Monetary policy with excess reserves: a lot like normal monetary policy

After two rounds of QE, banks’ balance sheets are awash in excess reserves:

Many commentators find this to be a frightening prospect—the Fed more than doubled the monetary base in just a couple years! Maybe banks are sitting on the money now, but at some point it’ll “get out into the economy” and cause rapid inflation, right?

Not really. Though excess reserves complicate implementation of the Fed’s policy, it has more or less the same options that it always does.

Excess reserves of this magnitude have one important implication: since they have zero value at the margin for making transactions or meeting reserve requirements, banks won’t be willing to hold them if they pay less than federal funds rate. In equilibrium, this means that as long as the excess reserves are on the Fed’s balance sheet, the fed funds rate will be the same as the interest rate paid on reserves.

The Fed still has the ability to set the federal funds rate and conduct monetary policy in exactly the way it normally does. The only difference is that now, rather than using open market operations to hit the target rate, it needs to change the rate it pays on reserves. Since it’s financing long-term asset purchases with short-term reserve deposits, raising interest rates is costly for the Fed. (This is one way that QE might be a useful commitment device for the Fed, as anticipated by Clouse et al. a decade ago.) But there is no question that it has the capacity to do whatever it wants, just as before—and relative to the government’s overall budget, the potential losses from any increase are very small.

Is there any other effect? Not really. Think about it this way: in addition to conducting its normal monetary operations, the Fed also has a traditional banking business on the side, where it accepts a few trillion in deposits at the prevailing interest rate and invests them in long-term assets. Unless the Fed’s solvency is in question (which it isn’t), there’s no reason why this should be inflationary, or tie the Fed’s hands in any way.



Filed under macro

13 responses to “Monetary policy with excess reserves: a lot like normal monetary policy

  1. Nathan Tankus

    this is an important post.

  2. Scott Sumner

    I basically agree. But couldn’t a sharply inverted yield curve raise some solvency issues for the Fed? (Assuming they held lots of long term bonds with low yields.)

    As a practical matter, I completely agree. The currency balance sheet is not a problem for the Fed, and is very unlikely to pose a problem in the foreseeable future.

    I think you are also correct in seeing the issue in terms of the government’s overall balance sheet, as any losses to the Fed from an upside move in inflation and long rates would be swamped by much larger gains to the Treasury.

  3. JKH

    This is quite good.

    Not novel, but good.

    The important point is that the existence of excess reserves presents absolutely no impediment to the Fed in implementing any desired tightening of monetary policy – by increasing the Fed Funds rate. The draining of excess reserves is a non-issue. That’s what most people get wrong about the situation.

  4. Tord Steiro

    Thanks for the post. I must admit that I have been far more skeptical to the effects of the QE policies and the build-up of reserves. But I don’t know, I consider it unchartered waters.

    Another issue I have problems to understand is how excess reserves can build up rapidly even after the deflationary spiral is broken. Facing inflation there should be other places to profitably invest the excess reserves.

    Either banks think it is necessary to keep these cash on hand, expecting a rainy day in the near future. Or, the banks thinks that there is not better investment available – which generally also indicates some serious rain in the near future.

    Or did I miss something?

    • wh10


      Banks can’t invest reserves in anything, with the exception of lending them in the interbank market or purchasing treasuries. The other thing reserves do is settle payments.

      That said, we have excess reserves because-

      1.the Fed has continued to flood the banking sector with reserves to keep the FFR near 0 (in the belief that this will stimulate the economy)
      2. the economy is in a “balance sheet recession” in which the private sector is deleveraging.

      Excess reserves can diminish through two means

      1. The fed sells conducts OMO to raise the FFR (sells treasuries to drain reserves)
      2. The private sector begins to demand more loans and banks are willing to lend

      However, remember, when banks lend, they don’t lend literally lend the reserves, as the money multiplier might falsely imply (see here for commentary and a paper from the Fed with empirical evidence ). Loans create deposits. Excess reserves are still reserves in the system, they will just become required reserves by label (according to the reserve requirement).

      • Tord Steiro


        Thank you so much for your response. From it, and other comments, I think the following is clarified:

        1. The banks ability to keep reserves have increased due to specific actions by the Fed.

        2. One reason to keep reserves is that they pay at least as high interest as, say, T-Bills or other safe investment (federal funds rate).

        3. One reason excess reserves has increased recently is that the spread between the interest rate paid on excess reserves and the FFr has increased.

        4. One other reason banks keep excess reserves is that their loan portfolio is diminishing (due to the balance sheet recession), hence there is a reduced need for reserves, and reserves become excess unless banks actively lend them out to someone else.

        5. Banks aren’t keen on lending. Which tells me that they don’t see lending as profitable business, relative to the interest paid on reserves.

        Again, I still find #5 puzzling, and potentially alarming. I hear what you say about the balance sheet recession – and what we do know about this type of recessions is indeed alarming (as, among others, spelled out by Richard Koo). Further, I don’t know what to think about #2 and #3 either. What happens when the Fed stop paying interest in excess reserves? Where will they go?

        I hear Matt’s calming opinion that it is not really a problem, but I still think we are in unchartered waters here, and that some caution should be directed tot he issue.

      • wh10

        Tord- no problem.

        “4. One other reason banks keep excess reserves is that their loan portfolio is diminishing (due to the balance sheet recession), hence there is a reduced need for reserves, and reserves become excess unless banks actively lend them out to someone else.”

        To clarify, just to make sure you understand the mechanics, banks make loans independent of their reserve position and *never* need reserves to lend. They will borrow reserves after the fact if they do not have enough, and they do not “lend” the reserves out. The loan creates a new deposit, which is backed by reserves afterwards, and the Fed must always supply banks the proper amount of reserves in order to defend their FFR target. Otherwise, if the system is short of reserves, the FFR will continue to rise.

