Email from JJ Lando, now at Nomura:
“THE LTRO DIDN’T DO ANYTHING. ALL THE MONEY WOUND UP AS DEPOSITS AT THE ECB” “QE DIDN’T DO ANYTHING. ALL THE MONEY BECAME EXCESS RESERVES BACK AT THE FED.”
(Apologies in advance to all who have heard me give this one ten times before)
1. Central Banks, whenever they buy any asset (eg lend eg grow balance sheet) create new reserves.
2. Commercial banks and people do NOT have the capacity to destroy those reserves. Once the Fed or ECB wires the money or creates that asset line item on its spreadsheet, there is an equal and offsetting liability on its spreadsheet called reserves. This spreadsheet cannot be broken.
3. All that commercial banks can do is lending, which moves some of those reserves from ‘excess’ to ‘required’ but they are still there.
4. Commercial Banks make this lending decision based upon regulatory capital and profit motives, not based upon reserves. They have a ‘captive audience’ in their Central Banks, who MUST create the necessary reserves (a floored amount) to prevent interest rates from going to infinity.
5. When a Central Bank does a lot of Balance Sheet expansion in a short time, it’s going to wind up as deposits/excess NO MATTER WHAT. If the Fed does 1T of QE, Banks don’t suddenly ‘find’ the regulatory capital to make 10T of loans. And even if they did, there would be the SAME AMOUNT OF TOTAL RESERVES.
6. Bank lending to a 0% risk weighted sovereign actually does NOTHING to diminish excess reserves.
7. Simplified Illustration: ECB does a very large unsterilized LTRO. They take a lot of sov paper on balance sheet (temporarily), and they wire NEW FUNDS to thie member banks. Those member banks take some of the money and buy paper from the ITalian government. That government spends the money by wiring it to its pensioners. Those pensioners take it to buy food from the local grocer. The local grocer DEPOSITS IT IN HIS BANK. SOMEWHERE DOWN THE CHAIN the money winds up on deposit in some member bank, be the chain long or short. WHATEVER MONEY THE ECB CREATES WINDS UP ON DEPOSIT IN ITS MEMBER BANKS, WHETHER OR NOT IT IS ‘USED’ TO BUY SOVEREIGN DEBT, ‘USED’ TO MAKE LOANS, OR NOT USED AT ALL.
8. Please. I never wish to read again that ‘Central Bank money went unused because it wound up as deposits.’ IT HAS NO WHERE ELSE TO GO. THE BANKING SYSTEM IS A CLOSED LOOP. With the possible exception of someone making a withdrawal, taking the paper, and making a bonfire (actually not feasible in the hundreds of billions anyway bec there are constraints)
9. And that is probably how Italy just managed to borrow at 1.64%
Good luck!
Yep, may help explain why front-end rates dipped below 25bps during that time.
I am a (relative) ignoramus, trying to figure things out. I assume that the Fed and the ECB are the only entities that can create Dollars and Euros (respectively). Banks can’t create dollars, but can create Promises to Deliver Dollars (PDDs), as I think of them. My checking account consists of promises to deliver a certain number of dollars as I order.
How do banks store Dollars (other than physical bills and coins)? I assume it’s more than an honor system — each bank having a register on its own facilities declaring the number of dollars it holds. Likewise, I assume that wire transfers involve something more than (for example) BankA increasing its register by $100, BankB decreasing its register by $100, and trusting both banks to do the Right Thing?
Perhaps instead, BankA and BankB hold accounts at the Fed, and the Fed increases BankA’s and decreases BankB’s numbers? If this is how it works, it would seem to me that (except for physical bills and coins), it would be literally impossible for actual Dollars to exist outside of the Fed.
So in this sense, talk of foreign banks holding tens of trillions of dollars off coast (outside the USA) is literal nonsense? They only hold either accounts at the Fed or PDDs from other banks?
