Tuesday, August 13, 2013

Honest Banking

Fractional Reserve Banking

 Imagine a brand new bank is formed and starts with no money.  Next customer A comes in and deposits $10,000 cash in a new savings account.  Next the bank loans out $9,000 to customer B who takes his borrowed money out as cash.

This loan increases any measure of the money supply that includes both "demand deposit accounts at banks" and cash.    That $9,000 is really counted twice.   Since both customer A and customer B think they can spend this money, economists have decided it is reasonable to count it twice.  Since there are 2 people that think they have this $9,000 it is inflationary. 

So when a bank makes a loan from demand deposits they are increasing some measures of the money supply.  In this sense people say, "banks make money".

This system of telling deposit customers they can take out their deposits on demand while at the same time loaning the money out for many years is called "fractional reserve banking".   This is because only a fraction of the deposit money is actually kept on hand, say 10%.   This type of banking has a terrible flaw.   If a higher fraction of the customers demand their deposits back at the same time than the reserve ratio the bank can not give all of them their money.   This is called a "bank run" and a "liquidity crisis".

The patch for this terrible flaw is to make a central bank that acts as a "lender of last resort" when a bank has a liquidity crisis and temporarily loans the bank some money.  The central bank has a real ability to make new money, either electronic or having it printed.  So the central bank can not run out of money.

A private bank can have a solvency crisis when the loans that it made are not paying.   In this case the bank has a bigger problem that just a liquidity crisis.   A temporary loan from the central bank will just delay the inevitable bankruptcy and not really fix the problem.    The central bank is supposed to decide which banks just have a liquidity crisis and which have a solvency crisis and shut down the insolvent ones.   It is sometimes not easy to decide if a bank is having a solvency crisis or a liquidity crisis.

If the central bank has the job of bailing out troubled banks then it opens itself up to being tricked by dishonest bankers.   It is possible to make it look like they have a liquidity crisis, get a bunch of money from the central bank, and then run with the money.  Of course the central bank tries not to get conned too often. 

The current system with this "bank made money" can have deflation when lots of people are just paying down loans and not getting new loans.   The overall money supply can go down.   Then there is a patch for this.  The central bank can print more money.   But this has risks of inflation or even hyperinflation if the government gets out of control debt and deficit.  So this deflation/inflation/hyperinflation trouble is another drawback of the current type of banking.

This type of bank can run into trouble if they make long term loans at current interest rates and then interest rates go up.  Imagine they were paying depositors 2% and loaning money for 20 years at 5%.  This is a healthy profit margin.  But then imagine that short term interest rates move up to 6%.  If they don't pay their customers this rate the customers can just take out their money and move someplace else.  But if they pay depositors this much then they will lose money on their loans.   There is a patch for this.   They can have adjustable rate loans.   But then if interest rates can change people who were credit worthy at the starting rate may not in fact be able to pay once rates have moved up.   So adjustable rates loans have their own bad side-effects and were part of the recent crisis.   So rising interest rates is yet another problem with this type of banking.


Honest Banking

An alternative to fractional reserve banking is to always match the duration of deposits with the duration of the loans.  To do this banks would sell 10 year bonds to raise money which would then be used to make 10 year loans.   They would sell 20 year bonds to get money for 20 year loans.   Since there are no "demand deposits" in this type of banking the banks are not "making money".

Since depositors can not just take out their money on demand, there is no danger that too high a fraction would all want their money at the same time that the bank would have a crisis.   Avoiding the regular banking crisis of fractional reserve banking seems a big win.

It is also possible to loan out all of the money taken in and not just 90% of it.   So this type of banking could in some sense be more efficient.    If we are matching duration on deposits and loans then we will have real market pricing information on what interest rates should be at different durations.  In the current system with most deposits short term and loans long term we are not letting the market tell us what long term interest rates should be.

I also think is more honest.  If you tell all your depositors that they can take money out on demand, anytime they want, but really your bank is not operating in a way that it can do that then there is a bit of fraud going on.   Even if the central bank can often patch up this problem it is still not clean.   If I were in charge it would be illegal to do fractional reserve banking.

This type of bank still has to be careful to make good loans.  But that is all.   It does not have to get lucky about never getting a "bank run".  It does not have to get lucky about interest rates not going up.  It does not change the money supply, so does not contribute to the inflation and deflation problems.   Avoiding the regular banking crisis of the current system can not be stressed enough.

Criticisms of my View

There are those who say that currently banks are not limited in their ability to loan by their reserves.  This is certainly true for many banks.   Since Bernanke came up with his new trick of paying interest on excess reserves, reserves are acting very different. 

There are those that think banks make real money "out of thin air".   To me these people don't seem to understand the simple example above where the bank making a $9,000 loan increases measures of the money supply.   They think there is some other power to make money that banks have.  So far these people have not been able to explain with sufficient clarity any other type of "bank making money" than my example above.  If anyone can give a very simple example, like my $9,000 loan, where a bank has some different type of money making power, please do.  If in your example you can not take out cash, then your bank is not like a real bank.   Given that my example clearly works, it is not even clear why they think a bank needs any other power to make money.   Isn't the ability to grow the money supply enough of a power?

I can not find a single example where "bank made money" is a real factor in hyperinflation.   If private banks had some magical power to make money, and huge numbers of private banks, why is it that hyperinflation always involves central banks?   Also, to get to run central bank you have to be far more qualified than to run a regular bank.  Also, there are far fewer central banks.   So if it was just a matter of making mistakes, we should expect private banks to do it far more often due to both larger numbers and lower qualifications.  But it is always the central banks.  

Another criticism of my view on banking is about statements I have made similar to "the $2 trillion in excess reserves flooding out into the real economy and cause inflation".    Excess reserves is money that the bank holds above its reserve requirements.  It could withdraw this money from the Fed as paper money if it wanted.   The bank might loan the money out or maybe use it to for some other type of investment.   I think that the reason the Fed has been able to make so much new money without causing inflation is that it has kept $2 trillion from entering the real economy by paying banks interest on excess reserves.  I think that if the current $2 trillion in excess reserves comes into the real economy then inflation will kick up.   People who are not worried about inflation have arguments about how this won't happen.   But none of these arguments seem at all tight.   Banks can loan out money.  Banks can take out cash.  Banks could invest in other things.  

