Tuesday, September 3, 2013

The convenience yield as epicentre of monetary policy implementation

Let's not get carried away by the idea that central banks set overnight interest rates. Central banks exercise direct control over the economy by pushing down on one shiny red button, the convenience yield on reserves. By modifying the convenience yield, a central banker nudges agents to either flee from reserves or flock to them. This causes a change in the purchasing power of reserves, the mirror image of which is the general price level.

So how do overnight interest rates like the fed funds rate figure into the picture? Reserves are scarce and convenient, so agents will only part with them if they are compensated with an adequate amount of "rent". The higher the marginal convenience of reserves, the more rent they require. The market in which reserves are rented out for very short periods of time, usually 24 hours, is referred to as the overnight market.

By conceptualizing monetary policy this way, it immediately becomes apparent that overnight interest rates are little more than a reflection of the underlying convenience yield on reserves. Fed funds rates move only in the instant after the convenience yield has been modified. Nor are overnight rates the first to move in response to changes in the convenience yield. They are just one of an almost infinite number of market prices and rates to react in the moments after the convenience yield has been diminished or improved. We should stop thinking in terms of overnight-rate exceptionalism. Price changes radiate out from the convenience yield.

I've gone into much more detail on all these ideas in an earlier post. In this post I want to focus on one point I made earlier -- the fed funds rate can be a bad reflection of the underlying convenience yield. There are several reasons for this. First, in renting out reserves, banks are effectively replacing the central bank as their counterparty with another private bank. In normal times, banks make excellent counterparties. But during times of stress, rentors may require an extra amount of rent that has nothing to do with the usefulness or scarcity of reserves, but the riskiness of counterparties. A spike in the fed funds rate may have nothing to do with the underlying convenience yield, and everything to do with credit risk.

Secondly (and more importantly), the funds rate ceases to be a good indicator when it falls to zero. When this happens, there is a temptation to view monetary policy as spent. After all, it may seem that a central bank can't reduce the convenience yield below 0, which is to say that it can't push the fed funds rate into negative territory. We should resist this temptation. The current fed funds is an estimate of the 24-hour convenience yield on reserves. But agents hold reserves not only to enjoy their present convenience but to enjoy their expected future flows of convenience. So even if the market puts no value on the immediate 24 hour convenience yield, that isn't to say that the market doesn't value their convenience 1 week from hence, or 1 month, 1 year, or 1 decade.

Just like there is a term structure to government bill/bond rates, there is a term structure of the convenience yield on reserves. We can get an inkling of the term structure by looking at longer-term fed funds deals. In the term fed funds market, for instance, banks will rent reserves for up to one year. Federal funds futures give an indication of the expected convenience yield several years from now.

When the overnight convenience rate hits zero, central banks still have plenty of traction over the rest of the term structure. By attacking the convenience yield on reserves one year from now, for instance, a central bank hurts the discounted value of expected streams of convenience thrown off by reserves in the present. As a result, today's forward-looking agents will be nudged away from holding reserves, causing a rise in the price level. As long as there are portions of the convenience yield curve that are still positive, a central banker can do their job.

If thinking in terms of short term rates is misguided, so is a focus on base money. Once the short term convenience rate on reserves has hit zero, present changes in the quantity of base money will have little effect on convenience yields further down the curve. One way to reduce convenience yields five years hence would be to promise that the then-supply of base money will be sufficiently broad so as to ensure that the marginal deposit yields no convenience flows. But a central bank will only be able to affect future convenience yields if the market believes it will have the gumption to follow through on its promises.

Monetary policy at low convenience rates is all about making promises about future convenience yields and ensuring those promise are credible. Conventional monetary policy, on the other hand, is relatively simple -- all a central banker need do is manipulate the current convenience yield in order to have an effect on prices.


  1. JP: I'm with you up to your very last sentence. Increasing the marginal convenience yield for one month, then reducing it again (or being expected to do this) would have very little macroeconomic impact. Isn't it always about the expected time-path of marginal convenience yield?

    (If the central bank can also pay interest on its monetary liabilities, this gives it a second shiny red button: the marginal pecuniary yield. And it's the sum of those two yields that matters for AD, I think, which means this doesn't violate the old "central banks have only one instrument so can have only one target" dictum.)

    1. Nick, you're right that it is still the expected time-path of the marginal convenience yield that counts.

      And good point on the sum of two yields, the marginal pecuniary and non-pecuniary yields, being the relevant instrument when a central bank can also pay interest on reserves.

    2. So... is the final sentence:

      "Conventional monetary policy, on the other hand, is relatively simple -- all a central banker need do is manipulate the current convenience yield in order to have an effect on prices."

      still the one that Nick disagreed with, or did you change it?

