The Two-Per-Cent Solution

inflation, deflation, ioer, fed, floor system, deflation bias, wicksell, monetary policy, excess reserves, NGDP
Modified, Adventures with Sherlock Holmes, Special Collections Toronto Public Library. Canada Public Domain, all other CC BY-SA 2.0.,_2012.jpg
inflation, deflation, ioer, fed, floor system, deflation bias, wicksell, monetary policy, excess reserves, NGDP
Adventures with Sherlock Holmes, Special Collections Toronto Public Library

The Fed’s persistent failure to reach its 2 percent inflation target ever since that target was first made explicit in 2012 has elicited a great deal of commentary in the last couple months, from economists, journalists, and some Fed officials themselves. And well it ought to, for whatever one may think of the Fed's choice of target, the fact that the Fed has been persistently falling short of it suggests that something is awry, if not with U.S. monetary policy, then perhaps with the U.S. economy more generally.

Although plenty of explanations have been offered for the inflation shortfall, none of them is even close to being satisfactory. Bias or "noise" in the inflation numbers? Bias would explain things if the Fed were targeting some supposedly "ideal" measure of inflation. But it isn't: it's targeting the rate of change of the Personal Consumption Expenditure (PCE) index, and so long as that measure of inflation falls below the Fed's target, the Fed isn't "really" hitting that target, even if the "real" inflation rate is higher than the PCE index suggests. As for noise, it should make the Fed just as likely to overshoot as to undershoot its target. Unemployment still isn't quite low enough? Despite what some foolish Phillips-curve reasoning suggests, putting more people to work doesn't make things more expensive. The public's long-run inflation expectations are down? No doubt. But surely that's a result, rather than a cause, of the persistently low values of actual (or measured) inflation.

Aggregate Demand is Part of the Story (But Only Part)

Of existing explanations, the least question-begging emphasize the fact that total spending, or its statistical counterparts such as nominal GDP, just hasn't been growing rapidly enough to achieve the Fed's inflation target. As Mickey Levy puts it in a paper he gave at last week's SOMC meeting,

Obviously, if nominal GDP had accelerated in response to the Fed’s aggressive easing as planned, both wages and inflation would have risen faster, and inflationary expectations would be higher.

Scott Sumner has made essentially the same point: "The only way to have low inflation despite low RGDP growth," he observes,

is if the Fed has such a tight monetary policy that NGDP growth remains slow. And that's exactly what they've done since 2009. If you produce 4% NGDP growth year after year after year [when the norm has been substantially higher] then why be surprised that inflation remains low?

In fact, as David Beckworth pointed out recently, using the chart reproduced below, since the third quarter of 2009 NGDP has grown at an average rate of just 3.4 percent, compared to 5.4 percent between 1990 and 2007 and 5.7 percent for the full "Great Moderation" period of 1985-2007:

According to Beckworth the decline reflects "a monetary regime change" consisting in part of the Fed's having failed to allow spending to "bounce back at a higher growth rate during the recovery" as it had done in past recessions.

But while explanations that attribute the Fed's failure to reach its inflation target to slow NGDP growth have the distinct virtue of taking the equation of exchange (MV = Py) seriously (which is more than can be said for some of the others), they still beg the question: why hasn't the Fed achieved higher NGDP growth?

To observe that it hasn't done so because it hasn't been directly targeting NGDP won't do. After all, a 2 percent inflation target implicitly calls for whatever NGDP growth rate it takes to achieve 2 percent inflation. Arguments to the effect that the Fed has failed to achieve 2 percent inflation because it has failed to achieve 5 percent NGDP growth (or some such number) instead of 3.4 percent growth do no more than invite us to restate the original question in a slightly modified form.

The Regime Change that Matters

It is in fingering "regime change" that Beckworth gets closest to the truth. Only the regime change that really matters isn't the shift away from level to rate targeting that he emphasizes. It's the switch, in October 2008, from the Fed's conventional monetary control arrangements, with a market-based fed funds target achieved with the help of  open-market operations, to its current IOER-based (leaky) "floor system," in which the fed funds rate is moved up or down by raising the rate of interest the Fed pays on banks' excess reserves, either alone or together with the rate it offers in its overnight reverse repurchase (ON-RRP) agreements.