From the comments, the best example so far is a new bank is formed, it borrows $1 million at short term rates from the Fed and loans out $1 million long term.   This would work but since there  there is no "demand deposit account" the bank does not add anything to the money supply.  The Fed probably just made the $1 million, so it probably added to the money supply, but the bank did not.  Perhaps the Fed already had that money sitting around, either way, the private bank did not make money.  Note this bank still has the interest rate risk problem.  The Fed has the risk that this bank was formed by banksters who are just out to rip off the Fed.  This example can work, but it is not a "private bank making money out of thin air" at all.   Nothing close to a magic trick in what the bank is doing.  It is getting money from the Fed and loaning it out.  Nothing to see here, move along.

More from comments.  Imagine it is after closing time and the bank makes two accounts for the same customer, who has put up his house as collateral.  In one account there is a $100,000 loan and in another there is $100,000 deposit.   No physical money has changed hands yet.  The bank needs to use up $10,000 of its excess reserves but it has created a $100,000 demand deposit account.  You could even imagine that the bank charged a $10,000 origination fee and used that for the reserve requirements.     But at the start it sort of adds $100,000 to the money supply out of thin air.   However, as the customer spends the money then the bank would have to get real money from the Fed or someplace else.  Still, you can think of it as making the money and then getting the reserves later.  This is the new best example of "making money out of thin air".  :-)

More comments from Tom.  Thanks so much Tom!   He points out that there are a few really big banks with branches all over the place.  So there is a chance that when the customer who got the loan spends money out of his checking account that the person he gives the check to actually uses the same bank.  In this case the bank can do the transaction still without getting "real money".  

Good Articles On "Banks Making Money"

James Tobin, Commercial Banks as Creators of “Money”


  1. Vincent. Let me start off with the good parts: I like your description of a bank getting into trouble with interest rates. I don't see a problem there (with your description).

    OK, now onto the problems. Let me start with some simple examples. First let's assume there's a 10% reserve requirement ratio (RRR) for required reserves (RR). Now the RRR applies ONLY to certain kinds of deposits the bank holds (deposits are ALWAYS a band liability), and those are demand deposits. An example of a demand deposit is a checking account. They DO NOT apply to time deposits (e.g. savings account and CDs). RRR does not have ANYTHING to do with a bank making loans or any assets it may hold on its balance sheet (BS). Capital requirements fold in both the assets and liabilities a bank holds on its BS. Capital requirements are what limit bank lending.


    Example #1: RRR of 10% and NO capital requirements (for simplicity). Say our bank starts off with an empty balance sheet: no assets and no liabilities. Now an extremely credit worth customer comes in requesting a loan for $1M in cash for 1 year. The federal funds rate (FFR) is 0.25% (as it is now). If our bank thinks that the probability of the FFR increasing is very low over the next year, they will happily oblige him. Say they give him a rate of 5%. But where do they get the cash? Well they can borrow it from other banks or the Fed at the FFR. So they do so, and their balance sheet looks like this:

    Bank A:
    Assets: $1M loan at 5%
    Liabilities: $1M borrowing of reserves (traded to the Fed for cash) at 0.25%
    Equity: $0

    This is a great money making potential for the bank! It's getting a spread of 4.75% for one year on $1M. Now did they require collateral for that loan? Sure, probably.

    OK, where does the RRR come into play here? Nowhere! The money walked out the door as cash... the bank doesn't hold a **DEMAND** deposit on their books, so the RRR doesn't apply! No bank run. Nothing! The interbank market provides the reserves. The loan is collateralized. The capital requirements are 0%. The RRR is 10%, but there's no demand deposit (DD) here so it doesn't apply! Make sense?

    But where did the Fed get the $1M in cash? They effectively have an infinite supply of it. They buy it from the Bureau of Engraving and Printing (BEP) of Tsy at production cost... so it didn't cost them NEARLY as much as $1M. When they sent it out the door the Fed recorded the following on their balance sheet:

    Assets: $1M loan of reserves to Bank A
    Liabilities: $1M in cash
    Equity: $0

    So it wasn't until the the Fed sent the $1M in cash to the bank that it took on value. This is not surprising because it is essentially a Fed IOU. Do IOUs you write for other people have any value either before you hand them out or once they are returned to you? Absolutely not! You are free to tear them up w/o any loss to yourself.

    Does this make sense? If you have a problem with this, there's no point in continuing to the rest of your post.

    1. "You are free to tear them up w/o any loss to yourself."

      ... should probably replace "yourself" with "anyone."

    2. Tom, I think sometimes MMT and MR people get so lost in the detail they get confused about what is going on. Sort of a can't see the forest for the trees thing. You can go on and on for an idea that is just "a bank could borrow money from the Fed and loan it out". Try to just get to the heart of the idea so it is easier for me to follow.

    3. Vincent, that was my 1st example because I wanted to show that you were wrong about some of the early assertions you made:

      1. Bank run would result (it didn't)
      2. It's dangerous to loan out the same deposit multiple times via FRL (actually no deposit was ever required ahead of time here, so it's silly to talk about the dangers of loaning out this non-existent deposit "too many times.")
      3. Clearly the FRL model is NOT what's going on here. So if no FRL and $1M is loaned into existence, then what's going on?

      A fourth point is that:

      4. The private sector leads, and the Fed follows. That $1M created had nothing at all to do with any action initiated by the Fed or Tsy (QE or deficit spending). That's the typical pattern: private sector leads then the Fed follows along behind facilitating. Notice how the loan had nothing whatever to do with any pre-existing excess reserves (ERs).

      Join the rest of us Vincent! The FRL is dead. The money multiplier is inactive. We live in an age of pure fiat currency and those ideas don't apply any more.

    4. "Tom, I think sometimes MMT and MR people get so lost in the detail they get confused about what is going on."

      "Try to just get to the heart of the idea so it is easier for me to follow."

