15 Jan 2011

Bank Reserves: Thesis, Antithesis, Synthesis

Economics, Federal Reserve 16 Comments

Ahh, I feel much better now. During the “office hours” for the Mises Academy “Anatomy of the Fed” class, I think I resolved something that had been bugging me for months.

We all know the standard textbook story of fractional reserve money creation: The Fed buys assets and creates new reserves (“out of thin air”). Armed with the new reserves, the commercial banks now have excess reserves, so they go lend it out. Thus, the causality goes like this: Fed creates new base money. ==> Banks have excess reserves. ==> Banks create new M1.

The only problem is, that’s not how it works in the real world. Doug French explained in our interview in the Lara-Murphy Report, when actual bankers are deciding on new loans, they never say, “Hey Jimmy, go check our excess reserves to see how much room we have to play with.” (I’m paraphrasing of course.) In fact, some critics of the textbook explanation go so far as to say that banks aren’t constrained by reserve requirements at all.

So this troubled me. It wouldn’t surprise me to learn that macroeconomics textbooks are completely wrong when it comes to explaining banking–they’re completely wrong about 19 other topics too. But surely the reserve requirements do something. If they didn’t, then why do banks pay interest on borrowed reserves in the federal funds market? And what’s the story with “sweep accounts,” if the official reserve requirements aren’t actually enforced?

So I think I came up with a resolution, and for those lucky students who attended the Office Hours session–they saw me do it in real time. Of course, this may be what (some of) the critics were saying all along, and I was too stubborn to listen…

(1) OK, suppose that an individual bank’s loan officers are granting loans based on the merits of each case; they don’t care what the bank’s “excess reserves” are, and in fact they don’t even know what that term means.

(2) If the bank’s lending decisions cause its total demand-deposit balances to rise so much that the bank is no longer satisfying its reserve requirements, it goes into the federal funds market and borrows the difference from another bank that has excess reserves.

(3) Now suppose a bunch of banks (in a bull market for example) are making loans, such that there aren’t enough excess reserves to go around. After all, if we were in an originally “fully loaned up” situation, then mathematically if banks make more loans, the system as a whole is deficient. The interest rate in the federal funds market shoots up, as the demand to borrow reserves increases and the supply to lend them decreases.

(4) Fed officials see that the fed funds rate has shot up, well above their target (announced at the last Fed meeting). Bernanke’s left eyebrow rises in tandem with the rate. They don’t know what to do, so they crack open Mishkin’s textbook. And yep, it says right there that when the fed funds rate is above the Fed’s target, the Fed will engage in “expansionary policy” by buying assets and adding more reserves to the system.

(5) Phew! Problem solved. The Fed’s injection of new reserves pushes the fed funds rate back to the Fed’s desired level.

Does everyone see how this story satisfies both camps? Yet the causality in this story is the exact opposite of the standard economist fable. In this new version, causality goes like this: Banks create new M1. ==> Banks have shortage of reserves. ==> Fed creates new base money.

16 Responses to “Bank Reserves: Thesis, Antithesis, Synthesis”

  1. Ryan M says:

    I thought the reserve ratio was generally non-binding in the same sense that the minimum wage is generally non-binding. Banks set their own “reserve ratio” based on liquidity concerns. This is why the fed doesn’t actually use the reserve ratio as a policy instrument; it’s because they are AWARE that it doesn’t work like the textbook model. I don’t think this is a grand mystery. While these facts weren’t in econ 102, they were at least in my undergraduate money and banking course at a mainstream university.

    I think the simplest explanation is that banks have that implicit (or probably even explicit) reserve ratio which is greater than the legally enforced reserve ratio, and they inform their loan officers what types of risks they should be taking and what interest rates they should offer in order to clear their reserves down to the desired level.

    What is profit maximizing, even in the short run, is not to leverage up to the point where all banks are fully loaned up. After all, the institutions doing the lending in the feds funds market are other banks, not the fed. So there is a disincentive to be completely leveraged both because banks don’t want to borrow at a high interest rate and because they would like the opportunity to LEND at a high interest rate.

    This creates an equilibrium of reserves and a fed fund rate. The causality should go Banks create new M1 because they were above their internal reserve target => Banks reach internal target => new equilibrium Fed Funds Rate established.

