Josh Hendrickson on IOR and demand for bank reserves

Josh Hendrickson has an important new (forthcoming) paper in the Journal of Macroeconomics. Here’s the abstract:

Over the last several years, the Federal Reserve has conducted a series of large scale asset purchases. The effectiveness of these purchases is dependent on the monetary transmission mechanism. Former Federal Reserve chairman Ben Bernanke argued that large scale asset purchases are effective because they induce portfolio reallocations that ultimately lead to changes in economic activity. Despite these claims, a large fraction of the expansion of the monetary base is held as excess reserves by commercial banks. Concurrent with the large scale asset purchases, the Federal Reserve began paying interest on reserves and enacted changes in its Payment System Risk policy. In this paper, I estimate the effect of the payment of interest on reserves (as well as other payment policy changes) on the demand for daylight overdrafts through Fedwire. Since Fedwire provides overdrafts at a fixed price, any fluctuation in the quantity of overdrafts is a change in demand. A reduction in overdrafts corresponds with an increase in the demand for reserves. I show that the payment of interest on reserves has had a negative and statistically significant effect on daylight overdrafts. Furthermore, I interpret these results in light of recent theoretical work. I argue that by paying an interest rate on excess reserves that is higher than comparable short term rates, the Federal Reserve likely hindered the portfolio reallocation channel outlined by Bernanke. Thus, the payment of interest on reserves increased payment processing efficiency, potentially at the expense of limiting the ability of monetary policy to influence economic activity.

And here Josh describes the role of overdrafts:

Reserves are held both to meet unexpected withdrawals and to settle payments between banks, with the latter motive being substantially more significant. Large-scale wholesale payments processed through Fedwire are done through a real-time gross settlement system. This means that payments to and from banks are credited and debited in real-time. If the bank has insufficient reserves to cover a debit in their reserve account, the bank is extended credit that is expected to be repaid by the end of the Fedwire day. Historically, a large portion of these daylight overdrafts have been backed by pledged collateral of the bank.3 It follows that banks can hold either highly liquid, short-term debt such as Treasury bills and other forms of short-term debt that can be sold quickly or used in repurchase agreements in the event of an overdraft or banks can hold reserves sufficient to fund payments. Prior to the paying of interest on reserves, banks faced a trade-off of holding short-term debt or reserves. If the bank held reserves it would enable more efficient settlement of payments at the cost of foregone interest on short-term debt. Holding interest-bearing assets provided a positive rate of return, but overdrafts due to insufficient reserve balances were subject to fees. With the payment of interest on excess reserves, this tradeoff is eliminated because the policy change allows banks to earn interest comparable to alternative assets while avoiding fees associated with overdrafts. The market for daylight overdrafts therefore provides an opportunity to analyze the effect of the payment of interest on reserves.

The paper confirms that the adoption of interest on reserves in October 2008 was indeed a serious (contractionary) policy error.  More broadly, it reconfirms the importance of looking beyond interest rates when evaluating monetary policy, and specifically the importance of the portfolio rebalancing channel as a transmission mechanism for changes in the quantity of base money.

 


Tags:

 
 
 

49 Responses to “Josh Hendrickson on IOR and demand for bank reserves”

  1. Gravatar of Iskander Iskander
    20. August 2017 at 13:19

    Scott, are you aware of David Laidler’s paper “Monetary Policy after Bubbles Burst: The Zero Lower Bound, the Liquidity Trap and the Credit Deadlock”? https://www.jstor.org/stable/3552306

    He mentions some of your work at one point, but the most interesting thing is that this paper was written in 2004 (Did Laidler predict the next couple of years?). It also criticizes using interest rates as an indication of the stance of monetary policy and looks at how to escape a supposed liquidity trap.

    I don’t know if you’ve covered this before but what do you think made monetary policy too tight prior to the great recession? A supply shock (oil prices?) causing inflation targeting central banks to tighten policy?
    Raising interest rates above the natural rate due to the natural rate falling more than expected due to the savings glut? (to use the interest rate terminology)

  2. Gravatar of Jerry Brown Jerry Brown
    20. August 2017 at 13:57

    When the Fed pays interest on excess reserves I view it as just one way they can use to hit their interest rate target, if that target is above zero. So I don’t think it should be any more contractionary than overall Fed policy is at the time in question. And some aspects of it might be considered expansionary- I mean the government is handing out money with no strings attached to bankers when the Fed pays interest on excess reserves. And they don’t have to do anything at all to receive this money, they aren’t even exposed to any risk that interest rates might change like if they purchased Federal debt, which otherwise is just as safe as reserves at the Fed.

    So I kind of think of it as a ‘preserve the banks’ strategy so they wouldn’t collapse. It provided them with a source of guaranteed income when they were too scared to lend and when people were too scared to borrow. I mean a collapsing banking industry is bound to be contractionary at least short term.