      Vincent, over in your FAQ section comments, you said my one page examples on my blog were too complex, so I thought I'd break it down and add some words to explain each step for you. Honestly you think this is too complex?:


      or this?:


      You wrote a considerable amount in this post... and it's getting longer!... it's going to take a few words to debunk each and every point that's wrong! There's very few parts that are right! ...and I already covered those in my first sentence on this page:


    5. Ok, I added another paragraph at the end of the post about making a loan and a deposit account at the same time. See if I am getting closer to understanding you with that. The way I think is in words and money flowing around and you are doing all kinds of debits and credits and it is just hard for me to understand. It seems like "borrow from the fed and loan it out" converted into a full page of math. I just don't think that is the simplest way to explain it, at least to someone with my way of thinking. Thanks for your time and patience.

    6. Vincent, sure no problem. Perhaps my way of explaining isn't the best for you. I was confused by the words (as I explain here):


      And that's why Scott Fullwiler's article was such an eye opener form me! (I could finally see, in concrete simple examples, what those words meant!):


      ... and probably an eye opener for Krugman (and other neo-Keynesians, neo-classicals, monetarists, and market monetarists) as well! You can find old posts of MMist Scott Sumner essentially arguing Krugman's pre-enlightenment position (the old "loanable funds" model that neo-classicals used to use) and now he's come full circle! Just today he wrote this:

      "And I certainly agree with your last point. 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." - S. Sumner

      That's miles from where he was just three years ago!

  2. Vincent, I know that 2010 Fed paper was a little long (58 pages), but it's worth reading I assure you. And I know my examples are a little complex to chew on right away. So let me recommend the following. Why don't you read this one article:


    Scott Fullwiler really knows this stuff. Cullen even agrees that's true, and you know that Cullen has a LOT of problems with MMT (Fullwiler is an MMTer). But this article is really just concerned with the basic fundamentals of banking. When I read it a light went off for me. All my blog posts incorporate what I learned from Scott in that article. I can't recommend it enough.

    Now Scott doesn't cover capital requirements in there. He mentions them, but doesn't cover them. That's where this next article comes into play by John Carney at CNBC. It's very good as well, and it's what I used as inspiration to build my three capital requirements examples from:


    Here again are my three examples of capital requirements and how they put a regulatory limit on bank lending:


    There's links to the other two inside that first one. I like to show everyone's balance sheets in my examples. You can imagine that the entities I show BSs for are the only entities in the world. They're like a miniature macro economy. But because I have both capital and reserve requirements in there and a deposit transfer (which is really the same as a purchase/sale but w/ one less person), it is a little complicated, so let me make it easy right here:

    Say we have 10% RRR and 10% capital requirements (this is expressed as a capital adequacy ratio (CAR) of 10% minimum).

    The CAR is expressed as follows:

    CAR = (Tier 1 + Tier 2 capital) / (sum of risk weighted assets)

    and CAR must be > or = 10%.

    I'll keep it simple and use an example with Tier 2 = $0 and Tier 1 capital = bank equity.

    Now the risk weighting in the denominator of the CAR ranges from 0 to 1 per asset. So a risky asset, like a car loan, might have the maximum risk weighting of 1. A risk free asset like Fed deposits or vault cash (the two components of "bank reserves") or Tsy debt would be weighted by 0. Something inbetween, like a mortgage might have a risk weighting of 0.5. Get it? The lower the risk of the assets, the smaller the denominator, and the higher the CAR and thus the better chance of meeting the 10% minimum threshold for the CAR. Get it?

    Now since I promised to keep it simple, and assume that Tier 1 + Tier 2 capital = Tier 1 capital = equity = assets - liabilities, we'll proceed.

    Say our bank starts off with an empty balance sheet:

    Bank A:
    Assets: $0
    Liabilities: $0
    Equity: $0

    Now say someone wants to buy a house for $95k and they want a mortgage. Can the bank lend them the money (assuming he's credit worth?) I think so. Here's how: They agree to loan the man $100k but take a $5k loan origination fee. Say they are required to assign a risk weighting to the mortgage of 0.5. Further assume that the house seller uses a different bank. Here's what the resulting balance sheet looks like for our bank:

    Bank A:
    Assets: $100k mortgage
    Liabilities: $95k borrowing of reserves (from Fed or banks)
    Equity: $5k

    So now let's apply the CAR formula:

    Tier 1 + Tier 2 capital = $5k
    Sum of risk weighted assets = $100k*0.5 = $50k

    Thus the CAR = $5k/$50k = 0.1 = 10%.

    1. (continued):

      Thus the bank is OK: it is meeting it's capital requirements. But what about its reserve requirements? Again, they don't end up with any DDs so they don't have any! the requirement is $0!

      Bank A is in good shape meeting both it's reserve and capital requirements.

      Now if Bank A was the only commercial bank in the system, they'd have to borrow the reserves from the Fed, offering the mortgage as collateral (if not, we know of at least ONE bank out there that has excess reserves they just received (via a Fed overdraft on Bank A's account) that they'd be willing to lend: namely the house seller's bank!... but they can't lend ALL their reserves because they have to keep some to meet THEIR non-zero reserve requirements (since they're holding the seller's deposit).

      Either case, Bank A should be able to borrow the reserves it needs to pay back the overdraft on it's empty reserve account when the house sale went through... and they should be able to do that by the end of the day! They'll have on their books the liability for this (the borrowing, for which they pay 0.25%), but they won't have any actual reserves in their account. Make sense?

    2. Here's a bit more information on Tier 1 and Tier 2 regulatory capital and how that relates to working capital, accounting capital and bank equity:


      Lots of good links in there too!

  3. Vincent, regarding your last paragraph about $2T. You say the bank might loan that money out. That's really a two step process: first someone takes out a loan and gets a bank deposit and then they are free to exchange their bank deposit for cash.

    Thus for every $ of cash that goes out of the bank as a withdrawn deposit, another $ of bank deposits are destroyed.

    The excess reserves (ERs) don't matter! All that matters is the bank deposits. Even if the bank had NO ERs, they could still provide the cash when someone wanted to make a withdrawal: they'd simply borrow reserves from other banks or the Fed (electronic) and then use those reserves to buy vault cash which they'd distribute. The presence of the ERs makes absolutely no difference! The only thing driving cash going out into the private economy are the loans (See my 1st comment above... the 1s comment you've got for this post) and the borrower's desire to hold these loans as cash.