        That is to say, this huge quantity of excess reserves is irrelevant to the potential for banks to make a loan. The only relevance is the cost of acquiring reserves when making a loan in order to meet reserve requirements (it’s relevant for the return on the loan to the bank and for attracting new borrowers- as you allude to in #5). The only effect reserves have on lending is their price, and in today’s environment, it’s the IOR. We could have 1/5 the amount of excess reserves, and as long as the IOR is the same, their is no effective difference.

        “Again, I still find #5 puzzling, and potentially alarming. I hear what you say about the balance sheet recession – and what we do know about this type of recessions is indeed alarming (as, among others, spelled out by Richard Koo). Further, I don’t know what to think about #2 and #3 either. What happens when the Fed stop paying interest in excess reserves? Where will they go? ”

        Again, reserves don’t go anywhere. The banking system as a whole cannot get rid of reserves. They just trade them amongst each other (thus the FFR) or use them to settle payments. Only OMO by the Fed or taxes can reduce reserves. If the Fed stops paying IOER, then potentially the FFR declines, and perhaps that leads to increased demand for loans (but again, the amount of excess reserves is irrelevant, it’s just the price). But I do not think a ~.25% drop would make a difference. But if it does, and the Fed begins worrying about inflation, they can sell treasuries or hike the IOER on excess reserves in an attempt to stem lending.

        The thing is, all the Fed can ever do to affect the money supply is change the cost of reserves, and many argue this is not a very effective tool since it is so indirect. This is always the case and it doesn’t matter how many excess reserves there are. When we eventually see increased AD when we come out of the recession, all the Fed can do is hike rates, and excess reserves don’t make this more difficult.

  5. Jeff Hallman

    First off, you should check the facts. Since the Fed started paying interest on excess reserves, the fed funds rate has consistently been significantly less than the excess reserves rate. Staffers at the Fed attribute the discrepancy to the fact that by law Fannie and Freddie cannot be paid interest on their excess reserves, so they have to lend them out in the federal funds market and the borrowers have some bargaining power. This sounds plausible, but there’s no real proof that it’s the reason. Why, for example, don’t other banks arbitrage away the difference?

    More importantly, you are implicitly assuming that anything a quantity-oriented policy can accomplish can also be accomplished by setting the overnight rate at the right level. There are two reasons this is incorrect. One is that you when you pay interest on reserves that is higher than the T-bill rate, you cannot grow transactions deposits at will by supplying reserves. If you think there are real balance effects, this could be a real handicap.

    The other reason is political. Here’s Paul Volcker:

    …the general level of interest rates reached higher levels than I or any of my colleagues had really anticipated. That, in a perverse way, was one benefit of the new technique; assuming that those levels of interest were necessary to manage the money supply, I would not have had support for deliberately raising short-term rates that much.

    (Volcker and Gyohten, 1992, p. 170) quoted in Lindsay, Orphanides and Rasche (2005)

    Volcker also doesn’t think he could have held the consensus together long enough to be effective if the policy had been stated as a funds rate target. If we ever have to drastically tighten again, we may regret having made quantity-oriented policy impossible.

    Finally, I think you are still confusing money and credit. Interest rates are the price of credit, not of money. Fifty years ago, Milton Friedman maintained that history strongly suggested credit booms and busts were inherent in the market system, an observation that Ken Rogoff and Carmen Reinhart would probably agree with. A good monetary regime (Friedman favored a 100 percent reserve requirement) should at least insulate the money supply from credit market fluctuations, rather than amplifying them. Scott Sumner’s expected NGDP target for monetary policy takes this a step further by attempting to offset shifts in money demand by via money supply adjustments. I think Friedman would approve.

    • A H

      “Why, for example, don’t other banks arbitrage away the difference?”

      Is there any evidence that the banks are not doing this?

      I’m still completely baffled as to why the interest on excess reserves is as high as it is. Can anyone explain the Fed’s reasoning?

      • Jeff Hallman

        The evidence is that the effective federal funds rate, which you can get data on at, has consistently been less than the interest paid on excess reserves. Currently, the funds rate is about 0.10 percent, while excess reserves are yielding 0.25 percent.

        This spread has recently gotten larger. This may be due to the new FDIC 0.10 percent premium levied against reserves, but that would only make sense if the premium were not levied against loaned-out reserves. I don’t know whether or not that is the case.

  6. Pingback: How Does Monetary Policy Work? « Observations of a Naive Undergraduate

  7. there’s no reason why this should be inflationary

    Indeed, it could be disinflationary (in the sense of reducing the risk that the Fed will allow inflation to exceed its target), as I argue here. If you want the Fed to be sensitive to the interests of bondholders, make the Fed one of them.

  8. flow5

    IORs are a perversion. They are the product of economists that don’t know the difference between money creating depository institutions and true financial intermediaries, between liquid assets & money, etc. Contrary to Milton Friedman legal reserves were never a tax. Fractional reserve banking is extrememly profitable to the participants. An expansion of reserves provided the basis for a multiple expansion of bank earning assets as well as interest which is recaptured by the Treasury.

    And then you have the crazies “banks never need reserves to lend”. LOL These are the real idiots of the world. Keynes’s liquidity preference curve is a false doctine. Greenspan’s performance between 2000 & 2006 demonstrates exactly how stupid “defend their FFR target” actually is. The effect of tying open market policy to a fed Funds rate is to supply additional (& excessive, & costless, legal reserves) to the banking system when loan demand increases.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s