This article continues the plan outlined in the previous article “Central Banks as Banks.” As I described therein, this is projected to become a chapter within my banking primer. I am not going to describe private banking — as that is the job of other chapters — but I will cover the issues of inter-bank transactions. This article is about fundamentals that we normally do not think about.
As the name suggests, central banks are at the centre of the banking system. The objective of a well-run banking system is that you do not have to worry about how it works. So long as everything under the hood is operating, you do not need to enquire how the money gets from one account to another. By not worrying about those details, we tend to only focus on the flashy bits of central banking (e.g., trying to hit an inflation target) instead of the banking system regulation part.
We need to step back and ask: why banks? If we assume that we have a monetary economy, why not have a system where individuals and firms directly transfer “money” to each other without intermediaries?
For some people, this sounds ideal — we have an economy where people pay for everything by handing each other lumps of gold, or directly transferring crypto currencies. However, for law-abiding citizens, this is unattractive — precious metals (and paper claims on safeguarded gold) present theft risk, and making large transactions would be unattractive. (Popular media like the movie Spartacus had the Romans making large commercial transactions with sacks of gold coins; they had the economic equivalent of banks — as discussed in “The Monetary Systems of the Greeks and Romans,” by W.V. Harris.) The inability to sue to recover crypto-currencies makes them also useless for large commercial transactions, as well as poses risks around losing passwords (or heirs not having access to a password).
Although economic theory suggests that everyone “wants to hold money,” in a sophisticated economy where there is trust among economic actors, people want intermediaries of some sort to hold most of the “money.” This means that most “money” ends up being a credit relationship.
The simplest possible system for intermediation is to have a single intermediary. Everyone would hold accounts with that one intermediary, as well as holding bearer certificates (e.g., banknotes) that are convertible to claims on that intermediary.
This intermediary is almost certainly going to resemble a bank, and the only question is who owns it.
Foreigners: you have adopted a foreign currency. Although this might be acceptable in some countries with poor inflation performance, it is not going to be popular in countries used to economic sovereignty.
The private sector. Although some free marketeers might like this, this is not a stable long-term arrangement. This private monopoly intermediary would stand in the way of war finance, and as soon as an existential war risk is faced, that entity would be nationalised.
It could be a public central bank — in a system where there are no private banks. This is popular with some people, many of whom comment on my website. I am not in that camp, as I believe that a well-regulated banking system (integrated with non-bank finance) provides the least instability in capitalist finance. (Finance is inherently destabilising, so all we can hope to do is keep the instability somewhat contained.)
Once we exclude these possibilities, we end up with a situation with multiple intermediaries.
Multiple Intermediaries Leading To Central Banks
If we have multiple intermediaries, we run into an immediate problem. What happens if the two sides to a commercial transaction use different intermediaries? (If they use the same intermediary, then the intermediary just adjusts the two balances without requiring any external transactions.)
We end up with the intermediaries facing the same problem that law-abiding citizens faced: they need to transfer the underlying “monetary asset” between themselves to allow transactions to go across them. If “money” is a precious metal, this means that there will be regular shipments of highly valuable metal, which poses theft risk.
The way to clear these problems up are to have “senior” intermediaries that clear transactions for smaller ones. Client transactions are cleared through a system of connected intermediaries. There is no requirement that there is a single “most senior” intermediary, but we will now get to reason why developed countries moved in that direction.
What is Money?
We now need to face the question: what exactly do our monetary units correspond to. There are two cases.
Pegged currency. The monetary unit is pegged to some external unit, either gold, or a hard currency. (Or possibly a gold-linked hard currency.) The value of the local currency is driven by the credibility of the peg, which typically requires holding backing assets (or generating a considerable trade surplus that allows it to credibly draw in backing as needed). Although there is a convenience factor to there being a central intermediary, it is not required.
Unpegged currency. The monetary unit is the unit of account on a senior intermediary. Since that senior intermediary can create the unit of account at will, it is going to end up as the “central bank.”