    3. It should read something like...

      Monetary policy is usually relatively simple since small changes in the current quantity of scarce reserves affect the entire time path of the marginal convenience yield (or at least a sufficiently large chunk of the path to have an effect on prices). Unconventional monetary policy is tricky because present changes in the base don't affect the time path much -- only commitments to bring down future convenience yields will be effective.

  2. Interesting stuff. Are you suggesting that the Fed should conduct open market operations in Fed Funds futures? That's very interesting as their actions would be more concrete than a vague suggestion like "0% for an extended period of time". Following up from your post a few years ago, on WCI, do you think the Fed has a legal loop hole to implement this?

    1. I hadn't made that leap, but now that you mention it, it would be an interesting strategy. I'll have to mull that one over. You're certainly right that it would be superior to vague forward guidance.

      The Fed was able to use broad lending authority under Section 13 to buy AIG's assets, so why not use the same to buy fed futures contracts?

  3. A central bank can't, or at least shouldn't, make any unconditional promises about future interest rates or money supply. That confuses the policy instrument with the target. If it's constrained by the zero bound, then it has to change its target path to one with higher interest rates.

    1. Max: I prefer to think of it like this: a central bank that lasts for 10000 periods has only 10000 degrees of freedom - one degree of freedom per period, *on average*.

      1. It can't chose both the target and the instrument (at least, not independently).

      2. It can "borrow" degrees of freedom from the future, and use tomorrow's degree of freedom today, and it's an interest-free "loan", but it can't keep on borrowing more and more, by using 2 degrees of freedom per period.

  4. JP, nice article. I have an O/T question, but I figured you'd be the one to point me in the right direction. Where can I find the best definitions for:



    1. Scott Sumner does not like to be bound by definition, so don't ask him.

      Look through my medium of account posts. That may help. Bill Woolsey is good for this stuff too.

    2. BTW Tom, I tried to reply to your question on HPE on ask-cullen, but I couldn't get it to work. Anyways, my suggestion was to come to this site and look at JP's HPE links (see right). They are some good posts.

    3. jt26: something is screwy w/ AskCullen... I answered one there, and the total answers remained zero. Plus it no longer allows me to edit my questions.

  5. "a central banker nudges agents to either flee from reserves or flock to them"

    Only some agents (banks) are counterparties to the CB. If all agents (general public) were included as direct counterparties policy would be effective because the general public would have a much higher convenience yield outside of reserves than in reserves.
    As a result they would flee reserves and pick the economy up.

    1. The fact that only banks can hold reserves doesn't alter the central bank "nudge" effect. All prices are denominated in terms of central bank money, from stock prices to debts to houses to bank deposits. So any change in the return on reserves will ripple through the entire network of prices.

    2. The problem is that you gotta have reserves in the first place to get out of reserves into other things. The general public don't have reserves.

  6. JP, ignore this if you want... it's O/T again (I wish you had an "AskJP" page :)

    I've been pestering Scott Sumner w/ this in his recent explanation of the HPE:


    Let me ask you (I'll try to keep this super brief!). Two cases:

    A. MOA NOT gold (like Scott's cases 5 & 6): Fed finds $1T of coins in the basement they didn't know they had

    B. MOA = gold (like Scott's case 4): Fed finds $1T of *GOLD* coins in the basement they didn't know they had

    In both cases they announce the discovery to the world, put all the coins (which are face value assets of the Fed) up for sale, and decide not to remit anything to Tsy (say they had that power).

    Is there a difference? Granted these represent simplified versions of what I first presented to Scott at two different times, and I don't think he ever understood either question, but in case A his answer was (paraphrasing):

    "It's not base money so who cares."

    While in B it was (quoting):

    “If I understand your question (doubtful) then I’d say that after the gold was sold it would be owned by the public, and prices would rise. Indeed if it was known the gold would be sold, prices would rise in anticipation.”

    So I probably confused Scott or unfairly simplified my cases to the forms I'm presenting to you here... but it still surprised me that I got two different answers. What say you?

    Here's my original version of case A:

    And my original case B:

    1. That's quite the detailed thought experiment.

      The Fed finding 1T worth of coins in its basement and spending them away is very much akin to the Fed increasing the quantity of reserves outstanding.

      To begin with, the economy won't need that many coins. Individuals will return them to banks for deposits, and banks will return them to the Fed for reserves. Which will increase the quantity of reserves in existence so that on the margin, reserves will be less rare. This will lower their return (more specifically, their convenience yield). Banks will try to rebalance their portfolios by spending away excess reserves, or lending them on the fed funds market. The price level will rise.

      Which is the exact same thing that happens with an exogenous increase in the quantity of reserves.