@John M,
All the Banks have accounts in the Central Bank and that’s how they transfer it from one side to the other. Not trust – Central banking.
bank loans create bank deposits. the deposits are denominated in the same currency as the loan, and the deposits are not the banks to spend, they belong to the depositor.
With the possible exception of someone making a withdrawal, taking the paper, and making a bonfire
Besides the point but I believe, this is illegal in many countries anyway.
@Tristan Lanfrey, By lowering long term interest rates, it does give some breathing room to those pre-crisis junk securities.
@pebird, sorry, comment was supposed to be reply to 4, Paul Krueger.
@pebird,
Even independent of interest rate effects, if those toxic assets are moved from the bank’s balance sheet to the Fed’s via QE, doesn’t that make it easier for the bank to meet Basel capital requirements? That’s really the question I was asking.
The Fed’s only been buying US govt and agency securities?
reminds me of a story. back in the early 1990’s i was asked how low the italian lira could go. I asked what the btu value of a 1 lira note was…
Warren,
I’ve often wondered whether QE has another purpose altogether, namely to replace riskier bank assets with reserves which have a 0 risk weight. Wouldn’t that have the effect of making it easier for them to meet capital requirements? Might this be considered a form of bail-out for banks with capital problems?
not sure how it might do that? If I’m a bank with, say, greek bonds in my portfolio, they don’t go away unless I sell them.
@WARREN MOSLER,
This is probably just my ignorance, but for some reason I was under the impression that in earlier QE, the fed was actually buying things like those risky mortgage-backed securities. When I read about QE it mostly says things to the effect that the fed exchanges reserves for longer term financial assets, which seems to me like it could be most anything including foreign bonds, corporate bonds, securities, CDS etc. But I don’t actually KNOW what they’re exchanging. If you’re saying that you know for a fact that when QE was done they only exchanged reserves for US bonds, then that certainly answers my question and relieves my concern. Is that what you’re saying?
@Paul Krueger,
I did a quick google search for QE and “toxic assets” and got LOTS of hits where people claim that part of QE was purchasing toxic assets from banks (done both here and in UK). At at least some claim that those assets were valued much higher than fair market value. So it still seems to me like that would have the effect of making it easier for the banks to meet risk-weighted capital requirements because they lowered their overall asset risk fairly significantly. Am I missing something?
talk is cheap
yes.
early on some other ‘risky’ financial assets were bought, but that wasn’t part of what became known as QE
@WARREN MOSLER,
You are selling them, to the central bank.
yes, if a bank sells some of its holding to anyone, it ‘makes room’ for something else.
and banks can sell their bonds any time, with or without the central bank buying them.
presumably the price is a bit higher with the cb buying, but usually not a lot higher, and sometimes a lot lower than if the sell waited until after the cb was done buying
@Paul Krueger,
Tom recently summarized some discussion on this topic:
http://mikenormaneconomics.blogspot.com/2012/01/more-on-mmt-and-risk.html
Short version… I think if QE involves buying risk assets such as MBS then yes it can “help” the banking system if those assets are (or would be) dropping in price. If QE buys treasuries, the effects are more complicated. Some interest rate risk is likely to be removed from the financial sector but it’s not clear (to me) whether that will have any meaningful impacts on the economy at a macro level. Perhaps in a world of rising interest rates and no regulatory forbearance, but even then it would depend on the capital raising environment IMO.
Price, not Quantity. I think this phrase turned the world upside down for me and I haven’t been the same since.
It was a similar moment to understanding the delta of an option is the probability it will finish in the money.
@TC,
Sorry, delta is NOT a probability that the option will end up in the money. 😮
@Peter, lol yes it is. For a plain jane call or put, the delta is ALSO the probability it the option will finish in the money.
It’s not “exactly” true, but its really, really close. Feel free to take it up with Natenberg if you like.
@TC,
No, it is not. It is a prob that it ends in the money in some fictitious universe, called 1″risk neutral world”, but not in our world. It is a common misunderstanding that it is a real-world probability and yes, it is close, most of the time, but mathematically it is not true.