    Regarding the bank using its ERs for "some other kind of investment." Sure... an individual bank can do that. But who are they going to buy this investment from? Who are they going to transfer those ERs too?

    Well, the way to conceive of this problem is to consider the banks in aggregate. As one giant consolidated bank with $2T of ERs. Now what seller can accept Fed deposits for payment? Other banks (whoops! there aren't any of those... that's just the same as passing the ER's around within the banking system!), or Tsy, or the Fed. Those are the choices! The aggregate banking system CANNOT use those ERs to buy anything from the non-bank private sector because the non-bank private sector does NOT have any Fed deposits!

    So then, what are Tsy and the Fed selling. Well if the Fed were selling (which it's not) then they could buy basically two things: Tsy debt and MBSs. If they did buy from the Fed what would happen to the Fed deposits? They'd cease to exist! The Fed would rip up its IOUs and that would be the end of that money.

    Now what about Tsy? What are they selling? They're selling just ONE thing: Tsy debt... and the banks are in luck because Tsy is actually selling some of that! But they'll have to be patient, because they're not putting up $2T in any one auction!

    That's it! Those are the only two avenues for "other investments" as you say.

    So now, how do the aggregated banks buy other stuff from the non-bank private economy (the economy that has no use whatsoever for Fed deposits)? Well, they can convert the $2T in electronic Fed deposits into cash by buying it from the Fed, and then they can go out and go crazy buying stuff, right?

    No, not really. They still have to meet their capital requirements! Remember those ERs were risk free!... if they are anywhere near the 10% CAR threshold, they'll need to buy other risk free assets from the public with their bag of cash. But what other risk free assets are there? Well, there's Tsy debt held by the public. That's about it!

    cont. below:

    1. cont from above:

      So, if that's all the banks can buy with cash from the non-bank public, how do they pay their electric bill? How do they buy paper for the office? How do they pay their employees and shareholders? And how do they make new loans to to public (i.e. buy loan agreements from borrowers).

      The answer is that they are chartered to credit bank deposits and bank deposits are defined by the Fed as "inside money" meaning the money is CREATED inside the private economy (as opposed to Fed created "outside" money created outside the private economy).

      Nobody else can create inside money except the banks. Is inside money real money? Absolutely! It's a medium of exchange just like cash. You can pay your taxes with it! It's denominated in dollars and is exchangeable for cash!

      So how do banks buy stuff with inside money? The exact same way the Fed buys stuff with outside money: They create it out of thin air! What's to prevent them from using this method to buy the whole world? Capital constrainst and CAMELS (in a regulatory sense) and just plain business sense (from a risk assessment perspective).

      If the banks in aggregate decided tomorrow to spend $10T on bags of dirt... I assure you there'd be hell to pay! The banks would all fail and the management would be imprisoned for fraud. All the banks would be make insolvent from such a bone headed move! ... as soon as those bags of dirt were "marked to market" and it was discovered that they had multiple trillions in negative equity!

      Banks buying stuff (including loans) by crediting bank deposits and banks having ERs have NOTHING to do with one another!

    2. One last thing, I'm anticipating an objection you're going to have: you're going to say that if the banks use their $2T to every so slowly buy Tsy debt at Tsy auctions (directly from Tsy) then despite the Tsy taking years to put up that much debt for auction, at least Tsy will have it and can spend it into the real economy (bit by bit, as the auctions occur). You are ABSOLUTELY correct about that!... Let's assume that Tsy spends whatever $ is raises immediately on the non-bank private economy: buying ships and fighter planes, etc.

      Now how does Tsy get the funds to the non-bank private sector? Well suppose Tsy sends you a check for services you've done for the gov. Then when you deposit that check two things happen: The bank credits your demand deposit (and now you're free to withdraw that as cash!) and the Tsy sends part of their Fed deposit balance to the banks Fed deposit (actually this happens inside the Fed computer: they debit Tsy's account (TGA) and they credit the bank's).

      So essentially, the banks in aggregate can't get rid of their Fed deposits buy buying Tsy debt from Tsy if Tsy spends the proceeds! They only get rid of it temporarily. At least if they buy stuff from the Fed those deposits are gone for good!

      Now, this process did create bank deposits though, right? Sure... either that or cash. But we've been through all this. I drew out the balance sheets for you once before showing how they would look given $T in total Tsy debt sold and $C withdrawn in cash by the public:


      Now your response was "what about bank loans? Those convert ERs to RRs."

      And my response to that was "sure, but reserve levels didn't change. And besides, bank lending has no relationship to the existence of ERs: even if ER = 0, the Fed is obligated to provide the RRs in the process of defending the FFR and ensuring the smooth running of the interbank payment clearing system. Plus, RRs will go right back to being ERs if loans are paid down."

      For a fuller formulaic treatment of what the various player's BSs will look like given that the Tsy spends every $ it gets immediately, and folding in bank loans, and Tsy debt owned by the Fed, the banks, and the non-banks, I worked that out here:


      There are three sets of BSs there: the 1st one assumes no cash, but the other two do allow cash: a small amount and a large amount. You'll see what I mean when you take a look.

      You're welcome. ;)

    3. You and Greg and others keep saying that private banks make money out of thin air. I gave a simple example where they increase the money supply by making a loan using money from a demand deposit. You gave an example where a bank loaned out money it borrowed from the Fed, but that is the closest we have to a private bank "making money". Please try to make another simple example where it "makes money out of thin air" in such a way that someone could actually pay their electric bill or something with the money.

    4. "Thus for every $ of cash that goes out of the bank as a withdrawn deposit, another $ of bank deposits are destroyed."

      But as the bank wants to keep a certain amount of cash on hand, if it has given out cash to a customer it will get more from the Fed. Or for real simplicity assume it does not keep any on hand and gets cash delivered as the customer is waiting. :-) As the bank gets cash from the Fed its excess reserves go down.

    5. The bank does NOT want to keep any cash on hand. It doesn't get paid any IOR for cash, so why not trade in any unused cash for IOR paying Fed deposits. Plus, Fed deposits don't require expensive vaults or guards, etc. Even if IOR = 0% the bank would want to minimize risk and expense, and maximize convenience (no need to sheep Fed deposits!) and electronic Fed deposits beat paper reserve notes on every score. They are there PURELY for the convenience of the banks' customers.