      As for gold, the Fed finding gold in its basement is just like a miner discovering a bonanza. Since prices are denominated in gold, the general price level will rise.

    2. JP, first, thanks so much for your response! So your conclusion in both cases: "price level will rise." That's very different than Sumner's conclusion for case A: "it's not base money, so who cares"

      Indeed, why should case A be coins? Why not have the BEP print up a $1T of paper reserve notes at production cost (as always) for the Fed and have them announce the new notes' permanent existence and put them up for sale? That was Case A version 1. Case A version 2 is linked to above (it's longer and addresses some of your concern about people wanting coins), and Case A version 3 is what I put in my last post (for ease of comparison with Case B). I'll copy my Case A version 2 from Sumner's blog here:

      Scott, building on my question here:


      Lets say that the Mint started minting a new $1M coin (production cost: $0.02), but that they are not permitted to do anything with the seigniorage other than use it to buy new raw materials (say a new Fed deposit is created, just for the Mint for this purpose)… in other words Fed funds go into this deposit but come out again only VERY slowly.

      Now let’s say the Fed starts buying these new coins (which, like all coins that have left the Mint are legally base money the minute they leave the Mint, and are assets to the Fed as long as the Fed holds them)… starting off at $85B a month and increasing their purchases 20% per month until they reach some target: could be NGDP, could be inflation, whatever.

      Also assume that the Fed announces all this and will gladly sell these coins to banks OR a new kind of paper note specifically backed by these coins (i.e. redeemable in these coins) in the usual face values: $1, $5, $10, $20, etc.

      Would this scheme work? Why or why not? What’s the substantial difference with your cases 4 or 5? Thanks!

    3. BTW, Scott pointed out that my assertion above is not true: the $1M coins would NOT be base money. He's correct. They're still money though!... and treated as face value assets of the Fed as all coins are while in possession of the Fed. Base money consists of reserves (thus money held by banks) or currency in circulation.

    4. "TallDave's" response to me regarding Case A was (paraphrasing): "So you have a magic box that makes gold. You announce to the world you're going to increase the world's gold supply by 20% each month, but you'll only sell the new gold for gold." I think that's a good response! So I'm curious how you'd counter that.

    5. My response to TallDave was:
      "TallDave, actually, my hypothetical Fed will accept standard Fed deposits as payment, or standard reserve notes. Or if the public wants to buy the new notes, I’ll let them trade their bank deposits (with banks acting as an intermediary: i.e. the banks debit the public’s bank deposits and the Fed in turn debits the banks’ Fed deposits)."

  7. The fact that only banks can hold reserves doesn't alter the central bank "nudge" effect.

    Of course it does. Only banks are doing the "nudge effect" as opposed to a broader spectrum of the economy directly doing the nudging.

    1. (also for lxdr1f7)

      The nudge effect will still play even if the broader spectrum of the economy doesn't hold reserves.

      Nor do they need to wait passively on the sideline for the effect to occur. If banks are nudged out of reserves and they choose to buy stocks/ETFs & other financial assets, the public can just as quickly sell their bank deposits for those same financial assets. Or they can raise the price that they are willing to accept for those assets.

      The key is the public announcement. Everyone can react to it.

  8. Everyone can react to the anouncement and move their current holdings. Only banks can move their new reserves into new holdings. Banks always have the edge in this imbalanced system.

  9. When you say 'reserves', do you mean the fed's monetary liabilities, i.e. federal reserve notes plus federal funds? Or do you mean just the stuff held by banks? or something else?

    1. Mike, excuse me for butting in: I have no idea what JP means, but reserves are defined in the US as base money (electronic Fed deposits and physical currency) held by the banks. (That's why it's impossible to "loan reserves outside the banking system" ... it's impossible by definition).

    2. ... but of course cash advances exist (i.e., they can lend base money outside of the banking system).

    3. I mean central bank book entries held by commercial banks and used to satisfy central bank reserve requirements and settle interbank payments.

    4. Central bank book entries? Does that include federal reserve notes held by commercial banks?

    5. I'm focusing just on deposits. My thinking here is that the Fed doesn't really control the return on notes. The economy withdraws whatever amount it wants or puts it back to the Fed. But it does control the return on deposits via open market operations, interest-on-reserves, and reserve requirements.

    6. Mike, I apologize... I sound like an ass there. Your name didn't register and I thought you were just wandering in off the street wanting to know what "reserves" were... when in fact I'M the amateur at all this )c:

      (My first clue should have been JP's use of the word "Sproulian.")

