing except the company announces that radar has detected the vast gold deposit, but it will take 2 years to dig down and get it out.  Again, prices will plunge right away, almost as much as in case 2.  (I hope everyone is still with me, this is just microeconomics 101.)

4.  Now let’s do the same three examples but assume gold is the medium of account.  How does that change things?  Obviously the price of gold can no longer plunge, as the price of the MOA is fixed by definition.  But it’s value will still fall sharply, as the price of other goods and services rise.  Now for the first key difference:  In cases 1 through 3, the fall in the value of gold was nearly instantaneous.  Now with sticky prices the fall in the value of gold (the MOA) will be more gradual.  But the long run effect will be the same.  Still with me?  The smartest people that don’t agree with me on the final steps (people like John Cochrane) would still be with me here.  As I recall Cochrane is essentially a “market monetarist” for a commodity money system.

Notice that so far these changes can be fully explained using the basic principles of microeconomics.  There is no need to resort to “transmission mechanisms” such as interest rates.  Just S&D.  However the following is important; if prices are sticky then other things will change to bring about a short run equilibrium in the gold market, before the price level has had time to fully adjust.  And obviously one of those “other things” might be a change in interest rates.  Other “other things” include changes in asset prices and real output.

But I keep coming back to the notion that no matter how important those other factors are, the HPE is still in some sense primary.  It’s what explains the long run effect of more gold on the value of gold.  It’s still econ 101.  The other effects are side effects that help nudge us toward that final equilibrium.

5.  Now switch the medium of account from gold to cash. I claim that changes nothing essential.  There are three cases:

a.  Positive interest rates:  The base becomes a hot potato, just as in the previous examples.  IOR changes that slightly, but less than you’d think.  Peter Ireland showed that the quantity theory still applies in the long run, even with IOR.  By which I mean than a one-time permanent increase in the base is still expected to lead to a proportional increase in the long run price level, even with IOR.  Money is still neutral.

b.  Nominal rates are zero and expected to stay zero forever.  Now open market operations are meaningless.  Bonds are essentially cash, and pay no interest.   It’s like swapping a $20 bill for two $10 bills.  Fiscal policy is powerful but currently inefficient, as the national debt is too small.

c.  Rates are zero but expected to be positive in the future.  Now a permanent increase in the base has an inflationary effect, for the same reason we discussed in the case of the gold reserve that would take two years to dig down to.  A temporary increase in the base?  Little or no effect, but then that’s true even if rates are positive.

6.  Now let’s assume the MOA is cash plus reserves.  There’s still no change.  Reserves are also a hot potato.  Indeed 100 years ago all reserves were cash.

7.  Now let’s assume a cashless economy where the MOA is 100% reserves.  Still no change; reserves are still a hot potato.  And as I said, IOR changes nothing fundamental.  Banks have X demand for reserves at an IOR of Y%.  If you double the quantity of reserves and keep the IOR at Y%, banks will suddenly be holding excess reserves.  The HPE kicks in.  Zero bound?  See case 5.

QED

Did anyone fall off the train on the way to market monetarist enlightenment?  I think I see a few MMTers in the ditch along the way, still scratching their heads.


Tags:

 
 
 

290 Responses to “The “hot potato effect” explained”

  1. Gravatar of Saturos Saturos
    1. September 2013 at 10:41

    But is the market price of gold the marginal cost of holding a marginally larger stock of gold?

    This might just be your clearest teaching post ever. (It also sort of works as a response to Quiggin, if he bothers to read it.)

  2. Gravatar of Martin Martin
    1. September 2013 at 10:58

    Let’s see if I understand,

    under 5b: if prices are sticky and nominal rates are expected to stay at zero forever, then fiscal policy “works” because nothing else can and will change real output? I.e. real output will not adjust in any other way to the new money?

  3. Gravatar of JoeMac JoeMac
    1. September 2013 at 11:04

    Scott,

    The problem I have with the HPE is that you always say that monetary policy works through expectations. You say that the monetary transmission mechanism is… 1) increased expectations of real income, and 2) increase asset prices.

    But what on earth do these have to the with the HPE? When you speak of the HPE you make it sound as if the Fed hands out cash to the public on the street and then those people start spending their cash. But the Fed does not do that.

    When the fed purchases bonds for base money from the commercial banks, where is the HPE there? After all, the banks don’t “spend” the base money. They increase M by making loans or buying bonds. How is that the HPE?

    And when the FED increases NGDP expectations thereby fueling Aggregate Demand due to “increased expectations of real income” and “increase asset prices,” how is that an example of the HPE?

    Similarly, let’s say hypothetically that the Fed lowering interest rates was an important part of the transmission mechanism (since I know you deny this), how would THAT be an example of the HPE?

    After re-reading your post I just realized that what I am saying is that I do not understand how #2 and #3 are examples of the HPE. When an individual has more money than he wants, he can get rid of it by directly purchasing goods, or by purchasing financial securities. I understand how the former is the HPE, but not the latter.

    In conclusion, in #2 and #3, when (asset) prices fall instantaneously in response to a new gold discovery, where is the HPE?

    I apologize if what I am saying makes no sense. This is very abstract.

  4. Gravatar of JAS JAS
    1. September 2013 at 11:42

    Please stop using acronyms and abbreviations. Please take the time to type out the full phrase. Abbreviations cloud the meaning of your explanation for the lay reader (like me), who otherwise might learn something. I am very interested in macroeconomics, but my only formal training in the field was a survey macroeconomics course thirty years ago.

    IOR Interoperable Object Reference
    IOR Istituto Ortopedico Rizzoli (Rizzoli Orthopaedic Institute, Italy)
    IOR Index of Refraction
    IOR Inclusive Or
    IOR Indian Ocean Rim
    IOR Improved Oil Recovery
    IoR Institute of Refrigeration (UK)
    IOR Inhibition of Return
    IOR Indian Ocean Region IOR Importer of Record (customs) IOR Increased Oil Recovery IOR Istituto Per Le Opere Di Religione (Vatican Bank) IOR International Offshore Rule (sailing) IOR Inspector of Record (construction term) IOR Independent Order of Rechabites (est. 1835) IOR Impact of Relationship (social media) IOR Institute of Operational Risk IOR International Orientation Resources IOR Ideas of reference IOR Immediate Operational Requirement IOR Interobserver Reliability IOR Input/Output Read IOR International Ocean Racing (Association) IOR Industry Optic Romania IOR Investigator of Record IOR Institute of Religion IOR Intelligent Office Repeater IOR Increment Operations Review IOR Issue on Request IOR Indicated Outcome Report IOR Isle of Rhythm (Belgian music event) IOR Issue On Requisition IOR Instructor of Record (various universities) IOR International Osteopathic Research (France)

  5. Gravatar of Tom Brown Tom Brown
    1. September 2013 at 11:45

    Scott, a few questions:

    A. Is case 5 b where you think we are now?

    B. “Fiscal policy is powerful but currently inefficient, as the national debt is too small.” Are you saying that fiscal policy would be powerful AND efficient if the “national debt” were larger?

    C. In your “cashless” case 7, you write: “Banks have X demand for reserves at an IOR of Y%. If you double the quantity of reserves and keep the IOR at Y%, banks will suddenly be holding excess reserves.” So you say HPE kicks in, but if we’re at zero rates, then we could experience case 5 b, correct? Where “open market operations are meaningless?”

    Thanks.

  6. Gravatar of Tom Brown Tom Brown
    1. September 2013 at 12:41

    Scott, in terms of expectations only, any benefit to having the BEP print up $1T in reserve notes, send it to the Fed for storage, and have the Fed announce to the world that they have another $1T in cash on hand, and they’ll keep it there permanently to meet future demand?

  7. Gravatar of Don Geddis Don Geddis
    1. September 2013 at 12:49

    Tom Brown (A): No, not case 5(b). Today’s economy is case 5(c). Rates being zero forever is only a hypothetical, used as a thought experiment to consider possible consequences and better understand the theory. What economist do you think is trying to make a strong prediction that rates in the real world will remain zero for hundreds, thousands, millions of years into the future?

  8. Gravatar of Tom Brown Tom Brown
    1. September 2013 at 13:08

    Don, I know of no such economists. It was a simple question. Scott didn’t make explicit any of the “thought experiment” part you brought up about 5b in particular (to set it apart from 5a and 5c), he simply presented three sub-cases as far as I could tell… but if it was implied it wouldn’t be the 1st time I missed something!

    Scott, what are your views?

  9. Gravatar of dtoh dtoh
    1. September 2013 at 13:25

    WRONG! WRONG! WRONG!

    It’s not a problem of understanding, it’s just that HPE is wrong.

    1. You assume that monetary policy is a one sided transaction like gold being discovered or a helicopter drop. It is not…I repeat it is not a one side transaction….it’s a two sided transaction…an exchange.

    2. If you assume that for some reason the production of extra gold causes the production of every other commodity to increase by an equal amount, then there will be zero changes in prices.

    3.That is what is happening with OMP, the financial asset price mechanism and the expectation of that mechanism cause a rise in NGDP. The rise in NGDP requires and perfectly correlates with the increase in money.

    4. The mistake with HPE is that because the increase in money and NGDP are so well correlated, you assume causality.

    5. The public is not forced to take the money in OMP. If they were (i.e. OMP were a one side monetary injection), an HPE would exist, but it would result in an increase in PY which is all P.

    HPE is a fundamental tenet of your monetary beliefs, but you need to go back and give this an honest reappraisal.

    Once you understand the financial asset price mechanism, understanding monetary economics is trivial. Just try it. For the next week, spend two minutes thinking about every problem in the asset price framework. You’ll be amazed by the clarity.

  10. Gravatar of JoeMac JoeMac
    1. September 2013 at 13:47

    dtoh, when you say “financial asset price mechanism,” can you please reference some academic paper or textbook on this…

    I think my question is exactly what dtoh is saying….

    However I’m not sure.

  11. Gravatar of dtoh dtoh
    1. September 2013 at 13:59

    JoeMac,

    I don’t know if there is an academic paper or textbook on this. It’s just my way of describing how monetary policy works.

  12. Gravatar of JoeMac JoeMac
    1. September 2013 at 14:09

    dtoh,

    I think Nick uses the same story as you in this post…

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/08/monetary-policy-as-asset-prices.html

  13. Gravatar of dtoh dtoh
    1. September 2013 at 14:42

    Prices and yields (or IRRs, or interest rates) are all exactly the same thing…. just expressed in a different way.

    I worked in the bond markets for a while so for me this is second nature. I don’t even think about it.

    For the purposes of explaining monetary policy, prices are sometimes easier to use because you don’t have to worry about artificial and false constructs like the ZLB. That said, if you’re not used to switching between prices and rates, it can be slightly confusing.

    Also, people tend to associate rates with bank borrowing. IMHO “financial assets” is a much more unified way to describe the mechanism rather than having to come up with separate mechanisms for the wealth effect, the interest rate effect, etc. etc. I look at all financial assets as the same thing for understanding monetary policy. I don’t make a distinction between Treasuries, stocks, bank borrowing, consumer loans, credit lines, etc.

    I also don’t distinguish between short and long positions. I.e. an exchange of financial assets for money can either be a sale of assets held in a portfolio, it can be a short sale of assets, or it can be an increase in indebtedness (e.g. taking out a loan and getting money).

  14. Gravatar of Kevin Donoghue Kevin Donoghue
    1. September 2013 at 14:48

    Scott,
    I think some of the confusion is due to lack of clarity as to what transactions trigger the hot potato effect. Suppose the central bank buys commercial paper worth 10m ducats. Does that, in your view, inevitably reduce the purchasing-power of the ducat? Note that the seller of paper is now more liquid than before, but not any richer. In contrast, the lucky gold prospector really has struck it rich.

    I’m not disputing any proposition, just wondering what’s being asserted.

  15. Gravatar of ssumner ssumner
    1. September 2013 at 15:32

    Saturos, I’m actually not sure. If there are non-pecuniary benefits from holding gold I think the value (price) can be greater than the opportunity cost of holding it. But my micro is rusty, perhaps someone else can answer.

    Martin, I should have been more specific. At zero rates (forever) bonds are effectively money. So OMOs do nothing, it’s swapping money for money. But fiscal policy increases the stock of bonds plus money. Since bonds are effectively money, fiscal policy is effectively monetary policy too. Helicopter drops.

    JoeMac, You said:

    “When the fed purchases bonds for base money from the commercial banks, where is the HPE there? After all, the banks don’t “spend” the base money. They increase M by making loans or buying bonds. How is that the HPE?”

    During normal times almost 95% of the new money goes right out into circulation, almost immediately. Then the HPE takes over.

    You asked:

    “And when the FED increases NGDP expectations thereby fueling Aggregate Demand due to “increased expectations of real income” and “increase asset prices,” how is that an example of the HPE?”

    The long run increase in NGDP is only due to the HPE—I think almost everyone agrees on that point. So it’s expectations of a future HPE, that is, expectations of where NGDP will be once real asset prices return to normal and money is neutral.

    On your last question, cases 2 and 3 are expectations of a future HPE that make current gold prices move.

    JAS, Interest on reserves.

    Tom, A No, 5c.

    B, Fiscal policy can move NGDP, but the current stance is too contractionary–the government is leaving $100 bills on the sidewalk, and should run bigger deficits.

    C. Yes, case 5b could also apply to 7, although of course I doubt it ever will occur. 30 year bonds would have zero yields.

    The rest will be answered after dinner.

  16. Gravatar of Daniel R. Grayson Daniel R. Grayson
    1. September 2013 at 16:13

    Perhaps instead of “Because before the discovery people were already in equilibrium, they held as much gold as they wanted to hold at existing prices” you could amplify that by saying that people know how much of their wealth, measured in dollars, they want to keep in the form of gold, and after the discovery, they will not want to increase that.

    The “hot potato effect” seems to be a misnomer, because it tries to describe the motivation only of the seller. But for each seller, there is a buyer. What’s the motivation of the buyer? Or at least, the motivation of those buyers who buy more than want to hold? (I’m wondering about this phrase: “Now someone else has the extra gold.” Why would anyone buy “extra” gold, i.e., more than they wanted?)

    What about the following alternative to step 1? Gold is discovered, doubling the world’s inventory instantly, causing the price to drop 50% instantly, because people know what has happened and what is about to happen. Now everyone has less of their wealth invested in gold than desired, so they all double their holdings by buying gold (from the discoverer). It’s the same net effect, but without a hot potato, and with everyone in the world, except for the discoverer, buying, not selling.

  17. Gravatar of JoeMac JoeMac
    1. September 2013 at 16:14

    Scott,

    You said, “During normal times almost 95% of the new money goes right out into circulation, almost immediately. Then the HPE takes over.”

    Let’s say the banks create new M on top of the other 5% by giving loans and buying bonds. Is that creation process of new money also HPE? And once the new money is created, is it ALSO part of the HPE when it is spent?

    Also, you said… “So it’s expectations of a future HPE, that is, expectations of where NGDP will be once real asset prices return to normal and money is neutral.”

    Does the future HPE ACTUALLY happen at some point? Or is it the case that the expectations boost velocity NOW, and so V takes the place of the future HPE that was expected to happen later. In other words, its a self-fulfilling process. Expectations of the future HPE cause V to go up NOW, and therefore the “future HPE” doesn’t actually happen because Velocity and Prices adjust.

  18. Gravatar of ssumner ssumner
    1. September 2013 at 16:24

    dtoh, You said:

    “The public is not forced to take the money in OMP. If they were (i.e. OMP were a one side monetary injection), an HPE would exist, but it would result in an increase in PY which is all P.”

    You are confusing unrelated issues. The way in why money is injected has nothing to do with the PY split between P and Y. That’s caused by sticky wages and prices.

    Again:

    Short run: NGDP rises because of asset prices increases and the expected HPE.

    Long run. NGDP rises in proportion to exogenous increase in M, solely due to HPE.

    I don’t think either is all that controversial. If my long run was false, then money would be non-neutral in the long run. And if it’s true, then expectations of the L-R effect obviously matter a lot.

    As far as “one-sided,” feel free to assume all the gold is sold for T-bonds. An OMP. The effects will be almost exactly the same.

    Kevin, If I understand your question correctly I’d say the “feel richer” effect is virtually zero. I don’t assume that is part of the transmission mechanism. Not to say it can’t happen, obviously monetary stimulus is boosting stock prices in Japan, for instance. But I don’t rely on that. I think stocks matter (mostly) because higher stock prices combined with sticky wages lead to creation of new corporate assets.

    In the long run I rely on a well-defined demand for money (ducats or whatever) and thus if you have more money (exogenously), its value will fall. Obviously money cannot be expected to be a permanent perfect substitute with the asset being purchased.

  19. Gravatar of ssumner ssumner
    1. September 2013 at 16:36

    Daniel, I can accept that, but I think the term ‘HPE’ gets at how the fallacy of composition lies at the heart of monetary economics. But I don’t deny that there are other ways of looking at the picture.

    JoeMac, Bank money (i.e. bank deposits) only contribute to the process if they reduce the demand for base money. However people like Nick Rowe would argue they probably do reduce the demand for base money (checking deposits instead of cash) and hence they probably do contribute to the HPE.

    The second question is sort of philosophical, and probably where I lose dtoh. Imagine you are trying to raise the floor of a building, and you inflate a giant balloon under it. But the balloon is too weak to raise the floor very fast, meanwhile other options are also used, but they automatically expire when the floor is fully raised. These other factors might be something like car jacks for flat tires. Once the floor is fully raised the balloon is just barely powerful enough to keep it there, and then you remove the more powerful items like car jacks. So the balloon didn’t do much of the heavy lifting, but maintains the new equilibrium. The balloon is kind of like the HPE, and the car jack is like the asset market channel and expected HPE.

  20. Gravatar of dtoh dtoh
    1. September 2013 at 16:49

    Scott,
    You said;

    “You are confusing unrelated issues. The way in why money is injected has nothing to do with the PY split between P and Y. That’s caused by sticky wages and prices.”

    I think you are wrong. If the Fed was doing repeated helicopter drops and announcing what it was doing, money injections would just translate into instant increases in P. In fact, I think that prices would come to be denominated not in units of the base, but rather as a percentage of the total base.

    I’ll think about your other points and respond after I eat breakfast.

  21. Gravatar of Steve Waldman Steve Waldman
    1. September 2013 at 17:22

    i think i agree with pretty much all of this post, but to be sure, i’d want a few clarifications:

    “a one-time permanent increase in the base is still expected to lead to a proportional increase in the long run price level, even with IOR…Banks have X demand for reserves at an IOR of Y%. If you double the quantity of reserves and keep the IOR at Y%, banks will suddenly be holding excess reserves.”

    This is all sensible on the assumption that IOR is held constant, at whatever level it is held. If the rate of interest on reserves varies in time, than the demand for reserves at a given price level varies with the variable interest paid, no? So that the equilibrium price level becomes a function of both the quantity of base and the level of IOR.

    “Nominal rates are zero and expected to stay zero forever. Now open market operations are meaningless. Bonds are essentially cash, and pay no interest. It’s like swapping a $20 bill for two $10 bills. Fiscal policy is powerful but currently inefficient, as the national debt is too small.”

    I’m glad to see you describe this case. I agree, and also agree that there are important differences between bonds and base money if base money pays no interest while bonds do pay interest, or are expected to pay interest in the future.

    But do we agree on a fourth case? Suppose that cash and short-term bonds are expected to pay the same rate of interest permanently, although that rate may or may not always be zero. Do we agree that in this case, OMOs remain meaningless, bonds are effectively cash, and fiscal policy is powerful?

  22. Gravatar of JoeMac JoeMac
    1. September 2013 at 17:23

    Scott,

    You said, “Bank money (i.e. bank deposits) only contribute to the process if they reduce the demand for base money. However people like Nick Rowe would argue they probably do reduce the demand for base money (checking deposits instead of cash) and hence they probably do contribute to the HPE.”

    But what if people purchase goods directly with their pre-existing checking accounts (debit cards or paper checks) or the newly created ones through the “bank multiplier”? Isn’t that the HPE?

    If people start spending money with their checking accounts, and NOT CB currency, at a faster rate then won’t velocity and AD go up. Isn’t this the HPE?

    In other words, let’s say expectations of future HPE go up. In response people start spending more money with their checking accounts….. isn’t that HPE?

  23. Gravatar of Steve Waldman Steve Waldman
    1. September 2013 at 17:25

    (oops. in the previous, in my last paragraph, i was loose with the term “cash”. i should have said “base money”, the variable controlled by central banks. we will as we do leave the division of base money between cash and reserves to the private sector, with only reserves paying any interest.)

  24. Gravatar of Franky Franky
    1. September 2013 at 17:29

    Scott, this makes your ideas much clearer, thanks. I just posted an example using apples in the other thread, but let me make the same point here.

    First off, note that in your description the role of money as MOA is irrelevant. Regardless of whether gold is the MOA or not people will try to sell it for other goods. If prices are sticky, this will have real effects either way. It doesn’t matter if it’s the price of gold coming down or the price of other goods going up. The implications for behavior are exactly the same.

    But you are missing two key issues in your description.

    First, you are implicitly assuming that money (gold, apples, whatever it is) is not a perfect substitute for other financial assets like bonds. That it is valued for something other than for the present value of its expected cash flows. Without this assumption, you end up in your case 5b. That is why the role of money as MOE is decisive. What makes interest rates fall to zero? Having enough money in circulation such that it loses its value as MOE at the margin, it’s that simple. Take away the MOE role of money and interest rates will have to be zero in any equilibrium where people hold any money. And in that case, as you point out, open market operations have no effect.

    Second, you are assuming that people will try to get rid of the extra money by buying goods, rather than financial assets. That is not what an optimizing consumer would do (unless money and goods are complements in the utility function). An optimizing consumer would try to buy other financial assets, causing interest rates to fall. And that fall in interest rates is what triggers increased consumption spending. And yes, I am talking in nominal terms here. Stickiness is not necessary for any of this to hold.

  25. Gravatar of Greg Hill Greg Hill
    1. September 2013 at 17:41

    Scott,

    I think this is pretty close to chap. 17 of the GT, on “The Essential Properties of Interest and Money,” and to Kaldor’s 1939 (or thereabouts) essay on “Speculation.”

  26. Gravatar of interfluidity » Banks and macroeconomic models interfluidity » Banks and macroeconomic models
    1. September 2013 at 17:44

    […] Scott Sumner — The “hot potato effect” explained […]

  27. Gravatar of dtoh dtoh
    1. September 2013 at 17:55

    Scott,
    You said:

    “Long run. NGDP rises in proportion to [an] exogenous increase in M, solely due to HPE.

    ….. If my long run was false, then money would be non-neutral in the long run.”

    I agree up to the point where you say “solely due to HPE.”

    I think you can equally argue that NGDP is neutral with respect to M and that M rises in proportion to an exogenous increase NGDP.

    Either approach is logically valid and neither proves causality. Where you (and everyone else) runs into problems is when you make the jump and assume the increase in NGDP is “solely due to HPE.” This seems to be an historical a priori assumption which has crept in as a underpinning to most monetary thinking.

    Tell me where I’m wrong.

  28. Gravatar of Max Max
    1. September 2013 at 18:16

    “Now switch the medium of account from gold to cash.”

    Cash (I assume this means central bank money) is the MOA for privately produced money. This is only an intermediate MOA. The ultimate MOA is the central bank’s target, e.g. a price index.

    Nick Rowe wrote a great post on the evolution from gold to modern central banks:

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2012/04/from-gold-standard-to-cpi-standard.html

  29. Gravatar of dtoh dtoh
    1. September 2013 at 18:26

    Steve,

    You ask;

    “But do we agree on a fourth case? Suppose that cash and short-term bonds are expected to pay the same rate of interest permanently, although that rate may or may not always be zero. Do we agree that in this case, OMOs remain meaningless, bonds are effectively cash, and fiscal policy is powerful?”

    In this scenario, you need to think of money as being dual purpose: both a medium of exchange and a store of value (i.e. a financial asset). While the public may be indifferent to holding bonds vs. money as a financial asset, they are not indifferent to holding financial assets vs. spending on goods and services. If wages and/or the price of goods and services are sticky, then OMP will raise the real price of financial assets through increased inflation expectations and this will cause a marginal increase in the exchange of financial assets for real goods and services….i.e. higher AD. Alternatively the Fed need not limit itself to fixed income assets. It can buy other assets as the BoJ is currently doing with REIT shares. This will have the same effect.

  30. Gravatar of JAS JAS
    1. September 2013 at 19:15

    Thank you for IOR = Interest on Reserves.

    The ‘hot potato effect’ with increased gold is just loss aversion driving down prices. People don’t have too much gold; they simply fear loss of wealth. Some people dump their gold quickly and relatively small amounts of transactions drive the price to a new equilibrium. The people who have not dumped their gold fast enough will not feel like they have too much gold, they will just feel that they have lost wealth. So the ‘hot potato effect’ is only active during the brief period while prices are adjusting to the new equilibrium.

    Increasing the gold supply is inflationary when it is the medium of account. (I seem to remember that the California 1849 gold rush saved the USA from a depression after Andrew Jackson destroyed the US bank.) But inflation doesn’t make us think we have too much money and therefore want to get rid of it, it is the erosion of value that makes us want to invest/spend faster. So the effect you describe in points 1 through 4 are simple and understandable, but I don’t find the term ‘hot potato effect’ helpful in describing the process.

  31. Gravatar of TallDave TallDave
    1. September 2013 at 21:09

    People don’t have too much gold

    Say you’re the person who mined all the new gold. Do you have more gold than you want? You probably want to pay your workers, suppliers, creditors, etc., so you exchange the mined gold for something that isn’t gold. Now someone else has the excess gold, so they exchange it… ultimately you’re all pushed to a new equilibrium.

    Prices and markets transmit this information very quickly, but the fall in price can’t be said to reside only in people’s fear of the lost value. That’s confusing what happens with how it happens.

  32. Gravatar of Jason Jason
    1. September 2013 at 22:43

    So in Japan they think that the 350% increase in the base over the past 20 years is just temporary leading to zero inflation over the past 20 years?

    http://www.adamsmith.org/blog/money-banking/chart-of-the-week-japanese-monetary-base-and-inflation

  33. Gravatar of lxdr1f7 lxdr1f7
    1. September 2013 at 23:09

    “Now with sticky prices the fall in the value of gold (the MOA) will be more gradual”

    Prices are sticky downwards but not as much in the upwards direction

    “There is no need to resort to “transmission mechanisms” such as interest rates. Just S&D.”

    There is a need to resort to transmission mechanisms. The supply of MOA wont actually enter the economy, therefore the S wont go up and affect prices. The transmission mechanism is what allows the supply to go up. If the new money supply isn’t transmitted into the system then the supply doesn’t change.

  34. Gravatar of Tom Tom
    2. September 2013 at 00:12

    The mistakes here are:

    Cash, unlike gold, is not a real good. Cash is one party’s asset and another’s liability. The production of real goods increases wealth. The production of cash deepens the financialization of the economy. Those are two very different actions with very different results.

    It’s true that when cash is created, people and/or organizations must hold that much more cash (net of the extent they’re extinguishing cash by deleveraging, as in 2009). QE creates bank deposits and base money simultaneously, it is not a mere swap of base money for sovereign. Those additional willing holders are found by bidding up prices of not only real goods and assets, including labor, but also of financial assets, domestic and foreign. The degree of price increases varies widely across those various kinds of goods and assets depending on the situation, and is never the only factor changing them. In the recent US case the most obvious results have been a rise in financial asset prices and compression of yields, and a rise in land prices.

    Or, put more simply, money supply is not spending.

  35. Gravatar of Anders Anders
    2. September 2013 at 01:18

    Scott – the MMT critique of the HPE surely complains that it omits consideration of balance sheets. The HPE may work well for financial assets via the behaviour of unlevered financial institutions; I don’t see it working as well for households and corporates, which are in many cases leveraged entities whose spending functions take account of their views on future prospects for income, as well as their starting net worth.

    How do you respond to this?

  36. Gravatar of lxdr1f7 lxdr1f7
    2. September 2013 at 02:30

    Tom

    “QE creates bank deposits and base money simultaneously”

    Not when the counterparty is a primary dealer

  37. Gravatar of Philippe Philippe
    2. September 2013 at 06:31

    “Not when the counterparty is a primary dealer”

    With QE the central bank buys assets from primary dealers, but the primary dealers also purchase assets from non-banks and then sell them on to the central bank. As such QE does lead to an increase in both deposits held by non-banks and base money.

    This paper looks at who the ultimate sellers assets to the Fed are in response to QE:

    ‘Analyzing Federal Reserve Asset Purchases: From whom does the Fed buy? (2013)
    http://www.federalreserve.gov/pubs/feds/2013/201332/201332pap.pdf

  38. Gravatar of ssumner ssumner
    2. September 2013 at 06:38

    dtoh, We’ll have to agree to disagree on that.

    Steve, I agree with most of your comments, but not the last one. If interest rates are positive but stable, there is still a HPE for cash injections.

    In addition, OMOs with longer term bonds still have some impact.

    JoeMac, Aren’t you just restating what I said. In your example demand deposits displace some of the previous demand for cash.

