Macro 101 – Reserves and interest rates

It has been quite some time since I did a post in the Macro 101 thread. There has simply been too much economic news to process lately. However, now that the stock markets seemed to have stabilised themselves at a lower base, and the western world has set itself up for a recession I think I have a spare few hours to revisit this thread.

Previously I have discussed banking credit, foreign trade and sectoral balance. Today I am going to try to unify these things through a conversation about reserves and interest rates.

So before I start I need you to ask yourself two questions.

  1. What does the term “reserve” actually mean in the title “Reserve bank of Australia” ?
  2. How exactly does the RBA control interest rates.?

If you know the answer to these questions then maybe you don’t need to read on. Otherwise let’s dive in.

I will be working through transactions step-by-step in order to slowly add complexity. I will obviously be over simplifying examples, and in some cases change the order of how things that actually occur in in order to explain things more simply. But none of these things change the actual functions/tasks that are occurring, it just makes it easier for me to explain the process.

In a previous post on banking credit I talked about “money of account” and “money of exchange”. If you get stuck with these concepts while reading this post please go back and read the previous post for more detail.

From the beginning then…

You decide to purchase a car from a dealer for $40,000, the dealer banks with CBA, you bank with ANZ. You approach ANZ for a loan and given your wonderful credit rating you are approved.  The steps of the transfer are as follows:

  1. The ANZ creates a new loan record and assigns it an outstanding balance of $40,000 under your name.
  2. At the same time it adds $40,000 to your account.

So far this entire transaction is internal to the bank. This is simply a change in the database that the bank uses to store account information. So although you think you have an additional $40,000 nothing has really happened apart from a few electrons have been re-aligned on a hard drive(s) somewhere inside the ANZ’s data-centre. This is “money of account”.

So next you want to pay the car dealer who banks at CBA. To do this you could use a bank cheque, personal cheque, cash, money order etc. It really doesn’t matter how what the medium of transfer is. However what you will note is that the electrons that were re-arranged on ANZs hard-disk(s) have no value to the CBA. So how exactly does ANZ transfer $40,000 to CBA in order to satisfy the transaction? Enter “money of exchange” otherwise known as “reserves” which are recorded on the reserve bank’s computer.

Banks transfer reserves to and from each other via their reserve banks accounts. In Australia they are called Exchange Settlement Accounts and you can read about them here.

In future examples I will refer to these as “reserve accounts” for simplicity but it must be noted that in fact an ESA is a special type of reserve account that is only available to banks that have access to the interbank lending markets. There are other reserve accounts that are used by foreign central banks to record their AU$ reserves.

Now let’s describe the transaction that occurs for you to purchase that car.

  1. You give the car dealer a cheque for your car.
  2. The dealer takes the cheque to CBA and deposits it in his/her business account.
  3. CBA processes the cheque and demands $40,000 of reserves from ANZ
  4. At the ANZ the balance of your account is deducted by $40,000 and requests the RBA move $40,000 from its reserve account to CBA’s reserve account.
  5. At the CBA the car dealers account is adjusted by $40,000

So now we have 3 sets of electrons being re-aligned. Firstly at ANZ the account database is once again updated to record that your account has lost $40,000. The database at the reserve bank is updated to record that $40,000 of reserves was deducted from ANZ’s exchange settlement account and added to CBA’s exchange settlement account. Finally at the CBA, their database is updated to show that your car dealer has an extra $40,000.

Functionally the same process occurs if you used cash. Because a $5 note is simply a plastic/paper version of a $5 reserve balance and are completely interchangeable.

Now obviously you are not the only one doing a banking transaction that day, there are in fact tens of millions of inter-bank transactions each day in Australia. Each one affects the account balance in each bank’s reserve account.  Over the period of the each day the balances of these accounts will fluctuate but the RBA specifies that these accounts must have a positive balance at all times ( more on that point later ).

The important thing to understand is that the amount of reserves required is much smaller than the total of the bank’s “money of account”. For example, if on the same day your loan and cheque were processed, a $40,000 loan was issued by CBA and a subsequent cheque given to an ANZ customer then technically there would be no change in either banks reserve accounts even though both banks issued $40,000 loans.