In the second case, there is no requirement that the central issuer of the currency be a government, but it seems unlikely it would be anything else. Historically, we had somewhat out-of-control corporations like The Hudson’s Bay Company that could get away with issuing tokens and maintaining the value of those tokens, but that was really only possible because the Company was acting as a de facto government. Otherwise, only the more gullible members of the public would take too seriously an unbacked private currency. (Although the crypto-currency craze has shown that such people exist.) Meanwhile, the spectre of war finance means that the government will sooner or later be the monopoly issuer of the (base) unit of account (leading into standard MMT primer topics).
My interest and focus is on developed economies that have non-pegged currencies (with the Franken-currency of the euro being a sort-of exception). Although the mechanics of a pegged and non-pegged currency might be superficially similar (e.g., there was no major domestic shock to operating procedures when Nixon closed the Gold Window), how the systems behave in a crisis is radically different.
Wholesale Payments Systems
I will now finish off with some general comments about wholesale payments systems. (Retail payments systems — how consumers pay for stuff — is not in my area of interest.) A payments system allows members (typically banks, but lobbyists are pushing for opening up to non-banks) to transmit large blocs of money to each other.
Each currency bloc has its own system, and there is also the issue of transmitting money to other currency blocs. The systems are complex, and most discussions are aimed at the handful of entities that interact with those systems. Given the variance across jurisdictions and my views on their economic impact, I will not attempt to delve into the subject.
The key observation is that the payments system is supposed to be a means of transmitting cash from Entity A to Entity B by the end of the day. Assuming everything goes well, all the payments into and out of the payments system net out to zero. As such, the balance sheet of the payments system is supposed to be effectively zero (beyond whatever infrastructure is on its balance sheet).
The risk that everyone worries about is that a big member fails, and then the payments might not net out to zero. Somebody owes a member money? Who? How is the debt resolved? Given that it is unclear what the bankruptcy judges will say, it is extremely likely that everyone involved would try to freeze transactions — causing a near-instant collapse of the system.
Since it is clear that a non-deranged central bank would view that as a very bad outcome, the payments system would end being bailed out. That is, the wholesale payments system is too big to fail — and properly should be seen as a contingent part of the central bank’s balance sheet.
This leads us to the simple (and standard) way of looking at inter-bank transactions: we assume that they are intermediated directly on the central bank’s balance sheet. If Bank A directly wires money to Bank B, we think of it as Bank A running down their settlement balance at the central bank, and Bank B raising theirs.
There is no need for the balances to be positive during the day. If we take the pre-2020 Canadian system as an example, the target for end-of-day balances was $0. (This is the “simplified system of government finance” that I described in Understanding Government Finance.) That is, the bank starts with a balance of $0, sends and receives money based on client orders (and its own transactions) during the day, and then the bank treasury desk needs to get the balance back to $0 at the end of the day (by undertaking some wholesale transactions). Unlike fairy tales spread by unreliable sources (mainly economic academics), banks do not wait for a positive balance before sending money out — if everyone did that, the system would be frozen at the beginning of the day.
This system is extremely useful for clarifying thinking: banks are sending “central bank balances” back and forth all day, even though their net holdings at the end of the day are expected to be zero. That is, we cannot look at the balance sheet entry (which are end-of-day) to infer anything about “transaction capacity.”
Reserves — Largely an Anachronism
One of the unfortunate side effects of American cultural imperialism is that the most popular undergraduate economics textbooks were in fact written by Americans. Balances at the central bank were called “reserves” for the very good reason that there were almost entirely required reserves. Banking regulations insisted that banks end the day with a target settlement balance, with that target based on the size of their deposit balances (based on arcane distinctions between types of deposits). This balance was a “reserve” allegedly against liquidity drains (not to be confused with loan loss reserves). However, since the funds were immobilised, all they really did was act as a tax on the bank back when reserve balances did not pay interest.
Eventually, the Americans followed the path of other developed countries and effectively abolished reserve requirements. Whether or not people will stop calling settlement balances “reserves” remains to be seen.
Although I might jump to a completely different topic, I think the next article will be about central bank operations and/or its balance sheet structure.