    Franky, You said;

    “First off, note that in your description the role of money as MOA is irrelevant. Regardless of whether gold is the MOA or not people will try to sell it for other goods. If prices are sticky, this will have real effects either way. It doesn’t matter if it’s the price of gold coming down or the price of other goods going up. The implications for behavior are exactly the same.”

    I strongly disagree. The MOA role is all important. In both cases the value of gold changes, but in only one case does the price level change significantly.

    You said;

    “Take away the MOE role of money and interest rates will have to be zero in any equilibrium where people hold any money. And in that case, as you point out, open market operations have no effect.

    Second, you are assuming that people will try to get rid of the extra money by buying goods, rather than financial assets. That is not what an optimizing consumer would do (unless money and goods are complements in the utility function).”

    You said you finally understood me, but I am afraid it’s back to the drawing board. I did not say OMOs are ineffective at zero rates, and I did not assume money is used to buy other goods and services, at least in the short run. You need to read what I say more carefully.

    Greg, It’s been a while since I read the GT, so I can’t comment.

    dtoh, The neutrality of money concept is partly informed by theory. After all, Hume had no data to work with, he was just thinking about the idea in the abstract. When the data confirmed Hume, that solidified the QTM, as economists were already inclined to accept it on logical grounds.

    Max, That’s true in a sense, but it assumes away the problem. People usually try to understand monetary economics in order to explain changes in the price level. But if you are correct, then the price level never changes. If so, then I agree that the MOA is in a sense the CPI basket of goods.

    If I am trying to explain real world inflation or deflation (i.e. 2009) then your explanation is not helpful.

    JAS, I agree that for gold the HPE drives changes in gold prices very quickly, almost immediately. When it becomes a MOA, it takes longer as prices are stickier.

    Jason, Of course they think it was temporary. If it was permanent then Japan would have hyperinflation when interest rates rose above zero. Of course, as always, I’m assuming zero IOR. With a positive IOR you can have a once and for all shift in the Cambridge k, and that might be the case in Japan.

    lxdr, I don’t think you understand the term “transmission effect,” it has nothing to do with the idea of the new money entering the economy.

    I agree with your second point.

    Tom, I disagree. Cash is a real good. Otherwise why would people hold cash at zero rates when market rates are 5%? The real value of a given stock of cash is X. The real value of the stock of cash is still X if you double the money supply. That’s why money is neutral.

    Anders, Balance sheets play no important role in this process.

  39. Gravatar of Anders Anders
    2. September 2013 at 07:16

    “Balance sheets play no important role in this process”

    Scott, yes, this is an assertion of the traditional (broad) monetarist position. But I don’t think you have addressed any MMT concerns here.

    If the authorities were to compulsorily purchase all my financial assets for cash, my natural response would be to use the proceeds to try and buy some replacement financial assets. MMTers don’t see why I would be expected to increase my consumption just because my bank account suddenly showed a lot more cash.

    I think this is the nub of what MMTers think broad monetarists don’t understand.

  40. Gravatar of ANon ANon
    2. September 2013 at 07:36

    Say in 1500’s Spain you have gold, goods, and a gold-linked zero coupon perpetual bond.

    Gold flows in to Spain from the Andes. However, it only flows into Queen Isabela’s coffers. Say she’s a spendthrift, so it just sits there. Say you know her descendants will keep it there too.

    What happens to goods prices?

    Queen then Isabela decides to hold less gold and more gold-linked bonds. She swaps with a group of gold merchants. The merchants now hold the equivalent of 20% more of the economy’s gold supply. They hold 20% less of gold-linked bonds.

    What happens to goods prices?

  41. Gravatar of Franky Franky
    2. September 2013 at 07:54

    Scott, I have understood you, and I also have a good grasp of mainstream modern monetary theory, which I’m not sure you fully understand (your comments about how using a representative consumer here doesn’t work make it seem like you don’t).

    The bottom line for me is:
    – You are implicitly assuming money is not a perfect substitute for bonds. Typically people argue this is because money is the MOE, and they stick money in the utility function as a shortcut. This is fine and in line with mainstream models.
    – You may be assuming money and consumption are complements. If you are, then yes, there is a hot potato effect separate from the interest rate channel. If you are not, then your model is the same as the mainstream model and monetary policy works through interest rates. Most people find that a high enough degree of complementarity to make this a relevant transmission channel is completely implausible, and I agree. Matt Rognlie once wrote a great post about this (“New Keynesian versus New Monetarist effects”).

    You may dismiss what I’m telling you, your choice. I think you’ll find it hard to convince mainstream macroeconomists that you have an alternative model if you don’t fully articulate your assumptions and figure out where the differences are. If any.

  42. Gravatar of JP Koning JP Koning
    2. September 2013 at 08:16

    Good post, I think I agree with pretty much all of it.

    “b. Nominal rates are zero and expected to stay zero forever. Now open market operations are meaningless. ”

    To me, this is the limiting case in which a central bank can literally buy up every asset in an economy and not have an effect on the price level. Are you ok with that?

    [Nitpick: I’m not sure why you always use the term “medium of account” instead of “unit of account” to describe the sign in which prices are expressed. In my opinion, we should reserve “medium of account” to refer to the medium that defines the unit. So for instance, in Hong Kong sticker prices are expressed in HK dollars (the unit of account) but the HK dollar is defined by the central bank as 0.13 US dollars, so $US is the medium of account. Of course, this is just a semantic point, but nevertheless we’ve got to be consistent in our use of words and underlying definitions in order to avoid confusion. I’ve made this point before.]

  43. Gravatar of Craig Craig
    2. September 2013 at 09:38

    The Fed controls NGDP expectations via the HPE only with permanent MOA injections.

    The Fed adds MOA until NGDP expectations overshoot.

    How does the Fed bring NGDP back to the level target?

    By shrinking the MOA, which creates impermanence.

    Therefore:

    1) Only an permanent MOA injection can be used to hit a target.

    2) Only a impermanent MOA injection can be used to hit a target.

    Both of the above statements must be true for the HPE to be an effective mechanism for maintaining an NGDPLT regime.

  44. Gravatar of gasman gasman
    2. September 2013 at 10:15

    Your use of medium of account is confusing. Medium of account is another way of saying price. Its the numeraire when you account for something. Without a MOA what you would put on a balance sheet might be 6,000. A reasonable question would then be 6,000 what? So $ is OUR medium of account here in the USA but in #5 where you say “Now switch the medium of account from gold to cash.” that is a nonsensical statement. Cash is a medium of exchange, checking accounts are a medium of exchange, credit cards are a medium of exchange. The exchanges are made in $ which is the MOA. What a price tells you is how much MOE it takes to buy said object.

    Gold would not be the medium of account but ounces or grams OF gold would be.

  45. Gravatar of Dan Kervick Dan Kervick
    2. September 2013 at 10:19

    Assume gold sells for $1200. Interest rates are low, so the expected future price is roughly the same, maybe a bit higher. Now assume a company discovers a gold hoard so vast that world output soars. Over a period of years the extra output causes gold prices to fall in half. But why? Because before the discovery people were already in equilibrium, they held as much gold as they wanted to hold at existing prices. The extra gold is a sort of “hot potato” that people try to get rid of. But obviously not by throwing it away! They get rid of it by selling it.

    The metaphors of “hot potato”, “trying to get rid of” and “willing to hold” seem to be doing zero work here, Scott. Why use that language?

    If all you are doing is explaining how an increase in the supply of any kind of something-or-other tends to reduce the per unit value of that some-thing-or-other, you don’t need to employ the above metaphors.

    It makes no sense to describe people either in the aggregate or at the invidual level as “trying to get rid of” their new gold. They are increasingly willing to surrender some of the additional gold only to get something they value more. The “hot potato” metaphor misleadingly suggests that the marginal units have negative value, and so people are not willing to hold it. But they are perfectly willing to hold it in itself; it’s just that there are other things they are more desirous to hold. Because the added quantities of gold have diminishing positive value, the per-unit value reaches a level where what previously looked like a bad trade starts to look like a good trade.

  46. Gravatar of OhMy OhMy
    2. September 2013 at 10:22

    SS,

    So now you are trolling the MMT bloggers, because you can’t get the MMT scholars to stop ignoring you as any time you did engage you couldn’t talk coherently about how the actual monetary system (where reserves are injected via banks) works.
    http://neweconomicperspectives.org/2011/07/scott-sumner-agrees-that-mmt-policy.html

    MMT agrees there could be a hot potato effect if you injected money-like assets via fiscal policy, but not when you limit yourself to monetary policy swaps – then and can only inject reserves when you change the price to the level at which banks want to buy and hold them (because unlike you banks know full well they can’t get rid of reserves via lending and any other process they have an influence on).

    Btw, there is no such thing as a “medium of account”. There is a “unit of account” and a “medium of exchange” *denominated* in the units. You got confused because “dollar” is the name of the unit and also colloquially used as the name for the “money object” denominated in the state-defined units. It is like “meter” vs “meter stick”. There is a metal bar in France that defines the meter unit but we don’t exchange metal sticks when we say “I bought 4 meters of silk”. We exchange some materials in quantity *denominated* in some unit. Saying that “meter” is a medium of exchange of cloth would be incoherent, it is a measurement unit.

  47. Gravatar of Philippe Philippe
    2. September 2013 at 11:19

    Is ‘OhMy’ Scott Fullwiler?

  48. Gravatar of Jim Glass Jim Glass
    2. September 2013 at 11:25

    Cash is one party’s asset and another’s liability.

    Fiat-money cash is an item of value, like a bar of gold, steel or lead. With its value set by the supply of and demand for it, like a bar of gold, steel or lead.

    Certainly a bar of steel can appear as an asset on one’s balance sheet, but who’s liability is it?

    A liabiltity can (must) be redeemed by an identified party, for something of specified value.

    If the cash in my wallet is a liability of the Fed, I can present it to the Fed to be redeemed for… what?
    ~~~~

    SS, So now you are trolling the MMT bloggers, because you can’t get the MMT scholars to stop ignoring you…

    🙂

    Good one!

  49. Gravatar of Philippe Philippe
    2. September 2013 at 12:48

    “If the cash in my wallet is a liability of the Fed, I can present it to the Fed to be redeemed for… what?”

    You can redeem it for coin, which is currency issued by the Treasury and an asset to the Fed.

    But in general the ‘liability of the Fed’ is really a liability, or obligation, of the US government. It’s a liability of the US govt because it has promised to accept it in payment.

  50. Gravatar of Philippe Philippe
    2. September 2013 at 12:55

    the Fed also promises to accept it payment… in repayment of loans or in exchange for assets the Fed holds (if the Fed wants to sell them) . But I’d say this is a subset of the overarching liability of the US government.

  51. Gravatar of Joe Eagar Joe Eagar
    2. September 2013 at 13:59

    I’m not sure what the controversy is here. Historically there most definitely was a HPE effect because interest rates were positive, and there was an opportunity cost to holding onto base cash.

    Even in our ZLB post-IOR world, there is still some opportunity cost to holding short-term debt relative to other, riskier assets like long-term debt and stocks. And once interest rates normalize, the opportunity cost will only increase.

    By the way, what’s the difference between the HPE and the portfolio-balance effect? They seem rather similar, to me.

  52. Gravatar of Geoff Geoff
    2. September 2013 at 13:59

    Many people, particular macro-economists, who tend to reverse causality in that which they observe, have trouble with the fact that economic phenomena is determined by individual activity. For some reason they just can’t grasp the fact that individual activity is the sine qua non of economics, and certainly monetary economics. For many years we have witnessed fumbled attempts from macro-economists seeking to explain money and prices, precisely because they ignore individual activity and believe the answers lie in mathematical relationships in aggregated data. I would have thought Goodhart and Lucas provided the establishment with what some economists have known for at least over 140 years that this is not the right approach.

    If one can’t explain a particular economic “effect” in terms of the individual, then it means the purported “effect” is either not an effect at all as stated, or it has some truth to it, but is in reality referring to some misunderstood real world process in a different context.

    Macro economists almost always want explanations at the “holistic” level. They want to connect various aggregated statistics together in a modelistic fashion. But there is nothing in this “holistic” sphere that relates to economics. An individual spends money. NGDP is just the numerical sum of dollars spent by all individuals in exchange. An individual buys and sells at a price. Price levels are just the numerical index of a basket of goods bought and sold by…all individuals in exchange. An individual can lose their jobs. Unemployment is the just numerical sum of all individuals who want jobs but don’t have jobs.

    The “hot potato” effect is a reference to only one side of the exchange coin, and intends to explain why spending rises in the aggregate. It includes only those individuals who want to spend more money on the basis that they own more money, and/or consistently earn more money. It ignores those who must want and accept more money if those money expenditures are said to have been made, and it ignores how individuals are capable of being able to spend more money (they earn, i.e. accept, more money from those whose incomes have been raised prior, due to yet still others having a greater income due to yet still others having spent more money on their goods/serves, and so on, all the way through the complex of exchanges until we end up (or rather start out) as money from the Federal Reserve paper money cartel.

    Part of the problem is that macro-economists want to think in terms of aggregates, which seems natural for the subject matter. After all, the economy is not Joe all by himself. It is not Bill. It is not Frank. The economy is Joe, Bill, Frank, and every other individual who participates in the division of labor. “Why are there more dollars spent in total, that has nothing to do with individual activity?” But that just won’t do. It’s impossible.

    ———————–

    “1. Assume gold sells for $1200. Interest rates are low, so the expected future price is roughly the same, maybe a bit higher. Now assume a company discovers a gold hoard so vast that world output soars. Over a period of years the extra output causes gold prices to fall in half. But why? Because before the discovery people were already in equilibrium, they held as much gold as they wanted to hold at existing prices. The extra gold is a sort of “hot potato” that people try to get rid of. But obviously not by throwing it away! They get rid of it by selling it. But notice that while that works at the individual level, it doesn’t work in aggregate. Now someone else has the extra gold. (That’s why attempts to understand money at the level of the representative consumer fail.) The only way for society as a whole to get rid of the extra gold is by driving down the price of gold until people want to hold the new and larger quantity. Assume the price falls in half. That also means the value of gold relative to other goods and services falls in half.”

    The bolded part is the error.

    When the new gold deposit is brought into market, you can’t interpret the buyers of this additional gold to have “too much” gold, or “extra” gold, or “excess” gold. The very fact that they accepted the gold shows us, by revealed preferences, that they wanted more gold than what they previously owned! There is no need for, nor existence of, a “society” to “get rid of the extra gold.” There is no “extra” gold. The fact that the increased gold supply is owned by someone proves the owners WANT that gold.

    You are smuggling in the assumption that there is some eye in the sky, aggregate need or desire for gold, which could conflict with individual desires and needs. If there is a doubled supply of gold, then there is somehow “too much” gold for “society” to handle, unless we, the individuals, who are tasked with reducing the dollar price of gold to half of what it was, so that “society” is back in equilibrium.

    Notice that when an “aggregate” explanation is sought after for why the price of gold falls in a context of a doubled supply, the preferences of the buyers are overlooked, dismissed, and we invariably end up conceiving of some “social” desire that has to be satisfied by individuals as means.

    But all that is happening with a decreased dollar price of gold is that individuals are placing a lower marginal utility to a unit of gold, because there is more gold. The law of marginal utility is a category of action. A lower marginal utility attached to gold means, ceteris paribus, that gold will trade at a higher ratio to other goods, for example dollars. That is to say: the dollar price of gold falls.

    ——————————–

    “But I keep coming back to the notion that no matter how important those other factors are, the HPE is still in some sense primary. It’s what explains the long run effect of more gold on the value of gold. It’s still econ 101. The other effects are side effects that help nudge us toward that final equilibrium.”

    You’re talking in circles. You keep going back to the HPE as “the explanation” because it is your a priori theory of choice that you are using to understand history, and yet you want to make it seem like that theory “pops out of the data.”

    But the theory itself is flawed, so it can’t be an explanation. It is flawed because it ignores all those individuals who want to ACCEPT more money from those you are referring to with the HPE. For every individual who wants to “get rid of money”, there must be another individual or individuals who wants to “hoard money”, in exactly the same offsetting numerical dollar amounts. The hot potato effect cannot coherently apply to those individuals who want MORE money, and yet those individuals are very much a part of “the economy.” Their preferences are real, just as real as the preferences of those individuals who want to “get rid of money.” So the HPE isn’t an explanation for “the economy”. It is at most an explanation for a subset of the population. But that isn’t an economics thing, it’s a subjective preference thing. “These” people want less money, while “those” people want more money. There is no excessive or deficient “aggregate desire for money.” It’s all offsetting.

    So why does spending rise with more money then? It’s because individual sellers of money AND buyers of money, attach a lower marginal utility to “a dollar” because there are more of them. So the exchange rate of goods to dollars falls, which is the same thing as saying the exchange rate of dollars to goods rises.

    The HPE in money is really a confused reference to the law of marginal utility in the context of individuals who spend more money, who were previously accepters of more money.

    —————————

    “Did anyone fall off the train on the way to market monetarist enlightenment? I think I see a few MMTers in the ditch along the way, still scratching their heads.”

    Enlightenment? I would say it is more a display of market monetarist curiosities. I don’t think people are smarter by having been taught the HPE. I think it just confuses those who have never been taught, or have never learned of, economics grounded on individual action.

    —————————-

    To summarize:

    There is no such thing as aggregate “excesses” or “deficiencies” of money or goods. There is only individual overabundences and underabundences of money and goods. What market monetarists are referring to when they say “there is an excess supply of money in the public”, what they are really referring to are individuals in one group, that excludes other individuals. This is quite tragic because it is those “other individuals” who must be present in order for there to be individuals who “want to get rid of money”.

    The whole society of individuals cannot “get rid of money” in a context of exchanges. If EVERY individual wanted to “get rid of money” in exchanges, then guess what? Nobody could do it! For there would be no corresponding group of individuals there to “hoard” all that “to be gotten rid of” money!

    Every exchange must be treated as unique, and separate from all other exchanges, if we are going to understand why exchanges take place, and why we find that there is more money traded in the average exchange as time goes on.

  53. Gravatar of Geoff Geoff
    2. September 2013 at 14:07

    Joe Eager:

    “I’m not sure what the controversy is here. Historically there most definitely was a HPE effect because interest rates were positive, and there was an opportunity cost to holding onto base cash.”

    Why do individuals agree to ACCEPT more money from those who want to “get rid of it”, and why are you ignoring them? You are ignoring them when you refer to the hot potato effect.

    The reason interest rates are positive is not due to any HPE. They are positive because of the fact that individuals prefer the same real goods sooner rather than later, ceteris paribus, and thus require a discount on future goods relative to present goods, which in a monetary economy results in individuals requiring to be compensated for giving up present purchasing power.

  54. Gravatar of Joe Eagar Joe Eagar
    2. September 2013 at 14:10

    Geoff, I’m not sure what you’re saying. The entire point of macroeconomics (going back at least two hundred years) is to describe situations where there clearly *is* too much (or too little) gold/money/aggregate demand/aggregate supply/whatever.

    “You are smuggling in the assumption that there is some eye in the sky, aggregate need or desire for gold, which could conflict with individual desires and needs.”

    That’s not really fair. Macro economists do make such assumptions, but not lightly; they base them either on accounting identities or government policies (e.g. perhaps the government has backed its currency with gold, and thus there really *is* an “eye in the sky.”).

  55. Gravatar of Joe Eagar Joe Eagar
    2. September 2013 at 14:26

    Geoff, how would these individuals know the money is “extra” to begin with? From their perspective, all they see are more customers buying their goods.

    I’m also not sure what to make of your comment on interest rates. Individuals are terrible at determining interest rates; they either save too little, or they save too much. That’s why corporate savings so outstrips individual savings in this country.

    For that matter, the U.S. government is itself heavily involved in determining the aggregate level of savings; if you graph fiscal policy and compare it to gaps in the savings rate, you’ll notice a striking correlation:

    http://cdn3.ricochet.com/var/ezwebin_site/storage/images/media/images/yfredgraph/4024196-1-eng-US/yfredgraph_lightbox.png

    That graph compares the federal deficit with the gap in private savings, offset by -3.5% (to approximate a crude fiscal “rule” of limiting the current account deficit to less than -3.5% of GDP).

  56. Gravatar of Joe Eagar Joe Eagar
    2. September 2013 at 14:28

    (Sorry, wrong link; here’s the original, FRED link).

    http://research.stlouisfed.org/fred2/graph/?graph_id=121102&category_id=0

  57. Gravatar of ssumner ssumner
    2. September 2013 at 15:00

    Anders, You said;

    “MMTers don’t see why I would be expected to increase my consumption just because my bank account suddenly showed a lot more cash.”

    Nor do I.

    ANon, Depends what the merchants do with the gold. If we assume they already were in equilibrium before she sold the gold, then prices would rise.

    Franky, I get really annoyed at people who tell me they fully understand what I am saying, and then proceed to describe it in a completely inaccurate way. If you really believe that then why aren’t you defending the statements of yours that I said were nonsense. You should either defend them or admit you were wrong. But to just move ahead insisting you were right doesn’t help your case with me.

    And I don’t need advice on convincing others, I’ve already had 10 times the success of my wildest dreams.

    Some Keynesians consider a rise in stock prices to be an “interest rate channel.” It seems to me that is moving the goalposts, because the actual interest rate channel is so weak. As far as “models”, anyone can write down models were this or that mechanism seems important, but the truth is that the interest rate channel is simply not useful in the real world, it doesn’t explain things. When you can explain 8700% annual inflation with a path of interest rates I’ll start listening.

    JP, I agree that there is some ambiguity in cases like HK, which in my view has dual MOAs.

    Craig. I don’t follow that.

    gasman, You are confusing MOA with unit of account (the $)

    Dan, Lots of people use the metaphor, so they think it must be useful. Tobin (a Keynesian) used it in 1960.

    OhMy, Throughout history there have been a number of examples of media of account that different from MOE. In 1930 the US MOA was 1/20.67 ounces of gold, but it was rarely used as a MOE.

    And currency injections are inflationary even if not accompanied by expansionary fiscal policy

    Phillippe, You said;

    “But in general the ‘liability of the Fed’ is really a liability, or obligation, of the US government. It’s a liability of the US govt because it has promised to accept it in payment.”

    If they promised to accept gold bars in payment, would that make gold bars a liability of the government?

    And the base actually includes coins.

  58. Gravatar of gasman gasman
    2. September 2013 at 16:07

    Geoff

    Are you not aware of the fallacy of composition? The reason macro came along is that micro only is insufficient.

    Your comment about prices and NGDP

    “An individual spends money. NGDP is just the numerical sum of dollars spent by all individuals in exchange. An individual buys and sells at a price. Price levels are just the numerical index of a basket of goods bought and sold by…all individuals in exchange”.

    Is a head scratcher. So you thinkNGDP is meaningless? It is itself an aggregation that cannot be understood by looking at what any one decision a person made. Price levels are a macro perspective too. A numerical index takes multiple data points and aggregates them. The Dow is a macro perspective. One point of studying macro is that we can only see ourselves It takes outside perspective, a MACRO view, to see our place in the big picture.

    GM can do what is right for GM but what is right for GM cannot be applied to anyone else. We cant all decide tomorrow to save 50% of our incomes when we were previously saving only 20%. Any efforts to do so will mean that that income you were expecting to receive wont be there because the spending of others that comprises your income is now reduced. We are all interconnected that way like it or not.

  59. Gravatar of gasman gasman
    2. September 2013 at 16:08

    Scott,

    Im not confusing anything you can call it unit of account or medium of account, same thing.

  60. Gravatar of Philippe Philippe
    2. September 2013 at 16:44

    “If they promised to accept gold bars in payment, would that make gold bars a liability of the government?”

    I think so, but gold bars are also real assets, so they have a real value beyond that of the government liability. In contrast pieces of paper, or base metal coins, don’t have much ‘real’ value beyond their use as ‘money’.

    I know the base includes coins, I didn’t say otherwise. I simply said that you can, if you want to, ‘redeem’ Federal Reserve notes for treasury coin.

    Coins are assets on the Fed’s balance sheet. In principle the Fed will always redeem its notes for Treasury coin. Whether this would always happen in practice I do not know.

    Anyway, these days Treasury coins are just as ‘fiat’ as Federal Reserve notes so no-one bothers to find out.

  61. Gravatar of Steve Waldman Steve Waldman
    2. September 2013 at 17:32

    Scott,

    I don’t think we’re disagreeing at all.

    I agree that if interest rates are positive but stable, there’d be an HPE effect from base money injections, although the magnitude of the effect would depend on 1) the opportunity cost of holding base money (if any); and 2) the degree to which the asset purchased is a substitute for base money, reducing the net effect of the purchase. So, with you I think, I’d say that traditionally the HPE was strong (there was an opportunity cost to holding base money), and there remains a (somewhat weaker) HPE when base money purchases long-term bonds, because long-term bonds are imperfect substitutes for interest-paying money or short-term bonds.

    Because I think markets believe that base money will pay interest at rates comparable to or greater than short-term bill rates indefinitely, I don’t believe there is a significant HPE when new money purchases shorter-term Treasuries. Under current market expectations, I think that base money and short-term Treasuries are near perfect substitutes. There might even be a slight inverse HPE, as base money pays higher rates than short-term Treasuries, so holding reserves implies a negative opportunity cost for banks.

    None of this conflicts with anything you’ve said. I think we’re just agreeing. If the Fed wishes to provoke a HPE, it ought to be buying longer-term bonds, agencies, maybe even munis or FX, things as far away from perfect substitutes for cash as it is legally permitted to buy.

  62. Gravatar of lxdr1f7 lxdr1f7
    2. September 2013 at 17:47

    Wont the HPE just send IOR up and eliminate the opportunity cost of holding reserves?

  63. Gravatar of Philippe Philippe
    2. September 2013 at 17:47

    possibly the difference between a specific liability and just vaguely promising to accept something in payment in the future, is that when you issue a liability you specifiy the amount you owe in nominal terms at least.

    So for example if the government were to stamp ‘$1’ on a bar of gold and promise to accept that bar of gold in payment for ‘$1’ owed to it, that ‘$1’ bar of gold would be a ‘$1’ liability. Whereas if the govt just said “oh you can pay me with a gold bar in the future”, but didn’t specify the nominal value of the gold bar that it would accept in payment, then it wouldn’t be a ‘liability’ in the same way.

    (Just my thoughts and are not based on any particular theory, by the way).

  64. Gravatar of dtoh dtoh
    2. September 2013 at 19:26

    Scott,
    You said;

    ” The neutrality of money concept is partly informed by theory. After all, Hume had no data to work with, he was just thinking about the idea in the abstract. When the data confirmed Hume, that solidified the QTM, as economists were already inclined to accept it on logical grounds.”

    Yes, yes and yes, but…. the QTM is still just an equality….it’s not a deterministic relationship. I’m not debating that certain types of increases in M can raise P or NGDP, I’m just saying that logically you can’t use the QTM itself as proof that increases in NGDP are solely due to increases in M (because of the HPE or any other reason).

    But in fact that is just the logic that you and others use when trying to explain HPE. Essentially what you are saying is that QTM is true in the long run therefore HPE must be true.

    As I said I think this is an erroneous a priori assumption that has crept into monetary economics. Probably because when people first started thinking about it, monetary injections were done in a way where there was in fact an HPE and therefore the causality could be assumed.

    Once you realize that you’re assuming HPE and that the assumption is flawed (again it may still be correct, it just can’t logically be assumed), then you can step back and take an objective look at causality and the HPE.

  65. Gravatar of Sumner’s Cold Potatoes | Rugged Egalitarianism Sumner’s Cold Potatoes | Rugged Egalitarianism
    2. September 2013 at 19:43

    […] Sumner attempts to explain the so-called “hot potato effect” which has played such an important role in the theories and […]

  66. Gravatar of Sumner’s Cold Potatoes | New Economic Perspectives Sumner’s Cold Potatoes | New Economic Perspectives
    2. September 2013 at 19:47

    […] Sumner attempts to explain the so-called “hot potato effect” which has played such an important role in the theories and […]

  67. Gravatar of Negation of Ideology Negation of Ideology
    2. September 2013 at 20:05

    5b – “Nominal rates are zero and expected to stay zero forever. Now open market operations are meaningless.”

    I assume this is the extreme case that you don’t ever expect to see -like your example of the Fed buying up the whole planet before NGDP hits a target. Still, do you mean all rates? Even if the 30 year Treasury is at 0% and expected to stay there forever there are still state and local bonds, Fed member bank bonds, etc.

    Also, since no one could expect interest rates less than zero, wouldn’t it would require every single person to expect zero rates for expected rates to be zero? For example, if 90% of the people expect rates to be zero and 10% expect them to rise to 5% that means on average, people expect 0.5% rates. (It’s late, my math may be off.)

  68. Gravatar of Franky Franky
    2. September 2013 at 20:50

    Scott, I am not giving you advice. You dismissed my comments without answering them, and I guess my frustration showed.
    I am also not making a case with you. All I am trying is to understand whether you have a model that’s different from mainstream NK monetary theory. Maybe mainstream theory is missing something, that is my motivation. So far I’m convinced it’s not.

