Cullen has agreed to respond to 3 of my points of my choosing. So here they are.
1) In Hyperinflation Explained in Many Different Ways I would like him to respond to the section, Modern Monetary Theory and Monetary Realism.
2) In my Hyperinflation FAQ, the comment, "Can't we wait till there are signs of inflation before doing anything?".
3) In my Hyperinflation FAQ, the comment, "How is hyperinflation stopped?".
Answer
Cullen replied here. Thanks Cullen!
Vince
Cullen, if you are rejecting the MMT view that government bonds are part of the money supply then how can you logically keep their claim that when the central bank makes new money and buys up bonds that it is “just an asset swap”. Their logic was that if bonds are already money then swapping for money does not change the quantity of money. If bonds are not money then making new money and buying bonds is “monetization”, replacing government debt with money. So if you reject their view on bonds, how can you keep saying it is “just an asset swap”?
Cullen, I really meant for you to respond to my answers for the FAQ questions, not just the questions themselves.
Cullen
It is just an asset swap. If you want to call it monetization
then fine. Just be sure to also call it unassetization or something
like that. Call it unprinting if you insist on calling it money
printing. Talk about both sides of the ledger. An asset gets printed
and another asset gets unprinted. If you want to claim that will cause
hyperinflation then so be it.
I’d love to go back and forth on this stuff, but I am pretty busy today so maybe another time. Take care.
I’d love to go back and forth on this stuff, but I am pretty busy today so maybe another time. Take care.
Vince
In my FAQ the parts that I wrote are the answers. So to respond to me should have been responding to my answers. Instead Cullen just looked at the questions. I called him on it and he ignored me.
He answered part 1 but when I point out a logical flaw in his position he does not address it and says, "I am pretty busy today so maybe another time". Cullen then responds to 7 other comments in that thread. I call BS. Cullen is not really debating with me.
There will always be buyers for bonds. Otherwise reserves must remain earning 0%. When the choices are 0% or >0%, people choose >0%.
ReplyDeleteSo there goes that theory.
If inflation is too high, create less money and become more productive, problem solved.
Hyperinflation is stopped by simply stopping the policies that got to hyperinflation.
That was easy.
Cullen's answer to you is good.
Deletehttp://pragcap.com/the-function-of-education-is-to-teach-one-to-think-intensively-and-to-think-critically/comment-page-1#comment-165789
There are plenty of historical examples of countries where their were no buyers for government bonds in that country. As you go into hyperinflation the value of bonds comes crashing down. So even if the interest rate on the bond is greater than zero, the return on the bond is far less than even. People will be much better off buying cans of tuna or anything else that holds value. So they don't buy bonds. The central bank is left buying the bonds with new money.
When the government is running a huge deficit and they need cash to spend they always get the central bank to buy their bonds. They make the laws the central bank operates under, they appoint the people who run the central bank, and they have the guns. The central bank always buys bonds when nobody else does, or the government would have to shut down. Even while central banks are making hyperinflation they can not stop. If the problem was so easy to solve you would not have the huge number of historical hyperinflation cases.
The politics that get you into hyperinflation are spending much more than you get in taxes. Japan is spending about twice what they get in taxes. This is not a simple or easy problem to fix. In practice the currency will be destroyed before they get around to fixing it.
If you think this is an easy problem then you do not understand it.
"Cullen, if you are rejecting the MMT view that government bonds are part of the money supply then how can you logically keep their claim that when the central bank makes new money and buys up bonds that it is 'just an asset swap'"
ReplyDeleteThe Fed exchanges excess reserves for t-securities and government guaranteed MBS. Fed reserves are balances which exist to help banks make payments (clear transactions) between one another. Thus, they are only a type of 'money' to the banking system, making them fundamentally different than bank deposit balances.
Non-banks predominantly use bank deposit balances in order to make payments to one another. The deposits can certainly always be exchanged for another form of money in physical fed reserve notes. This act does not increase the non-bank money supply, it only changes the composition of it.
