Monetary Realism

Understanding The Modern Monetary System…

Banks Are Not Mystical

The recent crisis has been beneficial in at least one way – it has begun to shed light on some of the myths of our monetary system that have poisoned economics and politics for many decades.  If I had to rank some of these myths I would almost certainly put the currency user vs. currency issuer myth at the top of the list.  But a close second is the myth of the money multiplier.  Students are generally taught that our banking system works through some sort of “loanable funds” market or “money multiplier” whereby banks obtain deposits so they can then loan them out.  There’s just one problem with these ideas – they’re not right.

These ideas have all come to a head in recent weeks when Paul Krugman and Steve Keen got into a bit of a back and forth about the operational realities of the banking system.  I won’t comment specifically on the ideas of either men, both of whom are fantastic economists, but I think this conversation exposes the degree to which most people continue to misinterpret modern banking and requires some brief discussion.

The standard banking model says that banks are reserve constrained and that the amount of loans a given bank can make is a multiple of its reserves.   But the recent crisis has shot king sized holes in this myth.  The Fed has substantially expanded the amount of reserves in the banking system, but lending has flat-lined:

This is a monumentally important chart so it’s important to understand a few points if you’re going to understand why the above chart looks the way it does:

  • Reserve balances are determined by the Federal Reserve who acts as the supplier of reserves to the banking system.  Banks can never “get rid” of reserves in the aggregate.  They can shuffle them among each other, but only the Fed can destroy or create reserves through open market operations.   The Fed oversees the payments system and in doing so must act to ensure that banks can obtain reserves in order to settle payments and meet reserve requirements as needed.
  • Bank lending is not reserve constrained (in fact, many countries don’t even have reserve requirements at all).  This means that banks do not need reserves before they make loans.  Instead, banks make loans first and obtain reserves in the overnight market (from other banks) or from the Fed after the fact (if needed).   New loans result in a newly created deposit in the banking system.
  • Banks are capital constrained.  Banks can always find reserves from the central bank so banks do not check reserve balances before making loans.  Instead, they will check the creditworthiness of the borrower and their own capital position to ensure that the loan is consistent with the goal of their business – earning a profit on the spread between their assets and liabilities.
  • Banks attract deposits because they want to maintain the cheapest liabilities possible in order to maximize this spread on assets and liabilities.   Banks are, after all, in the business of making a profit!
That pretty much sums up the above chart.  You don’t need to understand balance sheet recessions or liquidity traps to know what’s going on there.  You just need to understand how banking works.   Yes, it’s true that the balance sheet recession has been a truly unique period in American history.  But the above chart is only unique in that it exposed this great myth to the public.   When the demand for credit collapsed the Fed was nearly helpless in reviving the credit markets.   Despite a $1.6 trillion reserve injection the lending markets just didn’t budge.  This might have appeared like an anomaly to some, but to those who understood banking this made perfect sense.  More reserves were never going to result in more loans.  This was not because it had temporarily become true, but because this is the way banking works.  Not just inside a balance sheet recession or liquidity trap or whatever you want to call it, but always….


Mr. Roche is the Founder of Orcam Financial Group, LLC. Orcam is a financial services firm offering asset management, private advisory, institutional consulting and educational services. He is also the author of Pragmatic Capitalism: What Every Investor Needs to Understand About Money and Finance and Understanding the Modern Monetary System.

View all posts by

130 Responses

  1. Detroit Dan says

    I know Krugman has a good reputation, but in economics it seems that reputation is not a guarantee of brilliance…

  2. John Carney says

    This is a very important point.

    I really do think the MMTers didn’t handle this very well. I tried to give some of them advice on how to proceed rhetorically. Seems to have fallen on deaf ears.

  3. Fed Up says

    I’d like some comments on this “gem” from nick rowe.

    anon: yes, there are legal capital requirements. I don’t see how they create a multiplier. Central banks don’t normally (to my knowledge) use changes in capital ratios as an instrument of monetary policy (maybe China??). It’s more about financial stability, and so the government doesn’t need to bail out bust banks.

    Posted by: Nick Rowe | April 03, 2012 at 04:27 PM

    Does that mean he thinks there is no such thing as a capital multiplier?

  4. Cullen Roche says

    Agreed. That’s why MMR is trying to remove politics. We’re trying to make this as close to scientific as possible. If we can hammer out how the system actually works and offer people a better understanding then they can’t argue with certain conclusions on political grounds. That’s our approach at least. Clearly, it’s not an easy one to take, but it’s a fight worth fighting.

  5. Nell says

    I appreciate your position and it is off putting when an important point of debate descends into a playground fight. However, I think you are mistaken in the believe that this is an argument that can be won by logic. I think that mainstream economists, particularly well-known and highly regarded mainstream economists behave more like religious believers than scientists. There is no logic that can convert them, because to convert they have to renounce not only their models, but their position in academia and the business world. Yes we should behave politely, but not to convince economists like Krugman, but to avoid alienating those who are learning from the debates. For example, I found SFW’s post most illuminating. As I did your post above. The graph is very helpful.