    But since I’m accusing you of not answering my comments let me make sure I answer yours:

    “I strongly disagree. The MOA role is all important. In both cases the value of gold changes, but in only one case does the price level change significantly.”

    MOA matters for what happens to the price level, yes. What I meant is it doesn’t matter for the hot potato effect you described.

    “You said you finally understood me, but I am afraid it’s back to the drawing board. I did not say OMOs are ineffective at zero rates, and I did not assume money is used to buy other goods and services, at least in the short run. You need to read what I say more carefully.

    On OMOs: you said OMOs are “meaningless” when interest rates are 0 forever (case 5b). That’s what happens if money has no value besides its cash flows. It becomes a perfect substitute for bonds, and interest rates must be 0 in equilibrium. So what I am saying is fiat money having no role as MOE leads to your case 5b. Do you disagree?

    On money being used to buy other goods: if you are not assuming that, then let’s assume there’s an OMO and people try to get rid of the extra money by just buying bonds (or other assets, it doesn’t matter). Then nominal interest rates fall so that people are happy to hold the larger money balances. If you ignore the effects of the lower nominal interest rate, are you saying that consumption goods prices will go up anyway? They won’t, people are happy to hold the extra money because it’s not a hot potato anymore after interest rates adjust. I guess I can sum up my point by saying that the bond market cools off the hot potato before it hits consumption prices.

    “Some Keynesians consider a rise in stock prices to be an “interest rate channel.” It seems to me that is moving the goalposts, because the actual interest rate channel is so weak”

    Just a note on this. The reason macroeconomists focus on bonds and interest rates is for simplification. You only need money and one other asset to get the insights that the basic NK model can give you. It doesn’t mean other assets don’t matter for transmission, just that once short term interest rates move all other asset prices adjust by arbitrage. Here’s a good quote from Blanchard illustrating this (and how the financial crisis challenged that view): “Before I came to the Fund, I thought of the financial system as a set of arbitrage equations. Basically the Federal Reserve would chose one interest rate, and then the expectations hypothesis would give all the rates everywhere else with premia which might vary, but not very much. It was really easy. I thought of people on Wall Street as basically doing this for me so I didn¹t have to think about it.”

  69. Gravatar of Ralph Musgrave Ralph Musgrave
    2. September 2013 at 21:18

    Well done Scott Sumner and those who follow his blog. MMTers got there years ago.

    What took you all so long?

  70. Gravatar of Ralph Musgrave Ralph Musgrave
    2. September 2013 at 21:41

    Scott Sumner’s final sentence above is ridiculous: “I think I see a few MMTers in the ditch along the way, still scratching their heads.”

    The hot potato effect is inherent to MMT: MMTers have stressed over and again the importance of what they call “private sector net financial assets”. E.g. an increase in the monetary base, all else equal causes an increase in PSNFA. And that pretty obviously induces the private sector to spend more. It really doesn’t take a genius to work that out, although many of the commentators on Scott’s blog (e.g. the above) seem to have a problem with the idea. And there are plenty of so called professional economists who can’t work it out.

    To verify my point that the hot potato or “private sector net financial assets” idea is inherent to MMT, just Google MMT and “private sector net financial assets”.

  71. Gravatar of Tom Tom
    3. September 2013 at 00:23

    Sumner wrote: “Tom, I disagree. Cash is a real good. Otherwise why would people hold cash at zero rates when market rates are 5%?”

    I’m unable to interpret what you’re trying to say. You’re breaking with the standard definition of “real” as used in economics, which refers to material things and work. For example, take the definition of savings as real investment plus net lending. “Real” investment refers to all the material goods and assets produced that weren’t consumed. Net lending is the person or group’s net accumulation of financial claims, including cash. Because all financial claims are somebody else’s liabilities, on a global scale net lending always sums to zero, and savings always equals real investment. That’s the distinction I’m making. I’m just following the standard definition of economic terms as used in UN conventions.

    And my contention is that accelerating production of financial claims does tend to cheapen them relative to other financial claims and real goods, but not nearly as predictably, evenly and proportionally as accelerating production of a real good such as gold cheapens it relative to other real goods.

    “The real value of a given stock of cash is X. The real value of the stock of cash is still X if you double the money supply. That’s why money is neutral.”

    Here you are simply stubbornly clinging to an abundantly disproved hypothesis. M2 is up from $7.7 trillion five years ago to $10.7 trillion. If $7.7 trillion is X, the current money supply is 1.4X. And yet the real value of the current money supply is about 1.3X. Oops. Back to the drawing board.

    The crucial thing you’re missing is spending. Your hypothesis seems to fit evidence in a place like Zimbabwe, but only because the government there was spending the money that the central bank was issuing. For a central bank to cause inflation, it must not only create money, that money must somehow reach people in the real economy who are willing to increase leverage and spending. That can be the government or the private sector, and it can reach them directly or indirectly via commercial banks. But spending must increase, or the real value of money will not fall.

    To other commenters on whether QE creates bank deposits: the proof is there in the M2 growth. There hasn’t been nearly enough bank credit growth in the past five years to create an additional $3 trillion of broad money. QE did most of that job. To be pedantically exact, QE creates bank deposits except to the extent the net sellers of assets to the central bank are commercial banks. (In the US since 2008, overall, banks were actually net buyers.)

    Another pedantic point, since another commenter replying to me was confused about this: primary dealers are broker-dealers, not banks. Broker-dealers hold their cash in commercial bank deposits, not in Federal Reserve accounts.

  72. Gravatar of RebelEconomist RebelEconomist
    3. September 2013 at 00:32

    dtoh, For what it is worth, I agree with you – there is a problem when the “hot potato” is introduced by buying something, especially something newly created for that purpose, like debt. I dare say Mike Sproul would make the same point too. And I was also hoping for a more mechanistic explanation than this. Indeed, I can think of nothing more important in macroeconomics at this time than developing a mechanistic knowledge (ie understanding supported as necessary by detailed real world observations) of the monetary policy transmission mechanism.

    I am afraid though that you (and Franky) are wasting your time trying to iteratively reach a joint understanding with Scott. You will find that you just get to identifying the key points to get to grips with, and he has moved on to write two or three new posts, at least one of which will be on something outside his area of expertise. Very frustrating!

  73. Gravatar of dtoh dtoh
    3. September 2013 at 01:33

    Rebel,
    “I am afraid though that you (and Franky) are wasting your time trying to iteratively reach a joint understanding with Scott. You will find that you just get to identifying the key points to get to grips with, and he has moved on to write two or three new posts, at least one of which will be on something outside his area of expertise. Very frustrating!”

    Well actually I feel like I’m making progress. Scott has a lot of balls (or should I say potatoes) up in the air, so it’s hard to keep up with him especially since I have a day job and am usually 11 or 13 times zones ahead.

    Also, I’m not really aiming for a “joint” understanding. I feel like I understand the mechanism perfectly (took me a long time to get here) so I’m actually looking for complete submission. 🙂

    At this point, I think he’s close to conceding (although he’ll probably deny it) that the financial asset price mechanism is important and maybe from a short term perspective the most important mechanism.

    As I’ve said repeatedly, once you get this, the rest of monetary policy is trivial and completely obvious.

  74. Gravatar of lxdr1f7 lxdr1f7
    3. September 2013 at 01:49

    Tom
    “To other commenters on whether QE creates bank deposits: the proof is there in the M2 growth. There hasn’t been nearly enough bank credit growth in the past five years to create an additional $3 trillion of broad money. QE did most of that job. To be pedantically exact, QE creates bank deposits except to the extent the net sellers of assets to the central bank are commercial banks. (In the US since 2008, overall, banks were actually net buyers.)”

    Thats a good description of the QE mechanism. The problem is that deposits just go to the primary dealers and not the broader economy. Does the fed only do QE with primary dealers or are other entities included?

  75. Gravatar of lxdr1f7 lxdr1f7
    3. September 2013 at 03:58

    Tom
    Im pretty sure primary dealers have reserve accounts at the new york fed. Therefore QE doesnt directly affect deposits, Im pretty sure it does indirectly when primary dealers buy assets from from financial markets.

  76. Gravatar of DOB DOB
    3. September 2013 at 04:59

    Scott,

    You lose all the MMT-ers on #5. They specifically refuse to switch from gold to cash and expect the chicken to continue running with its head chopped off. I don’t agree with MMT-ers much, but that’s one of the few things they get right. Gold has intrinsic value (it shines) while Cash does not.

    Re #7, if you mean “cashless” in the Woodford sense, I believe he assumes base qty = 0 INCLUDING reserves. Money/reserves only ever exists as a medium of exchange instantaneously at the time of transactions as and both parties instantaneously buy/sell other assets at the time of the transaction. It exists throughout time as a unit of account though.

    You are by far the most prominent voice in pushing the NGDPLT cause forward. That is a noble cause and its importance can’t be overstated. I could be wrong but I think you turn off a lot of your readers with the old-school way of looking at quantity/HPE. 95% of the difference between QTM and interest rates is just vocabulary. The remaining 5% are in things such as the dynamics of the zero bound.

    Still awaiting moderation on a comment in “Interest rates versus the base” btw.

  77. Gravatar of ssumner ssumner
    3. September 2013 at 06:45

    Gasman, In 1930 the US unit of account was “US$” Our medium of account was 1/20.67 ounce of gold. Is that so difficult?

    Philippe, You said:

    “”If they promised to accept gold bars in payment, would that make gold bars a liability of the government?”

    I think so,”

    I don’t follow that at all. I guess I define liability differently.

    Steve, I mostly agree, but would add that even at at zero rates QE has an expansionary effect (clear in market reactions) even if not for HPE reasons. It could be the expected future HPE.

    I would oppose having the Fed buy risky assets, as there is no need. The best option is NGDPLT. If they don’t want to do that they should ramp up QE sharply, instead of tapering.

    lxdr, No, the OR is set by the Fed.

    dtoh, We are not saying money and prices move in strict proportion, we are saying an exogenous increase in M causes a proportional rise in P in the long run.

    Negation, I suppose you’d need the 30 year rate to be zero to get no immediate HPE, but as a practical matter the effect would get quite small if shorter term rates were expected to be zero forever.

    Franky. You still don’t get it. Saying monetary policy is ineffective if rates are zero forever (which will never happen) is very different from saying it’s ineffective if rates are currently zero, which is what you said.

    I don’t agree with the way NKs link asset prices to short term rates. Bonds often move in the opposite direction from stocks after a surprise money announcement. And if they really mean “stock prices” when they say “interest rates” then they should NEVER, EVER talk about the zero bound problem. Stocks are never at the zero bound.

    You said;

    If you ignore the effects of the lower nominal interest rate, are you saying that consumption goods prices will go up anyway? They won’t, people are happy to hold the extra money because it’s not a hot potato anymore after interest rates adjust.”

    That doesn’t follow at all. The main transmission mechanism is expected future growth in NGDP. And the future growth in NGDP is caused by the HPE. And expectations of that effect can cause consumption to rise today.

    Ralph, I’ve given up trying to figure out MMT. Some of them say OMOs are ineffective, others say they are effective. Some say the Fed can’t control the base. Whatever.

    Tom, You said;

    “Here you are simply stubbornly clinging to an abundantly disproved hypothesis. M2 is up from $7.7 trillion five years ago to $10.7 trillion. If $7.7 trillion is X, the current money supply is 1.4X. And yet the real value of the current money supply is about 1.3X. Oops. Back to the drawing board.”

    I hate to tell you this, but you are far behind the rest of the discussion here. The others have at least figured out that I’m not an old style monetarist. I don’t believe V is even close to be stable. Do your homework on market monetarism and then come back.

    And a MOE is a real good that makes shopping more convenient in exactly the same way a wallet or purse is a real good that makes shopping more convenient. That’s why people hold it at positive interest rates.

    dtoh, You said

    “At this point, I think he’s close to conceding (although he’ll probably deny it) that the financial asset price mechanism is important and maybe from a short term perspective the most important mechanism.”

    When have I denied that asset prices are important?

    DOB, Surely you know that in economics the term “intrinsic” is meaningless. All value is subjective. “Shiny” is no more valuable than “useful to shop with.” If a MOE did not have value in creating a more convenient shopping experiences then people would not hold a zero interest MOE during periods where nominal rates are 5%. Heck, they hold cash when rates are 100%!! It’s really that simple.

    A MOE has a value of roughly 1% to 2% of GDP. It’s that simple. The base is often larger because people also hold base money as a store of value (tax evasion, or zero nominal rates) In any case, because a MOE is worth about 1% or 2% of GDP, increasing the MOE 100 fold (exogenously) does not cause the value of the MOE to be 100% to 200% of GDP, rather the price level rises and the value of each unit of MOE falls by roughly 99%

    Sorry, but I don’t see your other comment in moderation. Can you try again?

    Correct me if I am wrong, but doesn’t the Woodford model assume a MOA exists, just that its net value is zero. Some banks have positive balances and some have negative balances. All monetary policy is done via control of the demand for the MOA, i.e. IOR.

    If not, then what is the IOR applied to?

    BTW, market monetarism differs from old style monetarism in that we focus both on the supply and demand for money. So IOR creates no problems for MMs.

    BTW, claiming a HPE for cash at positive rates is hardly controversial. It’s why Krugman doesn’t buy MMT at positive rates, even Tobin talks about a HPE. It’s not just MMs.

  78. Gravatar of Philippe Philippe
    3. September 2013 at 06:59

    Scott,

    I agree. I wasn’t thinking clearly. The gold bar itself isn’t a liability.

    But say for example a $10 dollar note were stamped on a thin gold plate rather than on a piece of paper. Would the $10 note still be a liability?

  79. Gravatar of Franky Franky
    3. September 2013 at 07:33

    Scott,

    “You still don’t get it. Saying monetary policy is ineffective if rates are zero forever (which will never happen) is very different from saying it’s ineffective if rates are currently zero, which is what you said”

    No, that’s not what I said at all. And you didn’t answer my question.

    “That doesn’t follow at all. The main transmission mechanism is expected future growth in NGDP. And the future growth in NGDP is caused by the HPE. And expectations of that effect can cause consumption to rise today.”

    Yes it does follow. If there is no HPE to start with there is no growth in NGDP (except through interest rates). And there is no HPE if people just try to get rid of the extra money by buying bonds and interest rates fall so that people are happy to hold that extra money without spending it. Which is what an optimizing consumer would do. If you refuse to discuss my claim and just repeat your statements, maybe we should call it a day and move on. I’m happy with my understanding of your ideas, and I don’t think you’re going to change your mind.

    “I don’t agree with the way NKs link asset prices to short term rates. Bonds often move in the opposite direction from stocks after a surprise money announcement. And if they really mean “stock prices” when they say “interest rates” then they should NEVER, EVER talk about the zero bound problem. Stocks are never at the zero bound.”

    It’s not the NKs that link asset prices through no arbitrage conditions. It’s the entire field of finance that’s based on that. Which is very different from saying that interest rates are the only thing determining stock prices. Do you disagree?

  80. Gravatar of JP Koning JP Koning
    3. September 2013 at 07:34

    Scott: So if nominal rates are zero and expected to stay zero forever, a central bank can literally buy up every asset in an economy and not have an effect on the price level, right?

    [You say HK has a dual MOA, yet you admit in your comment to Gasman that the US’s UOA in the 1930s was $US and the MOA was gold. You’re being inconsistent: either Hong Kong’s UOA is HK$ and MOA is US$ or US had two MOA’s in the 1930s.]

  81. Gravatar of gasman gasman
    3. September 2013 at 08:22

    “Gasman, In 1930 the US unit of account was “US$” Our medium of account was 1/20.67 ounce of gold. Is that so difficult?”

    This is simply wordplay Scott. When on a commodity standard a CB promises to give someone x amount of gold per $, Yen, Pound ….whatever. So in those situations (which virtually no one does anymore BTW since they are not flexible enough) when someone says they have a dollar they could also be saying “I have 1/35th of an ounce of gold”. Today someone COULD say “I have 1/1200th of an ounce of gold” and someone would look at them and say “Great, I’m happy for you”. They could also say “I have 1/3 of a gallon of gasoline”

    We could also say I have 10 dimes or 20 nickels too. Rewording things is not changing anything real at all.

    Today we have floating exchange rates and there is no promise to convert to anything so your silly point is pointless.

  82. Gravatar of DOB DOB
    3. September 2013 at 08:45

    Scott,

    “Surely you know that in economics the term “intrinsic” is meaningless. All value is subjective. “Shiny” is no more valuable than “useful to shop with.””

    I agree that all values are subjectives (though some can be somewhat constrained by arbitrages), but I do think that “intrinsic” is a meaningful term and that monetary analysis depends on whether money is commodity or fiat.

    I recognize the existence of liquidity preference (or as JPKoning calls it: convenience yield/non pecuniary return of money) and that it explains why people will hold money despite its returns being (in general) inferior to that of bonds.

    “Heck, they hold cash when rates are 100%!”

    How much and for how long? What’s the case you’re referring to?

    “A MOE has a value of roughly 1% to 2% of GDP. It’s that simple.”

    Is that another way of saying velocity should be 50x-100x but was ~ 15x pre-crisis in the US due to tax evaders? Out of curiosity, how do you get to 1-2%?

    “Correct me if I am wrong, but doesn’t the Woodford model assume a MOA exists, just that its net value is zero. Some banks have positive balances and some have negative balances.”

    By balances do you mean balances at the central bank? If so, that’s not how I interpret Woodford:

    Woodford writes: “Another case in which a monetary policy prescription would have to be specified in terms of an interest-rate rule would be if our advice were to be applicable to a “cashless” economy, by which we mean an economy in which there are no monetary frictions whatsoever. In a hypothetical economy of this kind, no central-bank liabilities have any special role to play in the payments system that results in a willingness to hold them despite yielding a lower return than other, equally riskless short-term claims.”

    You wrote: “If not, then what is the IOR applied to?”

    In microeconomics, the existence of a price vector that supports an equilibrium doesn’t necessarily mean that goods have to change hands. Similarly, the fact that there exists an interest rate policy that supports the right value for the unit of account doesn’t mean that there needs to be a non-zero base.

    Woodford adds:

    “A second reason why it is useful to consider policy implementation in this hypothetical case is that if we can show that effective interest-rate control is possible even in the complete absence of monetary frictions, it may well simplify our analysis of basic issues in the theory of monetary policy to start from an analysis of the frictionless case”

    And I think he is dead on the money there (pun intended). Not only do I find interest rate policy easier to reason with than QTM, but it also endogenously explains what goes on at the zero bound.

    “Sorry, but I don’t see your other comment in moderation. Can you try again?”

    Re-posted as suggested. I also removed the external link so it does not get moderated. Comment is here.

  83. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 08:56

    Scott, building on my question here:

    http://www.themoneyillusion.com/?p=23314#comment-272021

    Lets say that the Mint started minting a new $1M coin (production cost: $0.02), but that they are not permitted to do anything with the seigniorage other than use it to buy new raw materials (say a new Fed deposit is created, just for the Mint for this purpose)… in other words Fed funds go into this deposit but come out again only VERY slowly.

    Now let’s say the Fed starts buying these new coins (which, like all coins that have left the Mint are legally base money the minute they leave the Mint, and are assets to the Fed as long as the Fed holds them)… starting off at $85B a month and increasing their purchases 20% per month until they reach some target: could be NGDP, could be inflation, whatever.

    Also assume that the Fed announces all this and will gladly sell these coins to banks OR a new kind of paper note specifically backed by these coins (i.e. redeemable in these coins) in the usual face values: $1, $5, $10, $20, etc.

    Would this scheme work? Why or why not? What’s the substantial difference with your cases 4 or 5? Thanks!

  84. Gravatar of DOB DOB
    3. September 2013 at 09:04

    Tom Brown,

    Nice! Yet another corner case that interest rate-oriented thinking avoids 🙂

    Maybe the quantity theorists can create a new monetary aggregate called MBB4M (monetary base before mint)…

  85. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 09:17

    lxdr1f7,

    Take a look at this:

    http://www.fms.treas.gov/bulletin/index.html

    It appears that most Tsy debt is passed on to the non-bank, non-primary dealer private sector.

  86. Gravatar of Philippe Philippe
    3. September 2013 at 09:24

    Tom, I don’t understand how your scheme is supposed to work.

    1. Treasury mints $1M coins.
    2. Fed buys these coins at a rate of $85B a month or more.
    3. Treasury can’t spend the money it gets for the coins except very slowly.
    4. Fed offers to sell the coins to banks in exchange for reserve balances and other assets.

    How does any of this achieve anything?

    If the Treasury were allowed to spend the money it gets from the Fed straight away then it would have an impact. But seeing as it can’t, how does the scheme do anything?

  87. Gravatar of DOB DOB
    3. September 2013 at 09:31

    Philippe,

    That’s Tom’s point.

  88. Gravatar of Philippe Philippe
    3. September 2013 at 09:55

    Right. But it doesn’t really address Scott’s arguments in any way.

  89. Gravatar of gasman gasman
    3. September 2013 at 10:06

    Philippe

    Sure it does. Scott is of the belief that “base money” is important to prices and Toms example explicitly adds to base money and as you pointed out

    “If the Treasury were allowed to spend the money it gets from the Fed straight away then it would have an impact. But seeing as it can’t, how does the scheme do anything?”

  90. Gravatar of TallDave TallDave
    3. September 2013 at 10:07

    Tom,

    Let’s say you find a magic box that produces arbitrary amounts of gold (or you find some arbitrarily large quantity of gold). You don’t tell anyone, and you don’t exchange any.

    DOB — but if all values are subjective, then no values can be intrinsic. I think if you try to find meaning there, you can’t avoid the subjectivity problem — sure, gold has lots of industrial uses and various properties that make it more valuable relative to other things shiny and otherwise, but ask the median person which has more value, gold or the cash equivalent, and he’ll probably choose cash because he knows how to spend that 🙂

  91. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 10:41

    Philippe, I think if (in my example) the Mint were to instead be allowed to dump its seigniorage into the TGA, and Tsy spent that, this path would be covered by Scott’s Case 5b, specifically this part:

    “Fiscal policy is powerful but currently inefficient, as the national debt is too small.”

    Scott clarified when I asked (above) that it’s the “deficit” the gov is running that’s important here:

    “Fiscal policy can move NGDP, but the current stance is too contractionary-the government is leaving $100 bills on the sidewalk, and should run bigger deficits.”

  92. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 10:43

    TallDave, in my example the Fed announces loudly to the world that its vaults are overflowing with new base money and that that money is available to buy and that they’ll be permanently adding 20% more a month. No secrets.

  93. Gravatar of Ralph Musgrave Ralph Musgrave
    3. September 2013 at 10:43

    Scott,

    I suggested above that MMTers had in mind the hot potato effect when referring (as they often do) to “private sector net financial assets”. After a bit of Googling, I’ve discovered there is more disagreement amongst MMTers there than I thought there was.

    So point taken.

  94. Gravatar of TallDave TallDave
    3. September 2013 at 10:50

    Tom — OK, assume the magic box’s existence is well-known, and you plan to increase the world supply of gold by 20% a month, but you don’t plan to exchange any of the new gold, except for equivalent qtys of other gold.

  95. Gravatar of Scott Sumner, the hot potato effect and MMT. | The Jefferson Tree Scott Sumner, the hot potato effect and MMT. | The Jefferson Tree
    3. September 2013 at 10:55

    […] to Scott Sumner (who teaches economics at Bentley University, Mass) for explaining the so called “hot potato” […]

  96. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 11:08

    TallDave, actually, my hypothetical Fed will accept standard Fed deposits as payment, or standard reserve notes. Or if the public wants to buy the new notes, I’ll let them trade their bank deposits (with banks acting as an intermediary: i.e. the banks debit the public’s bank deposits and the Fed in turn debits the banks’ Fed deposits).

  97. Gravatar of Joe Eagar Joe Eagar
    3. September 2013 at 11:17

    Reserves and cash are liabilities of the U.S. government, at least in an accounting sense. They are backed by U.S. debt (and occasionally mortgages), debt the U.S. Treasury would otherwise have to sell on the open market if the public weren’t willing to hold U.S. dollars.

    Of course, in practice people don’t ditch their a currency for anything less than hyperinflation; short of that, the government can compel its citizens to hold base money in a variety of ways (reserve requirements being the traditional mechanism of financial repression).

    Thus, the “liability” of base money may be nothing more than a promise not to engage in hyperinflation. I’m not an expert on this sort of financial accounting, though, and I could be misreading the situation.

  98. Gravatar of Philippe Philippe
    3. September 2013 at 11:19

    “Toms example explicitly adds to base money”

    I doesn’t though… the coins sit in the Fed until someone wants to buy them for some reason. The monetary base does not increase in Tom example.

  99. Gravatar of Philippe Philippe
    3. September 2013 at 11:20

    *It doesn’t though…

  100. Gravatar of gasman gasman
    3. September 2013 at 11:33

    Philippe

    Toms own words

    “the Fed starts buying these new coins (which, like all coins that have left the Mint are legally base money the minute they leave the Mint, and are assets to the Fed as long as the Fed holds them)… starting off at $85B a month and increasing their purchases 20% per month until they reach some target: could be NGDP, could be inflation, whatever.”

    All coins that leave the mint are base money and Fed assets

  101. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 11:34

    Philippe, that’s like arguing that banks don’t loan out reserves because there’s no “out” they can be lent out to: by definition reserves are a kind of BANK OWNED asset. Scott has argued in that past that this is semantics, and of course he’s right: cash advances exist!

  102. Gravatar of Philippe Philippe
    3. September 2013 at 11:37

    coins are money the moment they leave the mint, but coins held by the fed are not counted as part of the monetary base, in the same way that the Treasury’s balance at the Fed is not counted as part of the monetary base.

  103. Gravatar of Philippe Philippe
    3. September 2013 at 11:39

    Tom, how is my comment like that? I don’t get your point.

  104. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 12:17

    “but coins held by the fed are not counted as part of the monetary base”

    Just as cash in the public’s hands is no longer “reserves” yet this cash was still lend out from banks via cash advances and the banks’ reserve levels dropped simultaneously. Semantics.

    I think the example I sketched out is similar to Scott’s case 4, only rather than a “company” discovering the gold, it’s the Fed itself: it was in the basement the whole time!… and by some legal oddity they DON’T need to remit to Tsy (say it’s considered to be like that bit of skim the Fed can keep when it otherwise remits interest payments). Case 4 says that gold is the MOA, and now the Fed owns (or is anticipated to own) a huge pile of it outright and EVERYONE knows about it!… and knows the inflation target, NGDP target etc.

  105. Gravatar of Jared Jared
    3. September 2013 at 12:57

    Scott,

    Following up on Steve Waldman’s line of questioning, how would the expected future HPE work if the market anticipates IOR = short-term rates INDEFINITELY (somewhere between your 5b and 5c)? The market may not believe that money and bonds will be near-perfect substitutes forever, but they believe they will be for a very long time.

    To put the same question in terms of your gold analogy, what would happen to the price of gold if NASA announced a huge discovery of gold on one of Neptune’s moons? We don’t currently have the technology to get the gold back to earth, but it seems reasonable to expect that at some point in the future the gold will be accessible. But that could take anwhere from 20 – 200 years. Who knows? How would such hazy expectations of the future supply of gold affect its current price?

  106. Gravatar of Philippe Philippe
    3. September 2013 at 12:59

    It’s not just semantics. If the coins are simply held by the Fed then they are basically outside the economy. They might as well not exist. No one in the private sector owns them, so they are not going to have any effect on the private sector.

  107. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 13:19

    Philippe, that’s why I’m asking Scott. I want to get his view.

  108. Gravatar of ssumner ssumner
    3. September 2013 at 13:19

    DOB, Regarding 100% inflation I presume you are aware of Latin American history 1960-90.

    Bennett McCallum says that Woodfords “moneyless economy” actually has money. I’ve generally found McCallum’s views to be pretty reliable. You said;

    “In microeconomics, the existence of a price vector that supports an equilibrium doesn’t necessarily mean that goods have to change hands.”

    I never said they did, I said the interest has to be paid to someone. Someone must hold the reserves that earn the IOR. That seems obvious to me. Who receives the IOR?

    BTW, It is much easier to imagine an economy with no interest rates, than no MOA. Just imagine a simple economy with no banks or bonds, just cash and consumer goods.

    The interest rate approach only causes confusion at the zero bound. It has led many economists to falsely assume that monetary policy is ineffective at the zero bound.

    Tom Brown, I don’t follow your example at all. The base refers to money in circulation. If that goes up you get inflation. If the Fed buys the coins then they aren’t in circulation. What did I miss?

    Franky, Stop denying you made mistakes in describing my views that are plainly visible in the comment thread. It makes you look like an idiot.

    The HPE is a long run effect. It’s explains why NGDP rises in proportion to prices in the long run, once the interest rates return to normal. Monetary stimulus only effects interest rates for a limited period of time. If that was the only thing pushing prices higher, then there would be no long run increase in NGDP. Only the HPE can explain why a 8700% rise in the base causes an 8700% rise in NGDP. Interest rates don’t work.

