Krugman’s Flawed Model of Open Market Operations

By Dan Kervick

In my recent post Escaping from the Friedman Paradigm, I noted the following remark by Paul Krugman on the way monetary policy ordinarily functions when interest rates have not fallen to the zero bound:

… people are making a tradeoff between yield and liquidity – they hold money, which offers no interest, for the liquidity but limit their holdings because they pay a price in lost earnings. So if the central bank puts more money out there, people are holding more than they want, try to offload it, and drive rates down in the process.

And I was very critical of this model of central bank open market operations.  As I put it then:

… what in the world can it mean to say the central bank “puts money out there” that people then try to “offload”?   How can that happen? The central bank doesn’t stuff money into people’s pockets, and it doesn’t force them to hand over their financial securities in exchange for money.  It offers money in the open market in exchange for securities.   So if people preferred the securities to the money, they would’t have traded the securities for the money in the first place.  It makes little sense to say that  financial institutions first seek money for their securities on the open market, and then having too much unwanted money hanging around seek to dump it by obtaining securities for their money.

Interestingly, Krugman offers up the very same flawed model in a piece in yesterday’s New York Times:

Now, think about what happens when the Fed makes an open-market purchase of securities from banks. This unbalances the banks’ portfolio — they’re holding fewer securities and more reserve — and they will proceed to try to rebalance, buying more securities, and in the process will induce the public to hold both more currency and more deposits.

Again, this model of Fed securities purchases makes no sense to me. The central bank doesn’t buy securities by exercising some kind of eminent domain.  It doesn’t force banks to sell their securities.   Rather, the Fed Open Market Desk announces it’s intention to buy securities and the primary dealers then submit offers.  In other words, the Fed offers to pay money for securities in the open market, and banks only sell those securities if they accept the price determined by an offer they themselves have made.  So it makes little sense to say that at the end of this process the banks find their portfolios in an undesired condition and therefore need to “rebalance” them.

When Krugman says that the banks then “induce” the public to hold more currency and deposits, I take it he means that the banks then lower their lending rates so that more people are willing to borrow at the new, lowered rate.  This they do, according to Krugman, in order to carry out the portfolio rebalancing he has described.  But I believe the process here works quite differently.  The Fed has no ability to push dollars and deposit balances out into the economy by forcing undesired money on banks which will then force the money onto the public, but achieves its aims by targeting interest rates.

Banks generally make their money on the spread between the rate they must pay for additional funds and the rates they are able to charge for the loans they make to the public, and the key rate in the market for funds is the rate paid in the interbank lending market (the “Fed Funds” rate.)   The Fed has shown that it has the ability to target this rate with very little volatility.  Thus it simply announces the new rate it wants to set, and participants in the market move automatically to that new rate.  If the rate is lower than it was previously, this will increase the banks’ willingness to loan at lower rates than previously and will thus build up aggregate bank deposit balances.  This will increase the banks’ aggregate demand for reserve account clearing balances to handle the larger volume of payment obligations that are the natural consequence of the expansion of deposits.  The primary dealers, who possess reserve accounts at the Fed and are themselves the key supplying participants in the interbank market, will attempt to satisfy that demand by selling more securities to the Fed in exchange for dollar reserve balances.  And the Fed then buys those securities via open market auctions.

So Fed open market purchases are not aimed to force money through the system and out into the hands of the public.  They are designed to support and accommodate the higher demand for reserves that the Fed itself has influenced by announcing a new target Fed Funds rate.  The Fed influences lending and expands bank balance sheets by targeting prices, not quantity.  And of course, none of this works any longer once the Fed Funds rate has fallen close to the zero bound, and the Fed cannot set the rate any lower.

Also, Krugman is still attempting in this new piece to defend the loanable funds model of credit markets.  He often seems to suggests that when banks want to increase their loans, they satisfy their increased demand for funds by attracting more deposits from the public, presumably by offering better rates for CDs and term deposits, better services etc.  But while that might make sense from the standpoint of some individual banks, it makes little sense from the standpoint of the banking system as whole, and cannot explain the function of bank credit markets in response to an increased demand for consumer loans. For the most part, when a bank customer deposits funds in a bank, those funds come from transfers from another bank account.  For example, your employer’s paycheck to you is a payment order issued against your employer’s own deposit account at some bank.  if you deposit your employer’s paycheck in your bank account, your bank will ultimately collect the funds by receiving a transfer into its reserve account from the reserve account of your employer’s bank.  The same sort of transfer occurs if you move your deposit account from one bank to another to take advantage of better terms. An individual bank can absorb deposits from its competitors and use those funds to expand its lending; but the banking system as a whole cannot in any significant way increase its lending by sucking up deposits.  Instead, banks extend their lending and deposit account liabilities first, which increases its subsequent demand for larger clearing balances in its reserve accounts, which the banking system as a whole then meets by absorbing injections of funds from the Fed as part of the open market operations described above.