QE insofar as it conducts this swap with banks does not impact the non-bank money supply at all. However, since the Fed works through the banking system/dealers in order to acquire t-bonds from non-banks, non-banks would see their bank deposit balance increase in equal dollar amount to the t-bonds sold. That said, to non-banks this is an asset swap where they sell t-bonds which are the government liability for bank deposits, a bank's liability. Since only a certain % of bank deposits are protected, some hold t-securities because they are perceived to be a safer store of value when compared to non-protected bank deposit balances.
Furthermore, the holder of the t-securities could always make this swap regardless of the Fed QE. They can always sell t-securities to a dealer/bank who may very well just increase the supply of deposits in order to accommodate that sale. In effect the bank would be making the decision to increase the size of their balance sheet.
If a bank sends an armored truck to a Fed bank they can withdraw their excess reserves as regular paper cash. So there is no fundamental difference or limitation on excess reserves.
DeleteIf a private party sells a government bond to another private party it just changes who has the cash and who has the bond. It does not increase the amount of money in existence.
"If a private party sells a government bond to another private party it just changes who has the cash and who has the bond."
DeleteIt depends. If the private party buyer uses bank credit (say a securities dealer) in order to acquire the t-bond, this would at least temporarily increase the supply of deposit (non-bank money). Essentially a bank would be accommodating the dealer by extending them a loan. The non-bank deposit (money) supply is dependent on the elasticity of the bank's balance sheet.
"If a bank sends an armored truck to a Fed bank they can withdraw their excess reserves as regular paper cash"
This would be an asset swap as well. The bank would exchange their reserve balances held at the Fed (one of the Fed's liabilities) in order to acquire another in the form of physical fed reserve notes. The bank would still be holding a Fed liability, just in physical instead of digital form. Banks can do this in order to accommodate their customers who wish to exchange their bank deposits for physical notes (again an asset swap). The bank customer wouldn't have more money, just a different form of money.
Aside from that it is logical to see that bank's would prefer to hold digital reserves since they pay .25% interest (IOER) rather than hold physical Fed notes which pay 0%.
Vincent writes:
Delete"If a bank sends an armored truck to a Fed bank they can withdraw their excess reserves as regular paper cash."
True, but that paper cash is still considered to be bank reserves. This won't change the reserve level at all. The only difference is that paper cash is non-interest bearing.
The point is that the idea that excess reserves are in any way tied down is just wrong. They are the same as cash because they can be converted to cash.
DeleteAnd cash can be converted to and from bank deposits (on a dollar for dollar basis). So using your logic, is there any difference between bank deposits and cash? And by the transitive property, is there thus any difference between bank deposits and reserves?
DeleteSo what's more important: the total dollar value of all reserves or the total dollar value of all reserves + bank deposits?
... or all reserves + bank deposits + non-reserve cash?
DeleteGood post hitting on some of this. (The charts are very good).
Deletehttp://www.forbes.com/sites/francescoppola/2014/01/21/banks-dont-lend-out-reserves/
For non-banks doesn't this all just come down to mostly caring about aggregate bank deposits and to a lesser degree physical Fed reserve notes(cash)?
Just because a bank is able to convert their digital reserves into federal reserve notes (cash), doesn't mean they will just hand them out to non-banks. It is still their asset. It is still the Fed's liability. I'm not sure why banks would want to hold most of their reserves as physical notes instead of digital reserve balances especially when the Fed pays IOER.
Maybe I'm bias considering I hate holding cash and would always prefer using a bank deposit balance and a debit card to make payments.
DeleteTom, I think all the bigger definitions of money supply are really just ways of moving ownership of the base money around faster. So each bigger definition of money supply has a lower velocity but in some way makes it easier to change ownership of base money. So a checking account or debit card can let you move money between accounts electronically. But all the different definitions of money supply have their uses.
DeleteThis comment has been removed by the author.
Delete"So a checking account or debit card can let you move money between accounts electronically."