But in reality that is not normally the case. When a bank runs a loan book it requires reserves in order to support interbank transfers and over the counter transactions because of the demand for the “money of account” that those loans created when they were issued.

So as a bank issues loans it must seek reserves to support its liquidity requirements due to those loans. So there will obviously come a point when a bank simply does not have enough reserves to support additional loans. The bank therefore has two choices; stop issuing more credit until some existing loans are repaid, that is some reserves are returned to the bank, or borrow reserves from someone else.

Obviously banks want to continue to grow their loan book because loans are profitable, so 99.99% of the time banks seek additional reserves.

But where from? I have previously heard some people state that banks can just issue a loan to themselves and use the created deposit to fund further lending. But if you have been reading carefully you will understand that this just creates more “money of account” in their accounts database in their own computer. It certainly doesn’t add to the balance of their reserve account in the reserve banks computer or suddenly create a pallet of $100 notes (remember cash and reserve balances are interchangeable). So in fact this would not help at all.

Banks can only get more reserves from two places, other entities that have an ESA or from the reserve bank itself.  Now I can hear what you are saying already. You are saying “hang on DE, Australian banks get most of their funding from overseas through bond issuance”. Yes they do! However you need to understand exactly what this means in terms of reserves to appreciate that both statements hold true. So let explain.

As I have been saying the RBA is a bank for banks. It is where they store and borrow Australian dollar reserves. That is why it is called the “Reserve bank of Australia”. (Question 1 above ). All across the globe there are similar institutions for each unique currency. The US has the Fed which records US reserves, China has the PBoC which records Yuan reserves, the EU has the ECB which records Euro reserves, and so on. At each of these institutions, foreign central banks also have non-EAS reserve accounts. These accounts are used to support international exchange and trade.

So let’s look at what happens when the CBA issues AU$1 billion 5-year bonds and a US superannuation fund purchases them. We will assume that the fund already has $AU1 billion equivalent of US$ in a US savings account at the Bank of America, and that the exchange rate is 1.00. Again this has all been oversimplified.

  1. The fund agrees to buy the bonds and instructs the Bank of America to pay the CBA $1billion Australian dollars.
  2. At the bank of America the balance of the savings account in their computer in the name of the superannuation fund is lowered by $US 1 billion.
  3. The Bank of America then transfer $US 1billion dollars of its reserves to the Reserve Bank of Australia’s US reserve account (non-ESA) under an order to issue the equivalent amount (governed by the exchange rate) of AU$ to the CBA.
  4. At the same time the Reserve bank of Australia adds $AU 1billion to the reserve account of the CBA in its computer.

So let us recap.  $US1 billion of “money of account” was subtracted from an superannuation funds’ account in the database of the Bank of America,  $US1billion of reserves was transferred from the reserve account of the Bank of America to the reserve account of the RBA in the database of the US Federal reserve, and finally, $AU1 billion was added to the balance of the Exchange Settlement Account of CBA in the database at the RBA. The CBA now has another $1 billion dollars worth of loan liquidity supporting reserves, and the RBA has gained $1US billion dollar of US reserves.

As you can see from that example the CBA actually got its reserve from the RBA because it is in fact the only place that could issue them. This is because reserves are simply electronic records in the RBA’s computer, which is why no one else can issue them. This may sound like a trivial statement, but it is in fact the basis on which modern economies are built.

Moving on…  Obviously over time the CBA must pay interest on those bonds, which will move a proportion of those reserves back in the opposite direction, but by this time the loans that these bonds support will have matured somewhat so new reserves will exist to make those payments. At the time of bond maturity the entire process will be reversed meaning that $AU1 billion worth of reserves would be removed from the bank’s ESA. You may now appreciate why banks must continually rollover funding and are therefore constantly susceptible to liquidity risk.

There is however a fairly large problem with the fact that a bank suddenly receives $AU1 billion dollars in reserves because it has suddenly flooded the interbank market with reserves. This would tend to drive down interbank interest rates which would undermine the RBA’s interest rate.

This leads nicely into a discussion about the interbank lending market and how the RBA controls interest rates through it. I was going to actually write about this myself, however it turns out that the RBA has a very good document available explaining it all for me. The one issue with that document is that it required you to have some background knowledge of reserves and exchange settlement accounts before you read it. I hope I have done a fair job of explaining these concepts to you. If that is the case then you hopefully will not struggle too much with the following document.