Paul Krugman seems determined to be the last dinosaur standing in defense of some outdated models of central bank operations.

Cross-posted from Rugged Egalitarianism

Follow @DanMKervick

30 responses to “Krugman’s Flawed Model of Open Market Operations

  1. Krugman is so annoying it’s hard to tell if he’s smart or not. But he certainly is clever:

    “discussion of these issue brings out the people who believe that they have discovered the hidden secrets of the monetary universe, somehow missed by generations of economists.”

    BTW, don’t all advances in knowledge result from people who have discovered the hidden secrets missed by generations of their predecessors? I guess Krugman has found the very last secrets of economics, everybody else can stop now.

    I asked on his blog

    “If the bank wanted to keep its portfolio “balanced”, why did it sell securities to the Fed in the first place?”

    • Yeah, he’s trying to inoculate himself against criticism by designating all critics in advance as kooks.

      And yes, doesn’t Krugman think that when he was a young blade working on his trade theories he was discovering things that previous generations of economists had missed.

    • Krugman smart? That’s a tough one indeed?

      I recall how in 2008 he claimed the rise (in the paper price) of oil was due to simple supply and demand – – evidently he’d never heard of the Baltic Index????) either completely ignorant or oblivious to the fact that the paper price was 13.8 times higher than the real price due to all the speculation on exchanges (especially the InterContinental Exchange, or ICE, by Goldman Sachs and Morgan Stanley) by a host of banks, investment banks and hedge funds, as well as speculating up the price of substances used in the refinery process, along with transportation prices (Freight Forward Futures, etc.).

      Krugman has repeatedly made similar “mistakes” or misinformational pronouncements.

      But, and Dan will probably support the simpleton status quo and disagree with me, but there is no one on the Group of Thirty ( who is anything other than a front man for the speculators and central bankers. That is the way it was set up by the Rockefeller Foundation, who originally financed it back in 1978.

      Krugman, Larry Summers, Trichet, Draghi, Timothy Geithner, William Dudley, Mervyn King,
      Ernesto Zedillo (whom the Mexican government recently attempted to extradite from the USA) et al., serve one primary purpose: support the speculators and central bankers!

  2. Note the role of CD’s in the final process … if banks extend lending they can also gain additional larger clearing balances in their reserve accounts if they can persuade depositors to shift funds from account with higher required reserve requirements to accounts with lower required reserve requirements.

    These kind of operations were a more important means of increasing clearing balances when the Fed was maintaining the pretense that it could effectively target monetary aggregates, and so the system periodically experienced constrained reserves.

  3. Thornton Parker

    A question, Dan. You write, “Banks generally make their money on the spread between the rate they must pay for additional funds and the rates they are able to charge for the loans they make to the public…”. But is this true? I was under the impression that commercial banks make most of their money by charging interest on credit that they create out of thin air, while borrowing and paying interest on enough money from other sources to maintain adequate reserves.
    Tip Parker

    • Thornton, yes more or less, and that’s where the spread comes in. When they create additional deposits via lending they may need to acquire additional reserves. What they have to pay for the reserves is a cost and the what they expect to earn from the loans is a benefit, and the money they expect to make is the difference between the two.

      • And right now, the banks in the aggregate have excess reserves, so for some of them the marginal cost will be zero.

      • But with a reserve ratio of 10%, the banks would only have to borrow (worst case) 10% as much as they lend, right? If they had just barely enough to start with, and the deposit came to their own bank. And for the banking system as a whole. (Clearly one bank could get in trouble if it were to end up with all the deposits and none of the loans.)

        So how come their lending rate (prime rate) goes up and down in the same absolute quantity as the Fed Funds rate? Seems to me if they borrow $1 at 5% and lend $10 at 8%, they’re making a lot more than the 3% spread on the rates. They’re paying only 0.5% of the loan amount and making 8. If the Fed Funds rate goes up from 5% to 10%, and they raise the prime rate to 13%, they’re making a spread of 12%. Seems to me that for every 1% increase in Fed Funds, the prime rate should increase 0.1%, if their margin is to remain constant.