DeleteThe electronic deposit balances to non-banks ARE money, in a sense that they are the medium of exchange. When I spend, my account is reduced and the receiver's account is increased. One person's bank deposit balance number decreases and the other person's bank balance number increases. These are record keeping devices.
If we both bank at the same institution, there is no corresponding base money (reserve) movement. Behind the scenes base money may be exchange if the two non-banks hold bank deposit balances at two different institutions. One bank's reserve balance is reduced, while another bank's reserve balance is increased. Again, digital reserves are record keeping entries that net out when payments are made between two banks.
I think it is important to appreciate this digital network and what that means. The "movement" for the most part is not physical, but has more to do with adjusting numerical balances up or down.
Vincent, if
Deleteelectronic Fed deposits = cash = bank deposits
because all can be converted from one to the other on a 1:1 basis, and if the equation of exchange is:
V = NGDP/M
for whatever way you want to define M, then which definition of M do you choose? How do you justify it?
You write "each bigger definition of money supply has a lower velocity" but that's not clear to me at all. For example, if you had a single commercial bank, and no cash, then reserve velocity would be zero, but if you looked at bank deposits, then there would be a velocity > 0.
Likewise if you increased the time between settlement dates in a multi-bank world: as the delta time increased the average reserve velocity would drop to zero (average payment settlements between banks would tend to net to $0, especially when normalized by delta-time), but bank deposit velocity wouldn't change.
I'm not sure where I'm going with this. I just questioned that bit about the velocity, and I'm also not sure at all which definition of M to use in NGDP/M. Each one produces it's own V, and it's hard for me to tell which to use, especially if you reason that convertibility implies equality between the different types of M.
Imagine M1 is $2.5 trillion and M2 is $10 trillion (for easy math). Then if V1=X then V2=X/4 since M2 = M1*4.
DeleteYou really can use whichever definition of money supply you want. I can imagine different problems could call for different definitions.
Convertability of dollars in checking accounts does not imply that a definition of money including that comes to the same total as a definition not including that. The different Ms have different totals. Really M2 is like $11 trillion and M1 $2.6 trillion.
http://research.stlouisfed.org/fred2/series/M2/
http://research.stlouisfed.org/fred2/series/M1/
Your explanation with M1 and M2: that's very clear, of course!! I'm wrong that V = 0 for the single commercial bank bank and cashless society case because by definition V = NGDP/M, and as long as M > 0 and NGDP > 0, then V > 0, even if M = bank reserves and not a single $ of bank reserves "moves" anywhere (all commerce in this case being done with bank deposits only). Kind of makes you wonder what the utility of that definition of V is though, doesn't it?
DeleteIn fact, in the caseless single bank case, as the reserve requirement approached 0%, and the bank lowers their reserve holdings proportionately, then V would approach infinity (assuming NGDP stayed constant)!
... in other words, just because some subset of "money" doesn't actually flow or move or change quantity in any way, this doesn't have anything to do with it's "velocity."
DeleteEach time something adds to the GNP some money changes ownership and that contributes to the velocity. This does not require physical movement and certainly not any change in quantity. It could theoretically all be on a computer at the central bank.
Delete"computer on central bank" right, but imagine the single commercial bank scenario again in a cashless society, and a period of time over which no Fed or Tsy activity takes place: all commerce is conducted with bank deposits (M1 but not MB) so NGDP > 0. No even an electronic Fed deposit entry need be changed, yet there's still MB > 0, thus VB > 0. But this "velocity of base money" is divorced from all concept of movement. Not one penny of electronic Fed deposit credit owned by the commercial bank (records stored at Fed) is changed at all: only commercial electronic bank deposits are altered to facilitate the commerce. This does not actually require a cashless society with a single commercial bank: this would be possible given our real system: imagine that over some period of time, the only transactions that happen are between electronic bank deposits at BofA. Then NGDP > 0, but again, not one penny of base money, electronic or otherwise moves in any way: no Fed deposit accounts are altered at all and no physical cash changes hands. Unlikely? Absolutely, but this would still give rise to a non-zero base money velocity (VB > 0). The fact that this can happen tells me that "velocity" really gives the wrong implication about what NGDP/M really means.