    6. should read "no need to ship (e.g. via truck) Fed deposits"

    7. Ok, so lets do a toy example. A bank with no cash on hand and a customer comes in and wants $10,000 cash. Logically it would then take out some of its excess reserves as cash, so it could give its customer some cash, right? Remember, the key to business is to take good care of your customers so you get more customers. Any real bank needs to keep enough cash on hand so that customers don't usually have to wait for an armoured truck to come. In my experience they do keep some cash on hand.

    8. Vincent, I didn't say they didn't keep cash on hand. They have to in order to please customers, as you say. But they want to minimize it as much as possible for all kinds of reasons that I already covered (also robbery and theft). Believe me, if technology advances more and there are STILL people using cash, I'll bet you that it's quite possible banks will be able to print their own again, using some kind of secure network. Hell, maybe INDIVIDUALS will be allowed to print their own (after they trade in their electronic bank deposit in exchange). That would solve the bank's problems with warehousing cash. The banks would LOVE it, and the Fed exists primarily to serve the interests of the Fed, so it could happen! We're already to the point where we can deposit checks ourselves w/o an ATM right on our cell phone.

      Someday in the not too distant future we may see tons of abandoned, boarded up ATM machines littering the landscape, in much the same way that pay telephone booths do now.

    9. But today, either a bank gets more cash when it gives some out, reducing their excess reserves, or they take a chance of not having enough cash when customers come in. So in practice they have to, on average, get cash from the Fed as they give out cash. So it is reasonable to think of a bank giving out cash as getting it from their excess reserves. Right?

    10. We can think of their "net cash outflow" resulting in withdrawls from their excess reserves, as clearly there can be some customers depositing as well.

    11. When a customer withdraws a $ of cash from their bank deposit and the bank had ER > 0, then yes, the ER stock goes down a $, all else being equal. That's what I was attempting to show here:


      But better than that, I've got a new post up explaining this and more:


      I've got it all laid out in a simple formula in each cell of each balance sheet in my simplified world. The major assumptions I'm making are no foreign contibution, no GSEs, and Tsy deficit spends. That last one is hardly controversial! It directly addresses your comment about private lending (except that RRR is 0% in this example).

      Plug in numbers for yourself and try it out. (unfortunately there are two separate cases, depending on how much cash is in circulation, but the Tsy's and public's balance sheets don't change between the cases!)

      So again, try it out! Bump "C" up by a dollar (the cash in circulation) and watch what happens: The money held by the public doesn't change and:

      Case 1: ER> 0:
      reserves go down at the bank (no ERs: RRR = 0%)

      Case 2: ER = 0:
      reserve borrowing goes up at the bank!

      That's it!

      I'll leave it as an excercise to add in a 10% RRR. It won't change anything as far as the public is concerned, and very little else either. And BTW, it doesn't matter what order of operations were performed to get you to that point (with everyone starting with blank balance sheets): Tsy auctions, Fed, banks, and public buy Tsy-debt. Fed, banks, and public buy & sell Tsy-debt to each other. Bank makes loans. public repays loans. Whatever! It doesn't matter, you end up at this same place.

      The main take away is that for the public to gain a $ of equity, Tsy has to auction a $ of debt.

      Now that's an oversimplified way of looking at savings and investment in the private sector. That's the whole point of Cullen's recent post here:


      What I'm looking at is essentially NET savings... and I'm looking at STOCKS not flows... (although you could argue it's flows since I'm starting from everyone having a clear balance sheet).

      Find all the comments in there by FlimFlam and Rafael. They are excellent: FlimFlam is an accountant and teaches me a thing or two, and Rafael is trying to learn, but he catches on fast!

      Here's Rafael summing up what he's learned and my response:


      But there are multiple threads in there with him that make good reading!

    12. Shoot, I made an error there:

      "reserves go down at the bank (no ERs: RRR = 0%)"

      Get rid of the "no ERs" part. Case 1 is ER > 0! and the "RRR = 0%" part applies to both cases, not just Case 1.

  4. Vincent, I posted a link to something like this before, but here's a version I made into a real post:


    It shows all possible balance sheets in a simplified world in which Tsy always spends all its proceeds.


  5. Tom, thanks for the "bank borrows from the Fed and makes a loan" example. I have added another paragraph to my post. I don't see this bank adding anything to the money supply or in any sense "making money" as there is no "demand deposit account" to make for double counting of money. This is not "a private bank making money out of thin air" as only the Fed does that in this example. I am looking for an example where a private bank makes money in any way different from my simple example.

    1. You don't see it because you don't see the two step process that I glossed over in that example:

      1 Loan and a deposit (inside money) created
      2 deposit (inside money) exchanged for outside money (cash)

      Notice this about that example:

      1 No bank run
      2 No concept that someone's pre-existing deposit was "dangerously" loanded out multiple times (there was no pre-existing deposit)
      3 Private sector leads and the Fed follows (nothing to do with QE or deficit spending: all activity here, including the creation of money, originated in the private sector)

      You want a different example? I've already started one above (the second example I gave you with the house and the capital constraints). Let's just modify that slightly:

      Reserve requirements ratio (RRR) = 0.1 = 10%
      Capital requirements ratio (CRR) = 0.1 = 10%

      A: $0, L: $0, E: $0

      Banks A:
      A: $0, L: $0, E: $0

      Person x:
      A: $0, L: $0, E: $0

      Person y:
      A: $95k house, L: $0, E: $95k

      x gets a $0 down mortgage to buy y's house, but with a $5k loan initiation fee. CAR risk weighting on house is 0.5. x & y bank at A:

      A: $9.5k loan of reserves, L: $9.5k reserves, E: $0

      Banks A:
      A: $100k mortgage to x
      A: $9.5k reserves
      Total A: $109.5k
      L: $95k deposit for y
      L: $9.5k borrowing of reserves
      Total L: $104.5k
      E: $5k

      Person x:
      A: $95k house, L: $100k mortgage, NE: $5

      Person y:
      A: $95k deposit, L: $0, E: $95k

      Now say y pays x $50k for some work. The Fed's balance sheet doesn't change (no more need for outside money). Here's the others:

      Banks A:
      A: $100k mortgage to x
      A: $9.5k reserves
      Total A: $109.5k
      L: $45k deposit for y
      L: $50k deposit for x
      L: $9.5k borrowing of reserves
      Total L: $104.5k
      E: $5k

      Person x:
      A: $95k house
      A: $50k deposit
      Total A: $145k
      L: $100k mortgage
      E: $45k

      Person y:
      A: $45k deposit, L: $0, E: $45k

      Now y pays x. And x pays y, etc etc.