Awesome. I remember a post from him here on QE when he was at GS.
Are you friends with/have taught him, Warren?
@wh10,
Lando and Hadden (MS) both formerly of GS are both from the same “school”.
Never met or talked to him, but he’s been on my mailing list and we’ve corresponded.
@WARREN MOSLER,
OT, but curious to hear your response to
http://pragcap.com/theres-a-difference-between-theory-fact/comment-page-1#comment-96133
and
http://pragcap.com/theres-a-difference-between-theory-fact/comment-page-1#comment-96164
sorry, no time to get to it today.
The banking system is not a closed loop. You can withdraw your deposit and never put it back into the system. Of course, under normal conditions, you can assume it works as a closed loop because someone down the payment chain will eventually deposit the cash back.
But the main problem with banks, that is bank runs, comes from the fact that the banking system is not a closed loop, or that closing the loop is beyond its control.
@MamMoTh,
Every loop has “leakages.” As for bank runs, aren’t we just talking about a loop with leverage?
@Art, then it’s not a closed loop or it doesn’t matter there is a loop at all? only electronic money and the disappearance of cash could close the loop. it seems to me bank regulations work to keep the loop as closed as possible, but that’s still beyond the control of the banking sector.
Warren,
Could you explain point 6. It seems incorrect to me.
If a bank has excess reserves and it buys govt bonds, then this excess reserves are diminished or not?
@walter,
If it has excess reserves then the central bank is not issuing the bonds.
Therefore the government bonds are being issued by the government to cover spending, so the reserves that come from the bank to buy the bonds end up back in the bank as soon as the government spends.
So whereas in a sensible MMT world the government would mark up the Bank reserve account and the customer liability account, we are marking up the Bank’s bond account and the customer liability account. The government is literally spending with bonds.
@walter,
I’d like clarification here too, but some thoughts.
When a bank lends, it creates a loan and a deposit. Reserves aren’t involved here. It seems this is how bank PDs ‘buy’ govt debt.
If a non-bank PD buys a bond, then reserves go down, but they may have been previously increased by the Fed to ensure enough reserves are circulating. So net, reserves might not change.
Lastly, if a govt sells debt, they intend to spend that amount back into the economy. So if reserves go down by amount of debt sold, they go right back up by amount of debt sold.
@wh10,
Re PDs:
http://www.newyorkfed.org/aboutthefed/fedpoint/fed04.html
http://www.newyorkfed.org/markets/pdcf_faq.html
Re “if a govt sells debt, they intend to spend that amount back into the economy.”
Technically, they’ve already spent it, but intentions for funds rec’d are determined by current spending plans and the interaction of revenues and expenditures. Treasury auction volumes reflect the net effects of the three.
“So if reserves go down by amount of debt sold, they go right back up by amount of debt sold.”
Depends what account(s) the proceeds end up in (see p 15 here: http://www.rayservers.com/images/ModernMoneyMechanics.pdf). Only a drain on reserves if proceeds wind up in a Treasury acct at the Fed, and only nets out to zero if proceeds go into its (or others’) non-FRB bank accts.
@Art,
Actually I tried to have that discussion with Edward Harrison from Creditwritedowns today. Maybe someone here can help out.
My query was around this step:
>Those member banks take some of the money and buy paper from the ITalian government.
I don’t know enough about the structure of the European banking system to understand if the italian treasury would have a reserve account at the ECB, or whether they have commercial bank accounts for these transactions.
If the treasury did have an ECB account then technically this particular action could remove reserves from the “closed loop” , obviously only temporarily.
That may not be the case in this instance, but it does raise the question whether there are certain financial transaction that can cause a drain on the reserves available to banks outside those of the treasury and/or reserve bank.
For instance what happens in the case where a foreign central bank sells treasuries in the open market in order to top-up its reserve account for some reason.
My understanding is that this would
1) transfer the treasury asset to a private sector actor
2) decrease the deposit amount in the associated bank of the private sector sector actor
3) transfer the corresponding reserves from the commercial bank’s reserve account to the reserve account of the foreign central bank.