    Again, you said;

    “Take away the MOE role of money and interest rates will have to be zero in any equilibrium where people hold any money. And in that case, as you point out, open market operations have no effect.

    Second, you are assuming that people will try to get rid of the extra money by buying goods, rather than financial assets. That is not what an optimizing consumer would do (unless money and goods are complements in the utility function).”

    Those two statements do not accurately describe my views.

    Gasman, I said both the supply and the demand for base money matters for prices. How many times do I have to say that before people get that through their heads!! If you print money, and Treasury demand for money rises equally fast, obviously nothing happens. Only an complete moron would think that is an interesting example.

    I’ll take the rest of them later. Please think before you comment here.

  109. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 13:28

    Scott’s already written the following:

    TB:
    “OK, so if you agree with David [Beckworth] here, and HPE is “very weak” when we’re at “zero rates” then why not ditch the QE and the ER > 0, and just promise that base money will be made available when needed. And then… if we do that, how is that different than what we have been doing prior to 2008? Except that now there’s an explicit NGDPLT?”

    http://www.themoneyillusion.com/?p=22948#comment-267056

    (The above was my “last point” BTW)

    SS:
    “And I certainly agree with your last point [above]. In a sensible system the base money is endogenous. You set the NGDP target, and the public tells you how much base money they want to hold. I’m all for that. But we don’t have a sensible system, the Fed uses QE to signal its target. That’s why it’s such a mess.”

    http://www.themoneyillusion.com/?p=22948#comment-267116

    I want to know why “it’s such a mess.” I also want to explore the possibility of other ways for the Fed to “signal its target” … not as a serious alternative but to flesh out my understanding of Scott’s view of one signal vs another. “Signals” are a kind of psychological tool in my view… and I’m wondering if that’s partly what’s at play here according to Scott.

  110. Gravatar of ssumner ssumner
    3. September 2013 at 14:20

    Joe, That’s about right.

    Jared, The further out in the future, the smaller the effect.

    Philippe, Yes, coins held by the Fed are not part of the base, at least according to textbook definition (cash held by the public plus reserves.)

  111. Gravatar of ssumner ssumner
    3. September 2013 at 14:23

    Tom, I agree that a NGDPLT policy (which is to provide base money as need) is better than the current QE oriented regime.

    QE is a mess because the Fed is very vague as to what it is trying to target. So it’s hard to know how much of the QE is permanent.

  112. Gravatar of Franky Franky
    3. September 2013 at 14:25

    Scott, when I say “Take away the MOE role of money and interest rates will have to be zero in any equilibrium where people hold any money. And in that case, as you point out, open market operations have no effect.”, any reasonable reader will understand I am saying interest rates will have to be zero forever, not just currently, when money has no role as MOE. In fact I had explicitly said right before that quote that I was talking about your case 5b.

    Otherwise I give up. I am happy to let readers decide who the idiot is.

  113. Gravatar of DOB DOB
    3. September 2013 at 14:46

    Scott,

    You said:
    “Bennett McCallum says that Woodfords “moneyless economy” actually has money. I’ve generally found McCallum’s views to be pretty reliable.
    [..]
    I said the interest has to be paid to someone. Someone must hold the reserves that earn the IOR. That seems obvious to me. Who receives the IOR?”

    I can’t speak for Bennett McCallum, or Woodford for that matter, but what I get from Woodford’s book is that in his cashless economy, the central bank stands ready to lend money into existence at a rate i and pays IOR on the existing balance of money at a rate i_m, and it must be that either the base is zero or i = i_m (in which case there is no opportunity cost to holding money as is the case now in the U.S.). Since the representative households are identical, it can’t be that some borrow from others: everyone’s money balance is zero. Nobody pays IOR to anyone.

    “BTW, It is much easier to imagine an economy with no interest rates, than no MOA. Just imagine a simple economy with no banks or bonds, just cash and consumer goods.”

    How is cash created in such an economy? Bitcoins? There is certainly a unit of account in Woodford’s cashless framework. All prices are quoted in terms of it. It’s just that there is not much need for the medium of exchange to exist for an extended period of time, and in the limit, at all.

    “The interest rate approach only causes confusion at the zero bound. It has led many economists to falsely assume that monetary policy is ineffective at the zero bound.”

    I can’t speak for other economists and I don’t think monetary policy is ineffective at the zero bound. However, I do think that it’s QTM that causes confusion and leads economists to believe that increasing base at the zero bound serves a purpose.

    In this other thread, you suggested that by tripling the base, the Fed somehow led the market to expect it would let it grow by 15% over 3-years, leading NGDP to grow by 5%/year. Why do they need to raise the base from 1 to 3 to get back to 1.15? What if they had raised it to 4 or 12? Why would they not raise it to 1.15 and leave it there? Or better yet by 5%/year since there’s no certainty we’ll be off the zero bound in 3-years?

  114. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 15:33

    Scott, thanks very much for your responses.

    To put my example another way: if in your Case 4 (gold = MOA) the Fed had found all the gold, announced it to the world, didn’t remit any to Tsy, and put it all up for sale immediately… that would NOT cause any inflation since because the Fed owned it, it would not be base money. Correct?

    I’m not arguing the point… I just want to make sure I understand. TallDave pointed out “what would people buy this gold with? More gold” … sure, I take that point. But that’s a *little* bit similar to your Case 7 (cashless) especially if we add a couple of extra details: like no reserve requirements and no Tsy spending. Then those excess reserves (w/o any excess bank equity!) would literally just sit in the banking system (passed around between banks, to be sure, at the IOR) with no place to go except back to the Fed someday. Or as TallDave might say “what are you going to buy those reserves with? More reserves?”… but would those reserves nonetheless have a psychological impact on expectations a little bit like an explicit NGDPLT? (although more messy, like QE is now… regarding Fed “signaling” purposes only).

    Or is the crucial difference with Case 7 (and QE for that matter) that in Case 7 the Fed likely conducted (using banks as intermediaries) the bulk of its OMOs with non-bank private sector bond sellers, and thus there are as many “excess” bank deposits as there are excess reserves?

  115. Gravatar of Philippe Philippe
    3. September 2013 at 16:38

    Tom,

    The only real difference between conducting OMOs with non-banks and conducting them with banks is that in the first case non-banks get the money, and in the latter case banks get the money.

    It doesn’t matter than in the first case the money received by non-banks is bank deposits, and in the latter case the money received by banks is reserves. That’s not an important difference.

  116. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 16:45

    Philippe, banks get the reserves in either case.

  117. Gravatar of gasman gasman
    3. September 2013 at 16:47

    Think before I comment?

    This coming from an “economist” who talks daily about monetary policy yet has openly stated he doesn’t know anything about banking and seems to think it is unimportant, even though it is via banks that monetary policy works its voodoo

    Didn’t you also claim recently that Japan’s price level was 100x ours and Koreas 1000x? Wow that is some serious inflation

  118. Gravatar of DOB DOB
    3. September 2013 at 17:00

    gasman,

    “This coming from an “economist” who talks daily about monetary policy yet has openly stated he doesn’t know anything about banking and seems to think it is unimportant, even though it is via banks that monetary policy works its voodoo”

    Having worked in a bank for a long time, I think I know quite a bit about banking. And, I agree with Scott that banks don’t need to have a special place in any monetary model.

    If the rate of return on nominal assets is higher than the Wicksellian natural rate, prices will fall. If it’s lower, prices will rise. It’s that simple, and the only thing banks can do is change the natural rate.. Just like a change in consumer preference, a productivity or supply shock, or fiscal policy. In every case, monetary policy can change the return on nominal assets to match the new natural rate.

    (Scott doesn’t like to think of it in terms of interest rates but I think he gets to similar conclusions for similar reasons.)

  119. Gravatar of Philippe Philippe
    3. September 2013 at 17:16

    “banks get the reserves in either case”

    Yeah I know, I should have been clearer. In the latter case bank assets and liabilities increase by the same amount, as non-banks deposit the money in banks. In the first case there is no bank balance sheet expansion, only a change in the composition of the bank’s assets (a reduction in bonds and an increase in money).

    The real difference between these two cases is that in the latter case non-banks receive money, whereas in the first case only banks receive money. The fact that non-banks get ‘deposits’ and banks get ‘reserves’ is not important.

  120. Gravatar of Philippe Philippe
    3. September 2013 at 17:36

    “If the rate of return on nominal assets is higher than the Wicksellian natural rate, prices will fall. If it’s lower, prices will rise.”

    Can you prove that this theoretical natural rate exists?

  121. Gravatar of lxdr1f7 lxdr1f7
    3. September 2013 at 18:17

    Tom

    It seems as if the demand for treasuries outside the banks is just because of risk adversity not QE. QE has probably diminshed the demand for treasuries.

  122. Gravatar of lxdr1f7 lxdr1f7
    3. September 2013 at 18:29

    The current method of monetary policy is not effective because it depends on banks so much. If the banks don’t move out of reserves then the returns on other assets aren’t attractive enough. Its like a negative loop. It may be broken by gov spending but if that fails then what?

    You need to have a third tool for when the banks enter the negative of loop of not taking their money out of reserves and the gov doesnt deficit spend enough to lift the economy. That third tool can be the fed directly interacting with the people to directly conduct policy when inflation is insufficient.

  123. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 18:49

    Philippe, fine call it money if you want. Sure the “bank deposits” part is not important. But Scott’s Case 7 was explicitly cashless, and that’s the case I’m interested in here… which kind of leaves bank deposits as the only option.

  124. Gravatar of Tom Tom
    3. September 2013 at 19:00

    To Sumner,

    re: “I hate to tell you this, but you are far behind the rest of the discussion here. The others have at least figured out that I’m not an old style monetarist. I don’t believe V is even close to be stable. Do your homework on market monetarism and then come back.”

    I’m not obliged to study everything you’ve ever written before I respond to what you write. You wrote that if the money supply starts at X and increases to 2X, the real value will still be the same. I corrected you. Now you admit you left out the role of V. Okay, indeed you did. But once you include V, the concept of neutral money falls apart, the hot potato effect works very differently than it does with gold, and my point, repeated below, begins to make sense.

    re: “And a MOE is a real good that makes shopping more convenient in exactly the same way a wallet or purse is a real good that makes shopping more convenient. That’s why people hold it at positive interest rates.”

    You can define “real” as you wish. I and the rest of the world will continue to define “real” goods as material goods and distinguish them from financial claims. My point is that production of financial claims has different results than production of real goods, as the latter are normally defined.

    To lxdr1f7:

    “The problem is that deposits just go to the primary dealers and not the broader economy.”

    That’s not true. Primary dealers pass the bank deposits on to the net seller of assets, which in the US case is usually the US government, which passes them on to the rest of the economy when it spends. See Fed release Z1 tables L204 and L205 for breakdowns of who’s holding the supply of bank deposits over time.

    “Does the fed only do QE with primary dealers or are other entities included?”

    Directly, only with primary dealers. However primary dealers are mainly market makers. They have prop books but they are rarely significant net sellers of the assets the Fed is buying.

    “Im pretty sure primary dealers have reserve accounts at the new york fed.”

    No, they don’t. You can find a list of who is allowed to hold reserve accounts at the Fed on the Fed site. Many broker-dealers are of course part of larger banking groups, but the broker-dealer is a legally distinct entity which holds its cash in a commercial bank account, not at the Fed. Anyway this is a pedantic technical point, of no real relevance to whether QE creates bank deposits.

  125. Gravatar of Tom Brown Tom Brown
    3. September 2013 at 19:00

    lxdr1f7,

    If you look here:

    http://www.fms.treas.gov/bulletin/index.html

    Depository institutions (banks) appear to hold no more than about $0.35T in Tsy debt, but the total held by the private sector is between about $5T to $10T… so when the Fed buys it’s trillions of Tsy debt on the open market during QE, I’m supposing that the non-bank private sector is the main net seller. Do you think that’s wrong?

  126. Gravatar of Tom Tom
    3. September 2013 at 19:18

    PS to Sumner, re V: Just to clarify a little, I would rather say you left out the role of spending. I don’t think of velocity as especially important and certainly not as a driver. It’s merely the ratio between money supply and spending.

  127. Gravatar of lxdr1f7 lxdr1f7
    3. September 2013 at 20:19

    “Depository institutions (banks) appear to hold no more than about $0.35T in Tsy debt, but the total held by the private sector is between about $5T to $10T… so when the Fed buys it’s trillions of Tsy debt on the open market during QE, I’m supposing that the non-bank private sector is the main net seller. Do you think that’s wrong?”

    I think the primary dealers are “non-banks” as in non depository banks. These primary dealers probably hold a large portion of treasuries since they are primary dealers in treasuries. The primary dealers must be the main net sellers. I think the SOMA accounts is how they interact with the NYfed and where they get reserves from the fed. The fed doesnt disclose the holders of SOMA accounts i dont think. I emailed and called them but no straight answer.

  128. Gravatar of lxdr1f7 lxdr1f7
    3. September 2013 at 20:26

    “That’s not true. Primary dealers pass the bank deposits on to the net seller of assets, which in the US case is usually the US government, which passes them on to the rest of the economy when it spends. See Fed release Z1 tables L204 and L205 for breakdowns of who’s holding the supply of bank deposits over time.”

    Yeah I understand that. The gov. doesnt spend enough or spend money effectively so the economy is just beholden to the interests or ignorance of the primary dealers or whatever administration is in charge of the gov.

    We need to create a mechanism whereby the fed directly interacts with the public as well to bypass all the gov and primary dealer blockages.

  129. Gravatar of Tom Tom
    4. September 2013 at 01:05

    @Tom Brown

    re: “Depository institutions (banks) appear to hold no more than about $0.35T in Tsy debt, but the total held by the private sector is between about $5T to $10T… so when the Fed buys it’s trillions of Tsy debt on the open market during QE, I’m supposing that the non-bank private sector is the main net seller. Do you think that’s wrong?”

    Sorry for butting in, but indeed that is wrong. The US government has been the biggest net seller of Treasurys during most of the QE era. The US government’s net sales of Treasurys are about the same as its deficits. The non-bank domestic private sector has, as a whole, been a net buyer of Treasurys during most of the QE era. It accumulated about $1 trillion of Treasurys between the ends of 2009 and 1q13. The foreign sector has bought about $2 trillion during the same period. The data is in Fed Z1 table L209.

    This has changed in recent months as the deficit has shrunk and the government has maneuvered to keep debt sales to a minimum, in order to push back the date when the debt cap will restrict spending. The government’s net sales of Treasurys have been negligible since Q1. The Fed data past Q1 isn’t out yet, but we know from TIC data that non-official foreign accounts have recently been net sellers.

  130. Gravatar of Philippe Philippe
    4. September 2013 at 01:28

    Tom,

    This has all the info on everything:

    Financial Accounts of the US
    http://www.federalreserve.gov/releases/z1/Current/z1.pdf

    Treasury securities are on page 93. (L.209)

  131. Gravatar of gasman gasman
    4. September 2013 at 02:32

    DOB

    You may have worked at a bank for years but if you think Scotts models that ignore banking are correct you have as much authority to comment on banking as our billing person has an authority to discuss medicine, and she’s worked in health care for years too.

    Phillipe took the words about “Wicksellian natural interest” right out of my mouth.

  132. Gravatar of DOB DOB
    4. September 2013 at 02:55

    gasman,

    In other words if I disagree with you I must be an idiot?

    I guess this conversation is over.

  133. Gravatar of DOB DOB
    4. September 2013 at 03:14

    Philippe,

    “Can you prove that this theoretical natural rate exists?”

    Hmmm.. I don’t feel the need to. It’s present in virtually every macro model as the real rate of interest.

    Intuitively it makes sense that, risk-adjusted, every asset would have to have the same expected return otherwise investors would buy the higher-yielders which would raise their yield back in line.

    That risk-adjusted expected real return is the natural real rate of interest. There’s a good article at the St Louis Fed here.

    P.S.: if you respond to one of my comment, pls put ‘DOB’ somewhere at the top so that it catches my attention. I nearly missed it

  134. Gravatar of ホットポテト効果、説明! — 経済学101 ホットポテト効果、説明! — 経済学101
    4. September 2013 at 05:35

    […] Sumner, “The “hot potato effect” explained”  The Money Illusion, […]

  135. Gravatar of Philippe Philippe
    4. September 2013 at 06:54

    DOB, that link doesn’t work.

  136. Gravatar of gasman gasman
    4. September 2013 at 09:51

    DOB

    Not idiot, but working at a bank doesnt give one anymore insight if they choose not to look. Its pretty clear that banks are not simple intermediaries that can be assumed out of any macro model of the economy. You really should get outside this monetarist bubble and see what others have discovered. JKH over at Monetary Realism is an excellent source as is Scott Fullwiler (you can google him and find many many banking discussions). The world is not a gold standard world and its not a world where your savings are the source of my loans. A saver can borrow too. Everyone with an income can be a borrower at the same time from a bank (Its hard for those without an income to borrow)

  137. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 10:14

    @Tom,

    “The US government has been the biggest net seller of Treasurys during most of the QE era”

    Does that include the Social Security Trust fund and Federal worker retirement funds? Does it include Tsy auctions of Tsy debt? Where is the SS in that data (I searched for Social Security and nothing came up… I thought they held a large amount of Tsy debt).

    @Philippe, thanks for the link.

    Either of you, why is line 41 in Table L.209 not $16.7T? Instead it’s $11930.1B. Are they subtracting off debt held by the US Fed gov itself, including the Fed?

    (Line 41 is Federal government debt)

    Also, what is the “Monetary authority?” (table L.209). Is that the Fed?

  138. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 10:27

    @Tom, so if you’re saying that Tsy itself has been the primary net seller of Tsy’s I’d have to agree. Then there’s Social Security and Fed gov worker retirement funds, GSEs etc (I call that stuff “intra-gov”). Those I count as “gov” but you could make a case that some of those funds (SS, Fed retirement funds) are owned by the private sector.

    But the Fed doesn’t buy (much) directly from Tsy, correct? So they have to buy on the 2ndary market… from intra-gov, from PDs, banks, foreign CBs, and the non-bank private sector (foreign and domestic). Of course they use intermediaries to buy from entities which do no hold Fed deposits (e.g. non-bank private sector).

    So given that, who is the Fed primarily buying from? Who’s a net seller to the Fed? Not Tsy, right? How much Tsy debt does the Fed buy from Tsy directly?

  139. Gravatar of Philippe Philippe
    4. September 2013 at 10:29

    If the US govt is a ‘net seller’ I assume that means the Treasury is selling bonds at auction. The SS trust fund doesn’t sell bonds to the public.

    L.209 leaves out ‘intra-governmental holdings’ of govt debt, but not debt held by the Fed.

    The Monetary authority is the Fed.

    ‘Households’ includes things like hedge funds (!).

  140. Gravatar of Philippe Philippe
    4. September 2013 at 10:32

    ‘Analyzing Federal Reserve Asset Purchases: From whom does the Fed buy? (2013)

    http://www.federalreserve.gov/pubs/feds/2013/201332/201332pap.pdf

  141. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 10:36

    @Tom, is it wrong to assume that the non-bank private sector is a main purchaser of Tsy debt from Tsy auctions (using PDs as intermediaries) and that it is also in turn a main seller of Tsy debt to the Fed (again using intermediaries)? This does not mean that it couldn’t also be a net buyer of Tsy debt overall (since it’s coming in from Tsy on one end, and going out to the Fed on the other: if the input from from Tsy is greater than the output flow to the Fed, they’ll acquire more Tsy debt over time).

    Perhaps you’ll claim this doesn’t matter because the end result looks like Tsy sold directly to the Fed. What I’m trying to do though is break it up into a two step process. 1. Tsy sells Tsy debt, 2. QE. I agree the end result looks the same as if the Fed bought directly from Tsy.

  142. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 10:47

    Philippe, thanks for that link as well. It’s never an easy answer… 🙁

    I’ll dig in. The two charts at the very end look they might be an easy answer, but I can’t figure out all the colors there. The left hand plot shows one curve WAY above the rest… is that foreign? I can’t tell. I’m not even sure what that plot means exactly.

  143. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 11:02

    I appears that households, pension funds, and investment funds come just shy of $1T apiece, and that “rest of world” comes in at about $5.5T. Everything else looks lower than that (for the left had chart in figure 1). I wonder if the “rest of world” includes non-bank private sector? I wonder if it includes foreign CBs? … Now I’ll try to reconcile this with all their tables.

  144. Gravatar of Philippe Philippe
    4. September 2013 at 13:06

    “What I’m trying to do though is break it up into a two step process”

    Why? These things are going on at the same time…

  145. Gravatar of Philippe Philippe
    4. September 2013 at 13:09

    “rest of world” means rest of world, whatever is out there in the rest of the world.

  146. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 13:25

    Why? Why not? I never doubted that Tsy is the biggest seller of Tsy debt. I guess what I meant was 2ndary market seller. Tom’s response left me confused until I understood that he was probably including Tsy itself in “biggest seller.” But you’re correct, in the end it doesn’t matter too much. Even if the Fed were to purchase direct from Tsy… in my super simplified model, public’s money stock still is L + B + F

    “rest of world” … well again I’d like to know what that consists of.

  147. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 13:29

    … and I guess I’m not the only one who cares: apparently so do Seth Carpenter, Selva Demiralp, Jane Ihrig, and Elizabeth Klee.

  148. Gravatar of DOB DOB
    4. September 2013 at 14:03

    Philippe,

    Sorry, correct link is here.

    gasman,

    “working at a bank doesnt give one anymore insight if they choose not to look.”

    Agreed.

    “Its pretty clear that banks are not simple intermediaries that can be assumed out of any macro model of the economy.”

    The opposite is clear to me. I went through a phase where I (briefly) believed in MMT and thought like you. Now I don’t anymore.

    “You really should get outside this monetarist bubble and see what others have discovered.”

    If there’s something in MR that’s not in MMT and that’s supposed to convince me that banks matter, I’m happy to hear it. I’m not going to read it all in hope of finding it though.

    “The world is not a gold standard world and its not a world where your savings are the source of my loans. A saver can borrow too. Everyone with an income can be a borrower at the same time from a bank (Its hard for those without an income to borrow)”

    None of these trivial tautologies require that banks be special in a macro model. In fact none of what you said is specific to banks.

  149. Gravatar of Philippe Philippe
    4. September 2013 at 14:11

    “rest of world” … well again I’d like to know what that consists of.”

    It’s everything outside the country, foreign governments, foreign banks, foreign households etc

    point is that if the Treasury and the public are selling treasuries to primary dealers at the same time it’s difficult if not impossible to know if there is a “two step process”.. it all gets mixed up together.

  150. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 14:29

    “two step process” … sure I agree, but isn’t sorting that out the point of the analysis in that paper you linked to above?

    http://www.federalreserve.gov/pubs/feds/2013/201332/201332pap.pdf

    “In this paper, we exploit Flow of Funds data
    (described in the next section) to identify
    which types of investors are sell
    ing to the Federal Reserve during
    four different asset programs:”

  151. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 14:39

    DOB, check out this article by Nick Rowe (who’s an MMist):

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/08/banks-and-the-medium-of-exchange-are-both-special-or-neither-special.html

    So, what does he think? Are banks “special?”:

    “So I think that commercial banks are macroeconomically important because, and only because, they influence the supply of media of exchange.”

  152. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 14:54

    DOB, not to convince you one way or the other, but it just so happens that JKH looked at the Tobin papers recently too (if you’ve been following that… Nick Rowe looked at it too… concluding there were two Tobins… Scott disagreed, said just one), but on the JKH:

    http://monetaryrealism.com/krugman-and-tobin-on-banking/

    http://monetaryrealism.com/bank-reserves/

    Just thought I’d point it out! … and one from Ramanan:

    http://www.concertedaction.com/2013/08/24/holier-than-tobin/

    I only read the 1st one myself, but just noticed they were there.

  153. Gravatar of DOB DOB
    4. September 2013 at 15:17

    Tom,

    I posted a comment on that article.

    Briefly, I think the unit of account is macroeconomically significant but the medium of exchange isn’t, and neither are banks.

  154. Gravatar of ssumner ssumner
    4. September 2013 at 15:49

    Franky, Sorry, my mistake. I didn’t know that when people say “interest rates are zero” they mean “interest rates are zero and will remain zero for the next billion years.”

    DOB, You said;

    “Nobody pays IOR to anyone.”

    So why is the IOR an important policy tool in Woodford’s framework?

    I agree with McCallum that without a MOA there is no anchor to the price level, but I’ll have to put off that debate as I simply don’t understand Woodford’s model well enough. When T-bonds come due and people are repaid the principle, what do the bondholders get?

    You said:

    “I can’t speak for other economists and I don’t think monetary policy is ineffective at the zero bound. However, I do think that it’s QTM that causes confusion and leads economists to believe that increasing base at the zero bound serves a purpose.”

    I half agree. QE at the zero bound is useful, as we see from the financial market reaction. However many conservatives take the wrong message from the QTM, and predict high inflation, which is wrong.

    You said;

    “In this other thread, you suggested that by tripling the base, the Fed somehow led the market to expect it would let it grow by 15% over 3-years, leading NGDP to grow by 5%/year. Why do they need to raise the base from 1 to 3 to get back to 1.15? What if they had raised it to 4 or 12? Why would they not raise it to 1.15 and leave it there? Or better yet by 5%/year since there’s no certainty we’ll be off the zero bound in 3-years?”

    No I did not, you misunderstood me. I don’t think they led the market to think that. I said if they wanted the market to think that they should announce a 5% NGDP target. If they did so they would not have had to triple the base, a point I’ve made many times.

    I favor letting the market determine the base. It’s up to the market to decide how much base money they want to hold when NGDP growth is expected to be on target.

    Tom Brown, You said;

    “To put my example another way: if in your Case 4 (gold = MOA) the Fed had found all the gold, announced it to the world, didn’t remit any to Tsy, and put it all up for sale immediately… that would NOT cause any inflation since because the Fed owned it, it would not be base money. Correct?”

    If I understand your question (doubtful) then I’d say that after the gold was sold it would be owned by the public, and prices would rise. Indeed if it was known the gold would be sold, prices would rise in anticipation.

    Gasman, You said;

    “Didn’t you also claim recently that Japan’s price level was 100x ours and Koreas 1000x? Wow that is some serious inflation”

    And so he proves my point.

    DOB, You said;

    “(Scott doesn’t like to think of it in terms of interest rates but I think he gets to similar conclusions for similar reasons.)”

    Yup, at least we are able to have an intelligent discussion. Here’s one additional thought. There is als o a Wicksellian equilibrium nominal price of zinc. That’s the price of zinc where NGDP is expected to grow at the target, say 5%. If NGDP is expected to grow faster than 5% then the market price of zinc will exceed the Wicksellian equilibrium price of zinc, and vice versa. I accept all that. And I believe the same about the Wicksellian equilibrium fed funds rate (in reverse). But I don’t think the price of zinc plays a particularly important role in the transmission mechanism. Undoubtedly the role of short term rates is a bit more than zinc, but I still don’t think it plays a major role in the transmission mechanism.

    I tend to prefer zinc to interest rates as an indicator, because easy money always tends to increase the price of zinc, but it doesn’t always reduce interest rates.

    Tom, you said;

    “Now you admit you left out the role of V.”

    Which statement did I make that assumed V doesn’t change. Provide an exact quote–in context. Maybe I did get sloppy, but I’d like to see the quote. Was I talking about long run effects?

  155. Gravatar of Philippe Philippe
    4. September 2013 at 15:58

    DOB

    “wage rigidity causes unemployment… without the rigidity part, there is no recession”

    Pure ideology.

    btw I left some comments at your blog.

  156. Gravatar of DOB DOB
    4. September 2013 at 18:01

    Scott,

    “[If nobody pays IOR to anyone], why is the IOR an important policy tool in Woodford’s framework?”

    Because it, along with the cost of funds, dictates the real return on nominal assets and therefore changes in the price level.

    “When T-bonds come due and people are repaid the principle, what do the bondholders get?”

    They get… hmmm… money 🙂 Here’s how (I think) it works: imagine in this scenario the Treasury pays the principal of its bond with some tax revenue and the rest with the issuance of a new bond.

    – At time t1, people liquidate assets in order to pay taxes to the treasury
    – At time t2, other people liquidate other assets to purchase the freshly issued bonds from the treasury
    – At time t, the treasury collects from these two and distributes the money to the holders of the maturing bond
    – At time t3, those bond holders purchase assets with the money they got back
    – (Optional) If the assets traded by the above 3 don’t match exactly, other investors get involved too, etc.

    Now, assume efficient electronic systems clear all this in a split second. In other words, take the limit as t1, t2, t3 -> t. In such a world, what’s your monetary base at the close of business? Zero.

    “I half agree. QE at the zero bound is useful, as we see from the financial market reaction.”

    I’d explain any effect seen in the market through the signaling effect of QE: “the Fed means business”. I’m starting to think we’re in agreement there. The problem with the signaling effect is that the asset purchases are no more useful than a peacock’s tail. However, unlike peacocks, central bankers can speak and announce what they will do. Again I think we agree and just spoke a different language.

    “Yup, at least we are able to have an intelligent discussion.”