    7. Tom:
      No offense given and none taken. Plus you made my day.

    8. Mike, good to know! Thanks. BTW, did you and Nick Rowe ever come to some kind of understanding on the Law of Reflux debate, or is that still a point of disagreement? I read these two:


    9. Tom:
      It depends on your definition of "understanding". If you mean in the sense that the Hatfields and McCoys came to an understanding, then yes. I do remember JP once commenting that Nick had posted a paragraph or two that could have been written by me, and Nick said that he had moved a little in my direction on that issue. It involved the idea that money can be considered convertible as long as the central bank uses open-market operations to target the price level, and that traditional gold convertibility is not required.

      The most common misunderstanding of the Law of Reflux (and I think Nick suffers from this) is the belief that reflux preserves the value of money by assuring that the quantity of money stays within certain bounds. The right view is that the value of money is preserved by the amount of backing that the money-issuer holds for each dollar issued. The issuer can hold 100 oz worth of assets as backing for $100 issued, or 300 oz worth of assets as backing for $300 issued, and either way $1=1 oz., regardless of whether dollars are refluxing to the issuer.

      The confusing thing about all this, and something that I can't seem to explain very well, is that if ALL channels of reflux are closed, then that is equivalent to a loss of backing, and the dollar will lose its value. In that case, reflux matters to the value of the dollar. But if, for example, each dollar is nothing but a claim to 1 oz. to be delivered 100 years from now, and reflux is impossible until then, then the dollar will have value today and for the next 100 years, not because it can reflux in any given year, but because it can reflux after 100 years. And the only reason that it will be able to reflux after 100 years is that it has backing.

    10. ... so, in the limit then, if a dollar is a claim to 1 oz to be delivered an infinite number of years from now, does that mean the dollar will have value today and for the next infinite number of years? But how is that different than ALL channels of reflux being closed? If it's not different, then if M is the number of years in the future when the oz is to be delivered for your dollar, and M lies somewhere between 100 and infinity, where along this timeline does the dollar start to lose its value... and where does it lose all its value? Thanks!

    11. Mike, also this: it's my understanding (from JP and Bill Woolsey) that with the current US $, we could make the following claims:

      UOA name = "dollar"
      MOA = CPI basket of goods
      UOA dollar = some sized slice of the CPI basket of goods which the Fed targets to diminish in size by approximately 2% each year

      However, the Fed does not promise to redeem your dollar for a physical slice of the CPI basket of goods. Not now, and not ever (as far as I can tell).

      Then do the Fed's dollar liabilities have "backing" in the way you mean in that it adds assets to its balance sheet as it adds liabilities (typically US Tsy debt) in an approximately 1:1 ratio? Are you saying the redeemability is not in the CPI basket but instead in the assets held on the Fed's balance sheet somehow? Does that mean you disagree with the UOA & MOA definitions above?

      I would assume that you probably have a problem with Nick's opening argument here that CB issued money NOT be counted as liabilities?


      or how about here:


      or here:


      ... and I suppose you'd probably have a problem with this "construction" of fiat money:


      (DOB, actually asked JP to review that in a comment below)

      Interestingly enough though, when I asked DOB about his views of the Law of Reflux, he had this to say:


      Does your view of the Law of Reflux cause you to disagree with what Nick wrote here?:


      particularly this bit:

      "So I think that commercial banks are macroeconomically important because, and only because, they influence the supply of media of exchange." - N.R.

      I looked at your comments there, but I couldn't tell. Also, it seems you are a bit of a historian w/ regards to money. What's your opinion of David Graeber's book "Debt: The First 5000 Years" particularly in regards to the period of time prior to the invention of coins (pre 600 BCE) and to modern primitive societies (Amazon rain forest, Aborigines in Australia) which he describes as "gift based" economies?

    12. And thanks for your reply above, BTW!! ... I know I threw a lot at you there, but even if you just answer some of it, I'll take what I can get. Thanks!

    13. OK, let me squeeze in one more: I know I should ask Nick, but perhaps you (or anyone here!) knows: When Nick write this (in the banks post):

      "The talented artist cannot create an excess supply of his product; banks (in aggregate) can."

      What do you imagine Nick's reasoning is for including the modifier "in aggregate" there?

    14. Tom:

      We don't get much mileage from talking about an infinite number of years, so suppose we have a bond that promises 1 oz payable in 100 years. That bond will sell for 1/(1+R)^100 today, and will grow at r%/year for 100 years. But if annual handling costs=C, then the bond will sell for 1/(1+R-C)^100 today, and if C=R, that means the bond will sell for 1 oz every year for 100 years, growing at zero. (People are willing to hold it because of its monetary convenience yield.)

      I have to run and take care of some family obligations. I'll be back in 2 hours, but I probably can't answer things that you posted as links. I'll have to limit it to things that you actually state in a paragraph or two.

    15. Tom:

      OK, I'm back (but it's getting late).

