Central bank funding government?

31 July – (some corrections 12 and 13 August)

Introduction

What inspired me to write this text, was an article entitled Oh Canada! Imposing Austerity on the World’s Most Resource-rich Country”. It was written by Ellen Brown (in full: Ellen Hodgson Brown, J.D.) and it is dated 1 April 2012 (which, I suppose, doesn’t mean it should be read as a joke).

I found Ellen Brown’s article because Dutch money reform activist Ad Broere in one of his web pages hyperlinked to a Dutch translation of Brown’s essay.

Summary and ideas

Ellen Brown stated in her article that until 1974, Canada was in a much better position financially than it is today. She thinks that was because the Canadian government then borrowed money from its own central bank, instead of from private, commercial banks. After 1974, encouraged by the BIS (Bank for International Settlements), Canada borrowed from commercial banks instead.

In Brown’s view, in both cases the money for the loans resulted from money creation, first by the central bank, later by those private banks.

An important difference, she feels, is in where the interest goes:

The difference is simply that a publicly-owned bank returns the interest to the government and the community, while a privately-owned bank siphons the interest into its capital account, to be re-invested at further interest, progressively drawing money out of the productive economy.

Evaluation

This is interesting. Could monetary financing, i.e., a central bank providing the money for its own government, have averted the debt crises that so many of the world’s governments find themselves in today? Could returning to such monetary financing solve those crises now that they are already there?

To investigate that, I’ll set up an example initial situation, involving a central bank and a non-central bank. Starting from that situation, I’ll look at three scenarios:

Looking at each of those scenarios, I will analyse the effects in terms of:

Initial situation

Central bank’s balance sheet

All amounts are in millions of euros, dollars (US, Canadian, Australian), pounds sterling, whatever. Million or maybe billions, really.

The actual amounts are not important, and they are not necessarily realistic either. They only serve as examples, to have concrete items to refer to, to see what may happen in the scenarios below.

Description Debit (assets) Credit (liabilities)
Gold 100
Capital 10
Banknotes in circulation 90
Government’s treasury 0
Banks’ current accounts (incl. minimum reserve) 110
Refinancing loans to banks 20
Securities 90

Non-central bank’s balance sheet

I also assume there is a non-central bank, which symbolises the many non-central banks that would exist in a real-life situation.

Description Debit (assets) Credit (liabilities)
Capital 100
Banknotes in vault 10
Current account with central bank (incl. minimum reserve) 110
Refinancing loan from central bank 20
Loans to households and businesses 1000
Current accounts of households and businesses (on demand) 50
Savings accounts of households and businesses (on demand) 950

Scenario A

Government borrows from central bank, spends by bank account.

Borrowing

I depict the government’s treasury as an account held by the government with the central bank. In my example situation the treasury is initially empty: any money the government received as taxes, it has already spent.

Now suppose the government needs extra money for an urgent infrastructure project. Some 20 million / billion pounds / dollars, whatever is a realistic amount in our semi-realistic fantasy world. It’s easy for the central bank to lend that money to the government: it just enters the amounts in its books!

(This is in a sense money creation, although the result is not part of M1, not even of MB or M0. It will become MB when shifted to bank reserves later.)

How does the central bank go about this? This is the transaction in the central bank’s books:

Description Debit (assets) Credit (liabilities)
Loan to government 20
Government’s treasury 20

We see here the phenomenon of increasing the balance sheet total, by adding the loan amount at the debit side, and at the credit side the corresponding availability of the amount to the borrower. What happens here between central bank and government is the same thing as what I described for a non-central bank and its borrowing client here.

Increasing the balance sheet total is easy and it seems cheap, because the interest is paid to the central bank. If the central bank is wholly owned by the government (which is often the case, e.g. Bank of Canada, Bank of England, central banks in the eurozone), the government effectively pays the interest to itself.

But. There’s a big ‘but’ to this.