    Thought we always were 🙂

    “Here’s one additional thought. There is also a Wicksellian equilibrium nominal price of zinc. [..] I tend to prefer zinc to interest rates as an indicator, because easy money always tends to increase the price of zinc, but it doesn’t always reduce interest rates.”

    So I never viewed overnight nominal interest rates as an indicator. Cost of Funds/IOR/the base are the control knobs, while the price of zinc/CPI/NGDP are the objectives.

    To this I know what you will say: that you can MAKE the price of zinc the control by having the central bank peg it and thereby get rid of the zero bound issue. (Like it would do with NGDP futures.)

    To this I would respond that I don’t like the idea of having the central bank hold a large and volatile inventory of zinc. Having large variations in the quantity demanded of zinc means that the zinc:aluminum price ratio and zinc:CPI and zinc:NGDP ratios aren’t going to be the same as they would otherwise have been. That means that despite the price of zinc being pegged, the price of other assets could drop. That’s how the zero bound shows up in the pegged-currency space.

    Negative IOR would solve the problem. In fact IOR could be used to have the Fed’s inventory of zinc kept small and constant.

    The only thing the Fed can buy to support the price of output at the zero bound is, well, output. That would be deficit spending..

    By the way, I just thought about what this means for NGDP futures. I don’t think you’ll agree with this but hopefully you’ll find it interesting. I think your NGDP futures peg is basically a fiscal operation artificially attached to an OMO:

    If the Fed (it should really be the Treasury) locks the NGDP futures market on a 5% growth path, firms/investors can effectively sell future output to the government at that price. That mean they can invest, hire, spend without fear of not finding a buyer for aggregate output. If they do fail to sell output at the promised price, the futures contract makes them whole and the government debt rises by the same amount. In a way, NGDP futures are a government promise to run a deficit contingent on the output gap (scaled down to the number of people who want to hedge it). Krugman should love it 🙂

    (Of course I’ve swept a bunch of details under the rug: market access, initial margin requirements, no one firm or even stock index provides a complete cross-section of GDP, etc. but we wouldn’t want the government to open this kind of loophole and rely such details to protect it would we?)

  157. Gravatar of gasman gasman
    4. September 2013 at 18:03

    Scott

    Your 100 x price level in Japan is the same as if we priced things in cents instead of dollars. Yes we would add two more zeros to every price tag but no one would pay a penny more than they are currently paying. Inflation would only be if prices were rising in the current MOA not an altered MOA which is 1/100th of the previous one.

    Epic fail

  158. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 18:08

    Scott,

    Your responses surprise me!

    “If I understand your question (doubtful) then I’d say that after the gold was sold it would be owned by the public, and prices would rise. Indeed if it was known the gold would be sold, prices would rise in anticipation.”

    To boil it down even simpler, I presented two cases:

    A. MOA is NOT gold (Like in your cases 5 and 6). Fed finds $1T of coins in the basement.

    B. MOA = gold (Like in your case 4). Fed finds $1T of gold coins in the basement.

    In both cases the Fed decides to keep the coins (rather than destroy them), puts ALL the coins up for sale, intends to keep all the profits, and announces all this to the world.

    In case A. (my 1st example) you said the coins are not “base money” because they’re not in circulation. I agree that they are not base money because they’re at the Fed, and thus not in circulation, but they’re still money and assets of the Fed (as all coins at the Fed are).

    In case B, the gold coins are at the Fed (so again: they’re money, but not base money because they’re not in circulation). But now you have a very different answer! Why?

    Sorry if my examples were not clear! I’ve tried to simplify them here and place them side by side for easy comparison.

  159. Gravatar of gasman gasman
    4. September 2013 at 18:22

    DOB

    Let me explain why I think my examples are not tautologies

    In order for banks not to matter all lending could simply be understood as me loaning my 1000 dollar savings to you. A zero sum game. You can spend that 1000 I can’t, until I’m repaid.
    What in fact happens is I have a 1000 dollars in a checking or saving account at a bank, you come seeking a loan, your income and collateral justify it and you go spend the 1000 dollars, and I can spend mine too. Your spending has taken no spending power form me. Its non zero sum. The bank creates the 1000 for you to spend and I can spend by transferring my deposit to another bank account
    And of course this applies to every person with an income. They can borrow and no one else needs to save.

  160. Gravatar of gasman gasman
    4. September 2013 at 18:26

    Scott
    You seem to believ that inflation is simply a bigger number not a bigger number in the SAME unit/ medium of account. Japan simply prices I pennies. If we did that that would not be inflation or 100 x price level it would simply be renaming what we price in

    Trivial

  161. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 18:28

    DOB, so it’s not just PKEers that think that banks might be special… apparently at least one MMist does too (Rowe). What about other MMists that think that MOE is “special?” Bill Woolsey?

  162. Gravatar of Tom Brown Tom Brown
    4. September 2013 at 19:50

    DOB, just the “unit of account?” Here’s my understanding:

    Say we’re under a gold standard and $1 = 1/35 oz gold:

    UOA = dollar
    MOA = gold
    MOE = bank deposits, reserve notes, etc denominated in dollars

    The specification of the UOA in terms of the MOA is what anchors it: i.e. $1 = 1/35 oz of gold.

    The only thing that’s different now (from the gold standard) is that the MOA = CPI basket of goods and the relationship is that the UOA represents a slowly shrinking (at 2% a year) portion of that basket of goods.

  163. Gravatar of Max Max
    4. September 2013 at 22:21

    TomB, “The only thing that’s different now (from the gold standard) is that the MOA = CPI basket of goods and the relationship is that the UOA represents a slowly shrinking (at 2% a year) portion of that basket of goods.”

    Yes, exactly.

    It may be clarifying to consider a proposal made in the gold standard days for stabilizing the price level. The idea was that instead of fixing the dollar price of gold, the dollar price of gold would be varied by the government to maintain a constant price level. Even though money is convertible into gold, gold is not the MOA; a price index is. It’s a fully “fiat”, floating exchange rate system, but with gold convertibility. A central bank is optional.

    What I think this ancient proposal clarifies is:
    (1) The equilibrium price level is determined by the definition of the unit of account, not by interest rates or quantity of money.
    (2) Lack of convertibility is not a defining characteristic of fiat money.
    (3) Fiat money doesn’t require a central bank, only a central authority that frequently redefines the money unit in terms of a commodity.

  164. Gravatar of DOB DOB
    5. September 2013 at 03:57

    gasman/Philippe/Tom, all comments on my blog have been answered.

    gasman,

    “And of course this applies to every person with an income. They can borrow and no one else needs to save.”

    In theory, how is you spending a bank-IOU (money) any different than you spending a gasman-IOU? I could at the same time spend a DOB-IOU. All the banking system does is it takes these “dirty” IOUs (in the form of mortgages or loans) and turns them into clean and fungible IOUs: money. That’s the definition of credit intermediation.

    The (financial) savers are those who accept the money (or the XYZ-IOUs) in exchange for goods and services. Note that MMT limits its definition of savings to financial savings; other economists call that savings minus investments. When Robinson cuts 10 coconuts, eats 9, and puts one away for tomorrow, that’s saving and nobody else needs to be borrowing for him to be able to save.

    Tom,

    “just the “unit of account?””

    Yes, being the unit of account is why money is macroeconomically relevant because that means it denominates labor contracts, the prices on display, and unfortunately it also denominates a lot of debt (I think it shouldn’t but that’s a different story).

    If all contracts were denominated in Gold units and physically settled in dollars at the then-prevailing exchange rate (for instance the sticker tags in your supermarket is in Gold units but when you get to the cash register they convert the total into dollars at the prevailing market rate so that you can pay with notes), then the dollar would not be macroeconomically relevant despite being the medium of exchange.

    Philippe,

    “Pure ideology.”

    I should be more specific and exclude supply-side recessions from the picture. There could also be small flare-ups in unemployments as labor rebalances from one industry to another. Unemployment insurance and minimum wage laws (supply side factors) also contribute to unemployment.

    Other than that, I stand by what I said, and I think many neo-Keynesians, Monetarists and even MMT-ers agree with this. If not rigidity, then what do you think they think causes unemployment?

  165. Gravatar of Tom Brown Tom Brown
    5. September 2013 at 07:34

    DOB, you write:

    “If all contracts were denominated in Gold units and physically settled in dollars at the then-prevailing exchange rate (for instance the sticker tags in your supermarket is in Gold units but when you get to the cash register they convert the total into dollars at the prevailing market rate so that you can pay with notes), then the dollar would not be macroeconomically relevant despite being the medium of exchange.”

    You’ve described something I didn’t touch on with either my gold MOA or my CPI MOA: a floating rate (and thus a floating definition) between the MOE and the UOA. What about my fixed definition case (gold) or my creeping peg case (CPI)?

  166. Gravatar of Tom Brown Tom Brown
    5. September 2013 at 07:41

    DOB, It seems to me that crawling peg case (MOA=CPI basket) is interesting. How do you know the stock of MOE doesn’t factor in to what the CB needs to do to keep the crawl going as planned? (i.e. to keep the UOA on track as an ever shrinking slice of the MOA?)

  167. Gravatar of Tom Brown Tom Brown
    5. September 2013 at 07:56

    DOB, here are Bill Woolsey’s and JP Koning on UOA vs MOA:

    http://www.themoneyillusion.com/?p=17412#comment-201444

    http://jpkoning.blogspot.ca/2012/11/discussions-of-medium-of-account-could.html

  168. Gravatar of Fed Up Fed Up
    5. September 2013 at 14:02

    Tom Brown said: “Say we’re under a gold standard and $1 = 1/35 oz gold:

    UOA = dollar
    MOA = gold
    MOE = bank deposits, reserve notes, etc denominated in dollars

    The specification of the UOA in terms of the MOA is what anchors it: i.e. $1 = 1/35 oz of gold.”

    Looking at the last comment of jpkoning’s post, it seems to me:

    Gold Standard

    UOA = 1/35 oz of gold
    MOA = gold

    $1 = 1/35 oz of gold defines a fixed (important) convertibility.

  169. Gravatar of Fed Up Fed Up
    5. September 2013 at 18:42

    “5. Now switch the medium of account from gold to cash. I claim that changes nothing essential.”

    I believe there is a difference. A gold miner probably won’t accept the gold it has produced back. The fed will accept currency the treasury has produced on the treasury’s behalf. Currency held at the fed or held at the treasury does not circulate in the real economy. It has a real economy velocity of zero.

  170. Gravatar of Fed Up Fed Up
    5. September 2013 at 18:55

    DOB, if banks and bank-like entities produce new demand deposits and demand deposits are both MOA and MOE, then do banks and bank-like entities now matter?

  171. Gravatar of Odie Odie
    6. September 2013 at 09:48

    Hello Scott,

    Sorry I am late to the discussion but I would really like to understand your post and where the divergence between Monetarists and MMT arises. It is my understanding that you want to explain with the HPE how the Fed affects the price level, correct? Can we then agree on the following:

    1. The Fed controls the amount of reserves held by banks.
    2. The total money supply M influences total spending (NGDP or PQ) and assuming V stays constant (I don’t want to debate V = const.).
    3. The Feds actions control the price level.

    I think it is point 3 that MMT does not agree with and you try to explain through the HPE that this is indeed the way the Fed increases M and with it P. When I got you correctly the HPE works by the Fed announcing their policies will change the amount of reserves held. The change in interest rate will give a lower value to the reserve money and therefore the banks will be forced to lend it out. Did I get you so far?

  172. Gravatar of Odie Odie
    6. September 2013 at 11:17

    Btw. What do you think of this statement from a former Fed president? Does that fit with your HPE?

    “It is true that the Monetary Base, which used to be considered high-powered money because it consists of currency outstanding plus the reserves of the banking system, expands with the expansion of bank reserves. But, with banks hoarding excess reserves as they have been, the Monetary Base has not had its historical impact on the public’s money supply. If one insists on calling the Monetary Base ‘money,’ then it is money that has gone only to the Treasury and the sellers of MBS. This has made the financing of our outsized deficit easier and cheaper. That will come to an end some time and financing the deficit will become more expensive; then holders of fixed-income securities will experience market losses if they sell them before maturity. But there are no inflationary pressures building.”

    http://www.forbes.com/sites/bobmcteer/2013/03/21/the-fed-has-not-been-printing-boatloads-of-money/

  173. Gravatar of DOB DOB
    6. September 2013 at 11:18

    TallDave,

    “if all values are subjective, then no values can be intrinsic.”

    The intrinsic value of an apple is that it can be eaten and enjoyed. How much exactly that it’s worth is subjective as it depends on consumer utility functions etc.

    In that sense, gold has intrinsic value and cash does not.

    “ask the median person which has more value, gold or the cash equivalent, and he’ll probably choose cash because he knows how to spend that”

    Some people must think that gold is worth at least as much as the cash equivalent otherwise it wouldn’t trade there.. No intrinsic value doesn’t mean no value.

  174. Gravatar of DOB DOB
    6. September 2013 at 11:20

    Fed Up,

    Depends what you mean by “matter”, but loosely speaking, no, they do not matter any more than car manufacturers as long as the central bank retains its control over the rate of return on the unit of account.

  175. Gravatar of DOB DOB
    6. September 2013 at 11:28

    Tom et al.,

    I generally don’t use the expression medium of account. I’ve never seen it used outside of blogs. To me there’s either a target, a peg, or some loose mandate.

    Qualifying the current Fed regime as a CPI MOA crawling peg certainly fails to convey the second part of the dual mandate.

  176. Gravatar of ssumner ssumner
    6. September 2013 at 16:59

    DOB, You said;

    “Scott,

    “[If nobody pays IOR to anyone], why is the IOR an important policy tool in Woodford’s framework?”

    Because it, along with the cost of funds, dictates the real return on nominal assets and therefore changes in the price level.”

    Sorry, but I just don’t follow this. I don’t see how an interest on reserves that is never paid can control anything.

    Regarding hoards of zinc, I agree that the government should not hold them. But under the gold standard countries actually did not need to hold stocks of gold (many did of course) in order to peg the price of gold. They didn’t even need stocks of gold to have a crawling peg with gold.

    Of course I’d oppose those policies, as I prefer NGDP futures targeting.

    Gasman, I was talking about the price level, not the rate of inflation. But since you brought it up, how do you suppose the yen and won got so small in the first place?

  177. Gravatar of ssumner ssumner
    6. September 2013 at 17:03

    Odie, I don’t have much time, but as far as I can tell the MMTers have nothing in common with market monetarism. I the other hand I can’t claim to have a deep understanding of their model, as whenever I tried to read the stuff people give me after 15 minutes I give up and throw something at the computer. Their arguments seem based on fallacies. Don’t they say the Fed can’t control the base because they target interest rates?

  178. Gravatar of Tom Brown Tom Brown
    6. September 2013 at 17:39

    Scott, not defending MMT here, but just to clarify, you write:

    “Don’t they say the Fed can’t control the base because they target interest rates?”

    previously, I’d asked you this:

    http://www.themoneyillusion.com/?p=23186#comment-271969

    and you wrote:

    http://www.themoneyillusion.com/?p=23186#comment-271999

    (The quote I gave there was a Nick Rowe quote actually)

    So I think I see what you’re saying: it’s incorrect to say the Fed CAN’T control the base, however they may very well chose not to (directly) and instead let it be a function of some other target (and they have done this frequently in the past). They may even chose to make this their publicly known policy.

    Am I reading you right there?

  179. Gravatar of Tom Brown Tom Brown
    6. September 2013 at 18:02

    Fed Up,

    Yes that appears to be a slight difference between Woolsey and JP (to my reading anyway!): JP incorporates the definition tying the UOA to the MOA into the UOA. I think I like JP’s better, but it’s a small difference (one less thing to keep track of!).

  180. Gravatar of DOB DOB
    6. September 2013 at 18:08

    Scott,

    “Sorry, but I just don’t follow this. I don’t see how an interest on reserves that is never paid can control anything.”

    Imagine that cost of funds is 5% and IOR is 4.99%. Let’s say payment systems are so efficient that even at this narrow spread, the base is 0. (Anytime someone makes a transaction they borrow from the Fed and pay it back intraday, therefore it never pays any interest).

    Now you might ask why do the cost of funds and IOR matter at all since nobody pays it?

    Well what’s the yield on short term treasuries? If it was higher than 5%, people would buy them, post them to the fed, and pocket the difference.

    If it was lower than 4.99%, nobody would ever buy them, they’ much rather leave cash at the fed.

    So despite the fact that the base is zero, the policy rates dictate yield on nominal assets and therefore control the price level.

    Here is my toy monetary model that I think is Woodford-compatible, though that would be up to him to decide. Would love to hear your thoughts.

    “But under the gold standard countries actually did not need to hold stocks of gold (many did of course) in order to peg the price of gold.”

    I agree. With the right policy rates you can target the price of anything without ever having to hold or trade any of it 🙂

    “Of course I’d oppose those policies, as I prefer NGDP futures targeting.”

    What did you think of my comment re this? In a way NGDP futures are just a cash-settled market on future output. If you forget the OMO that you attached to it for a second, you effectively want the government to commit to buy (or sell) future output at a fixed price.

    If it fails to get the market price high enough (NGDP below target), it compensates those who hedged by paying them the difference (which is essentially the same as buying their output at pre-agreed price and selling it back into the market at current price). This isn’t a prediction market, as you like to call it. It’s a REAL market for actual output (it’s just aggregated and cash settled..).

    Looks to me like contingent deficit spending.

  181. Gravatar of gasman gasman
    7. September 2013 at 05:13

    Scott

    First off here is your most recent quote, similar to something you said a year or so ago

    ” It tells us why prices are not 100 times higher than they are, or 1000 times higher. BTW, prices in Japan are 100 times higher than in the US, and Korean prices are 1000 times higher.”

    You were relating our price level to some metric you chose as the right one to base your price theory (QTM) on and postulating as to why it wasnt 100x more than it is now. You then claim Japans prices ARE 100x ours! And Koreas 1000x ours! Now you want to tell me you are not talkng about inflation!

    If I told you our prices would be 100x higher at Christmas, would that not be
    inflation?! You’ll have a hard time selling that one Scott!

    But that aside let me tell you why Im sure your claim is bunk. Japans Yen is the equivalent of a penny, not a dollar, so to compare prices and to know whether theirs is truly 100x higher they must be converted to the same unit.
    Otherwise all you are doing is telling a 72 inch person here that Japanese folks are more than 2x taller because they are 172.72 cm , conveniently omitting that 2.54 cm = 1 inch.

    By your reasoning, if we simply just started posting prices in pennies we would have 100x inflation……….. an obviously specious claim.

  182. Gravatar of gasman gasman
    7. September 2013 at 05:25

    DOB

    You said

    “I agree. With the right policy rates you can target the price of anything without ever having to hold or trade any of it”

    I disagree with both you and Scott here. If we were to go on any type of commodity standard and not keep that commodity in reserve to satisfy the desires of someone who wants to exchange their currency for it, that would be an example of the worst type of bad credibility. Why even go on the standard in the first place? You want to have people even more disenchanted than they already are, go on a gold standard but dont promise them any gold for their currency. They can already buy gold at a market price why have a govt controlled gold market which is exactly what you would have under your scenario.

  183. Gravatar of ssumner ssumner
    7. September 2013 at 05:55

    Tom Brown, One reason I have such a hard time understanding MMT is that I get lots of MMT commenters over here that apparently don’t understand their own model. When I tell them the Fed controls NGDP by adjusting the supply of base money they insist it cannot be done.

    Of course there are intermediate targets. The Fed adjusts the base so that the fed funds rate moves to a level where they expect NGDP to be on target. During the 6 week period where the fed funds rate is pegged the base is endogenous. Over the longer term you could consider it endogenous to the macro variable being targeted. Endogeniety is a matter of perspective.

    DOB, You said;

    “If it was lower than 4.99%, nobody would ever buy them, they’ much rather leave cash at the fed.”

    Isn’t this “cash at the Fed” the MOA? Again, I’ve always conceded the net balances of base money can be zero, as long as you allow for individual players to have positive and negative balances. Is that your point?

    The basic point here is that you need a MOA because you need an asset with a fixed nominal price. If you don’t have one, it’s not clear what the liability of bond issuers really is. What do the bond issuers owe people at maturity?

    My NGDP futures plan need not involve any fiscal effects because it can be set up so that the Fed never takes a net long or short positive. In some versions it does, but the fiscal effects are trivial compared to the monetary effects. Of course in a deeper sense all monetary policy involves fiscal effects, due to seignorage.

    Gasman, You said;

    “You were relating our price level to some metric you chose as the right one to base your price theory (QTM) on and postulating as to why it wasnt 100x more than it is now. You then claim Japans prices ARE 100x ours! And Koreas 1000x ours! Now you want to tell me you are not talkng about inflation!

    If I told you our prices would be 100x higher at Christmas, would that not be
    inflation?! You’ll have a hard time selling that one Scott!”

    Yes. It would be hard to sell to someone who doesn’t know the difference between location differences and time differences. I generally assume my readers do know the difference.

  184. Gravatar of ssumner ssumner
    7. September 2013 at 06:01

    DOB, Yes, that looks like a standard Woodfordian model to me. I have no objections at first glance. You simply do monetary policy by adjusting the demand for MOA to control its value, not the supply.

    I work with supply of MOA models because currency doesn’t pay interest.

  185. Gravatar of gasman gasman
    7. September 2013 at 07:12

    Scott, way to dodge the actual question here.

    You said this;

    “Gasman, I was talking about the price level, not the rate of inflation. But since you brought it up, how do you suppose the yen and won got so small in the first place?” in your response at 16:59 on 9/6

    Which was a response to this;

    “Your 100 x price level in Japan is the same as if we priced things in cents instead of dollars. Yes we would add two more zeros to every price tag but no one would pay a penny more than they are currently paying. Inflation would only be if prices were rising in the current MOA not an altered MOA which is 1/100th of the previous one.”

    and this

    “You seem to believ that inflation is simply a bigger number not a bigger number in the SAME unit/ medium of account. Japan simply prices I pennies. If we did that that would not be inflation or 100 x price level it would simply be renaming what we price in”

    So tell me what is price level vs inflation? Why should your claim that Japans price level is 100x ours be economically meaningful to anyone if it DOESNT mean that an American would have to actually spend 100x more if they were buying form a Yen denominated market? If it doesnt mean the previous and simply means Japan decides to divide their “dollar” by one hundred and list prices in that, what should the proper response be besides B…F…D !!

  186. Gravatar of DOB DOB
    7. September 2013 at 07:47

    Scott,

    “I’ve always conceded the net balances of base money can be zero, as long as you allow for individual players to have positive and negative balances. Is that your point?”

    No. In my view, base money can only be a central bank liability so as far as I’m concerned, no-one–other than the central bank–can hold a negative balance of base money. They can issue nominal debt but that’s different. In any case, this also works in the case of a representative agent so that everyone’s money balance is the same: zero.

    “The basic point here is that you need a MOA because you need an asset with a fixed nominal price.”

    Yes, that asset is money. The fact that the outstanding balance of money is zero doesn’t mean that money doesn’t exist or doesn’t have a price. (For instance, treasury bonds trade before the auctions that will set their coupons, in the form of “when-issued” (WI) derivative contracts. The market has a price of them before they are even created.)

    “If you don’t have one, it’s not clear what the liability of bond issuers really is. What do the bond issuers owe people at maturity?”

    Bond issuers pay money that’s instantaneously created and destroyed right around the time of payment (in the limiting case where monetary frictions are completely eliminated). Check out the example in my previous comment.

    “My NGDP futures plan need not involve any fiscal effects because it can be set up so that the Fed never takes a net long or short positive.”

    Interesting. Do you have a link for a place where you explain how the Fed can be net flat NGDP futures?

    “Of course in a deeper sense all monetary policy involves fiscal effects, due to seignorage.”

    Agreed, but with IOR = cost of funds and a crawling peg between physical and electronic currency, seigniorage is exactly zero 🙂

    I think IOR should be set some stable number of bps below cost of funds, which would yield a fairly stable monetary base, and therefore a constant revenue stream for the govt. That’s the revenue it gets in exchange for providing a convenient MOE and payment systems to society..

    “looks like a standard Woodfordian model to me. I have no objections at first glance.”

    I appreciate that you had a look. Thanks!

    “I work with supply of MOA models because currency doesn’t pay interest.”

    I think of physical currency as a ‘barbarous relic’ 🙂

  187. Gravatar of DOB DOB
    7. September 2013 at 07:55

    gasman,

    You can safely assume that Scott isn’t confused by the fact that 100 JPY = 1 USD and stop pressing the issue.

    “I disagree with both you and Scott here. If we were to go on any type of commodity standard and not keep that commodity in reserve to satisfy the desires of someone who wants to exchange their currency for it, that would be an example of the worst type of bad credibility.”

    The only reason why there would be a lack of credibility is that the fiscal govts have historically put pressure on the central banks to generate inflation. But if genuine, legally enforceable, politically untouchable, central bank independence was achievable, the central bank could set the return on money anywhere it sees fit, and that means it could target ANY PRICE IT WANTS (see methodology here)

    If it targeted the price of gold, that could be catastrophic for the NGDP trend path and nobody is suggesting that they SHOULD do it. But in theory, they COULD. Without ever trading any actual gold.

  188. Gravatar of gasman gasman
    7. September 2013 at 08:32

    “You can safely assume that Scott isn’t confused by the fact that 100 JPY = 1 USD and stop pressing the issue.”

    So now you run interference for Scott? Ill press him as much as I want. He hasnt answered the question…. remotely. WTF does “Japans price level is 100x ours” mean to us? If it doesnt mean that we would have to pay 100x more in Japan then its trivial. If it doesnt mean that Japanese people get 100x less for an equivalent amount of work then its trivial. Im asking him why I should care at all that Japan puts two more zeros after their prices.

    “But if genuine, legally enforceable, politically untouchable, central bank independence was achievable, the central bank could set the return on money anywhere it sees fit, and that means it could target ANY PRICE IT WANTS”

    Why would any group of 300 million want to make any individual or small cadre politically untouchable or give them legal standing above anyone else to insure their untouchability. How is their legal standing enforced btw.?…….
    ……… oh yeah by that thing called govt they dont want to have to answer to.
    Quite a conundrum. We dont want you interfering with us and we want you to allow us to legally prosecute you if you do. You just want to make the central bank THE govt, actually the king!

  189. Gravatar of DOB DOB
    7. September 2013 at 09:47

    gasman,

    “So now you run interference for Scott? Ill press him as much as I want.”

    Fair enough. It’s none of my business.. I was only saying this because you alienate him on a silly issue, he’ll stop answering those questions of yours that are actually interesting.

    “Why would any group of 300 million want to make any individual or small cadre politically untouchable or give them legal standing above anyone else to insure their untouchability.”

    It’s called separation of powers and it doesn’t mean that judges are “kings”. At a high level, I understand that congress has legislated a mandate the Fed to follow and have granted it independence and a relative freedom of interference from the political spheres. I don’t exactly know the details of how that works but it sounds like a good thing to me.

    (Btw if you want to see cases where govt try to interfere with its own central bank’s independence, you don’t have to look very far back: you can look at what the clowns in Argentina’s government did 3 years ago)

  190. Gravatar of gasman gasman
    7. September 2013 at 10:50

    So DOB

    What does Japans price level is 100x ours mean to you?

    When you hear that do you think “Gee i better bring a lot of money if I ever go to japan” or do you think “I guess japan just has a different Unit of account, what point is this man trying to make”

    Realize how ever that Scott was trying to say something profound here, he was using japan as an example in his point so how you answer this matters.

    Saying that japan has 100x our price level when it doesnt mean that it is 100x more expensive in Japan is saying exactly what? Scott was using this, I beleieve, in a post about how CBs can generate inflation.

    If our CB/govt just started using pennies as our UOA and changed nothing else would you call that higher prices (which is what inflation is, is it not?)

  191. Gravatar of Odie Odie
    7. September 2013 at 19:33

    Scott, I could also say that your HPE post is giving me a headache. 🙂 Somehow, I am not even sure what you want to explain with it but since I want to understand at least something it would be great if you can confirm what I am thinking it means. In essence, the HPE describes how excess reserves will find their way into the general economy to affect prices/inflation and NGDP. I, personally, have no problem with MV=PQ. I am just not sure if your way of thinking how the Fed and banks think and operate match with the real world. Some problems I have with your explanation:

    1. In your point 4, the price of gold drops because of that large find. When applying that to money you would say its purchasing power declined due to an increase in the money supply. The problem is: Banks do not care about the purchasing power of their deposits. Reserves are x % of their customer deposits; it is not the banks’ money. They will never spend it to buy anything. What banks worry about are their earnings which are (neglecting brick-and-mortar-stuff):

    Earnings = interest received – interest paid – asset (loans) write-offs

    Obviously, banks did not like excess reserves with 0 % return. For that reason they used to lend their excess to banks with less than the required reserves. The rest was acquired by the Fed through OMOs to keep the interest rate at their target level. Since the Lehman bankruptcy banks stopped that lending practice and only wanted to deal with the Fed. The risk of having to write-off assets was/is too big. As long as the spread between interest earned versus interest paid is positive banks will post positive earnings. Why doing risky loans then?