      The Fed's BS doesn't change. All payments are accomplished by the bank debiting x's deposit and crediting y's or vica versa. The Fed and outside money are no longer needed for this two person economy to continue indefinitely.

      The Fed created $9.5k of outside money (reserves) so Bank A could meet it's RRR. Bank A's retained earning of $5k means it meets it's CRR (with weight 0.5 on the mortgage)

      So who created the rest of the money in this miniature macro economy? Bank A! Bank A created $95k of inside money that is actually used as a medium of exchange here. Meanwhile the $9.5k of outside money does NOT function as a medium of exchange in the economy. It functions to allow Bank A to meet its RRR and nothing else.

    2. The $95k created by Bank A also serves as a store of value! Can the same be said for the $9.5k created by the bank?

      Now the Fed changes it's rules and drops the RRR to 0%

      Now no reserves are needed at all for this economy to buzzz along happily. Add persons a and b who also trade a house. Add whatever you want, all these happy Bank A customers can live their whole lives, w/o ever needed a single $ of outside money. The Fed never has to create a dollar of outside money.

      What if we add a different bank? Now the Fed needs create outside money temporarily every day to clear payments and facilitate balance transfers between banks, but each and every night it can turn out the lights w/o a single $ of outstanding Fed deposits (Oh, excuse me, BoC deposits) existing. They are created loaned out during the day and the are all repaid and thus destroyed each night. No permanent BoC outside dollars.

    3. Oh, shoot, ignore the BoC bit: I thought we'd moved to Canada in there somewhere (where RRRs = 0%). I forgot that the Fed (in my example) had dropped the RRR to 0%.

    4. Actually rather than move to Canada or have the Fed set RRR = 0%, we could instead just have person y transfer her deposit to a savings account:

      A: $0, L: $0, E: $0

      Banks A:
      A: $100k mortgage to x
      L: $95k savings deposit for y
      E: $5k

      Person x:
      A: $95k house, L: $100k mortgage, NE: $5

      Person y:
      A: $95k savings deposit, L: $0, E: $95k

      Either that happened right away, or later and then Bank A repaid the $9.5k loan: either way, it doesn't matter, we end up in the same place.

      Is there a problem on any of my balance sheets Vincent?

    5. I really have trouble following all your accounting entries. If a bank makes money out of thin air and a customer writes a check on this money to his plumber, and the plumber goes to his own banks and cashes the check, how is it that the plumber's bank is ok with this. The plumber's bank has to get money on the inter-bank network, right? If I think of the borrower's bank transfering some of its reserves to the plumbers bank I am about right?

      I will agree that a bank can make a loan and a deposit at the same moment and nothing is trouble as long as the customer leaves the deposit unused. I will add this to my post. I just think that if the money is going to be used then the magic ends. If the customer wants to spend that money or transfer it to another bank then this "out of thin air" trick does not work. But I do agree that it is possible to make the loan and deposit and worry about reserves later.

    6. "I really have trouble following all your accounting entries."

      Vincent, I think that's the heart of the problem here. I'm NOT an accountant. I learned double entry accounting (enough to get by here!) from Scott Fullwiler's article (read it!... it's written for the laymen and it's a LOT shorter that 58 pages) and from asking others, and even looking at a few simple videos on youtube. It's not that complicated once you get the hang of it: the Fed uses it in its publications, so it's valuable for following along. Also I have come to firmly believe that understanding double entry accounting basics is ESSENTIAL to understanding what money is and how our monetary system works! I'm MORE than happy to answer any of your questions about it to the best of my ability!

      "If a bank makes money out of thin air and a customer writes a check on this money to his plumber, and the plumber goes to his own banks and cashes the check, how is it that the plumber's bank is ok with this. The plumber's bank has to get money on the inter-bank network, right? If I think of the borrower's bank transfering some of its reserves to the plumbers bank I am about right?"

      Yes, your are right, but the borrowers banks doesn't have to have the reserves to start out with: they can borrow them or default borrow them by letting their Fed deposit overdraft: as long as they pay the Fed back by the end of the day (by borrowing from someone else) there's no problem with the overdraft. That's EXACTLY what this example illustrates:


      "I will agree that a bank can make a loan and a deposit at the same moment and nothing is trouble as long as the customer leaves the deposit unused. I will add this to my post. I just think that if the money is going to be used then the magic ends. If the customer wants to spend that money or transfer it to another bank then this "out of thin air" trick does not work. But I do agree that it is possible to make the loan and deposit and worry about reserves later."

      OK, good, I think we're making progress. Take a look at the example #1.1 I post above: it's one of my simplist ones and it directly addresses your concern here.

      Then go one step further and imagine that Bank A and Bank B are really the same bank (they're owned by a parent company, for example). Then you can consolidate their balance sheets, and Shazam! NO reserves required from the CB AT ALL even on a temporary (overdraft) basis. What I'm proposing here is NOT that far fetched! There are only five major banks in the US. Just pulling a number out of the air, that means there's maybe a 20% chance this is EXACTLY what happens you pay your plumber!

      But the broader point is it doesn't matter. The banks can logically all be aggregated together: the flow of reserves on the inter-bank market is not all lat interesting when it comes down to it! It causes more confusion than anything else, which is why for macro, assuming a single commercial bank is a valuable tool. That's what they do here on econoviz:


      Try it out for yourself! It's animated and interactive: the operations you can do are on a pull down list in the bottom left corner.

    7. And BTW, the whole house buying example I present here (in two different examples) I cover in the first two sets of balance sheets in this example:


      Just the first two! (1. and 2.). Don't worry about the rest right now.