    The Fed tries to inject M0 such that the equilibrium rate matches their target.

    There is no need for a vicious cycle to be going on here without ancillary public choice/knowledge problem assumptions.

    • Captain_Freedom says:

      You’re exactly right. Please consider yourself “lucky”. I had to learn this on my own.

  2. Jon O. says:

    You have it right. The process starts on the demand side and the banks/fed accomodate. This is why moentary policy can become impotent when demand for funds collapses for other reasons than the cost of those funds.

    This is also why alot of austrians are wrong when they draw a direct link between QE and future price inflation, it has more to do with peoples (and governments) desire to borrow and spend rather than the size of excess reserves sitting on the sideline. The latter is just potential for future inflation.

    • RS says:

      Jon O.

      You have it backwards. The desire to borrow and spend is the potential while the creation of fiat money out of nothing enables those desires to become real in the market, resulting in future price inflation.

      Anyone can desire anything he pleases, demand anything he pleases but that by itself has no impact on market prices. The implicit sub clause to ALL consumption is production, it is physically impossible to consume something before one has produced it (you can safely try this at home if you don’t believe me), so QE policy does nothing more than enable some people to consume what others have produced in a nefarious way. It does not stimulate new wealth, only wastes existing wealth.

      • Jon O. says:

        If someone desires to borrow at the prevailing rate and their bank doesn’t have the reserves the bank will bid for the funds in the interbank market. If the effective fed funds rate rises above the target the fed will conduct REPO transactions to bring the rate in line, increasing reserves and thereby money supply.

        The creation of fiat money is irrelevant if it is offset by an increase in the desire to hold cash balances or if it sits in reserves. Thats why an increase in the money supply only creates the potential for inflation. Velocity has to increase for price inflation to take hold. This is why there was no price inflation during the crisis even though the fed flooded the market with money and narrow monetary aggregates exploded; the demand for credit collapsed and there was no desire to spend or invest, just to hold cash.

        Bank reserves could increase by $100 trillion dollars tomorrow but if no one wants to borrow it you won’t get price inflation. Of course the government can and will take over to offset the collapse in private credit.

        Again, an increase in money supply doesn’t directly cause inflation, spending that money does.

  3. Bruce says:

    According to this article (http://libertarianpapers.org/2010/43-boyapati-why-credit-deflation-is-more-likely-than-mass-inflation/) the banks are constrained by capital requirements, not reserves.

    • Captain_Freedom says:

      Yes, that is another complication.

      The reconciliation of this fact with “traditional” frb analysis is that the “capital” in “capital requirement” is *itself* a product of fractional reserve lending and spending!

      Money is ultimately created by the Federal Reserve System, and when that new money hits stocks, bonds and real assets of banks, the monetary value of the banking system’s capital can increase, thus serving as a proximate basis for new loans. Those new loans then get spent and then re-spent again, which may then go into bidding up stocks, bonds and real assets of the banks once again, thus setting the stage for further loan expansion.

      It is true that the traditional minimum reserve requirement multiplier analysis has been, at least since the 1990s, no longer applicable in the US, at least for savings accounts, however this does not mean that it is completely worthless as a tool for understanding banking in general. According to current regulations, US banks have a minimum required reserve ratio on transactions deposits in the amount of 10%. In reality, in practice, most banks do NOT obey this regulation. They completely ignore it. Remember, it is the banks who control the Federal Reserve, not the other way around. If the banks have a real world reserve of 1%, then that IS the required reserve. The Fed won’t shut banks down for not obeying the minimum reserve requirement. The Fed works for the banks. This is why practitioners in banking look like you’re from another planet when you start talking about minimum reserve requirements. At the height of the real estate bubble, the banking system was “fractional reserved” up to around 100 to 1.

      In many countries, there are still reserve requirements for commercial banks that put a limit on lending. In China for example, the central bank often uses minimum reserves as a tool for monetary policy. They changed the minimum requirement 9 times in 2007 alone. Just recently China upped the minimum reserve requirement again for commercial banks.

  4. flow5 says:

    Nothing’s changed since 1965. The money supply can never be managed by any attempt to control the cost of credit. But the monetary transmission mechanism (interest rates) has been denigrated. Banks are no longer reserve constrained.

  5. Desolation Jones says:

    Where’s that AP Lerner guy? This topic was supposed to be his shtick.