Good luck.

Latest posts by __ADAM__ (see all)

Comments

  1. Very good summation DE. However, for m2c can you please follow up with how capital is allocated within a bank to restrain limitless borrowing?

    Whilst “Banks can only get more reserves from two places, other entities that have an ESA or from the reserve bank itself.” Capital must come from shareholders at some point, either directly or through dividends being lower than net profit.

    • >Whilst “Banks can only get more reserves from two places, other entities that have an ESA or from the reserve bank itself.” Capital must come from shareholders at some point, either directly or through dividends being lower than net profit.

      Well actually this is sort of a different discussion. Whether funding comes from Equity or Bonds or Deposits makes no difference to reserve accumulation.

      Having said that it is worth discussing the what backs use their different types of tiered capital for and exactly what happens in the case of a loan default. So I will definitely put this on my list

  2. A very good primer on the basics of our monetary structure for the laymen/conspiracy theorists/mouth breathers among us.

  3. The limit to banks I thought isn’t usually reserves – it is capital and their balance sheet. Banks keep a minimal amount of their actual reserves in the exchange settlement accounts that they have to in order to satisfy settlements. There are better places to put your money for the bank (yes loans is one of them but there are others).

    Otherwise a nice explanation

    • >There are better places to put your money for the bank (yes loans is one of them but there are others).

      I am not 100% sure what you mean by this statement. Reserves never leave the reserve banking system except through operations performed by the government sector or foreign exchange.

      Can you give me an example of what you mean by “better place”?

      • Maybe I’m mistaken but all over the place I have read that banks only hold a minimum amount of reserves in the system such that they are able to settle transactions.

        Otherwise they lend it to each other. Bascially there is every incentive to lend to each other and therefore reserves (given no regulated leverage ratio and these days there’s ways around this) do not really constain the amount of loans made as such. What constrains loans is capital normally as if they are capital constrained banks won’t lend their excess reserves to that bank.

        This was my understanding anyway (which could be wrong). Its sad that our financial system has gotten so complicated in all honesty. My point was is that I don’t think reserves are the ceiling on how much a bank can lend anymore as much as capital position is.

        • Found this on the web. Think it is related. Basically says what Steve keen has tried to prove – that reserves come after the loans normally.

          Also as mm participants need more bank money to keep the interest rate stable there is always the rba as a last resort..

          globaleconomicanalysis.blogspot.com/2009/12/fictional-reserve-lending-and-myth-of.html

  4. The article you link to regarding sectoral balances is misleading and readers would benefit if it was withdrawn or rewritten.

    The term “Private Saving” in the sectoral balances equation refers only to the nominal increase in bank deposits. It does not refer to growth in net wealth in real dollars, which is what is normally meant by private saving.

    The term “Private Investment” refers only to nominal lending. It does not include the adding of real value through the deployment of ones own funds and efforts, which is normally included in the concept of investment.

    You make the following statement:

    “If in an accounting period there is a trade deficit and a government budget surplus then there MUST be a … net loss of savings from the private sector. If this situation continues for a period of time then it leads to increasing indebtedness.”

    The statement is misleading.

    Firstly, you do not define “net loss of savings”. What you mean is a net loss of “Private Saving” – “Private Investment”, as miseladingly defined.

    Secondly, while it is correct to say that such a situation would result in an increase in nominal private borrowing, it is incorrect to say that this necessarily increases indebtedness. Real indebtedness only increases when the debt is used unproductively – i.e. not to create assets that can more than cover the cost of borrowing.

    • >The term “Private Saving” in the sectoral balances equation refers only to the nominal increase in bank deposits. It does not refer to growth in net wealth in real dollars, which is what is normally meant by private saving.

      Actually savings refers to the component of income that isn’t spent on comsumer goods. It is the same component used in the household savings ratio.

      >The term “Private Investment” refers only to nominal lending. It does not include the adding of real value through the deployment of ones own funds and efforts, which is normally included in the concept of investment.

      Actually Private investment refers to the amount spend on private sector investment goods. This includes macnufacturing equipment, housing , machinery etc.

      >“If in an accounting period there is a trade deficit and a government budget surplus then there MUST be a … net loss of savings from the private sector. If this situation continues for a period of time then it leads to increasing indebtedness.”
      The statement is misleading.