  3. Gravatar of Benjamin Cole Benjamin Cole
    20. August 2017 at 17:15

    The Bank of Japan pays interest on reserves…negative interest, that is.

  4. Gravatar of Max Max
    20. August 2017 at 17:41

    The Fed wanted to use its balance sheet independently of monetary policy, and IOR is one way to do that. But then, some people see the monetary base growing and say “the Fed is changing monetary policy”. It would be less confusing if the Fed expanded its balance sheet without increasing the base, if it doesn’t want to signal a monetary policy change.

  5. Gravatar of dtoh dtoh
    20. August 2017 at 21:46

    Scott,

    If I could summarize,

    The monetary transmission mechanism can be generalized as action causing changes to the amount of assets exchanged for money between the financial sector (i.e. the Fed AND commercial financial institutions) and the non-financial sector (firms plus consumers plus government). To the extent that these changes in the net amount of assets exchanged for money between the financial and non-financial sectors is not offset by changes in velocity, then monetary policy will be effective (i.e. it will impact aggregate demand.)

    The transmission mechanism can be thought of as a two step process. First, through Open Market Operations, the Fed exchanges money for assets (or vice versa) with other financial institutions. In the second step, these intermediate financial institutions exchange that money for financial assets with the non-financial sector (i.e. firms, consumers, and governments). The non-financial sector then uses the money for increased (or decreased) spending causing a change in aggregate demand.

    Under the zero bound condition, some economists argue that any increase in the exchange of money for financial assets will be offset by a decrease in velocity. More specifically, that the financial and/or non-financial sector will simply hold more money (decreasing velocity) and there will be no impact on aggregate demand.

    Bernanke argued that a special case of the transmission mechanism, the portfolio re-balancing mechanism, would at least partially mitigate decreases in velocity. This argument relied on the correct assumption that the financial institutions, who were acting as intermediaries in the second step of the transmission process, might after exchanging assets for money with the Fed choose to simply hold the money rather than exchanging it for new assets with the non-financial sector. The portfolio re-balancing theory postulates that, for various reasons, these financial institutions would choose to at least partly replace the assets sold to the Fed with a more “balanced portfolio” of longer term and/or higher risk assets rather than simply holding the money as cash or as deposits with the Fed. These assets were expected in part to be ones purchased from the non-financial sector thereby stimulating aggregate demand.

    Of course as Hendrickson concludes, if the Fed is offering zero risk high returns assets (i.e. interest paying ER), there ain’t going to be any portfolio re-balancing.

    There are three things I don’t understand, so maybe Scott or someone could elucidate.

    1. Why is it necessary to write an academic paper whose conclusion is so manifestly self-evident to any sentient being with an IQ higher than an aardvark.

    2. Why would anyone (presumably a tax subsidized institution) pay Josh Hendrickson to produce such tautological tripe.

    3. Is Bernanke duller than an aardvark or was he prevaricating about his real reason for bloating the Fed balance sheet.

  6. Gravatar of Matthew McOsker Matthew McOsker
    21. August 2017 at 08:25

    But the banking system cannot in aggregate get rid of reserves, and I am curious to see how the paper deals with that. IOR does add more reserves to the bank system, as QE also did. In totality QE + IOER is contractionary via the interest income channel, less longer term securities that pay higher income are circulating, and the Fed just remits those payments from their balance sheet to the Treasury.

  7. Gravatar of Matthew McOsker Matthew McOsker
    21. August 2017 at 08:45

    One more thing, let’s remember how the ER’s were created, and that was not necessarily the fed only soaking up a supply of t-bills/bonds. For example, a lot of mortgage backed securities were purchased.

  8. Gravatar of Matthew McOsker Matthew McOsker
    21. August 2017 at 08:56

    Sorry, one last thing for now. Look at this overdraft chart since 1986:

    https://www.federalreserve.gov/paymentsystems/psr_data.htm

    Currently, overdrafts look pretty low, and confused why he would say low overdrafts mean an increased demand for reserves? Look at 2008, now that is a time where there was probably a demand for reserves? Or was there simply not enough reserves in 2008 resulting in overdrafts?

  9. Gravatar of Jon Jon
    21. August 2017 at 09:08

    Scott,

    Let’s not forget the contemporaneous discussion was about whether we had a liquidity crisis or a solvency crisis. That is, should we extend credit to the banks and sterilize it or should we do nothing and let the insolvent institutions fail.

    IOR was introduced with the stated purpose of sterilization because the liquidity facilities had grown so large that the fed had exhausted its supply of trills— about half of assets before the crisis — with which to sterilize further expansion of the liquity auctions and swaps.

    mental gymnastics are required to sustain IOR and QE as sensible policies in combination!