      1. About backing and convertibility: Think of when the Bank of England suspended gold convertibility in 1797, and resumed it 24 years later in 1821. That was the period when the whole (fallacious) idea of fiat money originated, and it was a simple case of people mistakenly thinking that inconvertible=unbacked. All through that period, the BOE only ever issued pounds in exchange for a pound's worth of assets. The Fed has done the same thing, except that the gold suspension has lasted 79 years.

      2. Nick says federal reserve notes are not the fed's liability. I say they are.

      3. I haven't read Graeber's book. It's out of my area of interest. I mostly read about American colonial currencies these days.

      4. Nick says banks in the aggregate can create an excess supply of money. Yes, technically they can, but rational bankers won't, any more than artists will create excess supply.

    16. Mike, thanks!

      re: origin of fiat money: What about 12th century China?

    17. Mike, also re: Rowe/Woolsey on MOE being hot potato: I read through some of the debates at Glasner's site w/ you, David, Nick, Sumner, JP, Ritwik, and Bill. My thinking right now is that I don't buy anybody's position there 100%. I'm guessing that Nick is incorrect that state that bank money is a HP and that's there's no reflux AT ALL. Certainly loans CAN be repaid... and likewise other bank products (CDs, savings accounts, etc) purchased. But I can see some logic to he and Woolsey's point. I'm running at "warm potato" right now... or perhaps "sack full of potatoes of varying temperatures." Part of it depends on the "moneyness" (do you like that JP? Probably hate it I suppose) of the potatoes were talking about.

    18. Tom:
      It usually helps to think of stocks, bonds, or other securities, and ask if there is a hot potato effect for them. Of course, financial economists all agree that the value of those things is determined by backing alone, and not by any kind of hot potato effect. If, for example, bonds start being used as money, then we can at least imagine that the bonds might rise in value. The problem is that those bonds can be issued instantly and costlessly in infinite amounts, and any monetary premium would be arbitraged away.

    19. OK Mike. I'll mull it over. Thanks.

    20. Mike, I find this comment by Nick Rowe to be interesting:

      "This very old debate over the Law of Reflux is what is at the root of the very modern debate about whether Quantitative Easing can work. Those who argued for the Law of Reflux argued that an excess supply of money could not cause inflation [update: because any excess supply of money would immediately flow back to the issuer]. Banks could only cause inflation by lowering the price of money in terms of gold. Modern proponents of the Law of Reflux argue that banks can only cause inflation by lowering the rate of interest."


      He seems to imply that if you don't accept his HPE for MOE (law of reflux doesn't hold for MOE), that you can't accept that QE could work.

      Do you agree that's his logic? What do you think of that logic? Do you accept that QE can work [to raise inflation]?

    21. Tom:
      You've found a great example of why my discussions with Nick so often go off the rails. In just about every sentence, he mis-characterizes the law of reflux.

      Now for the hard part, actual refutation:

      "an excess supply of money could not cause inflation [update: because any excess supply of money would immediately flow back to the issuer]."

      Start with a bank holding 100 oz as backing for $100 that it has issued. Customers need another $20 to conduct their business, so they offer the bank 20 oz worth of IOU's in exchange for the bank issuing another $20. There's no excess supply of money here. The $20 wouldn't have been issued unless the customers wanted it badly enough to pay the bank 20 oz. for it. If business slows down, so that people don't need that extra $20 anymore, then the $20 will reflux to the bank as people pay off their IOU's, but this has nothing to do with the value of the dollar, which is always $1=1 oz. (determined by backing)

      "Banks could only cause inflation by lowering the price of money in terms of gold."

      You mean the bank, in spite of having 100 oz backing $100, suddenly declares that $1=0.6 oz? I suppose the bank could do that if it wanted, just like I could hit myself with a hammer if I wanted.

      "Modern proponents of the Law of Reflux argue that banks can only cause inflation by lowering the rate of interest."

      Well, sort of. If the going rate is R=5%, and my bank above starts lending at 4%, then the bank would lose assets with every loan and inflation would result. But we're back to the question of why would the bank do something tantamount to hitting itself with a hammer?

      "He seems to imply that if you don't accept his HPE for MOE (law of reflux doesn't hold for MOE), that you can't accept that QE could work."... "Do you accept that QE can work [to raise inflation]?"

      QE can work to increase the quantity of money, and as long as new money is issued for assets of adequate value, then QE won't cause inflation, but it can relieve any money shortage and end the resultant recession. In my world, the solution to recession is to issue more money for assets of adequate value. Inflation per se is not a solution for anything.

    22. Mike, thanks again! Very interesting.

    23. Mike, that makes sense. I don't follow this though: "If the going rate is R=5%, and my bank above starts lending at 4%, then the bank would lose assets with every loan and inflation would result." Could you elaborate? Are you saying that people would borrow at 4% from the bank, withdraw oz, and then loan them out to someone else at 5%?