    2. Banks DON’T SPEND money they LEND it. Even when cars are sold at 0 % financing it does not mean that everyone will be storming into the dealerships. Why? Because they have to pay back that money at some point unlike if something is been given away for free. Even if you tell people that rates will rise in the future and inflation is looming there is no way to force them to borrow more if they still have a hangover from the last debt binge. In lending decisions income is the deciding factor (at least now again). Banks as well as consumers/businesses decide based on their current and assumed future earnings if they can carry more debt. If both agree the loan is made. Expectations of inflation play no role IMHO compared with the bank’s desire to make a riskless loan.

    3. I also have the feeling you underestimate V. If M increases but the only thing it does is V goes down no pressure on prices happens. That is what we are seeing now, with savings increasing.

    4. This Fed paper explains quite well how the change in Fed policies affected reserve holdings of banks but has no inflationary effect: http://www.newyorkfed.org/research/staff_reports/sr380.pdf It also should become clear that the rise in the reserves did not change the total value of the balance sheet of banks only its composition.

    Overall, I don’t deny that changes in M can affect PQ but I doubt that increases in the money base will necessarily show up in M. In fact, while base money rose sharply in the last 5 years M2 kept its normal growth trajectory. Thus, the composition of M changed but the total money stock did not although everyone is talking about (hyper)inflation for the last 5 years at least.

  192. Gravatar of gasman gasman
    8. September 2013 at 03:36

    “Banks DON’T SPEND money they LEND it. Even when cars are sold at 0 % financing it does not mean that everyone will be storming into the dealerships. Why? Because they have to pay back that money at some point unlike if something is been given away for free. Even if you tell people that rates will rise in the future and inflation is looming there is no way to force them to borrow more if they still have a hangover from the last debt binge. In lending decisions income is the deciding factor (at least now again). Banks as well as consumers/businesses decide based on their current and assumed future earnings if they can carry more debt. If both agree the loan is made. Expectations of inflation play no role IMHO compared with the bank’s desire to make a riskless loan”

    Bingo Odie!

    Once again, not appreciating the role/function of banks is getting in Scotts way. He acts like banks will just start handing out money to customers at some point if the monetary base gets big enough.

  193. Gravatar of ssumner ssumner
    8. September 2013 at 06:25

    Gasman, If you don’t think relative price levels are interesting, that’s fine. I do think it is interesting to think about why Japanese prices are 100 times higher. Yes, they have a different MOA. But what explains the value of that MOA? Why does it differ from the dollar?

    The yardstick analogy is of course the one monetarists use all the time, it’s exactly our point. To understand inflation you need to understand changes in the (real length of the) monetary yardstick.

    DOB, I’m afraid I still don’t see your point, and the Treasury analogy does not help. The Treasury derivatives you cite are not Treasury bonds, they are Treasury bond derivatives.

    I also don’t follow the representative agent comment. Are you saying the public chooses to hold zero balances of money, or that they are simply unable to hold money even if they want to, because money does not exist? What if they wanted to hold some central bank liabilities, even for irrational reasons?

    And why does your model contain money? Why not a moneyless economy, if you think it works?

    Gasman, You said;

    If our CB/govt just started using pennies as our UOA and changed nothing else would you call that higher prices (which is what inflation is, is it not?)”

    I’ve argued that if the BOJ pegged the yen to the dollar at 200 per dollar, (effectively a half cent) that would create inflation in Japan. Do you disagree? In the long run, it would raise the Japanese price level by roughly 100%.

    Odie, You said;

    “I, personally, have no problem with MV=PQ.”

    That’s good to hear, as it’s a tautology. So at least we agree on tautologies.

    You said;

    “In your point 4, the price of gold drops because of that large find. ”

    No, the price of gold doesn’t change in #4, the value of gold falls. That’s the whole point, the price of the MOA never changes.

    The rest of your point one makes no sense. If the Fed is targeting rates then how can you have an exogenous increase in the base?

    2. In point two you are confusing money with credit. Monetary policy has nothing to do with credit markets, they are unrelated phenomena. When thinking about monetary policy it’s best to ignore banks entirely. Just think about the supply and demand for the MOA. If you insist in thinking about banks, then imagine they have a certain demand for reserves. Now inject more reserves. They will (by assumption) hold more than they want to at current asset prices. They will try to get rid of the excess. Asset prices will change first, then eventually goods and services prices will change.

    You said;

    “I also have the feeling you underestimate V. If M increases but the only thing it does is V goes down no pressure on prices happens. That is what we are seeing now, with savings increasing.”

    Ok, It is clear that what you said earlier is true, you did not understand my post. Please don’t say I don’t understand something until you understand what I am saying. BTW, “savings increasing is the wrong term ,you want to say “money demand is increasing” They are totally unrelated phenomena.

    As for your point 4, how does that conflict with anything I said in my post? You really need to reread the post and figure out what I am actually saying, not what you think I am saying.

    Let me give you an example. In case #4 I said:

    “Obviously the price of gold can no longer plunge, as the price of the MOA is fixed by definition.”

    And you characterized my comment as:

    “In your point 4, the price of gold drops”

    Would you say that’s an accurate reading of my post?

    BTW, if QE has no inflationary effect, then why did QE cause TIPS spreads to widen? Why are the markets horrified by the taper? You do understand (I hope) that the theory that claims QE has no inflationary effect also predicts no impact on asset prices.

    Gasman. You said;

    “Bingo Odie!”

    Why am I not surprised. Odie’s comment was a Godawful mess of misconceptions and misrepresentations, and you think it’s brilliant.

  194. Gravatar of gasman gasman
    8. September 2013 at 14:07

    “I do think it is interesting to think about why Japanese prices are 100 times higher. ”

    They are NOT 100x higher though. 100x higher would mean that they are paying 350$ for gas or 300$ for a gallon of milk, at their exchange rate. Which they are not. Your seeing a different unit of account and wondering “What is that all about?” Different nations use different units of measurement. We use inches some use centimeters. And if you disagree with this what exactly do you mean by Japans prices being 100x higher? What is a higher price?

    If Japan pegged to the dollar at 200 yen per dollar that very well could cause inflation but why would Japan do that?

    Pegging and commodity standards are bad ideas. They have been proven disasterous historically.

  195. Gravatar of Odie Odie
    8. September 2013 at 17:52

    Scott, I had read your post several times but when I asked you (twice!) whether my summary was correct I did not get an answer. But to get to your reply:

    You said:”No, the price of gold doesn’t change in #4, the value of gold falls. That’s the whole point, the price of the MOA never changes.”

    You are right, incorrect wording from my side. However, I think my next sentence is still correct: When money loses its value you would say its purchasing power declined. Agreed?

    You said:”The rest of your point one makes no sense.”

    What does not make sense? That banks don’t worry about the value of their reserves because it is not their money? Or that they only care about their earnings? The earnings equation?

    You said:”If the Fed is targeting rates then how can you have an exogenous increase in the base?”

    Because the Fed buys assets from banks and credits their reserve account thereby creating excess reserves. It then pays an IOR which takes the pressure of banks to lend those reserves out in the interbankmarket. If the Fed would not do that the Fed Funds rate would be close to 0 % like the current short-term yield on T-bills. I don’t deny that QE generated excess reserves but its main effect was to lower long-term interest rates. The Fed seems to agree with me:
    http://www.stlouisfed.org/publications/re/articles/?id=2258
    I also don’t see how the HPE gets banks to lend those excess reserves out.

    You said: “In point two you are confusing money with credit. Monetary policy has nothing to do with credit markets, they are unrelated phenomena. When thinking about monetary policy it’s best to ignore banks entirely. Just think about the supply and demand for the MOA. If you insist in thinking about banks, then imagine they have a certain demand for reserves. Now inject more reserves. They will (by assumption) hold more than they want to at current asset prices. They will try to get rid of the excess. Asset prices will change first, then eventually goods and services prices will change.”

    I don’t get your point that money is not credit. All money has been created through credit; for each monetary asset there is a monetary liability somewhere else. The Fed and banks swap money and bonds (which are a form of credit) all the time. The Fed and banks never spend reserve money on real goods or services; they use what they earn through interest payments for that (their true earnings).

    How do you think the banks will get rid off excess reserves? By spending? On what? If they buy a bond that is nothing else than lending out the money which implies there must be a willing borrower. If they don’t find one rates will fall further and at some point the Fed will intervene to keep their rate target and offer e. g. an IOR so the banks will keep their excess reserves. I am still lost on how you think the banks will get their excess reserves in the hands of consumers and businesses to affect prices of goods and services (Q). The only mechanism I know is making loans which implies there must be someone who wants to borrow.

    You said:” BTW, if QE has no inflationary effect, then why did QE cause TIPS spreads to widen? Why are the markets horrified by the taper? You do understand (I hope) that the theory that claims QE has no inflationary effect also predicts no impact on asset prices.”

    TIPS spread widened because people (wrongly) though QE would mean higher inflation. As the CPI shows that has not been the case; neither could the Fed confirm heightened inflation expectations (see link above). What assets do you actually mean, financial assets or real goods and services?

    “Odie’s comment was a Godawful mess of misconceptions and misrepresentations,”

    Thank you for those kind words about my comment. My parents used to say:”If you can’t say something nice, don’t say anything at all.”

  196. Gravatar of Tom Brown Tom Brown
    8. September 2013 at 21:51

    Hi Odie,

    Here’s a few things that may help you. I’m not an MMist… just an amateur pretending to be one here for a few minutes so take this w/ a grain of salt: I wrote this for someone like you who thinks about the banks (Scott, please don’t waste your time reading this if you are!)

    Imagine we’re in case 7 (no cash). No cash advances: no way to get base money into the hands of non-banks HOWEVER that doesn’t mean that the banks in aggregate still can’t buy stuff from non-banks: Tsy debt, financial assets, salaries, electric bill, dividend payments, loans (i.e. make new loans). How do they do it? By crediting bank deposits. Likewise they sell to the non-banks by debiting bank deposits. Again, in aggregate (forget about reserve & capital requirements).

    But even if they don’t trade w/ non-banks to any great extent the banks still hold excess MOA (after QE) and they still hold non-MOA assets… the prices of which will rise which will raise the prices of those kinds of assets amongst the non-banks. Financial prices up means yields lower means real goods and services prices up (eventually… if we’re not in case 5b… and Fed is doing a credible job of targeting a higher inflation rate or NGDPLT).

    You are right that ER > 0 means FFR = IOR. But long term rates can still rise. Take a look at this David Beckworth comment and post too (he talks about this very thing):

    http://macromarketmusings.blogspot.com/2013/08/a-permanent-expansion-of-monetary-base.html?showComment=1375901514604#c7523207141783165074

    Did David’s post help? Read all Jared’s questions and David’s answers there and in the previous post. Read both Jared’s and JP Koning’s questions here in this post (above), and Scott’s answers.

    http://www.themoneyillusion.com/?p=23314#comment-272734

    This might be helpful too. It was for me. Not all MMists agree about what’s more important: MOA or MOE. Sumner and Glasner tend to say MOA only, but Rowe and Woolsey say MOE. They also say MOE (which are bank deposits in Case 7) can have HPE. Here’s Nick on this:

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/08/banks-and-the-medium-of-exchange-are-both-special-or-neither-special.html

    Now I’ll give you some supplemental material. First here’s two cases for QE to consider:

    http://brown-blog-5.blogspot.com/2013/08/banking-example-41-quantitative-easing.html

    Also, a simple 0th order approximation of the public’s stock of money = L + B + F

    See this post (especially the bottom) for an explanation:
    http://brown-blog-5.blogspot.com/2013/08/banking-example-11-all-possible-balance.html

    So B + F of the public’s money stock is not “backed by” loans to the public. F represents the Tsy debt held by the Fed, so you can think of it as the result of QE. Why am I telling you this? I *suspect* it might help you… for one thing that portion not backed by loans (L) can’t be used to pay down loans and in that sense perhaps it’s more inclined to take on the Rowe/Woolsey asserted HPE for MOE? (i.e. diminished “reflux” capacity there). Not a necessary concept (HPE for MOE) but I like the Rowe/Woolsey take on that (which, BTW doesn’t relay AT ALL on my B+F story there! That’s my own add on to it). Search Rowe’s site for MOA vs MOE, Reflux, Hot Potato, irredeemable, and Chuck Norris. All good articles. Chuck Norris is especially important. Here:

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/10/monetary-policy-as-a-threat-strategy.html

    I’m sure none of that is interesting to Scott and he wishes you (and I) would just forget about banks altogether. I just provided the above to help you out if you “insist on considering banks.” 😀

    Hopefully I didn’t butcher the MM position too badly or lead you too far astray, but in all likelihood I did. Good luck!

  197. Gravatar of Fed Up Fed Up
    8. September 2013 at 22:13

    DOB said: “Fed Up,

    Depends what you mean by “matter”, but loosely speaking, no, they do not matter any more than car manufacturers as long as the central bank retains its control over the rate of return on the unit of account.

    DOB
    6. September 2013 at 11:28

    Tom et al.,

    I generally don’t use the expression medium of account.”

    DOB, I need a short explanation of what the central bank retains its control over the rate of return on the unit of account is.

    But first about the medium of account (MOA) vs. unit of account (UOA).

    See here:

    http://jpkoning.blogspot.ca/2012/11/discussions-of-medium-of-account-could.html

    Last comment:

    “A medium of account is the commodity defining the unit of account. A unit of account is a specific amount of the medium of account. For example, for the gold standard the medium of account is gold, while the unit of account might be one ounce or one pound of gold of specific purity.”

    And, http://worthwhile.typepad.com/worthwhile_canadian_initi/2012/10/medium-of-account-vs-medium-of-exchange.html

    “[Update: just to clarify terminology: in my model, gold is the medium of account; and (say) an ounce of gold is the unit of account.]”

    Let’s assume the MOA and MOE (medium of exchange) are the same to keep things simple for now. I believe UOA equals $1 of currency and $1 of demand deposits. MOA/MOE is the total amount of currency and demand deposits. Prices are determined by the amount of MOA/MOE circulating.

    ssumner is not picking up on the idea that since currency and demand deposits are 1 to 1 convertible (relative pricing) then demand deposits are part of MOA.

  198. Gravatar of Fed Up Fed Up
    8. September 2013 at 22:22

    “7. Now let’s assume a cashless economy where the MOA is 100% reserves”

    Let’s go to Canada where the reserve requirement is zero percent. The central bank of Canada makes no new announcements. The people decide to stop holding currency. They turn it in to the commercial banks. The commercial banks turn the currency in for central bank reserves. The commercial banks turn in all the central bank reserves for gov’t bonds. The balance sheet of the central bank of Canada now has zero assets and zero liabilities. I’m going to say not much changes because the MOA is 100% demand deposits. Demand deposits have been part of the MOA.

  199. Gravatar of ssumner ssumner
    9. September 2013 at 05:57

    Gasman, You said;

    “If Japan pegged to the dollar at 200 yen per dollar that very well could cause inflation but why would Japan do that?”

    First you tell me that it doesn’t matter, it’s just a yardstick or a ruler. Now you tell me that changing the yardstick causes inflation. Which is it?

    Odie, You said;

    “I don’t deny that QE generated excess reserves but its main effect was to lower long-term interest rates. The Fed seems to agree with me:
    http://www.stlouisfed.org/publications/re/articles/?id=2258
    I also don’t see how the HPE gets banks to lend those excess reserves out.”

    I don’t claim that the HPE gets banks to lend out ERs. I said it causes them to adjust their behavior in such a way that some of the reserves go out in circulation as currency. And I discussed the zero bound problem in my post, if the QE injections are temporary then there is little or no inflationary effect. They were mostly temporary but a tiny bit permanent, which is why QE increased inflation expectations by a very small amount. Nothing that has happened recently in any way conflicts with anything I said in this post. If you think it does provide a specific quote that you think is wrong.

    As far as credit, until 2008 the Fed never paid any interest on base money. They still don’t on cash. When gold was the MOA was gold “credit?” Cash has essentially replaced gold as the MOA. Why is cash “credit?” If I take my $100 bill back to the Fed to be redeemed, they’ll tell me “get lost!”

    You said;

    “How do you think the banks will get rid off excess reserves? By spending? On what? If they buy a bond that is nothing else than lending out the money which implies there must be a willing borrower.”

    The Treasury bonds were issued long ago. The government does not have to be a “willing borrower” today, banks can simply go out and buy bonds in the secondary market. What is the problem?

    Odie; You said;

    “TIPS spread widened because people (wrongly) though QE would mean higher inflation. As the CPI shows that has not been the case; neither could the Fed confirm heightened inflation expectations (see link above). What assets do you actually mean, financial assets or real goods and services?”

    There have been studies showing that asset prices responded positively to information relating to QE. Indeed the current talk of tapering QE is clearly affecting asset prices, I don’t think that is even a topic for dispute. These includes stocks, bonds TIPS spreads, foreign exchange, commodities, etc. The actual movements in the CPI are completely irrelevant, as you surely must know. Markets respond to new information.

    When the markets tell you that your model is wrong, it’s better to re-evaluate your model rather than assume the markets are stupid. If the markets were really that stupid it should be easy to get rich.

    Fed up, You said;

    “ssumner is not picking up on the idea that since currency and demand deposits are 1 to 1 convertible (relative pricing) then demand deposits are part of MOA.”

    Even if true, that doesn’t change my analysis at all. The Fed directly controls the supply of base money. Base money and DDs are not perfect substitutes. That gives the Fed control over NGDP via the base.

  200. Gravatar of Odie Odie
    9. September 2013 at 10:33

    Hello Tom Brown, thanks for trying to help me. I read most of the links you posted and also left a comment on your website. One misconception I want to point out:

    You said:”HOWEVER that doesn’t mean that the banks in aggregate still can’t buy stuff from non-banks: Tsy debt, financial assets, salaries, electric bill, dividend payments, loans”.

    Banks can buy with excess reserves: Tsy debt, financial assets, loans
    Banks CANNOT spend with ER on: salaries, electric bill, dividends

    Reserve deposits are the collateral for their customers’ deposits. It it not the banks’ money. For their businesses expenses they have to use their earnings which are made up from interest and fees. See here under “How banks make a profit”: http://www.federalreserveeducation.org/about-the-fed/structure-and-functions/banking-supervision/

    The only thing banks can do with excess reserves is purchasing other financial assets which they like to be interest bearing. (see also my earnings equation further above).

    Nevertheless, I am starting to understand what Scott wants to imply; I am just not sure if we have the same idea about OMOs and how the supply and demand there impacts interest rates.

    Thanks!

  201. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 10:47

    Odie,

    I think you missed my point. Again lets assume Scott’s Case 7 (cashless). And let’s add that the CB and all other non-bank Fed deposit holders (e.g. Tsy, foreign CBs, etc) are not buying or selling anything. Let me clarify some things:

    1. In this case (Case 7, w/ my added qualifications), the reserves (Fed deposits at the banks) are STUCK inside the banking system. They cannot leave. Look at my blog for the post entitled “The Three Places Reserves Can Go”

    2. INDIVIDUAL banks can trade reserves with other banks, but the overall stock of reserves does not change. Look at my Example #5 on my blog.

    3. And yes, that means that an INDIVIDUAL bank CAN buy (using its reserves) electricity, employee time, or donuts etc from the prvt. non-bank sector… by writing a check for those items from the prvt non-banks. Of course this is how they pay dividends to their shareholders as well.

    3. However, IN AGGREGATE, the banks DO NOT use reserves to buy items from the pvt. non-banks: Instead they credit bank deposits to do this. You can see this by aggregating the balance sheets of banks A & B in my Example #5. The aggregated bank (A&B) paid both person x and person y in the same manner: buy crediting their bank deposits.

    Make sense? (I’d provide links, but I’m afraid that would put my post “in moderation”)… I’ll provide the links in a follow up comment.

  202. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 10:51

    Odie, here are the links I referred to above:

    http://brown-blog-5.blogspot.com/2013/04/the-three-places-reserves-can-go.html

    http://brown-blog-5.blogspot.com/2013/03/banking-example-5-bank-spends-excess.html

    Hope that helps!

  203. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 11:24

    Odie, let me add a qualification to my item 3. above:

    Banks do have capital constraints, both regulatory and solvency related. Of course bank spending on non-financial assets is limited: they can’t just buy the whole world by writing a check: they have to consider their capital adequacy ratio (CAR) for one. Every dollar they spend on donuts is a dollar less equity for their shareholders! So in my world above, assume that there are no reserve requirements, but there are implicit capital constraints and legal requirements. When an individual bank trades excess reserves for Tsy debt, it’s swapping one risk free asset for another which does not affect it’s CAR. Look at my Example #3 and Example #7 (I added a search feature in the top right), but again, I’ll post direct links in a separate post.

  204. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 11:27

    Odie, here again are the the links I promised:

    http://brown-blog-5.blogspot.com/2013/03/banking-example-3-capital-requirements.html

    http://brown-blog-5.blogspot.com/2013/03/banking-example-7-calculating-capital.html

  205. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 12:40

    “Odie, let me add a qualification to my item 3. above:”

    That would be my 1st item 3! 😀

  206. Gravatar of Fed Up Fed Up
    9. September 2013 at 12:51

    ssumner said: “If I take my $100 bill back to the Fed to be redeemed, they’ll tell me “get lost!””

    And, “The Fed directly controls the supply of base money.”

    That is a yes/no answer. If you take $100 in currency directly to the fed, it will probably say ‘get lost’. If you take $100 in currency to a commercial bank, it will take the $100 in currency. You will get $100 in demand deposits. The commercial bank will probably send the $100 in currency back to the fed where it will be an asset on the fed’s balance sheet with zero velocity in the real economy.

    And, “Even if true, that doesn’t change my analysis at all. The Fed directly controls the supply of base money. Base money and DDs are not perfect substitutes. That gives the Fed control over NGDP via the base.”

    I’m going to need some expansion on the “Base money and DDs are not perfect substitutes.” part. Start here: $800 billion in currency, $200 billion in central bank reserves, and $6.2 trillion in demand deposits with a zero fed funds rate.

  207. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 13:13

    Fed Up,

    “The commercial bank will probably send the $100 in currency back to the fed where it will be an asset on the fed’s balance sheet with zero velocity in the real economy.”

    This isn’t true! Reserve notes at the Fed are notes which ARE NOT liabilities… that’s all! They are not assets to the Fed.

    And you’re forgetting the other side of that trade: if the bank sells a $100 note back to the Fed, the Fed will credit their Fed deposit, or send them another (crisper) $100, or five $20s, etc. The sum of the Fed’s liabilities do not change in any of these cases. Now if we bring in coins they will… but it doesn’t matter too much: it won’t change the Fed’s equity position (w/ coins).

    “I’m going to need some expansion on the …”

    I don’t get your example. But I would suggest to you that you go to Nick Rowe’s site (google Nick Rowe worthwhile) and then enter into his search box on the upper right the word “irredeemability.” You will find a host of good articles that may help you out here. I’ll post a link to the results of this search in a separate comment.

  208. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 13:15

    Fed Up, here you are (I included a link to the 1st article too):

    https://www.google.com/search?ie=UTF-8&oe=UTF-8&q=irredeemable&domains=worthwhile.typepad.com&sitesearch=worthwhile.typepad.com&btnG=+Google+Search+

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/09/interest-and-currency.html

  209. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 13:26

    Fed Up, also you might find the post I did called “Money Labels” interesting… especially the table below the Venn diagram. I left out an obscure form of money there called “US notes” which were/are paper notes issued directly by Tsy and direct liabilities of Tsy (although in a category which does not contribute to the “statutory debt limit.” Coins are face value assets of the Mint (part of Tsy) from the moment they are minted, and are not an official liability to any entity, but practically they are a contingent liability of Tsy in that Tsy will accept them in exchange for other currency, etc. Coins are money the moment they leave the Mint, and thus appear as assets on the Fed balance sheet after the Fed buys them, but before it sells them to the banks. When coins are sold to banks by the Fed they become “base money” and are removed as assets from the Fed’s BS. Reserve notes are different: the Fed buys those from the BEP (part of Tsy) for production cost. While at the Fed they are not assets or liabilities of the Fed, but once the Fed sells them to banks they become Fed liabilities… until they return to the Fed.

  210. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 14:06

    Fed Up, also there are a host of different views about “The Law of Reflux” (LoR). Rowe is what I’ll call a “zero refluxer” in that he thinks the LoR does not apply to MOE or base money (what Scott calls “MOA”… and what I’ll continue to call MOA in this comment). Woolsey agrees with Rowe. Glasner (and I suspect Scott too) are what I’ll call “partial refluxers.” They think, al la Tobin, that the LoR DOES apply to MOE but not MOA. I think that’s correct anyway, I could be wrong! Mike Sproul is a “full refluxer” thinking the LoR applies to both:

    http://mpra.ub.uni-muenchen.de/24813/1/MPRA_paper_24813.pdf

    I’ve seen others suggest/imply that it’s not that simple (as these three positions) and that varying degrees of the LoR apply in different situations.

    Rowe & Glasner agree on the “irredeemablity” issue.

    David Glasner has some interesting articles about this with debates in the comments section. One of them is called “Tobin vs Yeager” and has comments from Sproul, Koning, Sumner, Rowe, Glasner, and Woolsey (in addition to some other good comments from Ritwik and others).

  211. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 14:34

    Fed Up,

    Shoot! … including a JP Koning link means the message goes to the spam filter (sometimes).

    Mike Sproul just left me a response on this very subject at JP Koning’s: his article on “the convenience yield.” It’s towards the bottom: Mike is responding to me asking him about the following Nick Rowe quote:

    “This very old debate over the Law of Reflux is what is at the root of the very modern debate about whether Quantitative Easing can work. Those who argued for the Law of Reflux argued that an excess supply of money could not cause inflation [update: because any excess supply of money would immediately flow back to the issuer]. Banks could only cause inflation by lowering the price of money in terms of gold. Modern proponents of the Law of Reflux argue that banks can only cause inflation by lowering the rate of interest.”

    He dissects it line by line (he’s not in agreement w/ Nick).

  212. Gravatar of gasman gasman
    9. September 2013 at 14:54

    Scott

    “First you tell me that it doesn’t matter, it’s just a yardstick or a ruler. Now you tell me that changing the yardstick causes inflation. Which is it?”

    Remember Scott, the first point I was making was that what Japan uses as its UOA/yardstick…. RELATIVE TO US…. does not matter. In the scenario you describe with Japan pegging their currency to the dollar, the Japanese holders of Yen will experience a rise in prices if the UOA is essentially halved…… but to an American we would still get the same relative value. Our dollars would buy twice as many Yen as before so if prices in Yen doubled we would lose nothing. The Japanese standard of living would be determined by what their incomes did in the face of the new peg. Buying power and standard of living could be negatively affected in the scenario you describe. Yes Scott, monetary authorities screwing around with the UOA can change prices and standards of living (for those having to use that UOA), but that is a bad way to do things in my opinion. Better to alter the MOE levels and let the buyers make their choices imho. You know if you owed me 5$ I could say next week that you owed me 7$ and if I was enough of a bully I might be able to pull it off….. so what.

    The initial statement of yours that I had issue with was that “Japans price level is 100x ours”. That is clearly not true. If I told you that Costcos price level was 100x WalMarts you would expect to pay 100x more at Costco. . Maybe no for every single item but for the average of items and certainly many items. Thats what that statement would mean. We would pay 100x more for almost nothing in Japan than we pay here and certainly an indexed basket of goods would be much much less than 100x different.

    So once more, in what sense is Japans price level 100x ours, in your view? In a real sense or an altered UOA sense, which is of little consequence to us?

  213. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 17:30

    Odie, check out JP Koning’s latest:
    “The rise and fall (and rise) of the hot potato effect”

  214. Gravatar of Odie Odie
    9. September 2013 at 18:10

    Scott, you said:
    “The Treasury bonds were issued long ago. The government does not have to be a “willing borrower” today, banks can simply go out and buy bonds in the secondary market. What is the problem?”

    First, to make sure we agree on:
    1. Excess reserves will only be used to buy financial assets (e. g. bonds, treasuries, other securities).
    2. Purchase/sale of financial assets between banks’ reserve accounts and the Fed has no effect on prices.
    3. The government does not issue new debt; it just rolls over its old one.

    Thus, the excess reserves must be used to buy assets from non-bank financial institutions or the public. However, those must be willing sellers. If they have no use for deposits because they don’t want to buy anything else but rather keep their interest-bearing bond, the interest rate of those assets would plummet to zero. See current short-term T-bills. The demand for those is so high despite the fact that banks don’t invest their excess reserves anymore that the Fed could not keep their rate target through OMOs. That is the reason they implemented the IOR. Once the rate is at 0% banks will just keep the excess reserves. If the Fed wants to keep the target it has to supply enough treasuries to stabilize the yield at 1%. They could not do that at current rates and had to resort to paying IOR on the excess reserves. Why should it be able to keep the higher target then in your case 5c?

    As an aside, you said:”As far as credit, until 2008 the Fed never paid any interest on base money. They still don’t on cash. When gold was the MOA was gold “credit?” Cash has essentially replaced gold as the MOA. Why is cash “credit?” If I take my $100 bill back to the Fed to be redeemed, they’ll tell me “get lost!”