  10. Gravatar of Michael Rulle Michael Rulle
    21. August 2017 at 09:59

    I never could figure out what the Fed thought it was accomplishing, other than trading riskier assets for riskless assets with the banks. Was this a shadow “bailout” or capital builder for banks? In the meantime, we managed to destroy the securitization market, as if that, rather than irresponsible lending practices, were one of the causes of the crisis. Now they are oh so carefully going to retrade back reserves for securities. Was this the best the Fed could do? And for what purpose?

    The economy and markets did okay, stable and steady low volatile growth. Was this the intent or a random outcome? Perhaps they were successful. But was this ever explained by the Fed?

  11. Gravatar of Matthew Waters Matthew Waters
    21. August 2017 at 13:04

    “But the banking system cannot in aggregate get rid of reserves, and I am curious to see how the paper deals with that.”

    This is not correct. Roughly:

    Reserves + Notes in Circulation = Monetary Base.

    If the Fed’s one goal was to reduce the monetary base by 10%, it could be done. The Fed sells its assets on the open market. There would be buyers found for Agency MBS or Treasuries. If $1 million in Treasuries is sold, then the monetary base IS reduced by $1 million.

    How would the banks figure out the reserves? Well, that’s not the Fed’s problem. The banks and the market generally would figure it out. If excess reserves were reduced by the sale, then the bank’s don’t have to do anything.

    If required reserves were reduced by the sale, then banks will figure out some way to either reduce total deposits or attract notes in circulation. With higher interest rates, more notes are converted to deposits. If notes are exhausted, then deposits can be exchanged for bank assets by selling the bank assets. At some price, the banking system will find a buyer for its assets which was a depositor.

    To sum up, there is no THEORETICAL reason the Fed couldn’t dramatically reduce monetary base and reserves if that’s what it truly wanted to do. But the Fed’s ultimate goal is 2% inflation. In practice, a Fed hell-bent on reducing reserves would be successful but cause massive deflation in the process.

    But the theoretical exercise also shows the Fed has no real limits on creating inflation and deflation if it really wished. In the 70’s, economists had doubts that the Fed could really tame inflation. It seems preposterous in hindsight, as the Fed could reduce monetary base as much as necessary.

    The theoretical limits of a Central Bank at the ZLB are more understandable, where in theory the Central Bank is limited by its statutory authority. But simply charging a negative rate on excess reserves and restricting printing of new notes has unbounded ability to increase inflation. There is also the threat of far more QE should inflation expectations not increase. Such mechanics were deemed unthinkable, but like reducing reserves, there is no theoretical reason they wouldn’t work.

  12. Gravatar of Matthew Waters Matthew Waters
    21. August 2017 at 13:27

    @Michael,

    When the Fed bought a “riskier” asset, an interest-bearing asset was taken out of the market and a zero-interest asset was replaced. Or at least a near zero percent asset with excess reserves. Excess reserves only yielded 0.25%. It was thought by Fed members at the time that the interest was just too small to matter.

    Indeed, while 10-year Treasuries yielded a lot less than before 2008, they never went down to 0.25%. Certainly Agency MBS did not go down to 0.25%. At least some of the sellers wouldn’t accept the 0.25% and they would buy a longer-term or riskier asset. The other assets are bid up in price and rates are lowered, until more investment has positive NPV.

    This is a very naive explanation of QE, but it’s a good first step. Long-term rates should be reduced and increasing number of profitable investments. Without negative rates on the table, the Fed can push down long-term risk-free rates to zero and only then has it exhausted what it can do with risk-free asset purchases.

  13. Gravatar of Matthew McOsker Matthew McOsker
    21. August 2017 at 15:21

    Matthew Waters, it is correct unless the Fed drains reserves, and maybe I should have qualified that. But the banking system cannot get rid of reserves, unless the Fed does a drain via OMO.

  14. Gravatar of Jerry Brown Jerry Brown
    21. August 2017 at 17:33

    Matthew Waters, I think Matthew McOsker is correct. My understanding is that the private banks find it very difficult to get rid of excess reserves in the system by themselves. Unless they were willing to ramp up lending extraordinarily (by lowering their underwriting standards and interest rate margins, which is risky), and if at the same time, their customers were willing to increase their borrowing. For example, if there ten billion in excess reserves, banks would have to make 100 billion in new loans in order to have no excess reserves (if required reserves were 10% of deposits).

    But like you say, the Fed can conduct open market operations (as in selling securities the Fed already owns at favorable rates to the banks) to soak up excess reserves if the Fed wished to keep the Fed Funds interest rate above zero. Or it can just offer interest on the excess reserves at that rate they want to be around the minimum. So offering .5% on excess reserves is not a tighter Fed policy than targeting the Fed Funds rate at .5%. It is not much looser either, it just keeps the system awash in reserves.

    There are some other things I am not sure I understand about your statement. First off- having enough reserves in the banking system so that the system works is ALWAYS the Fed’s main priority- that is what the Fed was created for and it continues to be a primary function of the Fed. Can’t have the payments system crashing- its bad for business. The Fed will make reserves available at a price to any bank that is not insolvent. And the history of the last ten years shows they have a very flexible definition of insolvency.