  6. Michael Sankowski says

    This gets back to the fixed/floating fx through the management of the fixed fx rate.

    In a fixed fx regime, the central bank commits to defending an fx level. This means the cb will expand or contract potential reserves to defend this target. It must do this to defend the fx target.

    In a floating fx world, the central bank cares about defending a short term interest rate. To do this they must make reserves available when requested otherwise the short term interest rate would rise above their target.

    I think this is what Mosler means when he refers to hard endogenity. If you fix one, you need to let the other float. You can’t defend both at the same time.

  7. Tom Hickey says

    “To clarify, what your saying is that while the fed isn’t going to manage whether any one bank has enough reserves, they’re always going to make sure the system has enough as a whole to not be limited by the RR.”

    The Fed has to insure that every bank meets its need to settle and for RR. This is the lender of last resort function. If any bank comes up short at the close of period, the Fed provides the needed reserves at the penalty rate. Banks are aware of this and take steps to manage liquidity and RR so that they don’t have to pay the penalty, which cuts into profits.

    The Fed manages reserve quantity so as to hit the target rate with enough reserves in the system to settle and for RR. Banks should therefore be able to get needed reserves by borrowing from other banks in the interbank market at the rate the Fed sets. Or they can borrow from the Fed by putting up collateral. Generally, banks don’t come up short and get stuck with a penalty. They have a department dedicated to this. As Scott Fullwiler has said, all this involves minute by minute tactics to make sure that everything balances at the desired rate.

  8. Ramanan says


    I think Keen was quite right in taking Krugman to task and Krugman responded by saying things which proved that Keen was right about Krugman and his view of money exogeneity!

  9. Cullen Roche says

    Krugman isn’t pointing out how to fix the train. He’s describing how it works (incorrectly). Ray Dalio uses the machine analogy often. Not because we need to know how to fix the machine in case it breaks down, but because we need to understand how it operates so we will know how it navigates certain environments.

  10. vimothy says


    I’m not sure how I feel about that post–I’d have to go back and read up on everything that has gone before it and I don’t have time to do that. It might well be the case that Krugman is overreaching; alternatively, he might have a reasonable point that his opponents don’t see because they don’t understand where he’s coming from.

  11. Ramanan says


    True a train mechanic may not be up to date with what is happening with cars but if he understands it by Aristotelian principles then he neither knows how to fix the car nor the train.

    Worse is if he writes a blog on a teaching moment on how cars work!

    Sorry don’t follow your part regarding institutional features.

  12. wh10 says

    Yeah I’ll admit I’ve become a bit flustered.

  13. vimothy says

    wh10, Not sure. I’d guess that I agree with you guys on some stuff and with Nick on some other stuff. Haven’t followed this too closely though.

  14. wh10 says

    Vimothy, you aren’t on Nick Rowe’s side of the debate here?

  15. Jussi says

    I just wanna support Cullen’s approach. What is more important here, trying to push PK (which will prolly not happen / is not important as such) to admit that he was wrong or leverage the MMT/MMR approach?

    Overall I just like the way Cullen try to avoid ego clashes and push the well thought agenda. Thanks.

  16. vimothy says

    R, A mechanic who works on trains might not know how to repair a car, though. Why should he?

    One of the things that I haven’t seen in this debate is an argument for why Eggertsson & Krugman 2011 should have a different model of banking, and what that model should look like. Instead the arguments have centred around what specific institutional features actually are in reality, without tying those institutional features into any economic theory. I haven’t followed particularly closely, so could you perhaps point me at a blog where someone lays this out? Cheers.

  17. Cullen Roche says

    Banks make loans based on how profitable it is to make loans. Banks make money based on their net interest margin (well, at least pure banks do). IOR is not intended to bolster or reduce lending. It’s a policy tool implemented by the Fed that allows them to flood the system with reserves and still maintain control of the overnight rate. The Fed Funds Rate would sink to 0% if they didn’t pay IOR. So it’s best to think of this measure as a sort of emergency policy in case the Fed’s balance sheet is still large and they need to raise rates. In that case they’ll raise the IOR so this essentially serves as a de facto FFR. But it doesn’t stop banks from lending. What stops banks from lending is when it’s not profitable or excessively risky to make loans. IOR actually sucked about 70B in interest income out of the banking system so there’s an argument to be made that this policy hurts lending in that it makes banks less profitable. But there’s a counterargument that QE has added substantial stability to bank balance sheets by removing some toxic assets. So it’s probably a wash in my opinion. Regardless, banks make loans based on their capital position. Not their reserve position. And while 0.25% is a nice little subsidy it’s not making or breaking the banks. And in some cases, it’s marginally hurting since the Fed is taking a 3% int bearing piece of paper and replacing it with a 0.25% piece of paper. Bank would MUCH rather supplement this with higher interest earning loans since their net interest margins are declining….So they’re on the hunt for creditworthy customers just like they always are. The problem is getting the customers to walk in the door.

