Recent Posts

Thursday, November 30, 2023

The Central Bank And Government Finance

This article continues the sequence of articles on central banks as banks. This article was as brief as possible since it overlapped my book Understanding Government Finance (available for sale cheaply at online bookstores, and I emphasise that it would be an amazing Christmas present for friends and/or enemies (depending on what you think of my writing)). I might need to expand upon the less obvious points herein if this text does get into my book manuscript.

Central banking largely evolved the way it did due to the exigencies of wartime finance. The central government needs control over its financial operations in wartime, and any attempts to interfere by the private sector would be viewed as akin to sabotage. For a free-floating sovereign (and currency pegs are typically broken during major wars), the system guarantees that the financial flows will continue to flow.

As I discussed in Understanding Government Finance, the system is relatively straightforward, but it is unintuitive if you start with the misconception that the objective of government finance is for the government to “raise money” from the private sector. Instead, the system just allows the government to follow archaic accounting norms while ensuring that payment flows remain circular.

Wholesale Payments System Recap

I will assume for simplicity that we have a wholesale payments system that has counterparties that are only private banks, and the central government, which we divide into the finance arm (“Treasury”) and the central bank. (Since the central bank acts as the agent for the Treasury, it is really on the central bank that is the counter-party, but we need to break out the Treasury for this discussion to make sense.) If the government opened the wholesale payments system to non-bank entities for some reason, we just assume that the non-banks have to follow the same conventions as the private banks. We also assume that payments system balances for private banks will correspond to balances held at the central bank.

We will also assume for simplicity that the private banks are expected to keep their end of day balance with the payments system at $0 (the pre-2020 Canadian system). If the system featured positive reserve balances, then the target ends up being that positive balance, which does not really change the discussion herein — there is just a level shift of the target.

Under the reasonable assumption that the number of private banks is finite1, then the previous assumption implies that the total private sector net balance with the payments system at the end of the day is $0. Since the payments system is zero sum, by implication that implies that the consolidated central government balance is also zero (since it is the remaining counter-party to the entirety of the private banking system).

This then implies that the net money in-/out-flow of the consolidated central government is $0 every single business day. (It might be unbalanced during the day, but it has to be brought back to balance by the end of the day.)

Central Government has to Cancel Out the Treasury

The implication is that if the Treasury has a net financial flow during the day, the central bank has to undertake transactions that generate the opposite flow. (If the target for reserve balances is non-zero, the central bank has to make sure the change in the consolidated government balance matches the change in target level, if any.)

Let us imagine that the government had spend and tax transactions that net out to zero one day, but the Treasury issued $10 billion in bonds. This means that the Treasury has a $10 billion monetary inflow from the private sector — the bonds have to be paid for by the private banks, possibly under the instruction of clients who were the ultimate purchasers of bonds. By implication, the central bank has to have a $10 billion cash outflow for the balance target to be met. If the only assets the central bank holds are central government bonds/bills, it has to sell $10 billion in old securities (or do repo transactions) to match the $10 billion in new issuance.

That is, the central government did not draw in “new money” from the private sector, it just pushed out new securities in exchange for old.

This sounds weird, but it is just how the mathematics works.

No, the Private Sector Did Not “Create” the Money

One of the crackpot anti-MMT lines is that the private sector “creates the money” that pays for the bond issuance. Yes, bank lending creates banks, and those deposits might allow a non-bank client to put in an order at the bond auction. However, you could imagine a strange intermediary firm that allowed their clients to exchange wiener schnitzel for bonds at auction. This does not imply that government bonds are “paid for by wiener schnitzel,” rather, there is an intermediary deciding it likes wiener schnitzel. The intermediary itself has to wire “government money” on the wholesale payments network to the Treasury (technically, the central bank, since they typically run auctions) to pay for the bonds, and the intermediary then deducts the deposit (wiener schnitzel) from their client by agreement.

The private sector always starts the day with a $0 balance of government money — there is no pool of “money” to be drained to pay for the bonds; the central bank has to supply the “money.”

Well, Where Do Bonds Come From?

An alert reader will have noted that the bond auction created exactly zero net government bond holdings during the day — the old bonds that are sold to pay for the new one cancel out the issuance. So, where do bond holdings (which are non-zero) come from?

The answer is not along the lines of “When a mommy bond and a daddy bond love each other…,” rather, we need to drop the assumption of tax and spending cancelling out. If the Treasury runs a payments deficit during the day, it is sending money into hands of the private sector. The central bank needs to cancel that out — by selling bonds that it owns. This creates the net bond flow to the private sector.

As expected, the size of the fiscal deficit will determine the increase of government liabilities in the hands of the private sector (some of which might be banknotes, which need to be paid for by wiring money to the central bank).

Conventional Accounting

The analysis so far has just looked at the private sector, which is the counter-party to the consolidated central government. If we split the Treasury and central bank, we also need to track the balance of the Treasury at the central bank. A sensible society would argue that this balance is purely an accounting construct and can be ignored, but we do not live in that society. Instead, we see that bond issuance increases the balance of Treasury at the central bank. Subsequent deficit spending will tend to run that balance down, and so new issuance is needed to keep the balance positive.

As long as the central bank ensures that interest rate markets are orderly — literally its job — the Treasury will always be able to squeeze out new bonds to keep its balance positive at some price. (I.e., if bond yields go up, interest rate expenses go up.) Default is pretty much only possible if the central bank decides to force the default. (Why “pretty much” — we can imagine non-financial reasons for a default.)

Why Issue Bonds in the First Place?

Anyone writing on this topic on the internet is going to face comments from MMT fans arguing that government bonds do not need to be issued, allowing society to skip the elaborate financial game-playing described earlier. I will now just run through the conventional justifications for bond issuance without endorsing them.

  1. It is needed to meet the arbitrary accounting rules (a positive balance at the central bank). This is not an economic necessity — changing the rules makes it disappear. I note this as some people refuse to accept that governments can unilaterally change regulations.

  2. It creates a yield curve to allow the private sector benchmarks for its borrowing. Although this is useful, it is a niche concern.

  3. The yield curve allows interest rate policy to control the economy. The usefulness of interest rate policy is an ongoing controversy between MMT proponents and pretty much every body else.

  4. It provides credit risk free assets that are extremely useful for private pension and insurance provision. Given that governments have pushed a significant proportion of voters into private pensions (both defined benefit and defined contribution), this is not an easily reversed policy at this point.

  5. Having credit risk assets stabilises the financial system. Private sector bond prices melt during a financial crisis; only central government bonds have a hope of retaining their value.

  6. Ending issuance would put “bond vigilantes” and people who write about government bonds out of work. (Since I am semi-retired, not a major concern for me.)

Concluding Remarks

Central banks were not set up as a charity to employ economists with delusions of grandeur, they are there to ensure the smooth functioning government finance. The system is set up to allow extremely chunky transactions to take place. Ensuring that everybody can pay for cheesy poofs at the corner shop is an afterthought.

With this article out of the way, I can turn away from the structure of central banking and start discussing some of the many controversies about them.


If we have N banks, each with a balance of $1/N, the aggregate balance is $1 as N goes to infinity. 

Email subscription: Go to 

(c) Brian Romanchuk 2023

No comments:

Post a Comment

Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.