Now assuming I have got this correct, because I am still a little vague on whether foreign central banks do have true reserve accounts ( and not a commercial bank account ). But if that was the case then I was wondering whether this would technically transfer some reserves away from the banking system.
Warren states below that “euro states have their accounts at their central banks”. So does this mean that the italian central bank has an account at the ECB?
Anyone have any idea on this ?
reserve shortages and surpluses occur continuously due to what are called ‘operating factors’ including govts changing balances in their cb accounts, changes in cash in circulation, what’s called ‘float’ of uncleared checks, etc. etc.
Centrals banks do what’s called ‘offsetting operating factors’ as needed to keep rates at their targets.
@walter, I think part of the background here is that Lando is discussing the European system, in which the various national governments are just currency users like any other currency user. So I’m assuming buying an Italian bond is not much different than making any other kind of loan. The bank that buys the bond creates a deposit account for Italy’s government which is an asset of the Italian government and a liability of the bank. In return the bank receives a bond that is a liability of the Italian government and an asset of the bank. If the bank was carrying excess reserves, then the loan to Italy need not affect its reserve balances in any way. If the bank was not carrying excess reserves, then sometime during the calculation and compliance period they will have to acquire more reserves either through interbank lending of from a central bank lending facility.
If they fail to do so, the central bank will unilaterally credit the necessary additional reserves to the bank’s reserve account, at some penalty rate slightly above the usual CB lending rate.
@Dan Kervick,
Typo in:
“either through interbank lending of from a central bank lending facility.”
Should be:
“either through interbank lending or from a central bank lending facility.”
@Dan Kervick,a. Looks to me the mail is just as well about QE in US and FED.
b. Are Eurozone sovereigns’ bonds (e.g. greek) still o% risk weighted?
c. The Eurozone sovereign that the bank lends to (buys bonds from) does not always have its bank account at the buyer of the bond. (most likely not)
I agree with you that if the banking system has excess reserves and lends to other currency users then this does not diminish the excess reserves in the banking system. But again, I think that is not what is said in the mail/article.
@walter,
But again, I think that is not what is said in the mail/article.
Agreed. See my comment at 11 below.
@Dan Kervick
“The bank that buys the bond creates a deposit account for Italy’s government which is an asset of the Italian government and a liability of the bank.”
No, settlement is against payment into the TARGET2 account of the country held at its central bank.
for the bank, but not the banking system, unless the seller of the bonds has an account that’s considered ‘outside’ of the banking system (like the ECB)
@WARREN MOSLER,
Right , so where does the US Tsy has its bank account? in or outside the banking system?
And where do Eurozone member states have their bank account? in or outside the banking system?
the tsy has a fed account and accounts at commercial banks
euro states have their accounts at their central banks
@WARREN MOSLER,
All 17 governments have at least one TARGET2 account at their central bank. See eg. the 2010 annual report of the central bank of Italy. They list exactly 42,488,335,099 euros as a liability to general government. These euros don’t play in the fractional reserve circus until transfered from the MFI to a bank.
@Jacob Goense,
Jacob, there isn’t a fractional reserve system. But transfers of funds create some problems:
http://www.ecb.int/mopo/liq/html/treas.en.html
I still do not understand how Italy borrowed at low percentage.
Is it because ECB let private banks acquire cheap loans?
@Gary,
“I still do not understand how Italy borrowed at low percentage.
Is it because ECB let private banks acquire cheap loans?”
Yep (at least I think so). Banks have a guaranteed loan rate with relaxed terms from the ECB that goes way longer than their loans to Italy (3Y vs 3M). The lenders expect to pocket ~64 bps annualized over three months, which reflects (1) the ECB’s target rate and (2) the perceived risk of the current arrangement falling apart within three months, and maybe some other factors.