      In this case buyer and seller both bank at the same bank (JPM for example). I'll leave it as an exercise to draw out the balance sheets for a two bank setup (like my Example 1.1) and with the same assumptions (RRR = CRR = 0%, etc.)

      For writing out comments section balance sheets, I try to save space:

      A: = assets
      L: = liabilities
      E: = positive equitey (i.e. A - L > 0)
      NE: = negative equity (represented w/ a positive number)

      ... NE: = L - A

      Why not just put a minus sign on E and still call it "E"? You could. It's a choice. I'm being consistent w/ econviz here. They put NE on the left in the Dr column (debit column) and E on the right in the Cr column (credit column). E is NOT a liability (in the usual sense) but it needs to go on the right to balance the balance sheet. Similar argument for NE.

  6. Tom, thanks for asking Cullen my question. Did you notice that he did not say what could falsify his theory?


  7. OK Vincent, you ask me to look at your last paragraph in the post:

    "Imagine it is after closing time and the bank makes two accounts for the same customer, who has put up his house as collateral. In one account there is a $100,000 loan and in another there is $100,000 deposit. No physical money has changed hands yet."

    OK, I'm with you so far. It' unusual to me that you describe this as "two accounts" "for the customer" but I'll go with it. The loan is more "for the bank." I see what you're saying thoug, so I'll go with it.

    "The bank needs to use up $10,000 of its excess reserves but it has created a $100,000 demand deposit account."

    perhaps replace "but" with "because here?" You lose me a bit there.

    "You could even imagine that the bank charged a $10,000 origination fee and used that for the reserve requirements."

    Actually that would be used for the capital requirements. How does this work? You take a loan out for $100k but the bank only credit's you with $90k! Do you see... you didn't actually bring any money (reserves) to the bank! You "paid" by letting them destroy your money! This does nothing for the bank's reserve requirments... for that it actually has to hold "reserves" as an ASSET!... Thus it can borrow them, thus incurring another liability for itself (the borrowing or reserves)... this nets to zero extra capital, which is why meeting the reserve requirements does NOT help meeting the capital requirements.

    "But at the start it sort of adds $100,000 to the money supply out of thin air. However, as the customer spends the money then the bank would have to get real money from the Fed or someplace else."

    OK.. well maybe you've only got $90K, but close enough. And remember this is ONLY true if you're spending on someone who has a bank deposit at another bank! If they recipient of your funds banks at your bank, this is handled purely by a manipulation of electronic records at the bank and no reserves are required to do this!

    "Still, you can think of it as making the money and then getting the reserves later. This is the new best example of "making money out of thin air". :-)"

    OK, we're getting closer!

    1. BTW, even with two banks, as you've laid out: Bank A overdrafts their Fed account (because they don't have enough reserves), and the Fed credits Bank B anyway. Now Bank A is on the hook to come up w/ reserves to repay the Fed by the end of the day: Guess who's got excess reserves they're willing to lend now Bank B! The Fed just credited them! So they are happy to loan Bank A the reserves to repay the Fed with. Now Bank A has borrowings of reserves from B as a liability on its balance sheet, and Bank B has a loan to Bank A of reserves as an asset on its balance sheet. EXACTLY as I've laid out in this example:


      Is it always the case that A borrows from B like this? No... they may have plenty of ERs as you note and then not overdraft happens. Or they may borrow a little form B and some from C and more from D, etc. The rate will be the same everywhere because the Fed has set the rate and if you charge too much the Banks can simply borrow from the Fed! They will win the battle everytime. That's what it means for the Fed to defend the overnight rate they set! (I've oversimplified here, but that's the basic idea)

  8. Another convert to right minded thinking!:



    Now if only Schiff would read that. But what did Upton Sinclair say once?:

    "It is difficult to get a man to understand something, when his salary depends upon his not understanding it!"

    I sincerely believe that Schiff NEEDS to not understand so he can sound believable when he's talking with great confidence right into the cameras and trying to scare people into buying into his funds.

    I KNOW that's not you Vincent, since this is just a hobby to you. So I'm not trying to insult you, but I'm very skeptical of Schiff and his motivations.

  9. Vincent, I've been spending WAY too much time on this and have to turn my attention to other things for a bit! Looking through some of your comments and additions to your post, I think that I've got written down either here, in your FAQ or in your hyperinflation debate page an answer for every one of them. I'm not sure you make it though all the comments I wrote: I know there were a LOT of them. I urge you to read Fullwiler, pick over all my comments to see if there's something in there that helps, and don't hesitate to go to pragcap.com or monetaryrealism.com and ASK questions. There are plenty of nice people there that are willing to explain. Many of them are other commenters, such as Greg or Geoff or LVG or SS. If you find "FlimFlam" or "Joe in Accounting" or Frances Coppola there you are in for a special treat! (Frances has her own blog: coppolacomment): Those guys actually know something about the banking and accounting worlds. Joe is a bank auditor and I especially like him: He's often can fill in pieces of the puzzle that Cullen or others can't.

    And... of course there's my blog! Try going through it in reverse cronological order: it kinds of builds from simple to complex in general.

    OK, I wish you the best of luck: I posted a couple of links to our convo here on pragcap, so perhaps that will steer some traffic here. I'm confident that I can convince you about this banking stuff... and who knows you may convince me of a thing or two: I have no idea why you think that V goes up when interest rates go up (for example).. did I get that right? I'd like to see the argument for that... you didn't just restrict it to hyperinflation conditions as I recall... you said it was generally true.

    OK, like I say I'd LOVE to keep this going, and I will return someday, but I don't think you're going to see much of me here or on pragcap or on my own blog for a bit: need to get some other stuff done!

    All the best Vincent, and I hope you find the answers to the question your are looking for about this banking and inside money vs outside money stuff! I'll be back someday to see what's up!

    1. PS: Work your way up to my last post here, Example #11:


      See if you can punch a whole in it (given the assumptions I have made). See if there's a problem. I'd like to know about it if there is!

      I thought I'd mention it, because these conversations with you partly inspired it! That's "your post" Vincent! I hope it's of use to you. ;^)

    2. Thanks so much! You have written and linked to much more than I have digested. It is as if you are writing faster than I am reading. :-) So it is good that I will have some time.