What does the Fed's switch to a floor system have to do with the low inflation and NGDP numbers we've been seeing ever since? Bear with me, and I'll explain.

To do so I must first explain in a bit more detail how a floor system works. In so doing, I'm going to abstract from the role of ON-RRPs, for those aren't actually part of an orthodox floor system, in which the IOER rate alone serves as the central bank's policy rate. In the U.S., the situation is complicated by the fact that various GSEs keep balances at the Fed, but aren't eligible for interest on those balances. Consequently banks and those GSEs can mutually profit by having the GSEs lend their excess Fed balances to the banks for a rate somewhere between the IOER rate and zero.

To partially address this "leak" in what would otherwise have been a solid IOER-based federal funds rate floor, the Fed offered the GSEs and some other non-bank financial institutions the opportunity to undertake reverse repos with it, thereby establishing a solid ON-RRP subfloor below the leaky IOER-rate floor. This arrangement serves to limit the extent to which the effective fed funds can fall below the IOER rate, although it still allows it to vary between that rate and the ON-RRP rate, as can be seen in the chart below. Those two rates therefore serve as upper and lower "bounds" of the Federal Reserve's post-2008 fed funds rate target "range."

A Floor System Calls for Substantial Excess Reserves

Abstracting, then, from the "leakiness" of the Fed's IOER-rate floor, and the presence of an ON-RRP rate sub-floor, the basic idea of a floor system is that the interest rate on excess reserves displaces the fed funds rate as the key monetary policy rate. That's because, in an ideal (leak-free) floor system, banks would neither lend nor borrow federal funds, whether from other banks or from non-bank financial institutions. Instead, they find it more profitable to hold excess reserves. A necessary and sufficient condition for this is that the IOER rate be sufficiently high relative to equivalent private-market rates of interest. In that case, banks will find it more attractive to hold excess reserves than to lend in private overnight markets, including the fed funds market. So long as they accumulate sufficient excess reserves, they can also protect themselves against any need to borrow funds overnight to meet either their net settlement needs or their legal reserve requirements.

Of course the banks will find it especially easy to accumulate excess reserves if the Federal Reserves creates large quantities of fresh reserves, as the Fed did through its various rounds of quantitative easing. Still in the long run what matters most for the existence of a floor system is, not that any particular nominal quantity of reserves should be available, but that banks should have a robust demand for excess reserves, as they will only so long as such reserves yield a sufficiently attractive return.

In an orthodox floor system, such as is established when these conditions hold, the market for bank reserves functions as in the diagram below, taken from Marvin Goodfriend's locus classicus on the subject:

As the diagram shows, so long as banks hold sufficient excess reserves, monetary policy becomes a matter of making desired changes to the IOER rate alone. Through arbitrage those changes will influence other interest rates. Changes in the actual stock of bank reserves, on the other hand, are neither necessary nor sufficient to influence the general state of interest rates. Instead, banks' holdings of excess reserves increase and decline in lock-step with shifts in the supply of total reserves, leaving not only interest rates but lending, spending, and the inflation rate largely unchanged.

Whence the Deflationary Bias?

You're still wondering what all this has to do with the Fed's persistent undershooting of inflation. Trust me, I'm getting there!

It might appear that the switch from a conventional to a floor system should make no difference in the Fed's ability to pursue whatever policy it wishes. After all, all that has changed is the mechanism by which the Fed pursues its policy targets, rather than its ability to set and pursue those targets. Whereas before, to loosen (or tighten) policy, the Fed may have had to increase (or reduce) the available supply of bank reserves, now it only has to lower (or increase) the IOER rate. So, why shouldn't it be able to lower that rate enough to get inflation to 2 percent, or whatever other figure it prefers?

The answer is that, while in principle it could achieve a higher rate of inflation by lowering its IOER rate sufficiently (or, in the actual case, by lowering both its IOER and ON-RRP rates sufficiently), it cannot generally achieve any inflation rate that it wants while also maintaining a floor system. On the contrary: as I'll explain in a moment, although the masterminds behind the Fed's turn to a floor system don't seem to have realized it, maintaining such a system necessarily introduces a deflationary bias in monetary policy.