Not a bad deal at all if you’re a eurozone bank. The entire fiscal compact, in tandem with ECB policy shifts since Nov, now looks like it was aimed at buying 2-3 years time for Italy and Spain while allowing banks to do some B/S repair. What happens in three years is anyone’s guess, but I’d bet the whole thing gets extended. It will be interesting to see if any more dovish responses (beyond a lower ECB rate) arise in response to a recession though.
@Art,
thank you.
So then it is in Italy’s interest to do more short term bond sales, because short term rates are so low.
It wasn’t that low a yield. in fact, still pretty high. just lower than the time before.
@WARREN MOSLER,
ok, I see.
But the reason it is lower than before – is because of cheap ECB loans? In theory – since banks can always get ECB loans – that should not really make a difference…
What is wrong with this theory about the ECB LTRO implemented in December: it is potentially an unwind scenario for the currency itself. Essentially the ECB is now paying the various national banks in the Zone to re-patriate national debt: Italy’s banks get a juicy ECB subsidy to lend to Italy, Spain’s to Spain, Greece’s to Greece etc. As I understand it this subsidy is not limited to new debt either so a Greek bank could use virtually free money to purchase outstanding high yield Greek debt.
My first thought was, that won’t fix anything because Germany holds so much external (by which I mean non-German Euro zone) debt that re-patriation can’t possibly absorb it in three years so Germany can’t get itself off the hook for Southern debt in this time frame. But then it occurred to me that the austerity fad will put Germany’s balance of payments into reverse pretty soon, and pretty severely and either Germany will start running substantial deficits or what? export workers to Poland? Argentina? Where?
I think someone at the Bundesbank understands Wynne Godley’s stock flow consistent accounting and has submarined a mechanism his or her political masters obviously don’t understand into the ECB mix. There are so many different potential external shocks, who can possibly foresee where events will take us, but it is not at all improbable that a German deficit adequate to allow the orderly transition of the majority of debts held in Germany into German bank issued debt will occur over the next three years. This run up of German debt could expedite and complete a nationalistic debt repatriation within the union and eliminate cross border exchange rate risk ahead of a Euro event horizon imagined to be after this rebalancing and nationalizing of national balance sheets has run its course.
Once German held debt is mostly German, Spanish held debt is mostly Spanish, Greek, Greek etc. issue of national currencies re-introduces a currency float to balance international trade through the currency. Of course there is an equal possibility I have no idea what I’m talking about!
@john newman,
“Of course there is an equal possibility I have no idea what I’m talking about!”
Interesting hypothesis though! Is there actually an additonal incentive from the ECB for banks to ‘repatriate’ the debt of their HQ govt issuers? And if so, any evidence it’s occuring?
@Art,
Distributed political pressure? No one else wants to buy emu-zone debt.
While bowing to Bundesbank pressure? Not buying gov bonds directly.
John describes a staged retreat by the ECB. If so, who’s orchestrating it? Many bankers from the emu-periphery see the ECB as a career path. If Bundesbank wants to kill the ECB, then it’s also cutting loose other country CB staff, which will instill neither trust nor loyal cooperation.
@john newman,
During the 3 years you mention, peripheral countries will either have to join Germany, or leave. The scenarios you describe set the hook deeper, forcing those two options more & more as the only options.
It seems to me that there is a gap between point 6 and point 7. Up to point 6 it all sounded like something Warren or Scott F. would say. But then at point 7 Lando suddenly starts talking about deposits, not reserves.
a. Looks to me the mail is just as well about QE in US and FED.
b. Are Eurozone sovereigns’ bonds (e.g. greek) still o% risk weighted?
c. The Eurozone sovereign that the bank lends to (buys bonds from) does not always have its bank account at the buyer of the bond. (most likely not)
I agree with you that if the banking system has excess reserves and lends to other currency users then this does not diminish the excess reserves in the banking system. But again, I think that is not what is said in the mail/article.
@walter, Sorry this belongs sub 9 , see above, to Dan Kervick
@walter,
“b. Are Eurozone sovereigns’ bonds (e.g. greek) still o% risk weighted?”