    Also, customer deposits are not assets for a bank, they are liabilities for any bank. They may represent a less expensive source of loan funding in some cases, but they remain liabilities that are not discharged until the customer withdraws the deposit.

  15. Gravatar of dtoh dtoh
    21. August 2017 at 18:33

    “But the banking system cannot get rid of reserves, unless the Fed does a drain via OMO.”

    Not at all true. Banks can simply increase loans. Any portion of the loan withdrawn in cash will reduce reserves.

  16. Gravatar of Matthew Waters Matthew Waters
    21. August 2017 at 19:42

    Jerry,

    To clarify:

    1. Yes, deposits are liabilities. I meant the bank can sell its assets to non-banks.

    Let’s say the banking system in general has too low of reserves by $1 mil. The bank that has too low of reserves sells an asset. If it’s bought by another bank, then now that Bank has too low of reserves. Eventually, the banking system can find a buyer outside of it. That buyer has to pay for the $1 mil with a deposit at one of the banks. Overall customer deposits have now decreased.

    2. I meant that how banks actually get to their required reserves is not a concern of the Fed. Let’s say the Fed raises their interest rate targets and therefore has selling OMO’s. It’s also a pre-2008 scenario with almost no excess reserves.

    The OMO’s work through the system to reduce the monetary base. It may or may not be a reduction in reserves versus conversion of notes into reserves.

    It would be an interesting discussion about whether the Fed should be set up like a bank, with the discount window and reverse repo facilities. The Fed can be in charge of currency and wire payments without providing the loan facilities.

  17. Gravatar of Jerry Brown Jerry Brown
    21. August 2017 at 19:43

    dtoh,

    Banks do not generally make loans in cash as far as I know. And most people don’t borrow money and turn it into cash and put it under their mattress either. Most loans end up back as deposits in the banking system so it is difficult for the banking system to get rid of excess reserves in the system on their own without Fed intervention.

  18. Gravatar of Jerry Brown Jerry Brown
    21. August 2017 at 20:17

    Matthew Waters, typically a bank that does not have enough reserves will borrow at the Fed Funds rate from other banks. Banks excess in reserves generally were willing to lend (or do repos) overnight at this rate because otherwise they made no money on their excess reserves. If no other banks are excess in reserves, then a commercial bank could attempt to sell assets to the nonbank private sector (in which case the payment is not a deposit, but it does reduce overall deposits usually), or they could just borrow the reserves from the Fed, which will always make them available but at a higher penalty cost than the Fed Funds rate. The point is that the reserves will always be made available, since the Fed is in the business of keeping the business going. But this is all very different from the banking system getting rid of excess reserves without Fed help.

  19. Gravatar of dtoh dtoh
    21. August 2017 at 21:34

    @Jerry Brown
    Of course some portion of the loan gets converted to cash. That’s how OMO work.

    Typically a bank doesn’t disburse loans in cash, but some portion of the loan gets converted into cash directly by the borrower or indirectly e.g. through payments to suppliers or employers who then withdraw cash. Obviously most of the loan stays in the form of a deposit, which is why base velocity is something like 5 instead of 0.5 and why relatively small OMO can have a large impact aggregate demand.

  20. Gravatar of ssumner ssumner
    21. August 2017 at 22:06

    Everyone, dtoh is right. The banking system can get rid of reserves if it wants to, without any help from the Fed. If they make deposits less attractive, people will convert deposits into cash.

    Banks choose how much reserves they wish to hold.

  21. Gravatar of Jerry Brown Jerry Brown
    21. August 2017 at 23:38

    Scott, deposits are typically banks’ least expensive source of funding- they don’t want to get rid of them when people actually deposit money in their accounts. Banks make money by lending, which they have to fund when the deposit they create for the borrower gets transferred to a different bank. If they get rid of their depositors they have to pay more for the funding from elsewhere. What you say makes no sense if you are talking about all the banks.

    The other thing is that banks make loans but, as dtoh admits, most of every loan stays as deposits in the banking system as a whole. The loan creates deposits, which in a scenario where only ten percent of deposits become required reserves you are talking about loaning out something like 9 times whatever excess reserves are in the system to eat up the excess reserves. As of July 2017 there are over 2.1 trillion in excess reserves. One solution for banks would be to make like 18 trillion in loans. Like they have the capital or the customers to do such a thing. Another solution is the banks could buy back those mortgage backed securities that the Fed bought from them- not going to happen. Or there is your answer- Banks just return depositors’ money and stop making loans? That’s even less likely.

  22. Gravatar of dtoh dtoh
    22. August 2017 at 01:02

    And to add to what Scott said, the banks don’t necessarily need to make deposits less attractive. If the banking system as a whole lends more more money, this will boost AD increasing the demand for cash (assuming no change in velocity) thus reducing reserves.