  6. Robert Wenzel says:


    What you are doing here is very dangerous because it sort-of implies that the banks decide how much money is in the system and not the Fed. But that is not the case. The banks are always reacting in what they lend out based on the Fed funds rate, but this is set by the Fed. So the Fed has the first and last word.

    I am pretty much saying the same thing as you BUT I see the Fed control of the funds rate as the key factor and banks reacting off of that and not the other way around.

    • bobmurphy says:

      Oh right, I’m actually thinking of writing an article on this. I just want to see if I’ve captured the perspective of the people who say that banks aren’t constrained by reserves.

      Because even in my little fable, it’s the Fed’s adherence to a particular fed funds rate that drives it. Ultimately it is the Fed who decides whether to buy assets and inject new reserves; the banks can’t “force” Bernanke to do that.

    • Captain_Freedom says:

      >What you are doing here is very dangerous because it sort-of implies that the banks decide how much money is in the system and not the Fed. But that is not the case. The banks are always reacting in what they lend out based on the Fed funds rate, but this is set by the Fed. So the Fed has the first and last word.

      Not quite. Bob is right that the banks lend and then the Fed accommodates. This is why the Fed was created in the first place. To allow the banks to keep doing what they already were doing, and avoid bank runs and bankruptcies after engaging in frb.

      It is not “dangerous” to see that the banks have the ultimate power to create money, since it is they who make the decisions to expand credit and lend, or hold onto reserves as cash and/or loan it back to the Fed. It is the banks who are now deciding to hold over $1 trillion in reserves created by the Fed, with the Fed, rather than lend it all out in the market, generating even higher price increases.

      The Fed increases the money supply indirectly in practice, by constantly “bailing out” the banking system’s penchant for expanding credit who then scramble to satisfy withdrawals, inter-bank transactions, and daily financing. So you’re right that the Fed is the ULTIMATE creator of new money, for without the Fed, the banking system’s reckless behavior will eventually be punished by the market, thus stopping credit expansion, and with the Fed, the banks can continue engaging in credit expansion.

      But you’re wrong if you claim that in practice, banks “react” in their lending programs to prior changes in the Fed’s actions regarding reserve requirements, or fed funds rate. In practice, banks lend according to their capital requirements, not their reserve requirements. A lower fed funds rate sends a signal to the market that the Fed is creating new funds, and this will eventually increase bank capital, when is then the go ahead to make new loans.

      On paper, the current regulations state that banks must hold 10% reserve on outstanding transaction balances (0% on savings accounts). Most banks however do NOT obey this requirement. They lend according to risk/return, VAR, capital requirements, etc at the individual bank level. This is why an extreme jump in reserves does not necessarily lead to an equivalently extreme jump in bank lending. Yes, a long term sustained increase in lending is explained only with an prior increase in reserves, but operationally, day to day, banks do not use the fed funds rate to decide when to loan and when not to loan. The fed funds rate acts more like a fraudulent thermometer. The banks will lend according to capital requirements and risk/return, which will, all else equal, eventually make the thermometer rise as banks start to scramble for money to satisfy transactions, inter-bank transfers, and withdrawals, and operational financing, but the Fed always steps in and just changes the gradings on the thermometer, making banks think (correctly) that money is easy to come by in the overnight market, should they need emergency funds. This then reassures them to continue lending.

      I’ve worked in 3 major banks, 2 of them as a risk management officer, and I can tell you that banks do NOT “react” to past changes in the fed funds rate, or reserve requirements, two traditional pedagogical tools for bank theory. Banks make loans according to their capital requirement, and VAR analysis, at the individual bank level. I will even tell you that in all three banks I worked at, transactions balances were “reclassified” into two sub-accounts, one the daily transaction account (10% reserve), the other a savings sub-account (with no required reserves). Even if a client deposited $10,000 cash into their demand deposit (checking) account, we would lend out virtually the entirety of it by reclassifying the account into a holding and transactions sub-accounts.

      This “trick” was and is known by the Fed. They are complicit. The Fed never enforces a traditional minimum reserve requirement on banks who are below the 10% requirement. That’s why you never hear of banks being shut down by the Fed because of too low reserves. The 10% requirement is for Congress and the public to digest and believe makes their personal money safe. In reality, the percentage is virtually zero.