DeleteTom, great insight.
DeleteUsing a single commercial bank model/scenario really drives home the point that electronic fed deposit/reserve entries exist mainly to help commercial banks settle with one another. This function becomes unnecessary in the single commercial bank model.
If velocity = "the number of times one dollar is spent per unit of time" then wouldn't the "velocity" that matters the most mainly involve commercial bank deposits? Would velocity be defined not really as how often one dollar is spent, but how often dollar denominated electronic commercial bank balances are adjusted to reflect transactions?
Good questions Rafael, but I don't think that is how velocity is defined. It really is defined just as NGDP/M, where the M chosen matches the V result. e.g. VB = NGDP/MB.
DeleteA definition like you mention might be more useful and more in accordance with our intuitive understanding of "velocity." In that case it might actually be possible for V2 > V1. But the way it's defined, that can never happen since V2 is a superset of V1.
Thanks Tom.
DeleteWhat is the goal for trying to understand velocity? Is it to understand the change in nominal spending relative to the quantity of money?
Looking at the Fred website, what does the following graph tell us?
Gross Domestic Product / Deposits, All Commercial Banks
(GDP)/(DPSACBW027SBOG)
"In that case it might actually be possible for V2 > V1. But the way it's defined, that can never happen since V2 is a superset of V1."
Imagine if the system goes fully electronic where bank deposit balances are no longer exchanged for cash. How would this change our perception of all this? With advances in debit/credit card technology would it be a shock to see this happen voluntarily where the deposit/cash exchange ceases all together (or happens in negligible amounts)? If so, would the real economy lose its direct connection to the MB as reserves are then really only primarily used to settle between banks and nothing more?
Tom, if bank deposits are changing ownership that counts as money movement.
DeleteVincent, yes I know. Consider this situation though. Bank A owns $1 in reserves, and has two deposits, each of $1: one for customer X and one for customer Y. Customer X pays Y $1 for a service, and thus NGDP is $1. But now we want to calculate VB = NGDP/MB = $1/$1 = 1. The velocity of base money was 1, but yet Bank A's reserves (say consisting of a single $1 paper note in its vault) didn't move at all: only credit money moved (X and Y's deposits are part of M1, not MB). Now let's calculate M1 = $3. Thus V1 = 1/3. Curious that no base money moved at all, yet VB > V1. If both X and Y had savings accounts instead of time deposits, then bank A could have purchased a $1 T-bond and still met its reserve requirements, but MB would be $0. Now VB = infinity, and V1 = 1/2.
DeleteRafael, you write "What is the goal for trying to understand velocity?" ... well I think of velocity as literally NGDP/M, and try not to be confused by the terminology "velocity" since it has nothing to do with movement or change in ownership of the M you've identified. I confused myself on this point in a comment above!
DeleteLInk to the graph on Fed website?
I don't think it'd be a huge shock to go fully electronic. It wouldn't be smooth sailing either though (lots of political opposition to that I'm sure).
A country like Canada basically does have a system where bank reserves are somewhat divorced from the rest of the economy: they target overnight interest rates on a short term basis (for six week intervals). They do still have cash, so the tie isn't completely severed (although they did consider getting rid of cash at one time). The reserve requirement is 0%, but banks do hold small amounts of reserves overnight anyway.
But yes, you are correct, reserves would primarily be used to settle between banks and nothing more, although they'd also have a role in foreign exchange and any payments to or from Tsy, other gov agencies, and the Fed as well.
... so even with a single commercial bank and no cash, because of gov and CB payment clearing with the public, the 1:1 convertibility with bank deposits would be maintained.
Delete... and BTW, Vincent, in my example above, you can't say the $1 paper note in Bank A's vault belonged to anybody but Bank A: it only appears on Bank A's balance sheet (as an asset: and the CB's BS as a liability of course). This doesn't change when commerce takes place. X and Y only have their bank deposits on their BSs, no paper money.