    24. Mike, I think I'm starting to put some things together here.

      1. JP says at the ZLB the CB needs to commit to lowering the convenience yield of future reserves.

      2. In Sumner's recent post "HPE Explained" he proposes two cases, 5b, and 5c: both at the ZLB, but in 5b there's no expectation of future rates that are not zero, so you're stuck: bonds are money: no difference. In case 5c however, you're still at the ZLB but there are expectations of non-zero rates, so you're good: price levels will rise.

      3. Commentator "Jared" there, and JP here (in responding to me in his latest HPE post) says we're really between 5b and 5c. The more indeterminate that future non-zero rate is, the closer to 5b. Or perhaps the further out those expectations are of non zero rates, the close to 5b.

      4. In JP's post that he mentions the "Sproulian level" (I think that was #2 on his list of 4 levers), we again get into this idea of committing to keep rates zero (#3: the new Keynesian/Krugman version) or to keeping the marginal convenience yield zero (MM version: #4). But again, this seems to put us closer to Sumner's case 5b above: bonds are money: nothing happens. It's a bit of a paradox isn't it? The more you commit to keeping future marginal convenience yields zero... the *stronger?* the effect to raise prices now... but also in some sense the close we are to Sumner's case 5b where nothing happens.

      5. Nick Rowe makes a great comment right here (trying to shoot down the Sproulian lever and show that it's nothing but Chuck Norris still... just like the two "expectations" based levers: Krugman and MM). Well here's his comment:


      Now look at JP's response. That was my thought too! Almost exactly... but with regards a Nick Rowe article... explaining a concept to DOB (who's link is below) on his blog... only JP put it much better.

      See I was trying to say something like this: If our MOA = CPI basket, and UOA = slice of that basket, then what's the equivalent (under gold standard) of just halving the stock of gold? I say it's doubling the size of the TARGETED CPI slice. But they aren't really equivalent... since one happens right now (the gold) and the other is really a target that we have to get to.

      I think JP's argument there (in my link) in response to Rowe is kind of saying the same thing: he's refuting Nick's attempt to say that Sproulian level is the same as an expectations lever... it's immediate. It doesn't suffer from having to invoke Chuck Norris.

      Do you agree?

      Also, would you claim this is the Sproulian level: Have the Fed buy bags of dirt (and overpay) instead of market priced assets. What do you think? Cullen Roche has said this, but he says that's equivalent to having the CB do fiscal policy.

    25. Mike, do you have a blog? If you do it should be called "The Sproulian Lever."

    26. Tom,

      Yes, even if you hit a point at which the convenience yield is at 0 across all maturities, or Scott Sumner's 5b, the Sproulian lever (buying assets at very wrong prices) will still work.

    27. BTW, my example above w/ CPI and gold is backwards for what to do to get us out of a liquidity trap... what I described is how to get us INTO a recession: both with MOA = gold and MOA = CPI basket.

      The other way of course is double the gold stock or halve the CPI slice target. The other distinction between these two methods (time it takes, etc.) still remains.

      Thanks JP for the feedback!

    28. Tom:
      "If the going rate is R=5%, and my bank above starts lending at 4%, then the bank would lose assets with every loan and inflation would result."

      For example, the bank starts with 100 oz of assets backing $100, so $1=1 oz. The bank then lends another $200 for 1 year at 4%, even though the market rate is 5%. The banker then has an IOU promising 208 oz in 1 year. But if he had lent at 5% like he was supposed to, he'd have an IOU promising 210 oz in 1 year. The present value of the 208 oz IOU is 198 oz (and the PV of a 210 oz IOU is 200 oz.). So the bank has issued a total of $300 (=100+200), backed by assets worth 298 oz (=100+198), so each dollar is now worth about .993 oz. (=298/300). The loss of 2 oz worth of assets has caused the dollar to lose about 0.7% of its value.

    29. Tom:
      I haven't really wrapped my head around JP's concept of the convenience yield, and I get even more confused when folks start talking about the zero lower bound, etc. So I'm going to have to keep quiet on the subject for now.

      As for central bankers buying bags of dirt [or was it the other way around?], there would be (say) 10% more federal reserve notes laying claim to the same assets as before, so each FRN will lose 10% of its value. There will be no stimulative effect from this, since real money balances held by the public are unaffected.

      My wife is already teasing me about the "Sproulian lever", so I'll just try to steer the discussion back to the term "Sproulian purchase".