      Thanks again.

      In my FAQ there is a link to an article by Hussman where he has data showing that the velocity of money is higher with higher interest rates. To me it makes perfect sense. If the time value of money is higher (interest) you are going to move it along faster. If it is dropping in value faster (inflation) you are not going to want to hold it so long.

    3. Vincent, Thanks!: I need help! really, I just can't keep away... I really LOVE introducing people to this stuff. I'll put looking at Hussman on my todo. But it just occurred to me how to use this:


      to demonstrate that banks don't lend out reserves. First of all, if there are excess reserves > 0, then that's Case 1. Now look what happens if we move (independently) either of the two variables contributing to bank reserves: F and C (bank reserves = F - C)

      If F goes down, then the public's money (bank deposits + cash) goes down, all else being equal.

      If C goes up, then the public's money doesn't change.

      So if F goes down AND C goes up, then the public's money doesn't change.

      Make sense! Now you just have to show that those balance sheet formulas are incorrect (which I don't think you can do)!


  10. Bloody hell there's no way I can read through all those comments.

    First off, this is quite a good, clear simple explanation of how banks can become insolvent, from Positive Money:


    1. Ya. Tom can take a simple idea that should be explained in one sentence and make it so complicated I think very few people could read all the way through.

      That URL does a fine job of explaining how banks become insolvent. From the first sentence, "If banks can create money, then how do they become insolvent?" it seemed like he was going to address money creation. But really he does not talk about that.


    2. It's 'money creation' because we use bank deposits as a form of money. But those deposits are just bank debts, which are currently guaranteed by the state and supported by the central bank as you say.

  11. by the way, I'm not sure hyperinflation is really 'how fiat dies'. There have been quite a few hyperinflations around the world but fiat still reigns supreme, even in those countries that have been worst hit. Often a hyperinflation comes abruptly to an end when government changes its policy or introduces a new currency. And if you hold your savings in a form which appreciates in value as the currency hyperinflates you can potentially come out on top.

    Personally I wouldn't want to be anywhere near a hyperinflation but I doubt it's the apocalyptic end-of-the world scenario some people seem to imagine. Almost all prices go up at the same time so society keeps functioning.

    1. I just mean a particular fiat currency dies, not that fiat currencies in general die because of one hyperinflation. If they "introduce a new currency" I count the old one as dead.

      Yes, if you are holding gold through a hyperinflation and then buy land after everyone else is broke you can do well.

      Yes, it is amazing that most societies keep functioning rather well even with prices doubling every few days. Crime probably goes up a bit, but it is far from end-of-the world type thing, at least normal ones.

    2. Maybe your blog should be called "how fiat dies, and lives to fight another day".

      Fiat money is like Frankenstein's monster, you can't kill it. Not for good, anyhow. It just keeps coming back.

    3. If you only look at the last 40 years you might think you can't really kill fiat money. But if you go further back there are cases where fiat money was replaced in a country that "returned to a gold standard". They did not always return in a good way, and often it was just temporary. But it has happened.

    4. "there are cases where fiat money was replaced in a country that "returned to a gold standard"

      And then what happened?

      Robert Mugabe was musing a while ago that maybe Zimbabwe should go adopt a gold standard, so you never know. He's crazy enough after all.

    5. For example, both the North and the South used fiat money during the civil war. After the war they returned to a gold standard.

      During WWI most of Europe went to Fiat money. After most returned to a gold standard. Some tried to do it at the pre-war level but given how much money there really was after the war that was a bad idea. So there were some problems.

  12. "customer A comes in and deposits $10,000 cash in a new savings account. Next the bank loans out $9,000 to customer B who takes his borrowed money out as cash.

    This loan increases any measure of the money supply that includes both "demand deposit accounts at banks" and cash. That $9,000 is really counted twice. Since both customer A and customer B think they can spend this money, economists have decided it is reasonable to count it twice."

    Ok, there's a basic logical problem here. If customer A is depositing the money in a savings account, why does he suddenly want to spend his savings?

    1. there's a difference between savings accounts and demand deposit accounts.

      This FRED graph shows M2, M1, and MB.

    2. We don't really know if A will suddenly want to withdraw his money or not. There is some risk that he and other depositors want their money at the same time. Often things work fine for decades, then fail.

    3. "Definition of 'Demand Deposit'
      Funds held in an account from which deposited funds can be withdrawn at any time without any advance notice to the depository institution. Demand deposits can be "demanded" by an account holder at any time. Many checking and savings accounts today are demand deposits and are accessible by the account holder through a variety of banking options, including teller, ATM and online banking. In contrast, a term deposit is a type of account which cannot be accessed for a predetermined period (typically the loan's term). "


    4. Why does A suddenly want his money? Maybe his car needs to have the transmission rebuilt. Maybe his house was robbed and he needs to buy some new stuff. Maybe he suddenly got sick and his part time job does not have sick leave. The theory of money does not depend on knowing why A wants his money, only that there is some chance he may want it and that he has the right to get it on demand.

    5. phil, alright! ... I'm glad you came to pay Vincent a visit! ;^)

    6. The point is that the example is very unrealistic. Banks don't start out with no money. They need capital, lots of it, before they can even begin taking deposits. Then they need a base of savings deposits that can't be withdrawn immediately. Demand deposits can be withdrawn immediately, but normally only up to a certain amount. Beyond that amount, banks normally require notice. If they are short this gives them the time to sell some assets or borrow cash.

    7. If a bank sells a bond and then makes a loan, so the customer can not just demand their money back without notice, then the bank has not added to the money supply.

      The point of the tiny example is to illustrate how banks add to the money supply, not to be realistic. Also to show how there is a problem with lending long using demand deposits. It is a real problem.

  13. Vincent, there's been a lot written on those Tobin papers recently (check out all the links at the top of the 1st one)




  14. Vincent, here's a straight up neo-classical view (from an MMist, Nick Rowe) of banking. I recommend it!:


    Several of the processes he describes there I try to illustrate in my banking balance sheet examples. See... it wasn't an MR/MMT/PKE thing after all!... *normal* people believe it too. :D


Looking for polite debate on ideas. Never attack a person. Be nice.