Spending the loan

The government didn’t borrow this money just to have it sitting in the treasury. It borrowed it with a purpose: some programme or project or similar. In our example, the money was for an infrastructure project, so it is intended for paying design engineers and contracting companies for eventually building that bridge or road or tunnel or whatever it is.

(Confer article 11: same thing for non-central banks as what is done by the central bank here.)

So how do we get the money from the treasury, to the accounts of those engineering firms and building contractors? (For simplicity, clarity and brevity, I’ll summarise them all as ‘the contractor’ from now on.)

Depending on the country and payment infrastructure, there are several ways. For example MMM, for a different but similar situation, describes that “In today’s world of computerized financial transactions, the Federal Reserve Bank pays for the securities with an “electronic” check drawn on itself!

A check of course which the contractor will cash, i.e. have added to its bank balance with a (here: the, because there is only one) non-central bank.

In Europe, TARGET 2 could be used for this. Whatever the technical organisation may be, the end effect is a transaction in the central bank and one in the non-central bank.

To keep it simpler, I assume the amount the government borrowed, is for a down payment for just a first stage of the complete project. That’s why almost all (18) of the loan (20) is paid in full at once, not in smaller instalments.

At the central bank, the money leaves the government’s treasury and arrives in the non-central bank’s current account, to compensate for the payment the non-central bank is supposed to make on behalf of the central bank:

Description Debit (assets) Credit (liabilities)
Government’s treasury 18
Banks’ current accounts (incl. minimum reserve) 18

The non-central bank pays the contractor:

Description Debit (assets) Credit (liabilities)
Banks’ current accounts (incl. minimum reserve) 18
Contractor’s transaction account 18

After the lending and payment transactions described above, the balance sheets have become as follows.

Central bank’s balance sheet

Description Debit (assets) Credit (liabilities)
Gold 100
Capital 10
Banknotes in circulation 90
Loan to government 20
Government’s treasury 2
Banks’ current accounts (incl. minimum reserve) 128
Refinancing loans to banks 20
Securities 90

Non-central bank’s balance sheet

Description Debit (assets) Credit (liabilities)
Capital 100
Banknotes in vault 10
Current account with central bank (incl. minimum reserve) 128
Refinancing loan from central bank 20
Loans to households and businesses 1000
Current accounts of households and businesses (on demand) (not including the contractor) 50
Current account of contractor 18
Savings accounts of households and businesses (on demand) 950

Effects

Funding of the loan

Where does the money come from? From the above figures, it is quite clear where the money comes from. You say it isn’t? Time for some explanatory text then.

Money creation is a reality, yet the money for every loan has to come from somewhere. To balance the loan at the debit side, there has to be a corresponding item at the credit side. That is why the word ‘balance’ occurs in the notion of a ‘balance sheet’.

Initially, just after the central bank has granted the loan to the government, that corresponding item is the treasury. The treasury was filled with the loan amount, and that amount itself is the funding for the loan that provided it.

After part of the money has been spent, that is, paid to somebody’s account with a non-central bank, in the central bank’s books we see a shift: from the treasury, to the current account of that bank. So the central bank in fact now borrows the money for the loan to the government, from the non-central, commercial bank.

The non-central bank in turn borrows the money from the contractor. That’s because the money in the contractor’s account is a claim on the bank, in other words a debt of the bank vis-à-vis the contractor.

So effectively, taking all the links and intermediaries together and looking only at the beginning and end of the chain, the government is borrowing the money from the contractor it then pays that same money to!

I sometimes wonder if that is really so different from taxation: if the government did not borrow more money, but raised taxes instead, the contractor would pay the government and then the government would pay the contractor. Now on the other hand, the contractor lends the government money which the government uses to pay that same the contractor.

In fact, government borrowing is a kind of deferred taxation: the money already goes the same way, but later taxation will be necessary to be able to redeem the loan.

There is an essential difference of course: the tax payers are not always the same parties as those who receive tax money from the government: there is a redistribution effect.