      No I don’t believe it is.

      >Firstly, you do not define “net loss of savings”. What you mean is a net loss of “Private Saving” – “Private Investment”, as miseladingly defined.

      No, they are defined in the statement of GDP.

      >Secondly, while it is correct to say that such a situation would result in an increase in nominal private borrowing, it is incorrect to say that this necessarily increases indebtedness. Real indebtedness only increases when the debt is used unproductively – i.e. not to create assets that can more than cover the cost of borrowing.

      Yes this is correct, and you have a point in the short term. For instance there may be a period when a nation goes into debt because it imports foreign produced machinery to make itself more productive. If that was the case however you would see the effects of this over future acounting periods. Where exports would increase and the additional foreign income would pay down the previously accumulated debt.

      But what I said was

      “If this situation continues for a period of time then it leads to increasing indebtedness.”

      Maybe I was explicit enough about what I meant about “period of time”.

      • “Actually savings refers to the component of income that isn’t spent on comsumer goods. It is the same component used in the household savings ratio.”

        My understanding is that Sectoral Balances categorizes nominal dollar flows involving private sector actors into Sources and Uses.

        The Sources are: PI (borrowing), GS (government spending), C (consumption revenue), all nominal.

        The Uses are: PS (saving and lending), Tax, C (consumption spending), all nominal.

        PS is not the same as “household saving”. If it is, then the equations are incorrect. The equations only work out if you use the definitions I have included above.

        “Actually Private investment refers to the amount spend on private sector investment goods. This includes macnufacturing equipment, housing , machinery etc.”

        That is the definition used in the GDP calculation. It is not the same as the definition used for sectoral balances, which is part of what makes sectoral balances misleading.

        DE: “If in an accounting period there is a trade deficit and a government budget surplus then there MUST be a … net loss of savings from the private sector. If this situation continues for a period of time then it leads to increasing indebtedness.”

        PA: “The statement is misleading.”

        DE: “No I don’t believe it is.”

        Why?

        PA: “Firstly, you do not define “net loss of savings”. What you mean is a net loss of “Private Saving” – “Private Investment”, as miseladingly defined.”

        DE: “No, they are defined in the statement of GDP.”

        As described above, I don’t believe this to be correct. Perhaps if you can name the source you aer using for this information?

        PA: “Secondly, while it is correct to say that such a situation would result in an increase in nominal private borrowing, it is incorrect to say that this necessarily increases indebtedness. Real indebtedness only increases when the debt is used unproductively – i.e. not to create assets that can more than cover the cost of borrowing.”

        DE: “Yes this is correct, and you have a point in the short term. For instance there may be a period when a nation goes into debt because it imports foreign produced machinery to make itself more productive. If that was the case however you would see the effects of this over future acounting periods. Where exports would increase and the additional foreign income would pay down the previously accumulated debt.”

        This is due to debt not being used productively, and has nothing to do with “sectoral balances”. If sectoral balances made any difference, it would apply in the short term as well as long term. After all, the equations claim to represent the truth over any time period, not just over the long term.

    • Weimar Republic

      @Paul Andrews

      re: nominal vs real

      you may be outlining a semantic argument among economists but from a mathematical point of view if you have an equation:

      X = Y + Z

      If you want to draw conclusions about the relationship between one variable and another it doesn’t matter if they are all nominal or all real e.g. to get real values you deflate all terms so conclusions are the same.

      Delusional,

      In discussing the sourcing of offshore funding, and servicing of those loans, what you have described is a mechanism by which a credit expansion can add to a current account deficit which was the point I was making here:

      http://www.macrobusiness.com.au/2011/08/leigh-harkness-on-the-rba-cad-perspective/

      • I have been musing this point for a while actually. So let me collate my thoughts and maybe get some feedback.