    30. Re: your wife: Lol

      re: you bank example: Ah... so the bank causes inflation (dollar lost 0.7% of value), but the bank "hit itself in the head" while doing this (loss of $2 of (nominal?) assets for the bank). Basically the bank lost $2 of nominial equity, but those dollars all lost 0.7% of their value. Both things happened. Correct? So the bank lost $2*0.993 "real" equity (as measured from the time prior to the loan).

      re: "the other way around" Haha! :D

      you write:

      "there would be (say) 10% more federal reserve notes laying claim to the same assets as before, so each FRN will lose 10% of its value. There will be no stimulative effect from this, since real money balances held by the public are unaffected."

      I'm confused... I'm understanding that the goal is to cause the price level to rise over a period of time, and that happens due to sticky prices/wages and an instantaneous change in the stock of the base (MOA). Isn't the argument that sticky prices and sticky wages don't adjust instantaneously?... but if the Fed buys bags of dirt, isn't that the same as if we'd dropped 10% more gold into an economy on the gold standard? I.e. that in the long run, prices should rise 10% in both cases, but it'll take a while to get there? If we didn't have sticky prices & wages, then wages & prices would adjust instantaneously and we'd have just changed measuring sticks? I.e. nothing else would have changed? (I was going to write "we'd still be at the ZLB and in a recession" but if wages and prices were NOT sticky, I'm getting the idea that this would help to prevent recessions, so we wouldn't be there in the first place most likely.)

      Here's Sumner on the effects of a change in the long run value of the MOA (his Case 4) by an immediate increase in the stock of MOA w/ "sticky prices":

      "if prices are sticky then other things will change to bring about a short run equilibrium in the gold market, before the price level has had time to fully adjust. And obviously one of those “other things” might be a change in interest rates. Other “other things” include changes in asset prices and real output."


      So what's the key difference between Roche's-bags-o-dirt & Sproulian Lev.... uh, "Purchases" (with all due deference to your wife)?

    31. Sorry, the GOAL is to get out of a recession while at zero rates.

    32. Tom:

      I've discussed that nominal/real equity idea under the heading of "inflationary feedback" in a couple of papers: "The Law of Reflux" and "There's No Such Thing as Fiat Money". You've pretty much got the right idea.

      There are no sticky prices or wages in my world. That's an illusion created when a bank does an ordinary open market purchase and issues 10% more money while its assets also rise by 10%. A quantity theorist (Like Nick or Scott) sees the money supply rise 10% and expects prices to rise by 10%. But prices don't rise, because the bank's assets have risen in step with its issuance of money. Quantity theorists don't see this. They explain the stable prices by claiming prices are sticky, while the backing theory says that there was no reason for prices to change, since bank assets rose in step with money issuance.

      About stimulus: Here again, I live in a different world. A recession happens because there is a shortage of money. For example, the economy has $100 (=100 oz) of money in circulation, but people need 110 oz worth of money in order to conduct business efficiently, without being forced into less efficient systems like barter. Solution: The people most in need of money will offer the bank 10 oz worth of stuff in exchange for the banker issuing $10 of new money. Real cash balances rise from 100 oz to 110 oz, people now have enough money, and the recession ends. There's no inflation, since $1=1 oz throughout. But quantity theorists see that the 10% increase in the money supply did not cause a rise in prices, so they start talking about sticky prices.

      If the bank had instead printed $10 and bought bags of dirt, then there are 10% more dollars, each worth 10% less than before, so real balances are unchanged at 100 oz. Since there is no new money in real terms, there is no relief for the money shortage, and there is no stimulus.

    33. Mike, so how do Sproulian Purchases differ from Roche's-bags-o-dirt? It's not clear to me if you gave an example of a Sproulian purchase here.

    34. Tom:

      JP invented the term 'Sproulian purchase', and I think he meant it as "the central bank drastically overpays for stuff it buys". So it's a less extreme version of a Roche purchase. On backing theory principles, a Roche purchase would cause inflation and would not affect real cash balances and so would have no stimulative effect, while a Sproulian purchase would increase real balances while also causing inflation, so it would have some stimulative effect. But a true real bills purchase would have the bank issuing a new dollar for a dollar's worth of stuff, so there would be no inflation, but the expansion of real balances would be stimulative.

    35. What kind of stuff (for a true real bills purchase)? Tsy bonds? I'm guessing "no" but I don't know... because that's what we've got now with QE, but with apparently limited stimulation.

      Are you advocating Spouling purchases (mix of inflation & stimulation) for our current situation?

    36. Tom:

      Here's a classic statement of the real bills doctrine:

      "“The notes of the Bank of England, “the committee observes, “are principally issued in advances to government for ”the public service”…and in advances to the merchants upon the discount of their bills.” (Charles Bosanquet, Practical Observations, p. 53.)"

      so it would be a mix of government bonds plus "real bills". Personally, I'd say anything of value is potentially ok, from real bills to government bonds to lottery tickets. But historically, bankers preferred real bills, since they found that newly issued notes issued that way were less likely to reflux on them the next day.