Net interest costs

Ellen Brown’s idea was that the government can effectively borrow from the central bank interest-free. Any interest the government pays, contributes to the profit of the central bank, and because the government is the sole owner, any profit flows back into the treasury. The net interest cost is zero.

What Ellen Brown overlooks however, is that every loan has to be financed. Every loan requires funding. The money for the loan (despite the reality of money creation) must come from somewhere, usually at a price.

In my example, that cost is zero, because the central bank borrows the money from the non-central bank via its current account (which includes minimum reserves). The interest rate on such accounts is normally very low, often zero. So the central bank found cheap funding for its loan.

The non-central bank also found cheap founding: the current account (transaction account, on demand account) of the contractor likewise has a zero interest rate.

So far, so good, Ellen Brown is right. But is she?

The contractor received the money, but doesn’t have to spend it all at once: salaries of employees, invoices from subcontractors and materials suppliers are paid gradually, as the building project progresses. In the meantime, the contractor may wish to put the received amount (or a part of it) in a savings account, at say 2% annually.

The non-central, commercial bank, won’t be pleased with this situation: it is obtaining the funding for its loan at 2%, but itself receives 0%. The bank sustains a loss. To change that situation, the bank might decide not to renew the refinancing loan it took out from the central bank. Such loans (in the case of the ECB) have durations of 7 or 31 days, so they can be reviewed periodically.

Redeeming a refinancing loan means the bank’s reserves decrease. The bank can afford that, because its reserve ratio was already 11% (110/1000) and has grown to 12.57% (128/1018). Quite high, so using reserve money to reduce the central bank loan is affordable and wise.

This saves the bank some interest, to make good on the 2 percent it is paying to the contractor. Currently however, the ECB interest rate for refinancing operations is only 0.5%, so the net loss for the non-central bank is still 1.5%.

It does make some difference for the central bank too: it is now effectively funding the loan via the decreased refinancing operation, no longer via the minimum reserve. Assuming the government borrows from the central bank at 4%, the net interest cost for the central bank is now 4% – 0.5% = 3.5%.

That interest margin of 3.5% eventually flows back to the government, which is the only shareholder of the central bank. But the government pays 4%. So the net cost of the loan is 0.5%. That’s cheap, but not zero.

However, we must not overlook that the current extremely low ECB interest rates exist in the context of attempts to stimulate the eurozone economies in order to overcome the aftermath of the 2008 financial crisis. And the scenario under observation isn’t realistic because the Maastricht treaty forbids central banks loans to governments. I quote article 123-1:

Overdraft facilities or any other type of credit facility with the European Central Bank or with the central banks of the Member States (hereinafter referred to as ‘national central banks’) in favour of Union institutions, bodies, offices or agencies, central governments, regional, local or other public authorities, other bodies governed by public law, or public undertakings of Member States shall be prohibited, as shall the purchase directly from them by the European Central Bank or national central banks of debt instruments.

So it would be more realistic to look at pre-2008 ECB rates, for example 3.00% for the deposit facility in 2007. Still with 4% as the rate for the hypothetical central bank loan to the government, the net margin and profit contribution flowing back to the government would then be 4 – 3 = 1%. Net interest cost for the government would amount to 4 – 1 = 3%.

Effects on money supply

If the central bank loans money to the government, that extends non-central banks’ reserves. So there is a potential for extra loans, which would automatically extend the money supply (money creation by non-central banks). A larger money supply could have the effect of driving prices up.

Whether that will really happen, depends on a lot of factors: is there actually a demand for new loans, are there parties in the economy who need them, want them and can afford them? Are banks willing to take the risk? Do banks have sufficient capital reserves so they are allowed to grant more loans?

So perhaps inflation will not occur, but there is a real risk, that should be taken into account.

Scenario B

Government borrows from central bank, spends as cash.

Suppose for some reason, the contractor wants the full payment not in its account, but as cash, as banknotes. The non-central bank has 10 mbillion worth of banknotes in its vaults, but probably wants to keep that to serve any other clients that might want cash.