    That said, the banks are not going to get rid of 2 trillion in ER in a short period of time without hyperinflation or OMS by the Fed. Assuming constant velocity, you would need to triple the size of the nominal economy to create sufficient demand for that much cash.

    However, who cares if the banks hold a ton of ER.

  23. Gravatar of Matthew McOsker Matthew McOsker
    22. August 2017 at 02:58

    Scott, yes reserves can drain if people withdraw cash, or if the Federal government runs a surplus – assuming no action by the Fed. In practice the cash withdrawals don’t appear to happen in massive amounts (even in places with negative rates), and that cash easily find its way back to banks.

  24. Gravatar of Matthew McOsker Matthew McOsker
    22. August 2017 at 02:59

    Also if banks make loans excess becomes required reserves but still reserves in the bank system.on the flip side a loan creates a deposit which creates reserves. The thing the bank really needs to do is now meet capital requirements.

  25. Gravatar of Michael Rulle Michael Rulle
    22. August 2017 at 06:01

    Scott

    Thanks for clearing up the technical point on the banks ability to control reserves. Still, why would Banks want to do that? It seems like something which they have not done. Further, wouldn’t that have to be inflationary, as it breaks the relationship, or virtual equality between reseves and Fed held securities? I still would like to know what the Fed was seeking to accomplish with QE. Since reserves stayed even with Fed held securities—due presumably because of IOR, the intial forcast by some of inflation never happened. But I still don’t understand the overall impact on the economy the QE program provided, other tah making it easier for banks to be profitable, —- less profitable maybe, but less risky to. Plus it help weaken the secuitization market by subsidizing an alternative asset, reserves.

  26. Gravatar of Brian Donohue Brian Donohue
    22. August 2017 at 06:19

    When QE came to an end in 2013, the TIPS curve was really steep – 5-year TIPS yielded 0.06%, while 30-year TIPS fetched 1.64%.

    Today, 5-year TIPS yield 0.17%, but 30-year TIPS yield just 0.92%.

    Since (obviously) recent Fed policy has pushed inflation expectations down, the nominal Treasury yield curve has flattened even more dramatically.

    It seems to me that ending “above market” IOER can be financed by selling some of the FED’s longer-dated portfolio. This would push long-term real yields up (there is room for this) without stoking inflation and avoid the inverted yield curve that a continued pattern of hikes and IOER will produce.

  27. Gravatar of Matthew Waters Matthew Waters
    22. August 2017 at 09:00

    I think it’s clear what QE tries to accomplish. If you believe in the zero-lower-bound, then the next interest rate to push down is long-term rates. You do that by buying bonds. In practice, expectations muddy up the data, but the basic argument for QE is straightforward.

    What really doesn’t make sense is positive IOER alongside it. The explanations have never jived for me. It’s hard not to believe the conspiratorial explanation: IOER provides risk-free interest for banks.

  28. Gravatar of Brian Donohue Brian Donohue
    22. August 2017 at 09:25

    @Matt, AFAICT, IOER provided funds to the Fed that allowed the massive QE asset purchases without printing more money, which is why there hasn’t been any inflation.

    With the Fed tightening since December 2015, inflation expectations are 1.8% or less as far as the eye can tell. After seven years of falling short, the Fed seems determined to keep falling short.

    No more hikes, FFS. You can end IOER and either print some money (inflationary), or sell off some of the Fed portfolio (non-inflationary, increases long-term real rates), or some combination to finance it.

  29. Gravatar of Jerry Brown Jerry Brown
    22. August 2017 at 10:43

    Ok Scott, the commercial banking system can get rid of excess reserves all by themselves if they want to. What is the mechanism for this? They all just quit being banks at once? Sell off their assets for actual cash and just refuse to accept deposits or make any loans and pay back all their creditors in actual cash? Short of this scenario, what are the ways excess reserves can be eliminated? I only can think of three ways, but I’m not an economist and want to know the other ways that are possible. The ways I can think of are that bank customers can pay more taxes to the Treasury, or banks can buy bonds from the Treasury and hold them. Or the Fed can sell the assets it holds back to the banks in exchange for reserves. Or banks can drastically lower their lending standards and rates on loans and hope to lend out around nine times the amount of excess reserves that exist (2.1 trillion US)- and this assumes the public will be willing to borrow 17 or 18 trillion dollars, and that the banks can cook their books to fudge their capital requirements to extend that credit.

    Only the third way is possible for commercial banks to attempt on their own- the other two rely on the government, or it’s agency the Fed. Your post was about how the Fed is paying banks interest on their excess reserves, which you seem to think is a more contractionary policy than just targeting the Fed Funds rate at an equivalent interest rate. I disagree- I think it is slightly more expansionary because it prevented a bunch of banks from going bust and provided all the banks with a basic income guarantee. I personally disagree with providing banks an income guarantee for many reasons, but I do think it is sort of an easier monetary policy compared to the Fed using open market operations to target the same interest rate. And a gift to bankers from the rest of us. And they are a particularly unappreciative sort, those bankers.