    Tom Brown answered a lot of this already and I am pretty sure you know that but the difference with gold is we live now in a fiat currency not a gold-standard anymore. Money is created by someone taking on debt (loans, bonds, treasuries) and expansion of banks’ balance sheets. Each monetary asset has an equal monetary liability on someone else’s balance sheet. Cash is a liability of the Fed (it even says it on the front) which is only redeemable in itself. You may get 2 $50 or 10 $10 for your $100 bill. However, if you are to donate those $100 to the Fed they can destroy it and reduce their outstanding cash liability by $100. Since the $100 matter to you while the Fed does not care about the size of its balance sheet that remains a rather hypothetical possibility.

  215. Gravatar of Odie Odie
    9. September 2013 at 18:20

    Tom Brown,

    Thanks for the links; I will check them out. Regarding “spending” the excess reserves, I think we can agree on the following:
    Spending of banks happens from the owner’s capital. That is the same regardless how much reserves a bank has. Once that capital is exhausted the FDIC will come in and close shop due to illiquidity. A change in the base money supply does not affect a bank’s ability to pay its bills. Hence, the reserve account and the bank’s expense account(s) will be two different entries in its books. Ok?

  216. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 18:33

    Odie, sounds reasonable. But I’d add that *perhaps* prior to the FDID/Fed/regulators getting involved the banks’ shareholders might get upset… unless you get a Romer-Akerlof bankruptcy for profit situation going:
    http://www.jstor.org/discover/10.2307/2534564?uid=3739560&uid=2129&uid=2&uid=70&uid=4&uid=3739256&sid=21102633875603

  217. Gravatar of Fed Up Fed Up
    9. September 2013 at 19:41

    “This isn’t true! Reserve notes at the Fed are notes which ARE NOT liabilities… that’s all! They are not assets to the Fed.”

    I’ll have to check, but I believe federal reserve notes(currency) are physical items actually on the balance sheet. They will be both assets and liabilities. For example, Start here: $800 billion in currency, $200 billion in central bank reserves, and $6.2 trillion in demand deposits with a zero fed funds rate.

    Simple version. Now everyone turns in their currency for demand deposits. The banks turn in the currency to the fed. The fed now has $800 billion in currency as both liabilities and assets.

    It may be possible that federal reserve notes are still liabilities of the treasury, which complicates the scenario, but the point still stands.

  218. Gravatar of Fed Up Fed Up
    9. September 2013 at 19:50

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/09/interest-and-currency.html

    “The Bank of Montreal can refuse my request to open an account to convert my $100 Bank of Canada liability into a $100 Bank of Montreal liability. It is not obliged to redeem my money.”

    1) I’m not sure that is right.

    2) Even if 1) is right, out in the real world, a commercial bank does convert currency to demand deposits 1 to 1.

  219. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 20:10

    Fed Up,

    “One of the interesting things about the Fed’s liabilities is that some of your assets, like the green dollar bills in your pocket, get reflected as the liabilities of the Fed. Apart from this, the money lying in the reserve account of member banks and U.S. depository institutions also forms a part of the Feds’ liabilities. As long as the dollar bills are lying with the Fed, they would be treated neither as assets nor as liabilities of the Fed.”

    http://www.investopedia.com/articles/economics/10/understanding-the-fed-balance-sheet.asp

    You can check this too:

    http://www.federalreserve.gov/releases/h41/current/h41.htm

    Search for “Federal Reserve note”

  220. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 20:28

    Fed Up,

    “The fed now has $800 billion in currency as both liabilities and assets.”

    you are free to think of it like that… it does not harm, but that is an unconventional way to do it. The Fed doesn’t do it like that. Nobody does (as far as I know). You and I have been through this and you asked an accountant, remember? Joe?

    See the thing is if you think about it your way, there’s no difference between that and saying the Fed owns an arbitrary number of dollars to itself. Could be $1. Could be $0. Could be $1 googolplex. It doesn’t matter, because it owes it to itself! I could say the same for myself: I owe myself a billion dollars. I’m neither richer or poorer. You would even “owe” yourself negative amounts of money this way.

  221. Gravatar of Tom Brown Tom Brown
    9. September 2013 at 21:24

    Fed Up,

    “It may be possible that federal reserve notes are still liabilities of the treasury, which complicates the scenario, but the point still stands.”

    Federal Reserve notes are never liabilities of the Tsy. US notes are!… but they haven’t made those since 1971:

    http://en.wikipedia.org/wiki/United_States_Note

    Here’s an example of a Tsy debt outstanding statement:

    http://www.treasurydirect.gov/govt/reports/pd/mspd/2012/opdm122012.pdf

    Search for “United States Notes.”

  222. Gravatar of ssumner ssumner
    10. September 2013 at 09:03

    gasman, A price level is a nominal variable, not a real variable. It refers to the numbers you see on price tags.

    odie, You said;

    “Cash is a liability of the Fed (it even says it on the front) which is only redeemable in itself.”

    Interesting definition of liability. I hope it works when I tell my bank I plan to repay my mortgage by giving them another mortgage, and an interest free one to boot!

    I’m afraid I don’t follow the rest of you comment, or at least I don’t see how it relates to my post. The Fed never, ever has any problem finding assets to buy at market prices. So I don’t see what the issue is there. If you are focused on zero rates, then talk about zero rates. There’s no need to bring banks into the picture at all.

  223. Gravatar of gasman gasman
    10. September 2013 at 09:17

    Scott, you seem to be talking around the question I have or maybe Im not expressing myself clearly.

    When you say “Japans price level is 100x ours” it is necessary to convert both prices to the same UOA in order to make the comparison correctly. When you do that you see that Japans price level is nowhere near 100x ours its simply expressed in different units that have more zeros. So, what do you mean when you say “Japans price level is 100x ours” ?

  224. Gravatar of Tom Brown Tom Brown
    10. September 2013 at 10:11

    Scott, you write:

    “Interesting definition of liability. I hope it works when I tell my bank I plan to repay my mortgage by giving them another mortgage, and an interest free one to boot!”

    But it IS a liability of the Fed. People who hold reserve notes are creditors to the Fed. I know Rowe doesn’t like that description, but he’s in a minority there. The Fed treats them as liabilities … it says so right on their balance sheet! Every IOU makes one party a debtor and the other a creditor and reserve notes are no different. So yes in fact you could pay off your mortgage by trading your bank another mortgage for which you were the creditor. They wouldn’t be much interested in it if you were the debtor though!

  225. Gravatar of Tom Brown Tom Brown
    10. September 2013 at 10:52

    I wrote:

    “So yes in fact you could pay off your mortgage by trading your bank another mortgage for which you were the creditor. They wouldn’t be much interested in it if you were the debtor though!”

    And if it was a zero rate mortgage, they might demand some other benefit, like a premium or a convenience yield (non-pecuniary return). I think Mike Sproul and JP Koning are on board with that one,… along w/ the Fed itself (balance sheet link above).

    For anyone that’s interested, go to JP’s “The convenience yield as epicentre of monetary policy implementation” article, and you’ll see JP, Sproul AND Rowe all talking about the Fed’s liabilities in such a manner.

    Now you could argue that the accounting is kind of arbitrary: you could instead count reserve notes as assets of the Fed (while at the Fed) exactly as coins are. But perhaps the reason this is not done is because it would give the Fed a huge seigniorage beyond what it has now. Would they have to remit that to Tsy? I don’t see why not. Great way around the debt limit! If it was the Tsy’s seigniorage directly (like for coins) it would also be huge… something like the $1T coin concept (I think). As it stands coins provide limited seigniorage to Tsy… they actually lose money on pennies and nickles. And though not official liabilities of Tsy, coins are in fact practical ones, as they agree to accept them back at face value (e.g. when they’re worn out), just as the Fed does for reserve notes. They don’t HAVE to do that, but they do. Sproul argues that if all channels of reflux are permanently cut off, money loses it’s value (link above). I favor Sproul and Koning on that over Nick.

    Here’s a Fed paper describing all this in the context of defining inside and outside money (for anyone that’s interested):
    http://www.minneapolisfed.org/research/sr/sr374.pdf

  226. Gravatar of Odie Odie
    10. September 2013 at 10:58

    Scott, you say:”Interesting definition of liability. I hope it works when I tell my bank I plan to repay my mortgage by giving them another mortgage, and an interest free one to boot!”

    I actually just did that last year; it is called mortgage refinancing. Doesn’t mean that I payed it off. I just replaced one liability with another like the Fed would replace one liability with another when it would convert my $100 bill. From the Fed itself:”The major items on the liability side of the Federal Reserve balance sheet are Federal Reserve notes (U.S. paper currency) and…”
    http://www.federalreserve.gov/monetarypolicy/bst_frliabilities.htm
    Your answer goes back to the difference between buying/spending and borrowing/lending I mentioned earlier.

    You say:”The Fed never, ever has any problem finding assets to buy at market prices. So I don’t see what the issue is there. If you are focused on zero rates, then talk about zero rates.”

    Ok, let’s say the Fed wants to buy assets (bonds etc.). It means someone must hold them already and willing to sell them at the agreed price which is inverse correlated to the interest rate (lower interest rate means higher price). It also means at 0% interest rate there is a max price for an asset. Those sellers must not be a bank as the 3 conditions I mentioned. So, it would e. g. be a retirement fund holding securities. Why did they purchase those securities in the first place? Because they had money/deposits that people did not want to spend and therefore invested in interest-bearing assets. Now you want to give them a deposit for that asset. Does that change anything about the desire of those people to hold onto those deposits instead of spending it?
    People are not spending right now, they are saving and therefore interest rates are next to zero. In essence, the public took the hot potato and threw it to the Fed and they cannot get rid off it anymore.

  227. Gravatar of Tom Brown Tom Brown
    10. September 2013 at 11:14

    Odie, I wrote:

    “They wouldn’t be much interested in it if you were the debtor though!”

    imagining that I might own a mortgage as an asset like a bank typically does (which is possible, but not likely), but your point about re-finance is a good one: The effect for the bank is the same though: it effectively allows your old mortgage to be replaced on the bank’s BS by a new one that you provide (though still as the debtor).

  228. Gravatar of Odie Odie
    10. September 2013 at 11:40

    Tom,

    Yes, it is an asset swap like the Fed would swap your assets with the dollar bills. The difference lies in the interest. Banks should not like mortgage refinancing though because they change a high-interest asset for a low-interest asset. That is where Freddie Mac and Co. are coming in by taking those mortgages out of banks’ hands and taking away the interest risk which is then carried de facto by the government then.

  229. Gravatar of Tom Brown Tom Brown
    10. September 2013 at 12:28

    Odie, still I don’t think it’s fruitful to argue w/ Scott about banks. Read JP Koning’s “Scott Sumner ignores banks, so what?” article to see why I think that.

  230. Gravatar of Fed Up Fed Up
    10. September 2013 at 14:48

    “You and I have been through this and you asked an accountant, remember? Joe?”

    That was for demand deposits. The argument then was that the demand deposits get destroyed. It is actually hard to know if they are destroyed or not.

    This would be about currency/federal reserve notes. I don’t see them being destroyed. If the fed wants to burn them, then the scenario changes.

  231. Gravatar of DOB DOB
    10. September 2013 at 15:06

    DOB,

    “I’m afraid I still don’t see your point, and the Treasury analogy does not help. The Treasury derivatives you cite are not Treasury bonds, they are Treasury bond derivatives.”

    I made no mention of derivatives there. What makes you think that I did?

    Do you agree that in the case of a simultaneous settlement, 2 or more parties can exchange goods and assets without needing to actually hold any of the medium of exchange?

    For instance, Monday I buy a $100 worth of stocks (those settle T+3)
    Wednesday I sell $120 worth of T-bonds (those settle T+1)
    Thursday I buy $20 worth of apples (with the right electronic payment systems, that could settle T+0)
    If all 3 of these exchanges settle simultaneously on the same clearinghouse, then there is no need for me to hold any of the medium of exchange. The clearinghouse will compute $100-$120+$20 = $0 and clear that on Thursday.

    You can think of credit cards as another device to help carry out transactions without affecting base. One day a month you liquidate just the right amount of stocks to pay down your credit card and the merchant buys just the right amount of stocks when the credit card companies pay him. Once again, nobody needs to hold any of the medium of exchange there.

    Only in in the “limiting cases” does an entire economy manage to get by without holding any of the MOE, but those are just examples that illustrate that the volume of MOE in circulation isn’t all that relevant.

    “I also don’t follow the representative agent comment. Are you saying the public chooses to hold zero balances of money,”

    Yes! (if there are no ‘monetary frictions’)

    “or that they are simply unable to hold money even if they want to, because money does not exist?”

    No they chose not to hold any because they have electronic tools to make all the payments they want without having to hold any of the MOE.

    “What if they wanted to hold some central bank liabilities, even for irrational reasons?”

    Oh yes, they absolutely could. And until we have really fancy electronic payment and clearing systems, there are plenty of rational reasons to hold money.

    “And why does your model contain money? Why not a moneyless economy, if you think it works?”

    Nobody’s saying money need not exist, just that people could chose to hold little or none of it at any given time. Those are two different things. All this is suggesting is that we can do all the analysis in interest rate space rather than quantity space because interest rates are unaffected by any of this while quantity is.

  232. Gravatar of DOB DOB
    10. September 2013 at 15:07

    Scott,

    DOB should of course have been “Scott”. Copy pastes..

  233. Gravatar of Tom Brown Tom Brown
    10. September 2013 at 15:12

    Fed Up,

    “This would be about currency/federal reserve notes. I don’t see them being destroyed. If the fed wants to burn them, then the scenario changes.”

    Again, the way the accounting is done (check it out for yourself) paper reserve notes are a liability of the Fed once the Fed has sold them to banks. When they are held at the Fed itself (or the Bureau of Engraving or Printing prior to that) they DO NOT have their face value and don’t appear on the Fed’s balance sheets or anyone elses BS at all (face value or otherwise… I could be wrong about that regarding their production costs, but I don’t think I am). Henceforth I’ll just refer to the paper notes’ face values and ignore their production costs: paper $ at the BEP = paper $ at the Fed = nothing! No one’s asset. No one’s liability. It’s exactly the same as if you wrote out an IOU to your neighbor for a cup of sugar (“IOU: to the bearer of this IOU three eggs” … say you don’t have the eggs yet, but you’ve got a chicken). Before you actually hand that to your neighbor it has no value. When your neighbor or the person he transferred your note to brings it back to you and redeems it, it again has no value. In between times (when someone else is holding it), it’s a liability to you (you have become a debtor to the holder) and an asset to the holder. When it’s in your possession you are free to keep it for another time you need a cup of sugar, or burn it. It hardly matters… since the intrinsic value of the IOU itself (paper and ink) is negligible.

    So the Fed doesn’t necessarily “burn” or otherwise physically destroy reserve notes in its possession, but it does retire them when they’re old and worn out, and it does this for negligible cost since they don’t appear on the Fed’s balance sheet while in the possession of the Fed.

    The same is true for coins, but now coins are “outside money” to the Fed (see my link explaining this in the Fed document above). “Inside” and “outside” are relative terms. Coinage is an asset for the Fed while at the Fed and an asset for EVERYONE who holds it (it turns out)… but coins are a contingent (not official) liability of the Tsy, since the Tsy agrees to accept worn out coins and replace them with their face value… most likely by writing a check for them.

  234. Gravatar of Tom Brown Tom Brown
    10. September 2013 at 17:26

    Scott, I can’t believe I never noticed your “FAQ” page. Very nice. I wish I’d read that a long time ago. I especially liked this:

    “20. Aren’t you just a monetary crank trying to solve all the world’s problems by printing money?

    Yes, but like a broken clock the monetary cranks are right twice a century; 1933, and today. The other 98 years I am a Chicago-trained, libertarian, inflation-hawk. Twice a century I put on my Irving Fisher super-hero suit, and emerge from my deep underground bunker.”

  235. Gravatar of ssumner ssumner
    10. September 2013 at 18:01

    DOB, You said:

    “For instance, treasury bonds trade before the auctions that will set their coupons, in the form of “when-issued” (WI) derivative contracts. The market has a price of them before they are even created.)”

    Then I said:

    “I’m afraid I still don’t see your point, and the Treasury analogy does not help. The Treasury derivatives you cite are not Treasury bonds, they are Treasury bond derivatives.”

    Then you said:

    “I made no mention of derivatives there. What makes you think that I did?”

    Umm, maybe because you used the word “derivative?”

    Again, I’m not denying that you can get by without a MOE, I’m denying you can get by without a MOA.

  236. Gravatar of ssumner ssumner
    10. September 2013 at 18:08

    Odie, You got an interest free mortgage? Good for you.

    The rest of your comments are just warmed over liquidity trap arguments. They have no bearing at all on this post. Did you even read the post? If so, then don’t address me like I’m a moron.

    gasman, Yes, Japanese prices are 100 times higher than US prices, I noticed that right away in the Tokyo airport. That difference can only be explained by looking at the markets for the two MOA. There’s way more yen than dollars floating around. Roughly 100 times as many per capita.

  237. Gravatar of Fed Up Fed Up
    10. September 2013 at 20:11

    Scenario #1:

    The banks turn in $800 billion in currency to the fed. The fed holds the currency. The next day the banks take back the $800 billion in currency.

    Scenario #2:

    The banks turn in $800 billion in currency to the fed. The fed burns the currency. The next day the banks attempt to take back the $800 billion in currency. The banks fail.

    I see scenario #1 as:

    Assets of the fed = $800 billion in currency
    Liabilities of the fed = $800 billion in currency

    I see scenario #2 as:

    Assets of the fed = $800 billion in currency
    Liabilities of the fed = $800 billion in currency

    The burning destroys the assets (and the liabilities).

    Assets of the fed = $0 billion in currency
    Liabilities of the fed = $0 billion in currency

    I won’t guarantee my accounting is correct, but there has to be some difference in those 2 scenarios.

    Plus, one last response than I’m moving on to address ssumner on MOA being circulating currency and demand deposits.

  238. Gravatar of Tom Brown Tom Brown
    10. September 2013 at 21:26

    Fed Up,

    Please look at this:

    http://archive.org/stream/ModernMoneyMechanics/MMM#page/n16/mode/1up/search/notes

    It’s from the Federal Reserve Bank: it tells you exactly how the accounting is done. It confirms everything I’ve been telling you:

    1. Federal Reserve Notes are liabilities of the Fed when they go out into circulation and are removed as liabilities when they return.

    2. It confirms what I wrote earlier about coins too, although instead of “contingent liability” they use the word “obligation.”

    The link above goes straight to instances of the word “note” in the text. The forth book mark from the left (along the bottom) shows accounting tables walking you through the book keeping process. Page 17 in particular. Check it out!

    Regarding your Scenario #1 vs #2, here’s all that happens (aside from correcting your accounting): In scenario #2 the Fed calls up the Bureau of Engraving and Printing (BEP) and tells them they’re going to need some more reserve notes because somebody accidentally burned up their stock of them. The BEP gets on the ball, prints them up, and charges the Fed for their production cost. The Fed electronically credits the balance of the Treasury General Account (TGA) to pay for it (the BEP is part of Tsy). Here’s what the Fed has to pay for each type of note:

    http://www.federalreserve.gov/faqs/currency_12771.htm

    The banks pay face value when they in turn buy it from the Fed. When the notes are worn out they’re returned to the Fed and the Fed in turn returns them to the BEP which does this with them:

    http://www.moneyfactorystore.gov/5lbbagofshreddeduscurrency.aspx

    Hope that helps!

  239. Gravatar of DOB DOB
    10. September 2013 at 21:29

    Scott,

    “Umm, maybe because you used the word “derivative?””

    My bad. I forgot I brought up WIs and thought you were talking about the principal paydown. I brought up WIs not because they had anything to do with treasuries (could have been WIs on pre-IPO’d stock really) but because they are an example of how markets can put a price on something that does not “exist” yet. Likewise, the money can have a well defined price level despite the fact that people chose to hold zero units of it.

    “Again, I’m not denying that you can get by without a MOE, I’m denying you can get by without a MOA.”

    Why does anyone need to have a positive balance of the MOA outstanding? What’s the use in holding the MOA?

    And regardless of the use, why couldn’t they borrow it from the central bank at t-epsilon if it’s needed at time t, and repay it at t+epsilon, and let epsilon go to 0 since we’re assuming no monetary frictions?

  240. Gravatar of Tom Brown Tom Brown
    10. September 2013 at 21:35

    Change my next to last sentence to this:

    “When the notes are worn out they’re sold back to the Fed and the Fed in turn returns them to the BEP which does this with them:”

    I didn’t want to give the impression they were returned for free.

  241. Gravatar of gasman gasman
    11. September 2013 at 01:48

    So Scott, your telling me that in the Tokyo airport, something you would have paid 5$ for here you would end up paying about 500$ for? Like a magazine?
    Can you give me an example of something that cost 100x more there? Lunch 700$, shoe shine 500$, Starbucks cofffee 250$??

  242. Gravatar of Fed Up Fed Up
    11. September 2013 at 15:12

    Tom Brown, IF that accounting is right, then these points should still be correct.

    1) $800 billion in currency went to the fed where is it has a velocity of zero in the real economy for NGDP.
    2) $800 billion in currency is still at the fed. It was not destroyed.
    3) The fed does not refuse currency sent back to it.

    The problem with saying the $800 in billion in currency is not liabilities is that it gives the impression the currency was destroyed. It was not.

    Moving on. ssumner said: “I don’t claim that the HPE gets banks to lend out ERs. I said it causes them to adjust their behavior in such a way that some of the reserves go out in circulation as currency.”

    I’m going to say that does not matter.

    Back to here: $800 billion in currency, $200 billion in central bank reserves, and $6.2 trillion in demand deposits with a zero fed funds rate. Assume the $800 billion in currency and $6.2 trillion in demand deposits are circulating and velocity is 2. MV = PY. The M = MOA = MOE = circulating currency plus circulating demand deposits.

    $7.0 trillion times 2 = $14 trillion.

    Now assume something (whatever) happens so that entities do not want any demand deposits anymore. There are now $7.0 trillion in currency and $0 in demand deposits. MV = PY is still:

    $7.0 trillion times 2 = $14 trillion.

    Now on to currency and demand deposits being equivalent for NGDP.

    I bank at bank A. A grocery store banks at bank B. I buy $100 of groceries using a check. Part of bank A balance sheet:

    Assets bank A: $100 in central bank reserves plus $100 in vault cash
    Liabilities bank A: $100 in demand deposits (my asset in a checking account)

    Note there are other things on the balance sheet, and I’m not assuming a 100% reserve requirement. The check starts the transfer. Mark down bank A’s liabilities by $100, mark down my checking account by $100, and mark down bank A’s account at the fed by $100 in central bank reserves.

    Assets bank A: $100 in vault cash

    Now do the “opposite” at bank B. Mark up bank B’s liabilities by $100, mark up the store’s checking account by $100, and mark up bank B’s account at the fed by $100 in central bank reserves.

    Assets bank B: $100 in central bank reserves
    Liabilities bank B: $100 in demand deposits (store asset in a checking account)

    Bank A can now “borrow” the $100 in central bank reserves from bank B thru the fed funds market.

    Change the scenario slightly. I bank at bank A. A grocery store banks at bank B. Part of bank A balance sheet:

    Assets bank A: $100 in central bank reserves plus $100 in vault cash
    Liabilities bank A: $100 in demand deposits (my asset in a checking account)

    I take out $100 in currency. Mark down bank A’s liabilities by $100, mark down my checking account by $100, and mark down bank A’s assets $100 in vault cash. I buy $100 of groceries using currency.

    Assets bank A: $100 in central bank reserves

    The grocery store deposits the $100 in currency in bank B.

    Assets bank B: $100 in vault cash
    Liabilities bank B: $100 in demand deposits (store asset in a checking account)

    Bank B swaps $100 in vault cash for $100 in central bank reserves at the fed.

    Assets bank B: $100 in central bank reserves
    Liabilities bank B: $100 in demand deposits (store asset in a checking account)

    Bank A can now “borrow” the $100 in central bank reserves from bank B thru the fed funds market. Plus, bank A can swap $100 in central bank reserves for $100 in currency.

    The point is in both cases (one bought with demand deposits and the other bought with currency) the balance sheet of the banks end up the same.

    Assets bank A: $100 in central bank reserves (“borrowed”) plus $100 in vault cash

    Assets bank B: $100 in central bank reserves (“lent”)
    Liabilities bank B: $100 in demand deposits (store asset in a checking account)

  243. Gravatar of Tom Brown Tom Brown
    11. September 2013 at 16:06

    Fed Up,

    “The problem with saying the $800 in billion in currency is not liabilities is that it gives the impression the currency was destroyed. It was not.”

    It’s exactly like I wrote earlier about the IOU: You (you’re the Fed) borrow sugar from Bob and give him an IOU for eggs. You go home and mark in your ledger under “Liabilities” the following: “wrote IOU for 3 eggs” (you also write “1 cup sugar” on the assets side, but that’s beside the point)… you make sure you save those eggs from your chicken because you’ll never know who’s going to come back to you with that IOU. Turns out Bob used your IOU to buy flour from Sue. Now Sue knocks on your door wanting to redeem your IOU. You take the IOU, hand Sue the eggs, and toss the IOU in the fire (or put it in your drawer for next time), and then erase the “wrote IOU for 3 eggs” from your ledger. Deal done! Do you see? The money (IOU) doesn’t exist forever. If you need sugar from Bob or flour from Sue again, and you don’t have any eggs, you just write out another one. As soon as you hand it to somebody it becomes money. As soon as they hand it back it stops being money. Currency is created (sent from Fed to banks) and destroyed (sent from Banks to Fed) all the time! I’ll read the rest of your post later.

  244. Gravatar of Tom Brown Tom Brown
    11. September 2013 at 16:19

    Fed UP,

    In your specific examples, you could say that base money was not destroyed. That doesn’t have to be the case though!… The Fed creates and destroys the total stock of base money all the time. The part your forgetting in your example is this:

    scenario #1 (PRIOR to bank selling currency to Fed):

    Assets of the fed = $X
    Liabilities of the fed = $800 billion reserve notes

    scenario #1 (After bank selling currency to Fed):

    Assets of the fed = $X
    Liabilities of the fed = $800 billion Fed deposit
    (off-balance sheet): $800 billion in reserve notes: replacement cost to the Fed: $100 (just making that figure up!)

    scenario #1 (After Fed selling currency to bank):

    Assets of the fed = $X
    Liabilities of the fed = $800 reserve notes

    In your specific example no base money was lost or destroyed (unless you’re assuming the bank traded bonds for the money or took a loan from the Fed… in which case it would be created and destroyed like in my eggs example)

    So here’s what your scenario #2 looks like after the notes are accidentally burned:

    scenario #2 (After Fed selling currency to bank):

    Assets of the fed = $X-$100 (incl. replacement cost of notes)
    Liabilities of the fed = $800 reserve notes

  245. Gravatar of Tom Brown Tom Brown
    11. September 2013 at 16:21

    The last two “800” should be “800 billion” above.

  246. Gravatar of Tom Brown Tom Brown
    11. September 2013 at 18:29

    Fed Up, regarding the rest of your email:

    “Moving on. still:
    ….
    $7.0 trillion times 2 = $14 trillion.”

    First off, why don’t you add in the $200 B in Fed deposits to your $7T? Scott would say that you hand $1T = $800M + $200M of “base money.” OK, but whatever. You do realize that the exchange equation is normally used to calculate V, right?:

    V = NGDP/M

    Pick an M of your choosing (M0, MB, M1, etc) and calculate an associated V. Then MV = PY is true by definition.

    re: the rest of your email (groceries). OK, that looks fine. I haven’t been following Scott’s discussion with Odie or with you.

    Is there a question here for me? Do you want to know if I agree w/ Scott’s quote?

    You say it doesn’t matter (what Scott wrote)… OK, Scott might agree with you! That’s his Case 7. HPE CAN’T cause reserves to go out in circulation as currency if there’s no cash, yet HPE still applies. So I don’t think he’d dispute that.

    If there is cash (case 6), he’s probably correct that some reserves will go out as cash, but it doesn’t sound like you’re disputing that.

    I guess I don’t understand what you’re getting at.

  247. Gravatar of Tom Brown Tom Brown
    11. September 2013 at 19:27

    Fed Up,

    I haven’t done this, but I’ve heard good things. You should check it out:

    https://class.coursera.org/money-001/lecture/index

  248. Gravatar of Odie Odie
    12. September 2013 at 04:15

    Scott,

    Sorry if my answer offended you; my frustration in not being able to get your comments may have shown. I was going to give up our discussion as we did seem to talk a different language. However, another post of you where you mentioned the difference between money and credit gave me an idea what may be the problem. What do you consider the M in MV=PQ to consist of? Do you define that M only as base money (reserves+cash) or M1, M2, M3?