So the non-central bank will have to buy 18 million worth of banknotes from the central bank, which either produces new notes or takes them out of the existing stock of banknotes not currently in circulation. To compensate for the purchase, the bank’s current account with the central bank is debited. The banknotes go to the bank and then to the contractor.

I won’t detail all the transactions here, but the end result is that the non-central bank will be back into the initial state. The same is true of the central bank, except that the unspent 2 are in the treasury, and the value of banknotes in circulation has gone up from 90 to 108, accounting for the 18 in banknotes now held by the contractor.

Funding of the loan

The non-central bank has been taken out of the chain of loans. The government borrows from the central bank. The central bank borrows from the contractor, because that is what issuing banknotes entails: banknotes are claims of the holder on the issuing bank.

Net interest costs

Banknotes are a cheap form of loan funding in the sense that no interest is paid. However, their production is relatively costly because of the necessity of built-in security features against forgeries.

However, this scenario is unrealistic and unstable: even if the contractor wants full payment in cash (which is already strange), those he pays with that cash (employees, subcontractors and materials suppliers) are likely to put the banknotes in a bank again, and from there in savings accounts. That takes us back to the situation at the end of Scenario A.

Effects on money supply

The non-central bank obtained no extra reserves, no there is no potential stimulating effect on the growth of the money supply in this scenario, hence no added risk of inflation.

Scenario C

Government borrows from non-central bank, spends by bank account.

Provided the non-central bank has sufficient capital reserves and cash reserves, it can provide the loan.

The central bank is not effectively involved: where the treasury is a central bank account, the money needs to go to the central bank, but back later to make actual payments. So we can skip those steps. It may also be that the payment transactions are accomplished via government accounts with non-central banks, such as the TT&L accounts (Treasury Tax and Loan) in the US. European governments have similar accounts with non-central banks.

The resulting balance sheet of the non-central bank is as follows:

Description Debit (assets) Credit (liabilities)
Capital 100
Banknotes in vault 10
Current account with central bank (incl. minimum reserve) 110
Refinancing loan from central bank 20
Loans to households and businesses 1000
Current accounts of households and businesses (on demand) 50
Savings accounts of households and businesses (on demand) 950
Loan to government 20
Government’s transaction account 2
Contractor’s transaction account 18

Funding of the loan

In this scenario too, the government effectively borrows the money from the public (here: the contractor), with the non-central bank as an intermediary.

Net interest costs

The cost of the loan for the government is simply the percentage that the bank requires. If it is 4%, the same as what I assumed in scenarios A and B, this loan would be more expensive for the government: in scenario A, the net interest rate, assuming a realistic rate between central bank and non-central bank, was 3 percent.

The comparison may be incorrect, though: both a government and a central bank are trustworthy and creditworthy parties. If a central bank can fund a loan at 3 percent by effectively borrowing from a non-central bank, why would a government have to pay 4 percent?

So in practice, if both situations actually occurred (they don’t in the eurozone, because A is prohibited), net interest costs of scenario C would probably be the same as those of scenario A. In this example: 3%.

Effects on money supply

The non-central bank’s reserve amount with the central bank does not change as a result of the loan in this scenario. The ratio however does change: from 110/1000 = 11% to 110/1020 = 10.78%. Lending capacity is not extended, but slightly reduced. There is no risk of inflation.

Conclusion

It seems obvious that it is cheaper for a government to borrow from its own central bank, because the interest paid flows back into the treasury.

However, on closer examination, it is not that simple. Every loan, no matter by whom, requires funding, and funding normally comes at a price. That makes it likely that in accurately comparable scenarios, borrowing from the central bank costs just as much as borrowing from a non-central bank or other commercial investors.

There is a difference though, in the potential effects on the money supply and inflation: borrowing from the central bank is risky in this respect, and borrowing from commercial parties is not.

The bans on central bank government lending, by the Bank for International Settlements and in the Maastricht Treaty, are justified.


Copyright © 2013 R. Harmsen. All rights reserved.