To see why, recall that, to maintain a floor system, the IOER rate has to be high relative to corresponding market rates. Otherwise banks won't be inclined to accumulate and sit on substantial excess reserves. Instead, they'll increase their lending in wholesale and other markets that offer them better risk-adjusted returns. To serve as a floor (and forgetting about GSE-based leaks), the IOER rate must be an above-market rate.

That the Fed's IOER rate has in fact been kept above corresponding market rates is easily shown by comparing it to its close private-market counterparts.  Of those the closest is probably the Depository Trust & Clearing Corporation's GCF (General Collateral Finance) U.S. Treasury and Agency-Based MBS Repo rate index. The next figure shows how the IOER rate has generally been kept above, and often well above, that comparable market rate:

A comparison of the IOER rate and yields on various shorter-term Treasury securities tells a similar story:

As can be seen, whenever market rates have tended to increase, the Fed has made a point of having its IOER rate keep up with those changes. It has done so, moreover, despite the fact that it has persistently fallen short of its inflation target. Clearly the rate adjustments cannot be justified by appeal to the Fed's desire to stick to that target. They can, on the other hand, be accounted for as inevitable consequences of the Fed's determination to keep its shiny new (albeit leaky) floor system going.

A Wicksellian Perspective

I'm still not quite done, for I've yet to explain in the clearest way possible why a floor system introduces a low inflation bias into monetary policy. Doing that requires that I appeal to Knut Wicksell's understanding of how an economy's rate of inflation depends on the relation between its central bank's chosen policy rate and the economy's corresponding "natural" rate of interest.  According to Wicksell, a policy rate set below its "natural" level will tend to promote inflation, while one set above its natural level will tend to promote deflation.

One can quibble with the details of Wicksell's argument, as I myself have done by noting that a nation's inflation rate depends, not just on its central bank's monetary policy stance, but on the state of economic productivity, among other things. (For this reason I think it better to speak of an above-natural policy rate inevitably leading, not necessarily to deflation or disinflation, but to an excessively low rate of NGDP growth.) The point  remains that a central bank that tends to set its policy rate too high will also tend to generate less inflation than it wants.

To see that this is exactly what will happen if a central bank insists on preserving a floor system, imagine that we are back in the pre-2008 monetary control regime, with no IOER and a market-determined fed funds rate. Assume as well that the rate is both on target and at it's "natural" level — that is, the Fed's target is consistent with a modest (if not zero) rate of inflation.

Now suppose the Fed decides to switch to a floor system.  To do that, it first has to set an IOER rate at least as high as the established fed funds rate, and therefore either at or above the natural funds rate, so as to encourage banks to accumulate excess reserves, in anticipation of boosting the supply of reserves. These steps are needed to get the fed funds market onto the flat portion of the reserve demand schedule shown in Goodfriend's diagram above. Thenceforth, to stay in the new regime, as natural rates increase the Fed must increase the IOER rate as well. While a below-natural IOER rate will undermine the floor system by causing banks to shed their excess reserves, an above-natural IOER rate won't. A permanent floor system is, for this reason, a recipe for monetary over-tightening.

Obviously, if the Fed decides to introduce a floor system at a time when policy is already tight, so that the going fed funds rate is already an above-natural rate, the switch will tend to result in additional over-tightening, since it must involve introducing an IOER rate that's at least as high as the already excessively high established fed funds rate. Something like this appears, in retrospect, to be what happened in October 2008.

A Missing Piece

My explanation for the Fed's persistent tendency to undershoot its inflation target is still not quite complete. For it rests on the underlying premise that the Fed is determined, by hook or by crook, to maintain a floor system of monetary control. What proof have I of that?