  18. Dan M. says

    Ok… now I know Cullen has already tried to explain this, but why would paying interest on reserves induce banks to lend more as opposed to less? Interest on reserves, it would seem to me, would induce banks to want to hold reserves, or not loose depositors to other banks… I don’t see how it would affect lending though.

    I’ll get there, guys.. I promise.

  19. wh10 says

    remember that RR aren’t inherent to banking systems (e.g. Canada doesn’t have them). to simplify, take them out of the picture. then transactions just need to clear. the Fed can choose if they want to charge banks interest on such overdraft transfers.

  20. Dan M. says

    Tom… best explanation (for me) yet… Thank you!

    To clarify, what your saying is that while the fed isn’t going to manage whether any one bank has enough reserves, they’re always going to make sure the system has enough as a whole to not be limited by the RR.

    Is this correct?

    What about when the fed tries to enact a tight money policy like in 1981 or something.

    Thanks guys… I think this topic deserves its own blog post. I feel like this is both very important and very unintuitive/technical.

  21. Tom Hickey says

    I think this has a way to run, Cullen. Ed Harrison is just up with a drill down post at Yves’ place. I think that the debate is progressing, and I expect Nick will engage at WCI, too. Krugman? He is kind of in a bind here, it seems to me. His escape is to stay very theoretical and say he isn’t concerned with tactics (ops). He has already done that to a degree, and that will be his fall back position.

  22. wh10 says

    Cullen I internally really resist labeling Fullwiler as such. I respect him greatly, and I know you do too. To be frank, I felt personally attacked by Krugman as a crazy mystic when he was pointing to Rowe’s blog, since my involvement over there was prominent, extensive, but civil (at least I think so). So when I saw Fullwiler’s piece, my emotions got the best of me and I couldn’t help but feel like the side commentary did us all “justice.” That all said, I think you’re right. Putting the emotions aside likely could have only resulted in a better outcome. Maybe not though.

  23. Cullen Roche says

    They’re trying to change the point on you guys Tom. Don’t let them off the hook that easy. :-)

  24. Cullen Roche says

    Yeah, with fixed FX the money supply can’t just expand and contract because JP Morgan wants it to. Instead, JP Morgan has to fight CitBank for the reserves. But we obviously don’t live in that world. The Fed just supplies the reserves to meet the expansion in credit creation.

  25. Tom Hickey says

    Warren is not implying that banks create reserves directly. Only the Fed can do that, since it owns the spreadsheet. But loans creating deposits creating a need for the quantity of reserves in the payment system to expand to settle and meet RR. The Fed always insures that there are sufficient reserves in the payment system made available to clear interbank transactions, as well as to meet requirements. As lender of last resort the Fed will provide a bank that comes up short at the end of a period with sufficient reserves to clear as wellas meet RR, charging the penalty rate. So the quantity of reserves is determined endogenously by demand for credit, while the Fed sets the price (interest rate) exogenously to conduct monetary policy. The quantity of reserves is immaterial to monetary policy, Krugman not withstanding.

  26. Michael Sankowski says

    That’s what wh10 means. The overdraft “creates” reserves. An overdraft is a short term loan with a penalty (think “interest payment”) set by the fed.

    But I am having a hard time thinking through the fixed floating fx well enough to explain it. But it seems in a floating fx world the fed has to allow overdrafts making reserve requirements irrelevant, while in a fixed world, they enforce the quantity of reserves.

    You’ll either have endogenous interest rates or endogenous money, depending on the fx regime.

  27. Cullen Roche says

    Mosler’s always the best. He should just tell everyone else in the MMT attack dog gang to shut up and let him explain Mosler Monetary Theory. :-)

  28. wh10 says

    Which is maybe why Mosler’s way of explaining this is the best?

  29. wh10 says

    Right. Good point. Okay I am not sure about this but here goes. I think he is saying reserves are an artifice of the system. Eliminate the vertical channel. Pretend there is only a horizontal system of credits. Bank A creates a loan and deposit. Bank A then needs to transfer that deposit to Bank B. The transfer needs to clear on the books of the Fed. The transfer creates something artificial – we call them “reserves.” It’s the choice of the Fed, the master bank, to set the price of that overdraft resulting from the transfer to Bank B…

    Does that make any sense???

  30. Cullen Roche says

    This debate is silly. The monetarists are basically trying to change the point. The point Warren makes is that loans come first and reserves are supplied as a matter of necessity after the fact. The monetarists are trying to change the point to interest rates by claiming that rates are not set endogenously by the central bank. This isn’t an unreasonable point though I would argue that it’s kind of a semantic point. The market technically wields some will over what the CB does with rates, but that’s kind of the design. The Fed is a reactive entity. But that’s really a separate point from the loans create deposits point. The main takeaway is that reserves will ALWAYS be supplied because the Fed has to maintain a healthy payments system. It can be no other way.