DeleteRafael, some thoughts: I don't think getting rid of cash, or having a single commercial bank, or reducing the reserve requirement to 0% amount to huge changes from our current system. For Vincent that might not be true: if we don't have cash, does that seem like a big deal to you Vincent?
DeleteI also don't see the activities of the Fed with QE as being particularly Earth shattering. Why? Because the Fed only spends in a limited market, and a relatively small amount*, and they pay market prices. If the Fed purposely overpaid for ANYTHING, that could cause some inflation though. And of course deficit spending by Tsy always has the possibility of independently causing inflation.
*I wrote above "relatively small amount" ... here's some justification:
http://www.themoneyillusion.com/?p=25996#comment-314586
"So what’s eligible for purchase? Roughly…
Federal government Treasuries – $12.0 trillion
Agency bonds – $2.0 trillion
Agency guaranteed securities – $5.8 trillion
Bankers acceptances – zilch
Gold – $6.9 trillion
State and municipal bonds – zilch
Foreign bonds – $5 trillion
This comes to about $31.7 trillion dollars. Now, the Federal Reserve currently holds about $3.8 trillion in such assets so that leaves about $27.9 trillion in assets that they still haven’t bought.
(And I suppose they can buy up all the world’s foreign currency and overnight deposits which ought to be worth at least a few trillion dollars but let’s not get crazy!)
And I still haven’t touched on section 13.3 of the FRA yet.
Section 13.3 allows the Fed to lend to any individual, partnership, or corporation upon any collateral the Fed deems satisfactory."
-- Mark Sadowski
"I think of velocity as literally NGDP/M, and try not to be confused by the terminology "velocity" since it has nothing to do with movement or change in ownership of the M you've identified."
Delete"you can't say the $1 paper note in Bank A's vault belonged to anybody but Bank A: it only appears on Bank A's balance sheet (as an asset: and the CB's BS as a liability of course). This doesn't change when commerce takes place. X and Y only have their bank deposits on their BSs, no paper money."
Tom I 100% see the system in this way as well. I didn't mean to give the impression that I was implying that changes in velocity had to do with changes in the ownership of base money.
If velocity is defined as NGDP/M, my question was referring to which M to use and why. Why is using MB as M more informative than using bank deposits as M? Commerce is dominated by the use of electronic bank deposit balances not MB. Even when the Fed conducts QE by buying securities from non-banks, the most important effect is captured in the change in total bank deposits.
Good write up on QE, reserves, deposits.
http://www.forbes.com/sites/francescoppola/2014/01/21/banks-dont-lend-out-reserves/
So as a non-bank as long as I hold bank deposits, I am not simultaneously possessing ownership over any form of MB. The only way for me to possess ownership over MB is for me to exchange my bank deposit for physical notes. Thus any commerce that is done using bank deposits does not necessarily result in the changing of ownership of MB unless both parties transacting in bank deposits hold accounts at different banks. At that point the two different banks would need to settle in MB.
"I don't think getting rid of cash, or having a single commercial bank, or reducing the reserve requirement to 0% amount to huge changes from our current system."
Exactly. I don't either. Using a single bank example or a cashless system where all non-banks always transact in bank deposits isn't far from where we already are. Those thought experiments help drive home that point.
So when thinking about hyperinflation or high inflation, isn't it more important to focus in on the quantity of bank deposits relative to economic output/spending, and not changes to MB? Again, the important effects of the changes in the MB in QE manifest themselves in the changes in bank deposit totals anyways. Even here, it is always important to consider that the Fed is removing government guaranteed securities when conducting QE. Isn't it also important to focus in on the quantity of treasury securities held by the private sector as well?
This is the chart I was referring to earlier.
http://research.stlouisfed.org/fred2/graph/?g=rgP&dbeta=1
"So when thinking about hyperinflation or high inflation, isn't it more important to focus in on the quantity of bank deposits relative to economic output/spending, and not changes to MB? Again, the important effects of the changes in the MB in QE manifest themselves in the changes in bank deposit totals anyways. Even here, it is always important to consider that the Fed is removing government guaranteed securities when conducting QE. Isn't it also important to focus in on the quantity of treasury securities held by the private sector as well?"