It is a good question. My answer is, first of all, that the proof lies at least partly in the pudding. For, as I've stated, were it not for its determination to keep a floor system in place, it would be difficult to explain the Fed's insistence upon raising its policy rate despite falling short of its inflation target. Further evidence consists of the Fed's present normalization plan, which calls for eventual increases in the federal funds rate by close to 175 basis points. So far as Fed officials have indicated, that increase is to be achieved entirely by means of equivalent increases in the IOER rate. In short, if the Fed has any intention of ever returning to its pre-IOER operating system, or to a true "corridor" system in which the IOER rate serves not as an upper but as a lower bound for the fed funds rate, it has given no indication of it. On the contrary: having spoken to several Fed insiders on the matter, I'm assured that the Fed has every intention of sticking to the present system.

Why it should wish to do so is a different matter. I'm pretty sure that part of the reason is that Fed authorities are themselves unaware of the over-tightening bias present in a floor system. They applaud, on the other hand, the fact that a floor system allows them to separately manage interest rates on the one hand and the state of bank liquidity on the other. In recent Congressional testimony,I likened the Fed's gain in flexibility in a floor system — its being able to set its policy rate however it likes, while altering the supply of bank reserves however it likes — to the gain an automobile owner might secure, in being able to turn the wheel as much as she likes, while also stepping on the gas pedal however much she likes, by shifting from Drive to Neutral. The problem, of course, is that, while the driver seems to have more options, the car no longer gets her where she wants to go.

The Solution

What do you suppose it is? The Fed has to abandon its misguided floor system, the sooner the better, by moving to reduce its IOER rate, instead of increasing it as it presently plans to do. As the rate falls from above market to below market, the money multiplier will revive; and that revival will, believe you me, prove more than adequate to boost spending enough to get the PCE inflation rate to 2 percent — and beyond. To keep it from getting too high, the Fed will have to plan on more aggressive balance sheet reductions or resort to its Term Deposit Facility or both. All this is difficult, but not impossible. The Fed can certainly do it if it tries. What it can't do is stick to the present floor system and reliably hit its inflation target.

  • Ray Lopez

    By George, too many detours (at least four) before the main argument, which apparently is that the Fed Reserve artificially raises reserves. But it's well known, see below, that excess reserves are not a problem. Money is largely neutral. Think of it this way: if the Spanish crown has lots of 'reserves' of hard money like gold and silver in the New World, is that a problem in Europe? No, since as an economy expands, more precious metals are introduced into it (and the crown gets seigniorage profits) but since money is largely neutral, life goes on as before, just with a higher nominal price and NGDP. -RL

    How was the Quantitative Easing Program of the 1930s Unwound? Matthew Jaremski, Gabriel Mathy, NBER Working Paper No. 23788 Issued in September 2017 NBER Program(s): DAE
    ("Outside of the recent past, excess reserves have only concerned policymakers in one other period: the Great Depression. The data show that excess reserves in the 1930s were never actively unwound through a reduction in the monetary base. Nominal economic growth swelled required reserves while an exogenous reduction in monetary gold inflows due to war embargoes in Europe allowed excess reserves to naturally decline towards zero. Excess reserves fell rapidly in early 1941 and would have unwound fully even without the entry of the United States into World War II. As such, policy tightening was at no point necessary and could have contributed to the 1937-1938 Recession.")

  • Hu McCulloch

    George —
    The offering rate (and cutoff rate, which ordinarily is the same thing) on Overnight Reverse Repos (ON-RRPs) is clearly an important Fed intervention in financial markets, but I can't offhand find it, either on the NY Fed's website or on the St Louis Fed's FRED database. Where is it reported?

    • George Selgin

      FRED has it but labels it (confusingly) as "Federal Funds Rate Target Range — Lower Bound."

      (It took me forever to find it as well.)

  • Hu McCulloch

    The link you give to Marvin Goodfriend's article on the market for bank reserves has been shortened to an incomplete URL.
    I think this is the article you have in mind:…/0205good.pdf
    This still appears to be similarly truncated. If it still doesn't work, it can be found by googling "Goodfriend Interest on Reserves".

    • George Selgin

      Thanks, Hu. I believe that the link is now fixed.

  • John Hall
    • George Selgin

      Thanks, John. I've repaired the link. It seems I had a chronic missing link problem when writing this particular post!