  30. Gravatar of Michael Rulle Michael Rulle
    22. August 2017 at 13:40

    @Matthew Waters

    What you said makes sense of course. Thank you.

  31. Gravatar of Benjamin Cole Benjamin Cole
    22. August 2017 at 17:42

    Matthew Waters:

    What you are describing is not conspiracy, but rather old-fashioned industry-capture of a regulatory body, however genteel the participants and elevated the rhetoric and righteous the goals.

    Yes, the Fed has devised a plan worthy of the Department of Agriculture: Banks get money for doing nothing. In perpetuity, evidently.

    And the solution to the house price “bubble” (as Kevin Erdmann notes) was to cut off moderate-income people from credit to buy housing, but not to cut off upper-income people from credit to buy housing.

    Can you imagine the uproar if upper-income people were denied credit to buy housing?

    Nor has the Fed ever mentioned property zoning in its policy statements. After all, its favored constituency—the commercial banks—are heavily exposed to real estate. Zoned real estate, that is.

    And the Fed just happens to target a minimum unemployment rate–4.8%–that works out to 1.5 Americans looking for work for every job opening. How nice.

    In Japan, there are two job opening for every job hunter.

  32. Gravatar of dtoh dtoh
    22. August 2017 at 18:25

    @benjamin cole
    I somewhat agree with what your point. I often times find it easier to understand Fed action, if I think of the Fed and commercial banks together as black box rather than drawing a distinction between the “central” bank and the “commercial” banks. The Fed often seems to do what is “right” for the banking system rather than what is right for the economy. I suspect the Fed Governors would deny (and believe in their hearts) that this is not true, but I think people can’t help being influenced by their environments in ways of which they are not conscious.

    In the case of Bernanke, I wonder if the IOR policy may have been a very deliberate (but disguised) policy to insure bank stability and liquidity because after the Lehman shock he was worried that political pressure might inhibit the extension of credit to the banking system during a crisis.

    As for Japan, I’m not sure that’s an entirely equivalent comparison. Japan measures unemployment by asking if the respondent has looked for work in the last two weeks. The U.S. uses a six week standard. I’ve seen studies that suggest the male unemployment rate in Japan doubles if you use the six week measure.

  33. Gravatar of ssumner ssumner
    22. August 2017 at 19:39

    Everyone, Don’t confuse what banks can do with what they are likely to do. Obviously they are likely to do exactly what they are currently doing, as their current policies happen for a REASON. ERs are profitable.

    Whether banks convert reserves into cash depends on the situation. If the Fed suddenly adopted a 30% inflation target you can be sure that banks would be getting rid of ERs at breakneck speed. With a low inflation target, ERs are more profitable.

  34. Gravatar of Vaidas Urba Vaidas Urba
    22. August 2017 at 23:48

    “I argue that by paying an interest rate on excess reserves that is higher than comparable short term rates, the Federal Reserve likely hindered the portfolio reallocation channel outlined by Bernanke.”

    How can the Fed pay IOR at a rates not higher than comparable rates? Cutting IOR to zero won’t help, as other comparable rates would drop below zero. The reason other rates are lower than IOR is the convenience yield – i.e. other assets are even better than reserves, presumably because they can be held by a wider range of parties, creating market segmentation.

  35. Gravatar of George Selgin George Selgin
    23. August 2017 at 09:24

    “Ok Scott, the commercial banking system can get rid of excess reserves all by themselves if they want to. What is the mechanism for this? They all just quit being banks at once? Sell off their assets for actual cash and just refuse to accept deposits or make any loans and pay back all their creditors in actual cash? Short of this scenario, what are the ways excess reserves can be eliminated? I only can think of three ways, but I’m not an economist and want to know the other ways that are possible.”

    Usually bank lending, and associated deposit expansion,proceeds until excess reserves return to minimal levels (excess reserves are a function of total D and not just total R). Every textbook used to explain this: it too some clever Fed Board economists to forget it. https://www.alt-m.org/2016/01/05/interest-reserves-part-ii/

    “How can the Fed pay IOR at a rates not higher than comparable rates? Cutting IOR to zero won’t help, as other comparable rates would drop below zero.” Not so. The IOER rate doesn’t rule the roost once reserves again become scare. If it did, then “comparable rates” would always have been zero until October 2008!

  36. Gravatar of George Selgin George Selgin
    23. August 2017 at 09:50

    “Scott, yes reserves can drain if people withdraw cash, or if the Federal government runs a surplus – assuming no action by the Fed. In practice the cash withdrawals don’t appear to happen in massive amounts (even in places with negative rates), and that cash easily find its way back to banks.”