  249. Gravatar of Tom Brown Tom Brown
    12. September 2013 at 07:50

    Odie, I would suggest reading Scott’s FAQ page at the top. Seriously, it clears up quite a bit. I’ll go out on a limb and say for Scott M is what he calls “MOA” (some controversy there amongst his peers), and what he calls MOA is base money, or MB (see Wikipedia). Did I get that right Scott?… but he’d also say you can use any M you want… it just has the effect of changing what V is:

    V = NGDP/M

    Did I get that right?

  250. Gravatar of ssumner ssumner
    12. September 2013 at 17:09

    Gasman, No, they don’t price things in dollars in Japan.

    DOB, I don’t think we are going to be able to resolve this. I simply don’t understand how IOR can be applied to something that doesn’t exist.

    And there is no such thing as prices on stuff that doesn’t exist. The price applies to an option to buy something–that option does exist.

    And if I received reserves in payment for a service I provided, I might re-invest those reserves into something else. But I certainly wouldn’t do so in a nanosecond. I’d want to think about it.

    Odie. The M in MV=PY should be the MOA, which is cash plus reserves in our economy.

    Tom, Yes.

  251. Gravatar of Fed Up Fed Up
    12. September 2013 at 19:44

    “First off, why don’t you add in the $200 B in Fed deposits to your $7T?”

    Because currency and demand deposits are what circulate in the real economy and are what prices are quoted in.

    “You do realize that the exchange equation is normally used to calculate V, right?:”

    I don’t see what is wrong with estimating a realistic V and then finding what M is needed.

    ssumner said: “Odie. The M in MV=PY should be the MOA, which is cash plus reserves in our economy.”

    Odie, I’m going to say the M in MV = PY should be MOA = MOE = currency plus demand deposits and not central bank reserves in our economy.

  252. Gravatar of Fed Up Fed Up
    12. September 2013 at 20:21

    For everyone:

    $800 billion in currency, $200 billion in central bank reserves, and $6.2 trillion in demand deposits with a zero fed funds rate. Assume the $800 billion in currency and $6.2 trillion in demand deposits are circulating and velocity is 2. MV = PY. The M = MOA = MOE = circulating currency plus circulating demand deposits.

    $7.0 trillion times 2 = $14 trillion.

    ssumner would say M = $800 billion in currency plus $200 billion in central bank reserves.

    $1.0 trillion times 14 = $14 trillion

    Now add $2.0 trillion in central bank reserves. Have them all ($2.2 trillion) converted to currency. $3.0 trillion in currency plus $4.0 trillion in demand deposits.

    I say it is still $7.0 ($3.0 plus $4.0) trillion times 2 = $14 trillion.

    ssumner would say even with a little lower V of 10 that:

    $3.0 trillion times 10 = $30 trillion

  253. Gravatar of Fed Up Fed Up
    12. September 2013 at 20:28

    “7. Now let’s assume a cashless economy where the MOA is 100% reserves. Still no change; reserves are still a hot potato. And as I said, IOR changes nothing fundamental. Banks have X demand for reserves at an IOR of Y%. If you double the quantity of reserves and keep the IOR at Y%, banks will suddenly be holding excess reserves.”

    $0 billion in currency, $200 billion in central bank reserves, and $7.0 trillion in demand deposits with a zero fed funds rate. Now add $2.0 trillion in central bank reserves that are excess. Keep $0 billion in currency. I see $2.0 trillion in excess central bank reserves. I don’t get the hot potato effect here.

  254. Gravatar of Fed Up Fed Up
    12. September 2013 at 20:35

    Tom Brown said: “https://class.coursera.org/money-001/lecture/index”

    I’d probably get kicked out of class for being an “unruly student”. Ha!

  255. Gravatar of Fed Up Fed Up
    12. September 2013 at 20:41

    ssumner said: “Odie. The M in MV=PY should be the MOA, which is cash plus reserves in our economy.”

    Odie, I’m going to say the M in MV = PY should be MOA = MOE = currency plus demand deposits and not central bank reserves in our economy.

    I’m going to say this is possible:

    $0 billion in currency, $0 billion in central bank reserves, and $7.0 trillion in demand deposits. V = 2.

    $7.0 trillion times 2 = $14 trillion.

    I believe ssumner would say that can’t happen because his definition of M would equal zero here.

  256. Gravatar of DOB DOB
    12. September 2013 at 22:36

    Scott,

    “And there is no such thing as prices on stuff that doesn’t exist.”

    Alright here’s the best analogy I can cook up: it’s as if I stood there offering to sell a work of art (made-for-order, so it doesn’t “exist” yet) at $100/piece (I’m not a very good artist) and quantity demanded at that price happened to be 0. The price is $100/pc. The quantity is 0.

    “And if I received reserves in payment for a service I provided, I might re-invest those reserves into something else. But I certainly wouldn’t do so in a nanosecond. I’d want to think about it.”

    Your wanting to hold on to reserves and think about it is a violation of the no-monetary-frictions assumption. Which is OK: we know that assumption doesn’t hold true in the real world, but it doesn’t make monetary models particularly enlightening either.

  257. Gravatar of Tom Brown Tom Brown
    13. September 2013 at 05:44

    DOB,

    What do you think of JP Koning’s “convenience yield” concept? Doesn’t the convenience yield depend on there being a monetary-friction? Isn’t his convenience yield a direct manifestation of this monetary-friction? He does a good job demonstrating that w/o a convenience yield, then there’s no hot potato. So in your opinion does this translate into “no monetary friction” = “no hot potato?”

  258. Gravatar of Tom Brown Tom Brown
    13. September 2013 at 06:17

    Fed Up,

    I don’t think Scott would agree with your fixed V. He’d probably say that with NGDP > that as M = quantity of base money goes to 0, that V goes to infinity.

    Or perhaps he’d deny that you could ever have $0 base money.

    To me it seems possible if payment clearing systems were made super efficient and if cash were banned. This is because balances at the TGA are not reserves and thus not part of Scott’s definition of the MOA as “reserves plus currency.” Fed deposits outside of banks are not “reserves” and thus are not part of MOA. I think it’s possible to reduce the convenience yield on reserves (through technological innovation) to such an extent that reserve balances at banks are no longer necessary… that their convenience and marginal convenience = 0.

    Of course I’m assuming a 0% reserve requirement… and no QE. However, the CB could always FORCE the banks to carry reserves through OMOs… but technology could render their marginal convenience yield 0, thus eliminating the hot potato.

    Of course I’m ignoring risk. Perhaps risk always gives reserves a positive marginal convenience yield, even with efficiency in payment clearing at 1. (See JP Koning on this concept of convenience yield).

  259. Gravatar of DOB DOB
    13. September 2013 at 07:42

    Tom Brown,

    Yes JP Koning is, as usual, right on the money. The convenience yield is what people are willing to forgo to hold money and the only reason they wish to hold money (at a lower interest rates than interest bearing nominal assets) is due to monetary frictions.

  260. Gravatar of Tom Brown Tom Brown
    13. September 2013 at 15:16

    DOB, Nick’s latest on MOA vs MOE is “tying his brain in knotts” though!

  261. Gravatar of ssumner ssumner
    14. September 2013 at 06:35

    DOB, I have several problems with your argument. First of all, in the painting example you are selling a derivative–an option on a future painting. So it certainly does exist. If there was IOR on actual paintings, then once the painting was delivered it would earn interest. If the painting was always instantaneously destroyed then it would earn no interest, I agree there. But in that case changing the IOR would have no effect on the economy. It either exists, or IOR doesn’t matter. There is no alternative.

    On your second point, most money is held for reasons of tax evasion, not monetary frictions. In “models” it is sometimes held for reasons of monetary frictions.

  262. Gravatar of Tom Brown Tom Brown
    14. September 2013 at 08:15

    Scott, in your response to DOB about “frictions”: DOB and JP were referring to reserves I thought (and in JP’s case, in discussing the “convenience yield” concept, reserves in the form of Fed Deposits in particular). When you say “money is held for reasons of tax evasion” are you talking about physical currency? I don’t see how that statement would apply to reserves, either vault cash at banks, or bank Fed deposits (but especially the Fed deposits).

    I can see how it would apply to physical cash in circulation amongst the non-bank public.

  263. Gravatar of DOB DOB
    15. September 2013 at 15:31

    “First of all, in the painting example you are selling a derivative-an option on a future painting.”

    It’s the second time you mention options when there’s in fact, no optionality involved in the transaction. You must be confusing options with forwards. Under your definition, you’d classify most business transactions as derivatives: when a company buys something from another company, it generally doesn’t get instantaneous delivery of the good, nor does it instantly transfer the cash. Most people wouldn’t call that a derivative transaction though I agree it’s conceptually the same thing. In any case, just assume I can paint really quickly like those street artists that make portraits on demand.

    “If there was IOR on actual paintings, then once the painting was delivered it would earn interest.”

    IOR already exists in the painting analogy: it’s the price of the painting. In any case, if my painting paid IOR, it would certainly not have the same quantity demanded for a given price so IOR does matter even before I ever paid my first painting.

    “But in that case changing the IOR would have no effect on the economy. It either exists, or IOR doesn’t matter. There is no alternative.”

    I just don’t see how you can make that claim. If I put a 100% tax on sellers of french fries (that is any sales proceeds is taxed away), there would be 0 sales of french fries. I would get $0 revenue from the tax. In that sense, the tax wouldn’t “exist”. But it would still matter.

    The fact that the equilibrium base quantity is zero at a certain level of the convenience yield is completely irrelevant to how much impact the target level of rates have on the price-level/economy. By the way, all nominal securities would be impacted by it via no-arbitrage (although if agents decided not to trade nominal securities, there would be no less of an impact).

    If you still don’t agree, we’ll have to agree to disagree as I’m running out of examples and arguments..

    Actually one last piece of food for thoughts: imagine a world with no physical currency, 0% reserve requirements, and only depository institutions can hold reserves at the CB (that part is already the case). I think at a sufficiently high convenience yield (funds minus IOR), banks would never use reserves in cases when a customer from one bank wires to another customer in another bank. Banks would just transfer securities to each other instead. Base would be zero. Does that mean the CB has no control over nominal rates, the price level, etc?

  264. Gravatar of ssumner ssumner
    15. September 2013 at 17:24

    DOB, You said;

    “I just don’t see how you can make that claim. If I put a 100% tax on sellers of french fries (that is any sales proceeds is taxed away), there would be 0 sales of french fries. I would get $0 revenue from the tax. In that sense, the tax wouldn’t “exist”. But it would still matter.”

    But that’s exactly my point. Changing a prohibitive tax rate from one rate to another doesn’t matter at all. Quantity is still zero. So changing an IOR rate would not matter at all. But I thought you were claiming that the IOR rate could be used as a sort of monetary policy? Perhaps we are talking past each other.

    As far as your last question, yes, with no MOA there is no price level.

  265. Gravatar of Fed Up Fed Up
    15. September 2013 at 21:01

    TB said: “I don’t think Scott would agree with your fixed V. He’d probably say that with NGDP > that as M = quantity of base money goes to 0, that V goes to infinity.”

    That is why monetary base (currency plus central bank reserves) needs to go to zero.

    TB said: “Or perhaps he’d deny that you could ever have $0 base money.”

    ssumner said: “As far as your last question, yes, with no MOA there is no price level.”

    I am pretty sure monetary base (currency plus central bank reserves) can be zero with the economy still being fine. MOA would equal demand deposits here just like they are now because of 1 to 1 convertibility (relative pricing).

    I’m probably going to have to find an accounting and/or banking expert to back me up because ssumner is not going to believe me.

  266. Gravatar of Fed Up Fed Up
    15. September 2013 at 21:16

    DOB said: “Actually one last piece of food for thoughts: imagine a world with no physical currency, 0% reserve requirements, and only depository institutions can hold reserves at the CB (that part is already the case).”

    ssumner, do you agree with this part, only depository institutions can hold reserves at the CB (that part is already the case)?

    Let’s try it this way. Imagine a world where physical currency is available, but no one uses it. The currency is at the central bank. Also imagine a 0% reserve requirement (Canada and others) and only depository institutions can hold reserves at the CB (that part is already the case).”

    DOB, if two entities bank at the same bank, can a demand deposit settle the transaction so no currency and no central bank reserves are involved? For example, I swap $700 in demand deposits for a $700 notebook (there is an asset swap).

  267. Gravatar of Max Max
    15. September 2013 at 22:40

    DOB, “I think at a sufficiently high convenience yield (funds minus IOR), banks would never use reserves in cases when a customer from one bank wires to another customer in another bank. Banks would just transfer securities to each other instead. Base would be zero.”

    There has to be a non-zero (could be only $0.01) base for a FF-IOR spread to exist. If base is zero, then FF=IOR.

    Scott: central bank money is the MOA for private bank money. It can’t be the MOA for itself. That would be a circular definition, e.g. “One dollar is defined to have the purchasing power of one dollar” – how does that define the price level?

  268. Gravatar of Tom Brown Tom Brown
    16. September 2013 at 09:34

    Fed Up, only banks can hold reserves period. That’s how reserves are defined: base money at banks.

    Also, look here for an example of how payment could ultimately be settled w/o central bank involvement:

    http://brown-blog-5.blogspot.com/2013/02/banking-example-1.html

    the CB starts and ends w/ an empty balance sheet. So whey doesn’t bank A just directly borrow reserves from B w/o ever doing the overdraft in the 1st place?

  269. Gravatar of DOB DOB
    16. September 2013 at 13:00

    Scott,

    “So changing an IOR rate would not matter at all. But I thought you were claiming that the IOR rate could be used as a sort of monetary policy? Perhaps we are talking past each other.”

    You keep saying IOR and I keep saying cost of funds. If you have no monetary frictions, you can set IOR an epsilon below cost of funds and just forget about it (quantity will remain zero).

    Cost of Funds is the relevant control variable. Cost of Funds minus IOR only affects quantity and quantity doesn’t matter.

    “As far as your last question, yes, with no MOA there is no price level.”

    I strongly disagree with this and wrote a blog post in response.

    Tom Brown,

    “if two entities bank at the same bank, can a demand deposit settle the transaction so no currency and no central bank reserves are involved? For example, I swap $700 in demand deposits for a $700 notebook (there is an asset swap).”

    Of course: it’s just a bookkeeping entry that happens in the bank’s database without any involvement in the monetary base or anything outside of the bank.

    “There has to be a non-zero (could be only $0.01) base for a FF-IOR spread to exist. If base is zero, then FF=IOR.”

    FF-IOR is set by the central bank. Everything else held constant, base decreases when the spread is increased. You could certainly have a base of zero and FF > IOR and a non-zero base with FF=IOR (like now).

  270. Gravatar of Max Max
    16. September 2013 at 14:53

    DOB, “FF-IOR is set by the central bank. Everything else held constant, base decreases when the spread is increased.”

    I agree. But base demand is what enables setting FF independently of IOR. You can’t obliterate the demand and still use the demand as a tool to control FF.

  271. Gravatar of Fed Up Fed Up
    16. September 2013 at 18:32

    DOB said: “Tom Brown,

    “if two entities bank at the same bank, can a demand deposit settle the transaction so no currency and no central bank reserves are involved? For example, I swap $700 in demand deposits for a $700 notebook (there is an asset swap).”

    Of course: it’s just a bookkeeping entry that happens in the bank’s database without any involvement in the monetary base or anything outside of the bank.”

    That was actually mine.

    ssumner told me in a different post that transactions have to settle in base money (currency plus central bank reserves). I didn’t/don’t agree.

    So you are saying transactions don’t have to settle in base money, right?

  272. Gravatar of DOB DOB
    16. September 2013 at 18:34

    Max,

    I’m wrong and you’re absolutely right.

    Not that it’s any excuse but here’s how I got wrong: I used “cost of funds” in the rest of the comments above which–to me–is the rate at which the Fed lends funds into the market so those comments were correct.

    Generally, that’s not too far from FF so I leaped a bit too quickly when you jumped to FF in your comment. But as you point out, in a case with no base demand, the Fed will be unable to keep FF away from IOR. In fact effective FF won’t really be defined as there won’t be any transaction in the FF market.

    I can’t even blame you for jumping from CoF to FF since (a) those are often synonymous and (b) I used “funds” as an abbreviation in the comment you cited and that’s generally FF.

    Sharp eye, sir.

  273. Gravatar of DOB DOB
    16. September 2013 at 18:41

    Fed Up,

    “That was actually mine.”

    Sorry I got mixed up.

    “So you are saying transactions don’t have to settle in base money, right?”

    It’s like handing a casino chip to another guy. It only needs to be converted to cash if the guy wants a chip from a different casino. Otherwise, no cash is touched.

    I’ve never dealt with it myself so I can’t say 100% but I’m 99% convinced that no base money is involved. Why did Scott say there was?

  274. Gravatar of Fed Up Fed Up
    16. September 2013 at 20:07

    DOB said: “Why did Scott say there was?”

    I believe the M in MV = PY is currency plus demand deposits because that is what circualtes in the real economy. ssumner says the M is currency plus central bank reserves. See here:

    http://www.themoneyillusion.com/?p=21463

    One post there of mine:

    “ssumner’s post: “The crude quantity theory of money (assuming constant V) is a sort of multiplier model, where V is the multiplier linking the money supply and NGDP.”

    MV = PY

    Can we agree central bank reserves do not directly circulate in the real economy?”

    Another post of mine:

    ssumner’s post said: “Bank reserves are basically cash, and circulate between banks. (Banks are part of the real economy.) When they begin circulating outside the banking system they are called “currency.”

    That is what I thought your scenario would look like. Central bank reserves circulate between banks in the fed funds market. The fed funds market is not part of the real economy. The demand deposits are moving, and the central bank reserves are going along with the demand deposits if it is a different bank.

    For example, I buy a notebook for $700 from a retailer. I write a check for $700 from my account at bank A. The retailer deposits it in bank B. The central bank reserves end up going along with the check to bank B. It was the check that moved in the real economy (it has velocity), not the central bank reserves.

    A better way to think about it is to assume there is only one commercial bank or the retailer had an account at bank A. The demand deposits move (have velocity in the real economy) from my account to the retailer’s account to pay for the notebook. The central bank reserves did not move at all in those cases (velocity is zero).

    ssumner’s post: “If you are going to buy that $700 notebook, ultimately it must be paid for by moving $700 of base money from your account to the store’s account. The check is not payment, just a promise to pay. The store wants cash transfered into its bank account.”

    Another post of mine:

    “I buy a notebook for $700 from a retailer. I write a check for $700 from my account at bank A. The retailer deposits it in bank B. My checking account gets marked down by $700, and bank A’s account at the fed gets marked down by $700. The retailer’s checking account gets marked up by $700, and bank B’s account at the fed gets marked up by $700. I like to think of it this way. The demand deposits get transferred from my account at bank A to the retailer’s account at bank B, and the central bank reserves get transferred from bank A’s account at the fed to bank B’s account at the fed. Notice the central bank reserves are not in my account at bank A or the retailer’s account at bank B. The check/demand deposits settle the transaction. Assuming one commercial bank or assuming the check gets deposited in the same bank allows that to be seen easier. With either of those assumptions, the central bank reserves don’t move at all from bank A’s account at the fed. The vault cash does not move either. The demand deposits/check move (mark down then mark up) to settle the transaction and circulate in the real economy.

    I’m saying the check is payment.”

    ssumner’s post said: “Fed up, That’s wrong, base money does move from account A to account B.”

  275. Gravatar of ssumner ssumner
    17. September 2013 at 06:51

    Fed up. You said:

    “ssumner said: “As far as your last question, yes, with no MOA there is no price level.”
    I am pretty sure monetary base (currency plus central bank reserves) can be zero with the economy still being fine. MOA would equal demand deposits here just like they are now because of 1 to 1 convertibility (relative pricing).”

    Sure, but then DDs are the medium of account. BTW, what exactly is available “on demand” in a DD when there is no cash or reserves?

    Yes, I can imagine a case where all the base is held in deposits at the central bank.

    Max, I don’t know what “medium of account for itself” means. The MOA refers to the good in which prices are measured.

    Fed up. Obviously I shouldn’t have said transactions “have to” settle in terms of reserves or cash, you can do barter. Poker chips, whatever. I meant that when you buy something with say a credit card, you eventually pay the credit card company some bank reserves.

    DOB. It doesn’t matter whether you call it “cost of funds” or “cash” If there is nothing to apply it to, the rate doesn’t matter. You can change the rate, and nothing happens. It’s like taking a prohibitive tariff, and raising it even higher. So what?

  276. Gravatar of Tom Brown Tom Brown
    17. September 2013 at 14:37

    “Sure, but then DDs are the medium of account. BTW, what exactly is available “on demand” in a DD when there is no cash or reserves?”

    A small slice of the CPI basket of goods?

  277. Gravatar of Tom Brown Tom Brown
    17. September 2013 at 14:38

    … well I guess your bank wouldn’t provide that “on demand” so maybe that’s not a good answer.

  278. Gravatar of Tom Brown Tom Brown
    17. September 2013 at 14:39

    But your bank would make the credit available on demand for that purchase of the CPI slice (or any other goods) so in that sense it’s more “on demand” than a savings deposit I suppose.

  279. Gravatar of Tom Brown Tom Brown
    17. September 2013 at 14:46

    Given a period of time T, when over which no Fed or Tsy activity transpired, I can see where it might be possible, with a set of rules in place favoring it, that no base money would actually have to be used to settle payments: no cash, no reserve requirements, all banks are healthy. Then when the reserve transfers are netted out at the end of the day (e.g. Bank A owes bank B $1000 and bank B owes bank A $1020), then instead of B having to overdraft $20 and then pay it back, why couldn’t they just directly arrange with A to “borrow” the $20 of reserves (i.e. A becomes a debtor to B in the amount of $20) which they could do by simply crediting a demand deposit for B by that amount.

    Now as soon as someone pays taxes, or gets a check from Tsy, or the Fed buys or sells, reserves are needed again of course.

    I know my example is overly simple: all the banks in good health is not a good assumption, etc. But it’s conceivable it could world like that. It would be similar in that case to having just a single commercial bank. The Fed still controls the FFR because they offer to lend should the banks try to charge too much. Also, they are needed for those Tsy transactions, etc.

  280. Gravatar of Tom Brown Tom Brown
    17. September 2013 at 14:50

    Shoot, that should read

    “(i.e. B becomes a debtor to A in the amount of $20) which they could do by simply crediting a demand deposit for A by that amount.”

  281. Gravatar of DOB DOB
    17. September 2013 at 16:32

    Scott,

    “It’s like taking a prohibitive tariff, and raising it even higher. So what?”

    You’re confusing the spread between CoF and IOR, with the absolute level of such rates.

    The spread might indeed be “too high” if the base is zero, and raising it doesn’t achieves anything when that’s the case, we agree.

    But the absolute level of policy rates (i.e. not the spread) isn’t the opportunity cost of holding money, it’s the rate at which money is lent into existence. And by no-arbitrage, it’s also the rate of return on all nominal assets (after risk-adjustment).

    For that reason, it matters as explained here.

  282. Gravatar of Fed Up Fed Up
    18. September 2013 at 12:05

    ssumner said: “Sure, but then DDs are the medium of account. BTW, what exactly is available “on demand” in a DD when there is no cash or reserves?

    Yes, I can imagine a case where all the base is held in deposits at the central bank.”

    DD’s are the MOA, right. And DD’s are part of the MOA when there is currency too.

    Assume currency and central bank reserves can be created and destroyed quickly and without cost. Now they can be zero and people can still get currency whenever they want. The point is MOA = currency plus demand deposits with 1 to 1 convertibility. People can have more currency and fewer demand deposits so MOA is still the same.

    And, “I meant that when you buy something with say a credit card, you eventually pay the credit card company some bank reserves.”

    I have a checking account and credit card account at JP Morgan (JPM). I pay JPM both principal and interest on the credit card.

    Interest part: My checking account gets marked down, and JPM’s liabilities get marked down. JPM’s equity gets marked up.

    Principal part: My checking account gets marked down, and JPM’s liabilities get marked down. JPM’s assets get marked down.

    Let’s skip the barter part.

  283. Gravatar of ssumner ssumner
    19. September 2013 at 08:29

    DOB, I guess I just have no idea what you are talking about. I don’t see how an interest rate can be applied to nothing. If the asset doesn’t even exist, what difference does it make what the interest rate is? I don’t care if it is IOR, COF, whatever. The asset must exist for the interest rate to be meaningful.

    You were the one that raised the prohibitive tariff analogy. I am saying that prohibitive tariffs don’t matter in a world without trade, and IOR and COF doesn’t matter in a world without base money to apply them to.

    Tom, You said;

    “A small slice of the CPI basket of goods?”

    Transactions costs?

    You said;

    “Given a period of time T, when over which no Fed or Tsy activity transpired, I can see where it might be possible, with a set of rules in place favoring it, that no base money would actually have to be used to settle payments: no cash, no reserve requirements, all banks are healthy. Then when the reserve transfers are netted out at the end of the day (e.g. Bank A owes bank B $1000 and bank B owes bank A $1020), then instead of B having to overdraft $20 and then pay it back, why couldn’t they just directly arrange with A to “borrow” the $20 of reserves (i.e. A becomes a debtor to B in the amount of $20) which they could do by simply crediting a demand deposit for B by that amount.”

    Sure, but then reserves exist.

  284. Gravatar of DOB DOB
    19. September 2013 at 09:39

    Scott,

    “The asset must exist for the interest rate to be meaningful.”

    The asset exists, there’s just zero quantity of it outstanding.

    One more analogy:

    Imagine the electricity market is a monopoly. And let’s say:
    $3/kWh is the monopolistic optimal price, while
    $1/kWh is the cost of production

    The government in an effort to increase consumer welfare offers to sell electricity at a price of $2/kWh, but it is continuous current rather than alternating current, which is the market standard. (We assume that the cost of conversion from DC to AC is negligible, and that the government has the capacity to produce enough to satiate the market.)

    What’s the monopolistic provider to do? He’s gonna sell for $1.999/kWh of course! (and still come out with a decent amount of producer surplus)

    How much quantity of DC current is demanded by the market? Zero. Would you say that the government’s OFFER to provide electricity at $2/kWh does not matter just because nobody made use of it?

    (In this analogy DC current is base money while AC current is bank deposits. The only reason I distinguished AC and DC is because you would have said that they are the same good. If the providers of electricity deviate from the price, then the market will buy from the government instead. The standing offer matters, even if it’s not hit.)

    Likewise the Fed sets nominal interest rates even if base is zero. If nominal assets trade at a yield lower than IOR, then agents will exploit the arbitrage by holding base money (and then base won’t be zero). That threat COULD/SHOULD be enough to be bring yields back in line. People CAN hold base money if they choose. All I’m saying is that the Fed has control of rates (and the price level) EVEN IF in equilibrium agents happen to not want to hold any base money.

  285. Gravatar of ssumner ssumner
    23. September 2013 at 03:40

    DOB. OK, I see your point here. But that doesn’t tell me what asset banks transfer when a check is cleared. And you still need a MOA.

    I also disagree with the claim that control of rates implies control of the price level. That may be true in a world with a MOA, but not in a world with no MOA.

  286. Gravatar of ssumner ssumner
    23. September 2013 at 03:43

    I would add that a zero level of base money is not the equilibrium solution in the world you describe, unless everyone is identical.

  287. Gravatar of Tom Brown Tom Brown
    23. September 2013 at 10:21

    “But that doesn’t tell me what asset banks transfer when a check is cleared”

    We could have: A overdrafts (Fed deposit), B (Fed deposit) is credited, B then loans excess reserves back to A to repay overdraft…. OR we could just jump straight to A owes B (i.e. A establishes a deposit for B).

  288. Gravatar of Bitcoin Faucet Rotator Blog Rates or quantitites or both Bitcoin Faucet Rotator Blog Rates or quantitites or both
    10. April 2015 at 06:21

    […] bank to safely push rates below zero. As for the quantity side of the equation,  the threat of Sumnerian permanent increases in the monetary base may not be able to reduce the overnight non-pecuniary […]

  289. Gravatar of Why IOER, oh why IOER, oh why IOER, did I ever do IOER? | Historinhas Why IOER, oh why IOER, oh why IOER, did I ever do IOER? | Historinhas
    21. September 2015 at 15:20

    […] rates to wherever they settle, getting out of the way and letting the new cash work its “hot potato” […]

  290. Gravatar of Jeff Jeff
    12. January 2017 at 20:54

    Astonishing! This many people disagree with the original post and no-one actually disagrees with the part that is actually wrong!

    (Wrong here is a technical term meaning the part I personally disagree with).

    1, 2, and 3 seem reasonable. If you assume 4 then 5a and 5c seem reasonable. I would have to see a bit more of how you are using them, but at first glance I don’t yet see a reason to doubt 5b, 6, or 7.

    The problem with 4 is not what happens, but rather the implication that the value of gold with respect to all the other goods in society will move in sequence. A better model is that there will be a large movement of prices on assets and things rich people buy and very little movement of prices of things that things poor people buy. If there is a very large upper middle class that buys both in large quantities then potentially extra “dollars”, whether they be greenbacks, gold coins, or whatever else, could potentially drive consumer price inflation, but in a highly unequal society, increases in the supply of whatever is used for currency can be expected to drive heavy “asset inflation” but almost no CPI.

Leave a Reply