The word "stablecoin" is doing too much work. In a single conversation it can refer to a fiat-backed e-money instrument, a basket-referenced token that deliberately fails the e-money test, a tokenised money-market fund used as cash collateral, and a defunct algorithmic construction. These instruments share a chart shape near par, and almost nothing else. For anyone moving from a Web3 protocol seat into a tokenisation role at a TradFi institution, the first useful discipline is to stop saying "stablecoin" and start saying which of the four you mean. The regulatory categorisation is the load-bearing distinction, because it determines who can issue, what reserves must look like, who can hold the instrument on balance sheet, and whether it can move bearer-style on a public chain.
1. Definition: what stablecoin actually means
There is no single legal definition of "stablecoin" that travels cleanly across jurisdictions. The term is a market label, not a regulatory category. When a regulator wants to be precise, they reach for narrower terms: e-money token (EMT), asset-referenced token (ART), payment stablecoin, digital settlement asset, specified stablecoin, regulated stablecoin. Each of those carries a specific set of obligations and a specific perimeter.
The Web3 instinct is to define a stablecoin operationally: a token whose price targets a reference value, usually one US dollar (USD), achieved by some combination of backing, redemption, and arbitrage. That definition is fine for a trader. It is useless for a TradFi treasurer, who needs to know whether the instrument is bank money, e-money, a security, a fund interest, or none of the above, because that determines balance-sheet treatment, risk weighting, and which regulator picks up the phone when something goes wrong.
When a counterparty says "we want to settle in stablecoins," the right next question is which instrument, issued by whom, under which regime. The four answers covered in the rest of this chapter, payment stablecoins, ARTs under MiCA, tokenised money-market funds (MMFs), and algorithmic constructions, cover almost every case you will encounter in 2026.
The taxonomy is not abstract. Each category sits inside a different regulatory perimeter, with different authorisation routes, different reserve rules, different transferability profiles, and different roles in a portfolio. A payment stablecoin and a tokenised MMF can be priced within a basis point of each other and still belong on different lines of the treasury policy, with different counterparties, different custodians, and different supervisors. The price chart hides the structure; the structure is the work.
The BIS working paper by Garratt and Shin on the singleness of money is the cleanest articulation of why these distinctions matter at the macro level. Singleness, the property that any dollar is a perfect substitute for any other dollar at par, is what an institution implicitly relies on every time it accepts cash. Tokenised dollar-denominated instruments do not automatically inherit that property. Whether they do is the function of the regulatory and operational architecture covered below.
2. The four flavours
The next part walks through payment stablecoins, the category most people mean when they say "stablecoin," and the four regulatory regimes that have crystallised around them: MiCA EMT, the Hong Kong Stablecoins Ordinance, the GENIUS Act, and the Japan Payment Services Act (PSA) amendments. Read it as the foundation: the other three flavours (ART, tokenised MMF, algorithmic) are best understood by contrast with the payment-stablecoin baseline.