      All a bank need to do to operate at zero reserve is to ask the Fed for permission to reclassify transactions accounts. To get a “no objection” permission. Did you know that the Fed has granted this permission FOR EVERY RECLASSIFICATION REQUEST every bank has ever made? It’s funny because it’s true. The FDIC also know about this, and they too are OK with it.

      The Fed has all but abandoned enforcing minimum reserve requirements in the banks.

      From this article in 1993:


      The Fed writes:

      “Laws requiring banks and other depository institutions to hold a certain fraction of their deposits in reserve, in very safe, secure assets, have been a part of our nation’s banking history for many years. The rationale for these requirements has changed over time, however, as the country’s financial system has evolved and as knowledge about how reserve requirements affect this system has grown. Before the establishment of the Federal Reserve System, reserve requirements were thought to help ensure the liquidity of bank notes and deposits, particularly during times of financial strains. As bank runs and financial panics continued periodically to plague the banking system despite the presence of reserve requirements, it became apparent that these requirements really had limited usefulness as a guarantor of liquidity. **Since the creation of the Federal Reserve System as a lender of last resort, capable of meeting the liquidity needs of the entire banking system, the notion of and need for reserve requirements as a source of liquidity has all but vanished.** Instead, reserve requirements have evolved into a supplemental tool of monetary policy, a tool that reinforces the effects of open market operations and discount policy on overall monetary and credit conditions and thereby helps the Federal Reserve to achieve its objectives.”

      and then, further on:

      “Requiring depositories to hold idle, non-interest-bearing balances is essentially like taxing these institutions in an amount equal to the interest they could have earned on these balances in the absence of reserve requirements.”

      The Fed actually considers required reserves as a TAX on the banks! And we wonder why the Fed is not enforcing minimum reserve requirements on banks!

      In other words, the Fed uses the minimum reserve as a tool for their own policy analysis, not as an enforced regulation that generates real world effects.

      The reason why the Fed didn’t just lower the required reserve percentage outright, and instead did it in this loopy indirect way, is because lowering the minimum reserve requirement requires an act of Congress. The banks want to make money now and do not want to go through the hassle of getting the laws changed. It’s a chore to them.

      If all this does not convince you, then just ask yourself the easy question of how in the world could it have been possible for the ratio of checking accounts money to reserve money during the height of the real estate bubble to be in excess of 100 to 1? And then ask yourself why isn’t price inflation taking place like gangbusters even though the fed funds rate is near zero? It’s because the banks proximately control the money show, and the Federal Reserve gives them the power behind the scenes. The Fed could increase reserves by $100 trillion and not create any price increases….if the banks do not lend any of it. Of course in reality they do lend a portion of it, depending on the economic conditions, but the banks directly determine TMS (true money supply) in the short term. The Fed determines TMS in the long term.

  7. The Interest says:

    It is time for the Austrians to give up this notion that the Fractional Reserve System exists. Bob-you’re getting there. Read the previously sited article by Vijay Boyapti. (http://libertarianpapers.org/2010/43-boyapati-why-credit-deflation-is-more-likely-than-mass-inflation/) Boyapti sums up the arguments quite well not only about the FRS not existing anymore but also what Captain_Freedom is implying about the banking class making sure the system doesn’t crumble which would put the banking class out of business.

    I think what Captain_Freedom is explaining is Monetary Circuit Theory.

    • bobmurphy says:

      I’m not getting there. I did read Vijay’s article; that’s who I had in mind.

  8. AP Lerner says:

    Sorry I missed this post. As someone who understands the banking system, and understands the banking system is never constrained by reserves, it is good to see you addressing this subject. Unfortunately, its not so good so see you will not recognize the macro textbooks (and you) are wrong.

    What you fail to grasp in your example is loans create additional deposits. Once you recognize this, it becomes clear the system is never constrained by reserves as long as the inner bank market is functioning properly. The Fed has exactly zero control over the money supply. That power is strictly held by the fiscal authorities.

  9. David says:

    AP Lerner,
    I’ve seen you on many boards and I just have to ask, why do you give your resume on every post?
    “.As someone who understands the banking system….”

    It’s truly bizare. I’ve seen you do this at least two dozen times. There’s a couch in a psychoanalyst’s room with your name on it.