        If you say that their cost is the rate they pay on deposits, isn’t that the “loanable funds” theory?

        • That’s a good point. The spread in question isn’t the spread between the two rates, multiplied by the amount of the loan, but the spread between the cost of additional funds needed and the expected earnings on the loan. The cost of additional funds is the amount of funds needed times the rate one must pay for them.

          If no additional funds are needed, then the cost to make the loan (beyond the cost consisting of the added liability itself) is zero.

          One thing I think we need to bear in mind is that the costs of funds required to expand lending is not necessarily determined by the formal reserve requirement. It is determined by whatever the bank decides are its liquidity needs, beyond what the reserve requirement might mandate. If a bank has no excess reserves, and is making a type of loan that is going to place exceptionally high demands on its clearing balances to fulfill payment orders, with a more than usual quantity of reserves leaving the bank as a result, it will have to acquire more funds than merely meeting the reserve requirement would require.

          The reason this is not the loanable funds model, I would say, is that (i) the bank doesn’t need to acquire the funds first before making the loan and (ii) there is not a fixed stock of funds in the funds market from which the bank must acquire the funds by borrowing. If the banks in the aggregate require more funds, those funds automatically are made available by the Fed so that it can maintain its interest rate target.

          • OK, I see now. If the loaned funds end up in another bank, then when the check clears they have to come up with reserves in the amount of the loan, not just 10% of the loan. Reserves have to be 10% of deposits (oversimplification, I know) not 10% of loans. And even today, with excess reserves, an individual bank would still have a cost of funds which could be up to the (foregone) interest rate on their excess reserves times the full amount of the loan.

            • Yes, that seems right to me. Although as we are both aware, we are creating a misleading picture by looking at single loans as though they were dealt with as a single atom of activity in isolation. The bank has liquidity managers who use various tools to estimate the volume of payments it is going to have to make in any given time, and the volume of payments it is likely to receive in the same period of time.

              • Mr. Kervick. Excellent post. Your posts on the monetary and banking system provide incredible insights and understanding. I’m from India. My question is why do some central banks like India’s keep a reserve requirement for banks (they call it here cash reserve ratio or CRR). Don’t the banks bring reserve balances equivalent to the amount of equity capital brought? I was wondering how much of reserve balances as a % of total deposit balances does a system require on an average to support the total bank deposits in a banking system.

                Am I correct when I say that when a bank purchases government securities, there is an equivalent amount of outflow of reserve balances, however the same may not be so when a loan is created as some of the payments from the deposit account created from the loan creation may end up with the same bank? Indian banks are required to purchase government securities equivalent to 23% of their deposits which is termed as SLR (statutory liquidity ratio) here . Could that be a reason why India’s central bank would want a reserve requirement?

                Also when the Indian central bank increases the CRR or SLR, the established view here is that it sucks out liquidity from the banking system. As per MMT, given that money supply is really not determined by the central bank and that it can only increase the interest rate to dampen credit (money supply) growth, wouldn’t you say that the established view here is erroneous? Here, the Indian CB sets a money supply growth target and when it exceeds that, it increases the CRR, SLR, or decreases the amount of reserves a bank can borrow from the repo window. Wouldn’t you say that the Indian central bank is operating on a misconception that it can target money supply?

                • I was wondering how much of reserve balances as a % of total deposit balances does a system require on an average to support the total bank deposits in a banking system.

                  Excellent question. Offhand, my intuition is that the aggregate amount needed is roughly 0%. Individual banks are losing reserves as depositors make various types of payments to depositors at other banks, but they are gaining reserves in the same way as depositors at other banks make payments to depositors at their bank. Similarly, banks are losing cash reserves as depositors make withdrawals of cash, but they are gaining cash reserves at the same time as depositors make cash deposits. So unless the public suddenly feels the need to withdraw and hold more cash, there is no net change. It seems to me that what the effect of reserve requirements is more important at the level of the individual bank. It allows the bank to experience short-term volatility in its net payment obligations without having to overdraw its account at the central bank or liquidate other assets.

                  Some systems, such as Canada, Sweden and New Zealand, have no reserve requirement.

                  Am I correct when I say that when a bank purchases government securities, there is an equivalent amount of outflow of reserve balances, however the same may not be so when a loan is created as some of the payments from the deposit account created from the loan creation may end up with the same bank?