DeleteYes and yes. :D
Although not just hyperinflationists would disagree: MMists wold probably disagree too... and they make an interesting argument. Sumner argues that the hot potato effect still exists even in a cashless society with fiat money (see his case 7 here):
http://www.themoneyillusion.com/?p=23314
From the Sumner post:
Delete"Now let’s assume a cashless economy where the MOA is 100% reserves."
Has he ever addressed what would happen in a cashless system where there is only a single bank? Would there be zero need for Fed reserves or the Fed? Would the single bank issue bank deposit balances when swapping for private (loans, securities) and public assets (t-sec/agencies)? Would those bank deposit balances serve the function of MOA and MOE? I would say yes.
Am I essentially describing a merger between the CB and Banking system? In the story above, would financial intermediaries exist to receive bank deposits as their assets in exchange for their liabilities? What if their liabilities start to act like bank deposits?
Stepping back to where we stand today, bank deposits (at least guaranteed deposits) trade at par with reserves.
The fact that,
$1 bank deposit = $1 Fed reserves is incredibly under-appreciated.
In the story, even if the liabilities of the financial intermediaries acted like the single bank's deposits, they wouldn't be because their prices are not pegged to one another. The fixed exchange could eventually break down.
In our system, I don't think the peg between bank deposits and reserves will be allowed to clear at anything less than 1 to 1. A deposit/reserve fixed exchange rate is only possible because the Fed's balance sheet is elastic in a multi bank system. If it weren't and bank deposit balances were allowed to float relative to $ reserves, then physical federal reserve notes would reassert themselves as the real MOE.
I'd love to get your thoughts on this
http://ftalphaville.ft.com/2013/09/05/1623992/gold-paper-scissors-lizard-e-money/
Here is another post which explores "trade at par" or par clearing.
Deletehttp://www.themoneyillusion.com/?p=23314
In order for 1 bank deposit to always trade at par (fixed exchange rate) with 1 reserve, does this imply that the Fed/treasury is on the hook for all bank deposit liabilities? If there was a systemic issue with assets held by banks where the fixed exchange rate broke down, wouldn't the Fed have to swap out the toxic assets with reserves or t-securities? Otherwise, the bank could either fail or their deposits would start to trade at a discount (sort of like Cyprus where depositors faced a haircut).
Rafael, I'll try to get to that FT Alphaville piece a bit later. And your other questions too. Keep in mind I'm really no expert... I haven't even really made up my mind what I believe yet... the more I learn the more questions I have, etc. (I'm used to being the one asking the questions actually!). But I did notice this the other day: an MMist highlighting both Miles' and Sumner's proposals:
Deletehttp://macromarketmusings.blogspot.com/2014/01/miles-and-scotts-excellent-adventure.html
I think Austrians would scoff at both!
A long time ago I did read a piece by Miles on that subject, but I don't recall 18 steps! I'll check it out.
re: the link to the Sumner post: that's the one I gave you on HPE. Is that what you meant to post?
Delete... Sumner and banks: Sumner actually understands the basics of banks and the accounting that goes with them, but he still doesn't think they are important for macro. He once famously stated that he doesn't understand banking, but he understands at least as much as I do. He is focused on the MOA which to him is only base money. Bank deposits he calls "credit." Not all MMists agree with that. I think he makes an interesting argument for his point, but I'm not fully on board with him there. I think Vincent and Sumner perhaps agree about the importance of banks and base money, but they have very different views on hyperinflation, QE, and the wisdom of central banks targeting anything (inflation, NGDPLT, etc). Nick Rowe has a lot of good pieces exploring the convertibility issue between base money and bank deposits. Rowe likes to create crazy imaginary worlds to illustrate his points. Try looking up things like "paintings" in his search tool (if my memory serves me) for some good ones. Rowe and Sumner don't completely agree on the MOA issue.
