> Regarding your theory of gold deposits being behind money creation is on Wikipedia as Metallism.

Huh, what? There's no gold deposits in a fiat system, and they still have money creation. And in eg in 19th century Scotland gold was the unit of account, but they scarcely had any gold in the country (yet alone any gold deposits), and they still had plenty of money creation.

> Regarding how banks worked before such laws existed, we could look at the period immediately before the creation of fractional reserve, I would predict that immediately before the rule was put in place, the banks were creating these 1000X (maybe 100X? I don't know) which caused a bank run.

Huh? When would that be? There was no banking system 'before' fractional reserve banking. And what bank run are you talking about? Historically, bank runs basically only happen when a bank has solvency troubles (or when a law pretty much compels bank runs).

> I think a more standard term is endogenous, as in an emergent behaviour of private individuals, rather than exogenous, as in set by a central bank. https://en.wikipedia.org/wiki/Endogenous_money

Sure, I agree that private money creation in commercial banks is endogenous in that sense.

> You are right these are not standard terms, I meant gross money creation as money that is created with a corresponding liability. For example a loan, or a loan with a mortgage on land. The money base increases, but so does the size of the economy represented by that currency in equal increments. Endogenous money seems to map quite cleanly to gross money. Although not all exogenous money is "net" (as in creating currency without an underlying asset).

OK. Technically even the central bank money is created with offsetting liabilities, at least for the mainstream economies and their central banks like the Fed, ECB, Bank of Japan, Bank of England etc.

I say 'technically', because the treasury can just create a T-bill from thin air, sell it to the market, and then the Fed will loan you newly created money against that T-bill.

There's an interesting debate to be had about seignorage and inflation, and about fiat currency vs fiduciary currency.

First, you don't need to print money without a liability to get seignorage income. When the central does a classic open market operation where they buy a bond yield eg 5% for newly created money (naturally yielding 0%, if it's cash), they earn a 5% interest rate differential for as long as that new money is in circulation. That's very similar to how any old commercial bank earns an interest rate differential.

Now this gets a bit more complicated with inflation and/or when the central bank pays interest on reserves held with them. At the moment the Fed pays 3.65% interest on reserves. (Source: https://www.federalreserve.gov/monetarypolicy/reserve-balanc...). But generally they still have a positive seignorage income.

Money created by the Fed and money created by commercial banks contributes to inflation in the same way. The Fed just measures the total inflation that's occurring (and tries to forecast future inflation), and then adapts its own policy instruments accordingly. They automatically take private money creation into account.

About inflation and seignorage: you might think that it's very straightforward and printing more money would lead to more seignorage income. However that's not true in general.

As we noted above, central banks mostly get their seignorage income from the interest rate difference between their assets and liabilities. (Cash and 'reserves' being the primary central bank liabilities.)

When inflation is high, people really cut back on how much cash they hold, at least in real terms. For an example: as the money printing in Zimbabwe really kicked off, people might have held more cash in nominal terms, but in real terms they wallets got lighter and lighter.

In contrast, Japanese people are famous for holding oodles of cash, partially because their years of mild deflation made holding cash relatively more lucrative.

(Turkey is perhaps a good example to follow along, because they want from relatively low inflation in the 2000s to much higher inflation since, but not nearly as catastrophic as Zimbabwe. Or you can look at real money balances in the US through the Great Inflation in the 1970s vs before and after.)

If all you want to optimise for is seignorage income, the optimal rate of inflation is something low-ish and stable. But details depend on how big your economy is, and how easy it is for people to switch to other currencies. So eg the optimal value for Singapore is probably lower than for the US. (Optimal just in the seignorage income maximising sense. Monetary policy typically has other goals.)

About fiat vs fiduciary money: you know fiat money already. Fiduciary money is where you promise to redeem the tokens in some underlying asset on demand. A classic example could be paper money that can be exchanged for gold on demand.

Confusingly, both intermix freely. What I mean is that in a fiat system, the central bank issues fiat money. But the private banks create fiduciary money: your bank typically lets you withdraw your deposit (fiduciary money) as cash (central bank fiat money).

Perhaps fiduciary money is a better term for what you describe as 'net money'?

https://cdn.mises.org/rae9_2_5_4.pdf has some good points on the distinction. (Though keep in mind that the target audience is 'Internet Austrians' who see everything but gold or silver coins as suspicious, and the authors are trying to convince them not to be so dogmatic.)

You might also like 'Those Dishonest Goldsmiths' https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1589709 about the myth that goldsmiths invented fractional reserve banking.