  • Dan Thornton

    There is another explanation for the Fed’s
    inability to effect inflation that is more viable. Specifically, macro economists have two theories of inflation—inflation is determined by
    excessive growth of the money supply and inflation is the consequence of a
    narrowing of the output gap (the Phillips curve theory). However, neither of these
    theories has had any meaningful predictive power for inflation for at least 50
    years in spite of the fact that economists have considered a wide variety of
    alternative measures of the money supply and considered many alternative
    estimates of the output gap and more or less continuously re-estimated potential
    output and/or the Phillips curve. If economists don’t understand what causes
    inflation, it should be no surprise to find that they don’t know what to do to make
    the inflation rate rise or fall, let alone hit a specific numerical target. We
    should quit talking like we understand the inflation process when there is
    overwhelming evidence that we don’t.

    • George Selgin

      While I'm the last to insist that Fed economists understand inflation, they should at least understand that a lowered policy rate is usually a way to boost it! All the Phillips Curve nonsense is partly (though not entirely) just an excuse for not messing with their beloved above-market IOER rate. In any event even if Fed officials have the theory down they will tend to undershoot in a floor system, for reasons I've tried to explain.

      And leave me out of that editorial "we"! I don't bother with Phillips curves. Give me MV=Py (and that old-time Monetarism) any day!

      • Dan Thornton

        Yes Fed economists do believe that reducing the policy rate is the way to boost inflation. But that's because they believe in the Phillips curve theory of inflation. But I'm glad to hear that you believe the Phillips curve story is "nonsense" too.

  • Milton_Hayek

    The discussion misses the obvious – why on earth is inflation (presumably in the Feds collective mind) assumed equivalent to economic prosperity?

    I'd much rather live with a stable money supply and long term mild productivity driven deflation than to buy into the fantasy that any inflation is good policy…

    • George Selgin

      This criticism makes sense, assuming you didn't get quite as far as the second sentence! I know it would help if my blogs were shorter, but you could at least try to read the whole first paragraph, which tells you that the particular inflation rate target isn't what I'm concerned with here.

  • Michael Byrnes

    I think you've been getting at this point for a while now, in your writings, but this is the first time I think I understand it.

    • George Selgin

      Thanks, Michael–I can at last relax now!

      More seriously, this topic has proven an especially challenging one for me: gaining my own, present understanding has been hard enough; but communicating that understanding has been still harder. There are many moving parts to the argument, more than a few of which tend to be contentious. I plan to keep on plugging away: there are many more minds I'm hoping to change!

      • Michael Byrnes

        There is still much more I (and I'm sure many others) don't understand, especially regarding implications of this new monetary policy world, so I fear that your work is not yet done!

  • Fed Up

    "As the rate falls from above market to below market, the money multiplier will revive; and that revival will, believe you me, prove more than adequate to boost spending enough to get the PCE inflation rate to 2 percent — and beyond."

    I am not sure about that. Let's assume new loan demand is zero. What will the commercial banks buy and from what entity?

    • George Selgin

      But of course new loan demand isn't "zero." At present the marginal loan yield, net of expenses, must equal current marginal yields on alternative assets, including excess reserves, the last of which is 1.25%. As IOER falls, banks acquire other assets until the marginal rates coincide once more.

      • Fed Up

        "As IOER falls, banks acquire other assets until the marginal rates coincide once more."

        Let me try it this way. Which other assets would the commercial banks buy and from what entity?

        By new loan demand is zero, I mean no entity issues a *new* bond for the commercial banks to buy. That means no business or household goes into debt.

  • "…for whatever one may think of the Fed's choice of target, the fact that the Fed has been persistently falling short of it suggests that something is awry, if not with U.S. (Fed) monetary policy, then perhaps with the U.S. economy more generally."

    A good parallel to draw with the Federal Reserve System is OPEC. So, we could say, "…for whatever one may think of OPEC's choice of (oil price) target, the fact that the OPEC Conference has been persistently falling short of it suggests that something is awry, if not with OPEC oil policy, then perhaps with the global economy more generally."