DeleteThe one on HPE is the one I read. The quote "Now let’s assume a cashless economy where the MOA is 100% reserves" comes from that.
DeleteAlso read the Macro Mark Musings post yesterday.
My adventure in all things money and banking actually started when reading Mises, The Theory of Money and Credit. After a few years of experiencing the world post undergrad, it was not possible for me to stick with what I once thought to be true.
In that book, Mises sees bank liabilities which trade like reserves as counterfeit. Contemporary monetary thinkers do a much better job of exploring the relationship between the CB liabilities and bank deposits. Perry Mehrling's work does an awesome job exploring this while incorporating the parallel/shadow banking system's role.
Katherina Pastor’s paper “Towards a Legal Theory of Finance” is also fascinating because it explores the nature of financial securities especially securities which are forms of money.
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2178066
Really I guess we are both learning. That said, I've gained a lot from reading your work which is why I ask you these questions. If anything I think they make for interesting thought experiments.
This comment has been removed by the author.
Delete"He is focused on the MOA which to him is only base money. Bank deposits he calls "credit."
DeleteThis view seems very similar to that of the Austrian economists. This is also why I would like to see how he would describe a cashless system where there is one universal bank whose liabilities in the form of bank deposits serve the role of the MOA and MOE. The deposits would come into existence as the bank acquires other securities/loans.
Would this essentially be describing a system where the banking system and Fed have merged into one entity?
In the FT Alphaville piece she entertains the idea that the Fed would essentially become the universal bank whose liabilities would be the primary MOE for every other entity. In this world, I can see why the monetary base would matter a lot. Other financial intermediaries would accept these Fed deposits as their assets while issuing their liabilities (which could be called bank deposits). What would happen though is that Fed deposits and bank deposits would not be guaranteed to trade at par.
In my mind, the fixed exchange rate between bank deposits and fed reserves make bank deposits money, not credit. I can't see this not being the case as long as the Fed and Treasury are there to backstop bank balance sheets as they did during the crisis.
If you find any of those old Rowe articles, he argues that the BoC (bank of Canada) and the Bank of Montreal (BoM) have an arrangement where the BoM dollars can be traded at par with BoC dollars, but the relationship is asymmetrical. It's a good article, and I think it could be used to justify a position like Sumner's, at least in part. But overall I find it unconvincing.
DeleteDid you follow Sumner's full progression there? Through the gold standard, etc in the other six cases? The implication is that the CB leads and any other banks are thus compelled to follow, otherwise arbitrage kills them.
You should just go ask Sumner: he'll take off topic questions. :D
I did get through the progression and will work through it again.
Delete"The implication is that the CB leads and any other banks are thus compelled to follow, otherwise arbitrage kills them."
Is that more consistent with what we have witnessed or...
The implication is that the Banking system leads and the CB/Treasury are thus compelled to follow, otherwise systemic failure kills them.
Has the banking system consolidated its power enough to be the dog wagging the tail? The Fed does not providing the real economy with the primary medium of exchange, banks do. Right?
Good questions Rafael! I don't know the answers, although I do believe that bank deposits are probably the current primary medium of exchange (MOE).
DeleteSumner and Rowe part ways on the importance of MOE (MOE which is not also MOA) a little bit. Rowe thinks that the MOE is important in it's own right, while Sumner disagrees. Woolsey tends to agree with Rowe on this, but not 100%. I think Glasner tends to back Sumner. But this importance of MOE is strictly a short to medium term affair in Rowe's view... in the long run, Rowe and Sumner agree I think: it's the MOA that matters.
BTW, here's one of several posts Rowe did on MOA vs MOE. I remember his weird invented world used paintings for money, so it was easy to find based on that search word. There's another word he's used, which escapes me at the moment... which is the key to finding more on this. Anyway, here's the article:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/09/moe-vs-moa.html
Ah, "asymmetric" did the trick I think. Again, this is one of many: he has many little stories that repeat some of the same themes from slightly different points of view. It's an interesting way to form an argument... and appealing to those of us not interesting in pouring over pages of math or empirical data ... but I tend to take it with a grain of salt. Anyway, here it is:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/10/what-makes-a-bank-a-central-bank/comments/page/2/
BTW, I had to laugh, once I saw where Rowe was disputing something that Stephen Williamson had written (Williamson's recent article showing how QE could be deflationary), and Rowe used one of his little stories in the comments section to argue his point. Williamson was not amused and basically told Rowe to grow up and start acting like an economist. Haha.. it was a little harsh I thought, but funny.
