In this section, the debt-money theory is explored.

The debt-money theory, developed notably by economist Alfred Mitchell-Innes and later expanded by David Graeber, asserts that money did not originate from barter, but rather from debt and credit relationships present from the earliest human societies. According to this theory, money appears when social obligations (debts between individuals or groups) become quantified and transferable. This concept introduces the notion of the quantification of exchanges and transfer. Thus, money is above all a recognition of debt, backed by mutual trust and often guaranteed by political or religious authority, rather than a commodity serving merely as an intermediary in exchanges.

An even more advanced form of this theory is that all money is a debt. Someone who possesses money has a claim on society in general—a claim on all those who accept that money. And since money, in this vision, must be accepted in a compulsory way, otherwise it is not money, because if this money can be refused, the system does not work. In a given group that has identified a currency, all those who possess it are as if they have a credit to all the other members of that society who accept that currency, and those who accept the currency have a form of debt to the person who possesses that monetary unit. This, in essence, is their theory.

I will now offer a critique of this theory. I accept that money is a kind of social consensus. It is a trust that a group places in a monetary "sign", whatever it may be. So I subscribe to the concept of mutual trust, which we'll take up again later with the concept of the monetary premium. Where it gets complicated is when the proponents of this theory add a notion of obligation and insist on the notion of debt. For them, when someone possesses a monetary token, they hold a claim on the community as a whole, and the community would therefore owe the holder a debt. There would also be an obligation to settle, with group members obliged to accept the token; otherwise, in their view, the system wouldn't work. I think the system described above in ancient China represents this theory well.

For my part, I can easily imagine a system where money is a social consensus where the chosen token is generally accepted as a means of payment, with no formal obligation. I can see a system like that working if people accept it of their own free will. It can work. On the other hand, it often comes with obligations, legal tender, and violence, which we'll see later in a section. But theoretically, I don't have the impression that this system can quite work. I think Bitcoin is an example. Bitcoin has value because it has a monetary premium, because there's demand for it to be put to monetary use, i.e., a store of value, perhaps also a medium of exchange or protection against the fiat system. No one is obliged to accept it, and the more it is accepted as a means of payment, the more it can be used as a means of exchange. We can already use it as a store of value without any problem. There's no doubt about that. What I mean to say is that Bitcoin seems to me to be an example, a kind of currency that is a form of social consensus where we think we'll be able to get our money back if we buy some later, we acquire it to be able to exchange it later, to preserve our purchasing power.

This less restrictive view of the theory helps to explain how two distinct societies can trade with a monetary token such as gold, without creating a claim or debt between said societies. To force acceptance in society, I find, doesn't work in trade between two tribes, because the other tribe is not subject to authority. Tribe B is not subject to tribe A's authority. But yet, it accepts the same monetary token—let's say, gold. Now, people might say to me, "Yes, but that's because tribe B has also decided that gold is their currency", so it works. Yes indeed, it works, but there's no debt creation at that point. It's a final transaction. So, when tribe A exchanges a good for gold at tribe B, there is no creation of debt; the transaction is completed. And clearly, the person in tribe B has not acquired this gold coin or nugget in order to be paid later by people in tribe A. He accepts it to be exchanged for gold. He accepts it to be exchanged later, probably within his own tribe. So, in this example, there is no creation of debt, there is no latent claim, it's simply a direct exchange, a direct payment. Proponents of debt currency will tell us that this is barter. And I disagree, because barter is the exchange of a direct good. Direct barter is the exchange of one good for another good, for use. Goods are exchanged to be used, whereas in our case, gold is clearly used as a currency because it's not going to be used to do anything else; it's going to be used later as a means of payment. So, in my opinion, it doesn't fall into the barter category.

Owning a monetary token is not a "claim" on the community, but rather a hope that the token will be accepted at a later date at a comparable value.