    I would suggest that it is OPEC Conference policy that can only be awry, and not the global economy for which it has influence, though no control over whatsoever. For all of the sophistication that goes in to the models they use to develop policy, OPEC, nor the Fed, has enough influence over variables outside their control to accurately predict their behavior. As a result, their ability to "set" the price of oil and money, respectively, is limited to the degree to which they can adequately control all variables. No small feat!

    Consider the challenge to OPEC oil price policy in assuming and predicting the global economy, a required given for modeling purposes. Non-OPEC oil producer production/reaction to OPEC policy, non-oil energy production/reaction to OPEC oil price policy, energy (oil and non-oil) consumer reaction (demand) to production and pricing decisions of all energy producers, non-energy goods and service production, national and state government politics, reaction and policy, legislation, weather, at a minimum, are all in play.

    Switching to the Fed exclusively, it would be one thing if the U.S. economy was closed off economically to all foreign interaction, but it is not. They need be concerned with the global economy, no different than OPEC. They must control for competing (or cooperating?) central banks, financial markets, politics, etc., etc. George, I have no doubt your suggestions for managing short-term interest rates are good ones, but I'm not sure we are any closer to guaranteeing your policy will transmute to U.S. consumer prices than the Fed's current policy. Your policy prescriptions will set off a series of unknown reactions, not to mention completely unrelated and completely non-predictive events could arise that render even the best model toast.

    Switching back to OPEC, we must ask the following question; has the formation and operation of said cartel been beneficial, not to the producers of oil in those countries (though we might even question that), but to the average citizens of those countries? I'm not sure to the degree they are even acting under that pretense, but to draw a complete parallel to the Federal Reserve System, we would need to determine whether or not State-privileged monopolization or cartelization of production of any good is in the public interest. Without going in to the details, I would have to say that the a priori case for mass prosperity due to monopolization and cartelization of industry is a weak one and that evidence to support this conclusion can be found in many historical examples. So, assuming our current system, if we are better served achieving targeted consumer price inflation using George's recommendation, it should be considered for immediate implementation. This should not, however, be a substitute for complete repeal of state-sponsored monopoly money and cartelized banking in the United States.

  • gf

    "whenever market rates have tended to increase, the Fed has made a point of having its IOER rate keep up with those changes."
    Or more likely, the market anticipated the rise in IOER and T-bills sold off.
    What evidence is there that the causality goes from market –> Fed when the opportunity cost for banks holding short maturity bonds is IOER.

    • George Selgin

      The market is (still) bigger than the Fed, and we know that the Fed had to postpone rate increases on many occasions when it had planned on them. Although Fed officials sometimes pretend otherwise,it is mainly they who do the anticipating, looking ahead for evidence of market developments that themselves are tending to raise "natural" rates, and (given current inflation) nominal ones, and planning their own rate changes accordingly. (For more on this see

      By the way, this account is also consistent with what central bankers _ought_ to do, according to Wicksell and others, except that in endeavoring to maintain a floor system they tend to err o the side of over-tightening.

      • gf

        I agree completely that in general, it is the market that sets interest rates and I am sympathetic to your arguments against IOER. But as someone who trades short term interest rates, we are buying and selling based on the perceived probability of Fed action. If I know for certain that the Fed will raise IOER in a months time, I will be selling front-end rates regardless of whether it is the optimal policy from a Wicksellian perspective! It often only becomes clear with a high probability that a particular month is the month that action will occur, in the immediate lead up to a meeting – which is I think is illustrated by the graph.

        Obviously this is circular when the Fed is "data(prices included)-dependent".
        And these two explanations are observationally equivalent in your chart – but I base this on market behaviour I observe and engage in.

        • George Selgin

          Of course you are correct that, from a trader's perspective, markets anticipate Fed movements. But such movements are not the only source of rate changes traders must try to anticipate. Moreover, traders apart from the Fed are themselves, not individually but collectively, the "bigger" player. Consequently, if you were at the Fed, you would see that they regard, not themselves but "the bond market" as the real force to be reckoned with. The Fed in this sense is like any one "big player" in the stock market. Powerful, to be sure; but ultimately not decisive.