Not sure what you ask, but the Eurosystem Credit Assessment Framework (ECAF) still lists many Greek bonds as eligible for collateral.
What is MMT’s overall take on the 2008 fall-out, the 2011-12 Euro crisis, the respective governments’ (including central banks) monetary and fiscal responses, and the coming effects?
Specifically (I apologize if some of these are off-base):
Are the QEs increasing the monetary base relative to the contemporaneous economic expansion/growth (or lack thereof)? To what extent? Is this even relevant to the MMT? Is this how MMT measures inflation?
Is it possible that there is at some point no (or little) demand for the non-convertible currency (gov’s monopoly product) in excess of what’s needed to pay taxes? If so, what would be the effect of that given the current monetary base (and its trends), and how much control would the govt have over such lack of demand (given that taxes cannot realistically be raised over a certain level, etc.)?
Has there been any growth in real wealth attributable to production and creation of real wealth (increased standard of living) and not attributable to investor psychology (nominal growth)? Is this even relevant?
How much specifically is the gov constrained by inflation in MMT’s view? Can the gov always contain interest rates? If not, what is the specific breaking point (beyond the abstract “inflation”)?
In MMT’s view, is there currently real economic inflation in excess of what is being stated by the government and pundits?
lots of questions! start by reading through my blog posts over the last few years?
@Nick Boyd,
1. From what I’ve read, QE changes nothing in practical. Maybe except lowering interest rates for bonds. The measure of inflation remains the same – a consistent increase of price levels over time (wikipedia knows definitions by heart, I don’t)
2. At some point, in fact, in most occasions, there will be more currency in circulation than what is needed to pay taxes. If inflation arises (which will happen when economy is in full employment) the Government should reduce it’s deficits/increase superavits. That can be made either by cutting taxes or by cutting expenditures. Some argue that interest rates on bonds also have an effect.
3. It’s probably all in some statistic somewhere.
4. I don’t get how the control of interest rates is connected to breaking point. The Government is constrained by inflation and full employment, so when the country is at full employment (like 1 or 2% unemployment) and if prices start to rise, that means the private sector cannot increase production to meet additional demand. Thus, if the Government keeps spending, it will give more money into an economy which cannot consume more products, and prices will rise.
5. Statistics for inflation are tricky, that is sure. Some take in account taxes, others don’t, etc. For example, here, the INE (National Statistics Institute) takes in account consumer prices which are weighed in a basket according to the consumption pattern of the average family. So for you to know if statistics are off, just check if they have all the consumer prices, if they are weighed or not, etc.
On the closed loop of banking:
What if, as a result of a changed investor psychology, many will take their deposits out and put them in stashed physical gold, silver, land, guns, oil, etc.? What responses to this trend are available to a gov (that has a proper understanding of MMT)?
Thanks, guys…
@Nick Boyd,
For somebody to be able to buy gold, somebody wiser had to sell it for cash.
From a macro point of view that transaction doesn’t exist. Somebody used to have gold and now has cash, somebody had cash and now has gold. Nothing has changed except the ownership labels.
@Neil Wilson, Well, each transferee of cash does not deposit that cash into any bank? What if no one wants to accept cash in exchange for gold or any other thing of value? Wouldn’t that be a different macro situation?
@Nick Boyd,
There would have to be an extreme political situation for the population to no longer value the currency.
yes, like no enforcement of tax liabilities
@Nick Boyd, Close the gold window at the central bank.
Warren,
I refer to the comments and your replies above sub 9 regarding point 6.
Point 6: “Bank lending to a 0% risk weighted sovereign actually does NOTHING to diminish excess reserves”.
From Lando’s mail it is unclear to me whether his point 6 is meant exclusively for the Eurozone (EZ).
Do you agree that point 6 is true for ez member states, but not for the US (and other monetarily sovereign states)?
In other words:
Accounts of ez member states at their national Central Bank are considered inside the banking system.
Accounts of the US Tsy at the FED are considered outside the banking system.