        If we break down sectoral balance into its sectors we get this

        P = G + X

        Where P is the difference between private sector investment and private sector non-consumer spending. In other words this is whether the private sector is paying more for investment than it has in savings from income. In Australia we know this is negative because we continue to pay more for housing ( and investment good ) than we have in savings from income after consuming. To do this requires the private sector to use debt created in the banking system. (incidentally this is proof that housing is not productive, but I will leave that for another time )

        G is the net government position ( Spending – Tax )
        X is the net export sector position ( Exports – imports )

        Leigh’s theory tends to support this because he states that if the banking system is issuing net credit then you will see a trade deficit. However his research suggests that the government position is somewhat irrelevant, I am still struggling with that one, but it maybe because governments continue to issue bonds to balance their position which means that the transfer net to zero.

        Your point about CADs from financing banking credit is an interesting one, and my gut feeling is that it must be a component especially the interest paid. However the question is whether you would have a CAD if the banks where not allowed to source foreign reserves and were forced to use the RBA facilities. It seems obvious to me that you still would, but maybe you have a different opinion.

        What is also interesting for me is which comes first. Do you get a trade imbalance because you are issuing credit in the banking system? or are you issuing credit in the banking system because you have a trade imbalance? You can approach this argument from either angle and come to different conclusions I think.

        Something to ponder I guess.

        • Weimar Republic

          To reiterate, I don’t disagree that credit creates demand for imports. I am just a bit flummoxed on the apparent focus on this exclusively. As I have said previously, I don’t understand why anyone would embrace one action resulting from expanding credit — such as higher demand for imports, and not another action — such as overseas borrowing. Both actions result from expanding credit and, admittedly without having gone through the national accounts, I believe that interest payments on overseas borrowings are substantial as a proportion of GDP. Didn’t the RBA quote 2-3%??

          If the banks had of sourced locally then impact on the CAD would depend on the %GDP of the overseas interest payments. For example if interest was 2% GDP but the CAD was 4% then clearly foreign interest is not in itself sufficient to create the CAD. Nevertheless at its supposed current magnitude it is substantial and you might ponder whether it exceeds the credit inspired increase in imports (i.e. imports over an above what would have occured had credit not been loose).

          Banks presumably must source overseas because it is more profitable to do so (lower borrowing rates). Bottom line is that all of this is easily testable empirically. It should be all there in the national accounts.

          As to which comes first I seems that all commentary on this seems to have causality running from credit -> CAD.

          Also what we are talking about is exclusively the effect of credit on “M” in the national accounts, since you can have expanding credit and have current account surpluses (because “X” is big in that case). Also one might want to study correlations of retail sales with the CAD if imported goods are the main driver (rather than foreign interest payments).

      • “re: nominal vs real; you may be outlining a semantic argument among economists but from a mathematical point of view if you have an equation: X = Y + Z If you want to draw conclusions about the relationship between one variable and another it doesn’t matter if they are all nominal or all real e.g. to get real values you deflate all terms so conclusions are the same.”

        This is not correct – inflation does not affect real economic factors uniformly. If the real value of assets grow through productivity improvements, this may affect real savings differently to real lending.

        • Weimar Republic

          can you show me where “assets” appear in the equation that DE presented?

          If not please refer to my comment in which I wrote only about the variables present in the equation.

          apart from anything else an asset is a stock and would appear on a balance sheet. DEs equation is for flows, it is (basically) a P&L.

          • Real private savings (as opposed to nominal) includes accumulation of real assets. For instance shares in super fund accounts, adjusted for inflation.

            You are correct – the equations presented do not include accumulation of real assets. That is my point. The equations only refer to nominal dollar values, not real values, and therefore are of little use in understanding the real economy.

            In addition, the equations use terms misleadingly. “Private Savings” does not usually mean “All private lending”.

          • “apart from anything else an asset is a stock and would appear on a balance sheet.”

            An asset is a stock. An increase or decrease in the value of an asset is a flow.

            An increase in the value of an asset, or the creation of a new asset, is part of real private savings as usually understood.

            Increases and decreases in asset values appear on the P&L.

  5. Even though this isn’t anything new to me, I still read it our of interest of how well you would explain it. I must say, it’s very well done, and I actually learnt a couple of things I hadn’t known before. Thanks Delusional Economics!

  6. I often struggle with economic concepts on this blog so really appreciate the 101 on this topic.
    If you can explain fractional banking that would be good too.

  7. I struggle with Reserve Accounting but my limited understanding is the RBA document gets at least one thing wrong. ESAs do not have to be positive at all times (although ideally they should be), they just need to be positive at the accounting period.