        • Fed Up

          gf, I would say the fed sets a price inflation target. It probably also has other unstated indicators. The fed then basically says to the "market", take the gov't bond yield curve to where you think it needs to be to meet that price inflation target. We will follow along with the fed funds rate to meet short term gov't yields. Thoughts?

          Can you say what short term interest rates you trade?

          Pre-2008, do you think the fed followed an asset buying system for the indirect price inflation target or a system where there is a fixed exchange rate with solvent banks and interest rates were used to meet the price inflation target?

          What if IOER and RRP had the same rates?

  • Japan and the Big EZ don’t pay interest on reserves and they can’t get their inflation erect either.And the US has seen inflation in assets, especially stock and bond markets and real estate.

    • George Selgin

      The Japan an Eurozone stories are different, but perfectly explicable. (Not in a comment, though.) And yes, U.S. stock prices and some other asset prices have risen. But the question is why isn't the Fed able to reach its 2 percent PCE rate target?

      • Do you have other posts where you discuss the Japanese and Big EZ failures? I would guess that the Fed can't reach it's PCE target because all of the new money is going into assets and not consumption? I'm not trying to argue for the sake of arguing, just trying to understand.

        • George Selgin

          I discuss Japan's case in the testimony I link to in my post above. As for the Eurozone, the ECB also payed positive interest on bank reserves until 2012' it was zero for a while, and did not go negative until 2014. European bank excess reserve holdings have in the meantime fallen considerably, from about Euro 900 billion to about Euro 100 billion. Unfortunately the ECB responded to the decline in excess reserves by–you guessed it!–dramatically reducing the monetary base! So in the end, the stimulus to lending from the reduction in IOER was trivial. Still, everything there was nothing mystery about what happened, except so far as the monetary geniuses running the show are concerned!

          It's remarkable but true: central banks are just as capable of screwing up by overtightening as by overloosening.

  • Fed Up

    Let me try it this way.

    Assume there are excess (central bank) reserves systemwide at the commercial banks and each commercial bank has excess (central bank) reserves.

    Assume new loan demand is zero. By new loan demand is zero, I mean no entity issues a *new* bond for the commercial banks to buy. That means no business or household goes into debt.

    Which other assets would the commercial banks buy and from what entity?

    • George Selgin

      "Assume new loan demand is zero. By new loan demand is zero, I mean no
      entity issues a *new* bond for the commercial banks to buy. That means
      no business or household goes into debt." I'm sorry, Fed-up, but I cannot grant you this assumption, which assumes that there isn't an ordinary, downward sloping demand schedule for loans, that is, a schedule that suggests that the quantity of credit demanded increases as the interest rate declines.

      Banks can also acquire assets from non-banks, including both non-bank financial firms and the general public. They can, for instance, acquire commercial paper and securities from other holders, and increase their own holdings relative to non-bank holdings, without there being any need for the total outstanding quantity of the assets in question to increase.

      • Fed Up

        "They can, for instance, acquire commercial paper and securities from other holders"

        That is what I was getting at. Now I would call those *existing* bonds, not *new* bonds (going into debt). So there is probably *new* demand deposits for *existing* bonds.

        Next, are you going to assume some kind of "hot potato" effect for prices of goods/services from the *new* demand deposits?

        "a schedule that suggests that the quantity of credit demanded increases as the interest rate declines."

        It seems to me that near the beginning of a recession or slightly after the beginning of a recession, that can happen (where no entity borrows).

        Also, does that mean lower interest rates, from the market(s) with the fed following along, are about more debt or at least an attempt at more debt?

        • George Selgin

          This is getting too much in the weeds, Fed-up. My concern isn't with special cases; I will for the sake of argument allow that you can come up with conditions such that whatever you wish to maintain is correct. I do not believe, in any case, that such conditions held at any time during or since the recent crisis.

          • Fed Up

            "My concern isn't with special cases;"

            But that allows to show how the system actually works.

            "I do not believe, in any case, that such conditions held at any time during or since the recent crisis."

            You probably can find better data than me, but it *seems* to me that private borrowing went near zero right around the financial crisis.