    “Excess Reserves” (ER) are not the same as Reserves (R). R does not have to decline, via cash withdrawals or otherwise, in order for ER to decline. An increase in reservable deposits also reduces ER, ceteris paribus. Until October 2008, ER remained minimal despite all sorts of growth in R, because deposit expansion kept it so. IOER put a stop to by making them more attractive than before relative to other assets. That is the point made in Hendrickson’s paper.

  37. Gravatar of Matthew McOsker Matthew McOsker
    23. August 2017 at 10:17

    George, thanks for the reply. Oddly before I read your post above, I had just read a piece by you as I was going back doing some reading on this sometimes controversial topic.

    That piece is here:
    https://www.alt-m.org/2017/06/01/ioer-and-banks-demand-for-reserves-yet-again/

    Of course, I was looking for evidence that banks would not hold ER absent IOER, and trying to understand the reasons they would “demand” ER. One interesting item that popped up was the amount of ER held by foreign banks, which were in worse shape than US counterparts, and those banks seemed to certainly wanted to hold those ER versus some other “shaky” financial instruments. US banks close to the 2008 crisis certainly would prefer ER instead of some shaky MBS security. So still feels like it begs the question is the demand related to quality versus what they gave up, or the IOER, or a combo?

    So far have not got so far as looking at ER in countries with negative IOER rates (obviously the neg rate will drain some in and of itself).

  38. Gravatar of George Selgin George Selgin
    23. August 2017 at 10:51

    Matthew, the subject of foreign banks’ holdings of high reserve ratios is one deserving of more research. I looked into Japan, and the story there is not inconsistent with the general story that IOR is an important driving force.

  39. Gravatar of Matthew McOsker Matthew McOsker
    23. August 2017 at 11:02

    And to be clear above I am referring to Foreign bank holdings of US$ ER.

  40. Gravatar of George Selgin George Selgin
    23. August 2017 at 11:08

    Well, Matthew, that one is actually somewhat easier. US branches of foreign banks that do not have insured US dollar deposits are not subject to FDIC premiums (assessed against total assets, including reserves), and so profit more than most from IOER. Also, foreign bank net interest margins are relatively low, while their operating costs are relatively high. A large chunk of total USD excess reserves of foreign banks is for these reasons held to take advantage of the Fed’s IOER policy. I have talked to employees of banks in Australia and Europe, among other places, who have affirmed all this with me. See also https://www.clevelandfed.org/newsroom-and-events/publications/economic-trends/2015-economic-trends/et-20150811-who-is-holding-all-the-excess-reserves.aspx

  41. Gravatar of Jerry Brown Jerry Brown
    23. August 2017 at 12:18

    George Selgin @9:24, thank you for the link to your article. I read it and don’t disagree (I think) with your conclusions. I read some of the comments there also and I see you are familiar with the very informed ‘JKH’ who I have learned to have a lot of respect for. I will defer to the explanations given by JKH rather than my previous attempts here to explain what I meant. Although I really don’t see that anything you wrote there or that JKH wrote there contradicts what I wrote here.

    I understand that new loans made by banks will reduce the amount of excess reserves in the system- they just reduce them at a much slower rate that depends on the required reserve ratio for deposits. At least in my understanding, at a 10% required reserve ratio on deposits, eliminating 2 trillion in excess reserves through new lending alone would require somewhere around 20 trillion in new bank loans, which I doubt there is either demand for or ability to produce by banks.

    Also, I don’t really understand why Scott says that IOER is any more contractionary than the Fed targeting that same rate via open market operations as it had done previous to 2008. In fact I think it is somewhat more expansionary, since it keeps the system very liquid and provides income to banks. And also because I seem to believe that not all institutions with access to the Fed Reserve system were eligible for the interest payment on their excess reserves, which had the effect of causing the rate on lending reserves between financial institutions to be slightly lower than the stated return on excess reserves provided by the Fed. Anyways, thank you again, and don’t be afraid to point out my mistakes, because I make a lot of them and it is helpful when they get pointed out.

  42. Gravatar of George Selgin George Selgin
    23. August 2017 at 15:20

    “Also, I don’t really understand why Scott says that IOER is any more contractionary than the Fed targeting that same rate via open market operations as it had done previous to 2008.”

    It is the move from below market (including zero) IOER to above market that is highly contractionary, because it represents a regime switch. At IOER > r* reserves dominate other short-term assets, and the multiplier collapses. That is what happened in October 2008. From then on, the old monetary control regime was dead. That is why trillions in reserve creation could no longer do what billions would have done before. This is THE story of how the Fed went from situation somewhat normal to SNAFU, where we remain to this day.

  43. Gravatar of dtoh dtoh
    23. August 2017 at 17:49

    Everyone is getting their knickers twisted thinking about rates. The only thing that counts is the amount of financial assets (e.g. loans) exchanged for money between the financial sector and the non-financial. Changes to the amount of assets exchanged (e.g. increased bank lending, etc.) is what causes expansion or contraction to the growth rate of the economy.