DeleteI probably didn't represent Rowe's differences w/ Sumner on MOA and MOE very well. Here's a recent Rowe comment in his own words:
Deletehttp://worthwhile.typepad.com/worthwhile_canadian_initi/2014/01/one-good-thing-about-bitcoin.html?cid=6a00d83451688169e201a51154c135970c#comment-6a00d83451688169e201a51154c135970c
Vincent, you should update your post with Cullen's responses to your response: (i.e. his bit about if monetization, then you hve to agree it's also un-printing bonds, etc.). Well, it's easy to do, here's Cullen in his own words to Vincent's 1st point:
ReplyDelete"It is just an asset swap. If you want to call it monetization then fine. Just be sure to also call it unassetization or something like that. Call it unprinting if you insist on calling it money printing. Talk about both sides of the ledger. An asset gets printed and another asset gets unprinted. If you want to claim that will cause hyperinflation then so be it." -- C. Roche
Oh well.
DeleteI was hoping for a more interesting debate out of Cullen but Tom found me a fun debate for today. And happy birthday Tom!
ReplyDeletehttp://worthwhile.typepad.com/worthwhile_canadian_initi/2014/01/is-the-macroeconomic-importance-of-finance-an-artefact-of-monetary-policy.html
Thanks Vincent!
DeleteVincent, Sumner has now picked up Rowe's ball and run with it... giving you license to jump in over there and drag the conversation off into the hyperinflation pit as well!
Deletehttp://www.themoneyillusion.com/?p=26025
BTW, it looks like you have replies on Rowe's site from 1) Too Much Fed, 2) Ralph Musgrave, 3) anon (who provides a great link to another Rowe article):
Deletehttp://worthwhile.typepad.com/worthwhile_canadian_initi/2014/01/nominal-loss-aversion-and-its-consequences.html
... and 4) Peter N.
O/T: Looks like Cullen's favorite MMist (David Beckworth) also made a recent appearance on Erin's RT show:
ReplyDeletehttp://macromarketmusings.blogspot.com/2014/01/my-interview-on-rt.html
The little write up before the video is a pretty good summary. I liked the preceding article too. Nothing there directly pertaining to hyperinflation, but I thought you might like it anyway since some of the ideas that he briefly summarizes there are probably in direct conflict with aspects of Austrian theory (aspects which may underpin your hyperinflation theories).
BTW, I think I mentioned that I did a brief poll amongst the MMists to see if they'd be OK with an algorithm replacing human judgement at the CB to implement NGDPLT. A bit of a mixed reply, but overall they'd be OK. Sadowski actually says the existing Taylor rule can be used for that purpose and that furthermore had that been implemented in late 2008, the CB BS would be much smaller than it currently is.
Your thoughts? Would you accept a robotic CB performing NGDPLT?
Vincent and Rafael,
ReplyDeleteThis is the kind of mini Sumner speech I was hoping to find (his reply to John Becker):
http://www.themoneyillusion.com/?p=26030#comment-315209
Now I don't necessarily agree with him, but I find that to be a very nice little summary. So Vincent, that sounds like the cue for you to pay him a visit and start "waving your hands." :D
Actually that's a good question for you Vincent: if you were made Fed chair tomorrow, and dissolving the Fed was not an option, how would you control the money supply? (OMOs, IOR rate, etc: What would your decisions be on how to control these instruments?). What would you tell the public your goal was? What would be your big picture strategy?
I asked Geoff/Major_Freedom what he'd do as Fed chair, and his response was that he would refuse the job because only a "psychopath" would take a job like that. Nice.