      I suppose it's strange to say, but I'd never advocate Sproulian purchases. The only kind of purchases I advocate are purchases at the prevailing market price. That would work for our current situation, but since I'm a free banker at heart, I should point out that private banks, left to themselves, would naturally follow the real bills rule of only issuing new money in exchange for stuff of adequate value. No central bank is needed.

    37. Ah, thanks for the clarification. So is QE a "true real bills purchase" even though it's with a central bank? Or is non-central banking (Free Banking) required?

      If QE is a true real bills purchase, and true real bills purchases are stimulative, do you regard QE as stimulative? Why or why not? Is it too much or not enough stimulation?

    38. Tom:

      The real bills doctrine has been around long enough, and has been stated and mis-stated in enough ways, that it's hopeless to try to define a "true real bills purchase". But if we ask the question differently, and just ask "What's the right thing to do?", then it's easy to see that as long as any bank, central or otherwise, gets assets worth $100 for every $100 that it issues, then that bank will not cause inflation or recession. Since QE seems to be doing just that, I have no complaints about it.

      Will such purchases be stimulative? If the economy is suffering from a tight money condition, then yes. The new money will relieve the tight money condition, and people who previously couldn't trade, or were reduced to inefficient things like barter, will be able to trade, and business will pick up. Note that if the economy is short of cash to the tune of $100, then people will eagerly bring $100 worth of their stuff down to the bank, asking the bank to issue $100 in exchange. A well-functioning bank does not need to actively try to inject more money. It only needs to accommodate its customers' requests for money. That's the argument in favor of free banking: Banks will naturally issue the right amount of money, just like farmers naturally grow the right amounts of apples and oranges.

      The other side of the question is, what if the economy does not currently have a money shortage? In that case, new money will do no good, but of course in that case, customers would not be bringing in their stuff to the bank and asking it to issue money.

      I think of 'stimulus' like taking a drink of water. If you're thirsty, a drink can have miraculous effects. If not, the drink does nothing.

    39. "Banks will naturally issue the right amount of money, just like farmers naturally grow the right amounts of apples and oranges."

      Mike, so the scenario you describe sounds a lot like people coming into the bank bringing their signed loan documents in exchange for money. Does that count? A swap of IOUs. But of course there's the collateral too... the house for example, that the new borrower is purchasing. How does that fit into your description? Both the mortgage and the attached collateral (house) have value. Do they count together as "stuff" people bring into the bank in exchange for an issuance of $?

    40. Tom:

      Counting both the mortgage and the house is double-counting. Someone who needs $100 will bring his $100 IOU to the bank, and the IOU is backed by a $100 lien against his house.

  10. Interesting stuff.

    Random question:

    How do you define convenience yield without referring directly to the inconvenience of paying the discount rate for borrowed reserves from the Fed – which itself is set as a direct function of the target fed funds rate?

    1. (thinking randomly about pre-2008 in that question)

    2. The relationship between the convenience yield on an asset and its rental rate is governed by arbitrage. If the rental rate is too high, no one will bother to borrow the asset since the expected convenience yield is outweighed by borrowing costs. Rentors will have to reduce the fee at which they are willing to rent out their asset until they attract borrowers.

      The old discount window offered to rent out reserves at above market rates. There was also an implicit cost to using the window -- it signaled weakness. As long as borrowers were of good credit, it made sense for them to secure the convenience of reserves in the lower-cost fed funds market. Only when they were locked out of the fed funds market would a bank turn to the more costly discount window.

  11. The effect on prices as a result of the convenience yield of banks is disproportionate when compared to non banks and causing monetary policy to be ineffective. Banks create more purchasing power when they take greater on risk (new deposits through lending and they get new created reserves). Others just use their existing purchasing power when they take on greater risk.

    Therefore most of the effect on markets is caused by the banks not just economic agents in general.

  12. Hi JP,

    I've rewritten my fiat currency framework and was curious to get your thoughts on it. I do mention convenience yield in it as means to control the quantity.


    Let me know,

    1. Neat. I think granting a central bank the ability to set negative interest rates make a lot of sense too.

  13. JP, I'm not sure I understand this sentence:

    "One way to reduce convenience yields five years hence would be to promise that the then-supply of base money will be sufficiently broad so as to ensure that the marginal deposit yields no convenience flows."

    Could you elaborate?

    1. If the central bank guarantees that it won't withdraw base money five years from now, then base money at t+5 years will be plentiful. Therefore the marginal convenience yield on reserves will fall to zero at t + 5 years.