    Rates are only a guidepost for policy makers. What’s important is the impact on the exchange of money for assets between the banks and non-financial sector. If the banks are buying assets from the Fed (i.e. making deposits with the Fed in the form of ER,) then they are not buying assets from the non-financial sector. All else being equal this is contractionary.

    This is not rocket science!

  44. Gravatar of ssumner ssumner
    23. August 2017 at 21:30

    Vaidas, You asked:

    “How can the Fed pay IOR at a rates not higher than comparable rates?”

    They paid less than T-bill yields prior to 2008.

  45. Gravatar of Major.Freedom Major.Freedom
    24. August 2017 at 23:11

    Sumner you are spewing falsehoods on this blog, without shame at this point, as there has already been corrections provided multiple times.

    You wrote:

    “The banking system can get rid of reserves if it wants to, without any help from the Fed. If they make deposits less attractive, people will convert deposits into cash. Banks choose how much reserves they wish to hold.”

    That is false, because even if banks tried to “get rid of reserves”, the central bank will top those reserves right up again to what the central bank wants, by sending the banks money in return for whatever garbage the banks are selling, government debt, or worthless complex structured products. Banks cannot exist without income. The income banks earn become the very deposits you claim can be reduced unilaterally without limit by the banks AND yet the banks do not all go bankrupt? Not even close.

    Central counterfeiters have taken the requirement of banks to earn an income to exist, and by coercion from government, established themselves as the monopoly issuer of that which is earned as income, namely money. If the public tries to hold more cash, which would presumably decrease reserves (this by the way is not the main transmission mechanism, but to give Sumner a break we can all play make believe as if it is), then guess what? The central counterfeiter will just send more reserves to the banks in exchange for what the banks are already selling in the open market that you say are “less attractive”?

    How attractive are worthless mortgage backed derivatives? That would surely dissuade people from paying for them, and hence dissuade the public from sending banks money. But did that stop the banks from getting billions of dollars for them anyway from the counterfeiters? Of course not! This is just one way in which your misunderstsnding of banking and of the monetary system is exposed.

    The truth is that the central counterfeiter determines how much reserves are in the banks. If there is less than what the central cointerfieters want, it will just print more rapidly, and if there is more than what it wants, it will just print less rapidly. More inflation from the central counterfeiter is what drives the banks to extend credit beyond what consumer-savers are providing in the form of abstaining from current consumption and saving a portion of their incomes. This is what generates inter-temporal discoordination, the fundamental characteristic of the repeating business cycles we see in every country in the world, including Australia. Aggregate measurements for “employment” and “GDP” are not the correct indicators of business cycles. Never has been, never will be, by the way.

  46. Gravatar of Vaidas Urba Vaidas Urba
    25. August 2017 at 01:20

    George, Scott,

    so in the current circumstances we could keep IOR, reverse QE, and make reserves as scarce as T-bills. Per Friedman rule, T-bill rates which are equal to IOR are preferable to T-bill rates which are much higher than IOR which is permanently frozen at zero.

  47. Gravatar of Jose Jose
    25. August 2017 at 04:50

    Forget interest rates, at the zero lower bound, if aggregate demand falls below some perceived “good” target, do QE, but scrap IOER, ignore banks, buy a diversified basket of assets, therefore increasing the impact of portfolio rebalancing effect.

  48. Gravatar of Benjamin Cole Benjamin Cole
    26. August 2017 at 06:43

    OT but interesting:

    Another benefit of free trade is global capital flows.

    “According to the Asian Real Estate Association of America, Seattle is currently the sixth most popular destination for Chinese immigrants in the world. And there’s good reason for it; Seattle is the closest mainland U.S. city to travel to from Beijing and offers things that really appeal to the Chinese, like clean air, quality education, and employment opportunities with several Fortune 500 companies,” said OB Jacobi, president of Windermere Real Estate, covering the Seattle market. “This is why Chinese buyers are now estimated to represent upwards of 40 to 50 percent of all real estate activity in Seattle’s most expensive neighborhoods, of which at least 75 percent are paid in cash. While some call this a trend, I believe it’s an emerging aspect of our market that’s here to stay for the foreseeable future.”–

    http://www.realtytrac.com/news/company-news/share-of-chinese-speaking-buyers-paying-cash-for-u-s-homes-increased-229-percent-in-the-past-10-years/

    Several nations, including Hong Kong, Great Britain, Canada, Australia, New Zealand and part of the US, are experiencing extremely strong housing appreciation.

    Usually, people talk about interest rates when discussing house prices.

    But trade deficits are more important, along with property zoning.

    https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2991876

  49. Gravatar of ssumner ssumner
    26. August 2017 at 09:30

    Vaidas, I don’t understand the question.

Leave a Reply