Before we argue about stablecoins, tokenized deposits, or CBDCs, we have to be precise about what money already is. Almost all of it is a ledger entry. The dollar in your bank app is not a thing you hold. It is a row in your bank's database asserting that the bank owes you that amount. Get this right and the rest of programmable money stops being mysterious. Get it wrong and every later debate turns into a fight about magic internet coins.
Today's money is account-based
The money supply we use is overwhelmingly a hierarchy of claims recorded as account balances. Your deposit is a liability of your commercial bank. That bank, in turn, holds an account at the central bank, and the balance there is a liability of the central bank and an asset of the bank.
When a payment moves, no object travels. Someone debits one account and credits another. The "money" is the agreed difference between two ledger states before and after.
This is why a wire feels instant to you but is operationally a coordinated edit across several books. Your bank decrements your balance, the receiving bank increments theirs, and the two institutions settle the net between them on a higher ledger they both trust.
A worked example: a Fedwire transfer
Say Bank A sends $10 million to Bank B over Fedwire, the Federal Reserve's real-time gross settlement system. The Fed debits Bank A's master account and credits Bank B's master account on its own books, in central bank money. Fedwire settles each transfer individually and the credit is final the moment it posts.
Notice what makes this safe. Both legs of the entry happen on the same ledger, the Fed's, so there is no moment where the money exists in two places or neither. The Fed processes on the order of $4.5 trillion across these books on a typical day, and the entire apparatus is double-entry bookkeeping with a trusted operator in the middle.
That operator is the point. Account-based money works because we agree on who keeps the book and we trust them to keep it honestly.
What a ledger has to verify
Here is the reframing that the whole course turns on. The Bank for International Settlements draws the line cleanly: account-based money and token-based money differ in what you have to verify to accept a payment.
Account-based systems verify identity. To move money from your account, the system must confirm that you are the rightful account holder. That is what your password, your card PIN, and your bank's KYC file are doing. The security model is "prove you are who the ledger says you are."
Token-based systems verify the object. To accept a token, the recipient must confirm the token is genuine and has not already been spent. With physical cash that means checking the note is not counterfeit. In the digital world it means checking the token is real and not double-spent. The security model is "prove the thing itself is valid," and it cares far less about who is holding it.
Account-based money asks "are you who you claim to be?" Token-based money asks "is this object valid?" Programmable money is mostly a shift from the first question to the second.
From accounts to tokens
A token is a unit of value that carries its own validity, so it can pass from holder to holder somewhat like a bearer instrument. Cash is the original bearer token. Whoever holds the note can spend it, and the payee checks the note, not your ID.
A stablecoin built on a public blockchain pushes this model into software. The blockchain is still a ledger, but it is a shared one where the rules of validity are enforced by the protocol rather than by a single trusted operator deciding whose identity checks out. Value can move peer to peer because the recipient can verify the token is genuine and unspent without phoning the issuer.
This is the structural reason a stablecoin behaves differently from a bank balance even when both claim to be worth a dollar. Same denomination, different verification model, different operator, different settlement assurance. We will pull each of those threads apart in later modules.
Why the reframing is load-bearing
Most confusion about programmable money comes from carrying an account-based mental model into a token-based system, or the reverse.
People expect a stablecoin to behave like a deposit, with someone who can reverse a mistaken transfer or freeze a stolen balance on request. Often no one can, because the validity lives in the token and the ledger, not in a customer relationship. Conversely, people sometimes assume tokenized bank deposits give them the open, send-to-anyone freedom of a public token, when in practice they stay account-based and closed, settling only among the bank's known customers and approved counterparties.
The denomination on the front of the instrument tells you almost nothing. The verification model and the operator tell you nearly everything: who can move it, who can stop it, who you are trusting, and what "final" means.
The takeaway
Money has always been a ledger. The interesting question is never "is it on a ledger," because it always is. The questions that matter are which ledger, who keeps it, and what you have to verify to make a payment stick: an identity or an object.
Hold that distinction. When we look at how a stablecoin settles, what backs it, and how regulators wrap it, we are really asking the same three questions of a different ledger each time. The four instruments we cover next are four answers to those questions, not four kinds of magic.