                  Yes, that seems right to me. Also, even if the borrower uses the borrowed funds to make payments entirely to depositors at other banks, that takes some time, so the corresponding amount of reserves do not leave the bank immediately.

                  • Mr. Kervick,

                    Thanks for your reply. I probably didn’t put the question properly when I asked how much of reserve balances as a % of total deposit balances does a system require on an average to support the total bank deposits in a banking system. I didn’t mean the reserve requirement that a bank needs to maintain with the CB but the total stock of CB reserves in the system. Or maybe I haven’t understood the concept of CB reserves correctly. I’m not sure about the terminology : Is central bank reserve balances the same as central bank reserves?

                    I’m not sure about this but I think the monetary base of an economy consists of central bank reserves and physical currency in the form of notes and coins. So suppose that total deposits balances in the US are $15tn and monetary base is $1tn out of which say physical currency is $250bn. Wouldn’t the balance be $750bn of central bank reserves, which would be 5% of total deposit balances? And over time, as deposit balances increase with economic growth, wouldn’t CB reserves increase in aggregate to accomodate loan and deposit growth?

                    I had another question. Since loans create deposits and banks only need reserve balances according to the level of the likely payment obligations, are most of the deposits that a bank holds created out of the loans it created, and deposits obtained from the public more or less at the level of reserves required to settle likely payment obligations?

        • Thornton Parker

          This is the reason for my question.

  4. Most open market purchases are called “repos.” The Fed buys a security at a price with the promise to sell it back after a certain time for a lower price. The owner of the security makes money on the difference. In essence the Fed is borrowing the security and paying interest on the loan. “Repo” is from the word “repurchase.”

    There are “reverse repos” where the Fed sells one of its securities with a promise to buy it back at a higher price in the future, effectively borrowing money from its member banks.

  5. “So if the central bank puts more money out there, people are holding more than they want, try to offload it, and drive rates down in the process.”

    People arent holding more money than they want and they aren’t trying to offload it. hahaha.

    “The central bank doesn’t buy securities by exercising some kind of eminent domain. It doesn’t force banks to sell their securities.”

    I think the market makers are expected to make bids to maintain their market making status.

  6. Krugman has a blog advising on how to manage the economy because he has somehow discovered the hidden secrets of the monetary universe.

    Does anyone know if banks actually adhere to their reserve requirements?

    • Yes they do, if only because if they haven’t met them by the required date, the Fed will automatically credit the reserve account with the needed amount and charge the penalty rate for the funds.

      But there is a two-week calculation period and a two-week maintenance period that intervenes between the expansion of lending the date when such a penalty would be imposed, and during that time there are all sorts of tricks the bank can use to come into compliance. Scott Fullwiler is the expert on this stuff. Google his paper on Central Bank Operations.

      The Fed recently changed its policies to bring all banks into a single, synchronized maintenance period schedule.

      • ok thanks will look up Scott Fullwiler’s paper.

      • In Australia there is no reserve requirement at all just capital requirements:

        “It is important to recognise that operating procedures for monetary policy in Australia – or in any country for that matter – do not conform to this traditional model. There are no reserve requirements on banks in Australia; their demand for central bank funds comes from their need for settlement balances, not reserve requirements. The Reserve Bank’s operations focus on establishing a price at which these are made available, rather than on the quantity of liquid funds – i.e. they operate to affect the cash rate, not the money base.”

        • Canada has a similar system of no reserve requirements. Is the Australian banking sector also dominated by a mere handful of banks? (Canada’s banking sector effectively has 5 banks, while the US banking sector is not nearly so concentrated.)

  7. Pingback: Dan Kervick: Krugman’s Flawed Model of Open Market Operations « naked capitalism

  8. I wish that the US would skip the reserve requirement altogheter. It is completely irrelevant and obviously creates a lot of confusion.

  9. krugman is flawed period!

  10. “Paul Krugman seems determined to be the last dinosaur standing in defense of some outdated models of central bank operations.”

    I think that’s unfair. Krugman is a prominent Mainstream Economist but he is gradually changing and updating his models. Sure he’s clinging to the past and moving slowly but most of the profession isn’t moving at all or is going backwards. I would expect governments and their bank appointed advisers will be maintaining the old orthodoxy long after Krugman has given it up. If he’s a dinosaur he’s closer to the front of the pack than the back.

  11. I’m torn between sympathy and wanting to change the situation but I try to be polite when I comment. I really do.