Blockchain without the jargon — a ledger, not a legal wrapper

Singapore – “Blockchain” arrived in this city the way many fashionable nouns do: first as a punchline, then as a slide title, then as a word people put in front of an ordinary business in the hope that the ordinary business would look like infrastructure. A grocery voucher became a “ledger play”. A private database with three banks and a vendor became “the blockchain”. The slang was doing more work than the software.

This briefing takes the word apart. It is not a recommendation to use, build or avoid any network. It does not tell you which chain is “better”. For the path a payment actually takes after you tap send, see how on-chain transactions actually move. For the statutes that sit on top of the plumbing, see how MAS regulates digital payment tokens. Adjacent explainers live in the explainers hub.

A ledger is a book of claims

Start with the dull object. A ledger is a list of who is recorded as having what. A bank’s core system is a ledger. A hawker’s order pad, if the stall is honest, is a ledger. CPF records are a ledger. The useful questions are always the same. Who is allowed to write a new line? Who is allowed to read the old ones? What happens if two people write conflicting lines? Who do you call when a line is wrong?

A blockchain is a particular answer to those questions. Copies of the same book are held by many computers. New pages — blocks — are added in a chain, each page pointing at the one before it. Changing an old page without everyone noticing is designed to be expensive. That is the whole trick, stated without a keynote.

Two consequences follow, and both are easy to overclaim.

The first is transparency of a sort. On a public chain, a determined stranger can usually see that address A sent a token to address B. They cannot, from that fact alone, see that the sender lives in Bedok. An address is not an NRIC.

The second is rigidity. A confirmed, valid line is hard to unwind. That is attractive if you are tired of counterparties who “amend” history. It is unattractive if you typed the wrong destination. The network does not have a branch manager.

A company that says it “uses blockchain” has told you almost nothing until it answers the four questions above. Many such companies are running a vendor database with a marketing noun attached. That can be a decent database. It is not a reason to change how you think about the law.

How computers agree, without the theology

The industry word is consensus. The household version is: someone proposes the next page, and the network has a rule for whose page counts.

On some public networks the rule has been a race that burns electricity — proof of work. On others it is closer to a weighted vote among people who have locked up tokens — proof of stake. This desk will not rank the recipes. What a reader in Singapore needs is the social fact. No clerk in Raffles Place stamps the page. A rule, implemented in software that many people run, picks a winner. If enough of those people later accept a different history, the first winner can be displaced. Short displacements happen. Deep ones are rare and, when they are not rare, you are no longer in a calm market.

What consensus is not

Consensus is not a court. It decides that a message was valid under the rules and that a particular ordering is the one the network will treat as canonical. It is not MAS approval: the Authority does not run bitcoin, does not run ether, and does not keep an approved-chain list for retail. A fashionable chain can still carry a capital-markets product; a quiet chain can still carry a digital payment token. It is also not a customer-service desk. An unconfirmed transfer is not “being processed”. It is sitting in an auction for scarce space. The companion piece on movement is the map for that auction.

A blockchain is a shared book with a rule for the next page. It is not a company type, a licence, a court, or a promise that the thing recorded on the page is worth anything in Singapore dollars.

Public, permissioned, and the slide that mixes them

Public (sometimes called permissionless) networks let anyone who can run the software read the ledger and, usually, submit a transaction. Bitcoin and ether, in their ordinary descriptions, live here. You do not apply to a consortium. You also do not get a manager.

Permissioned networks restrict who may write, and often who may read. A group of banks, a shipping line and a software vendor agree on membership. The ledger still chains pages. The politics of membership replace the politics of an open set of anonymous operators. Project-era experiments in this city have often looked like this, because a known membership is easier to reconcile with existing bank controls.

A third, sloppier category appears on pitch decks: “our blockchain”, meaning a database the founder controls, with a block explorer skin. If one party can rewrite history without an expensive fight, you have a conventional administrator.

None of these designs answers a regulatory question by itself.

• A public chain does not make a token legal tender.
• A permissioned chain among MAS-regulated banks does not make a consumer app a bank.
• A private database with a hash published once a day is not, by that hash, a public good.

The useful habit is to ask who can write, who can freeze, and who you sue. Those are legal and operational questions. The word “decentralised” is a slogan until those three have answers.

Tokens sit on ledgers; they are not the ledger

A chain can record more than one kind of object. A native unit — bitcoin on Bitcoin, ether on Ethereum — is built into the rules. A token created by a smart contract is a guest: a balance sheet living in a program that the chain is willing to run. A collectible identifier, a stablecoin, a “governance” coin, a lookalike of someone else’s ticker — all of those can share a chain and share nothing else that matters.

This is why “it’s on the blockchain” is a poor due-diligence sentence. On which chain? Under whose contract? With what freeze function? With what rights in a Singapore court? A transfer you see in an explorer is a fact about a message. It is not a fact about copyright, about a share of profits, or about a claim on a reserve account at a bank.

People collapse these layers because the interface collapses them. One icon, one ticker, one green tick. The law does not. MAS looks at function. A digital representation of value used as a medium of exchange can be a digital payment token. A token that behaves like a share or a unit in a fund can be a capital-markets product. A closed-loop voucher with a chain costume may be neither. The database did not decide. The rights did.

Why this is not a magic legal structure in Singapore

Singapore already has legal structures: a private company, a variable capital company, a trust, a limited partnership, a sole proprietor at the hawker centre. Each is a creature of statute or of equity, with a method for owning assets, owing debts, and being sued.

A blockchain is none of those. You cannot “incorporate on-chain” and skip ACRA. You cannot put a prospectus on IPFS and skip the Securities and Futures Act because the file is content-addressed. You cannot tell a customer that a PSA licence is unnecessary because “the code is the company”. The code is not the company. The company, if there is one, is the company.

What the technology can do is change the operational facts that lawyers then have to describe. Assets may be held at an address rather than in a CDP account. Transfer may be irreversible. Records may be public. A key may be the only practical way to move the asset. Those facts matter for custody, for insolvency, for AML, for consumer disclosure. They do not dissolve the statutes.

A few local confusions deserve a short list.

• “We are a blockchain company, so we are fintech, so we are exempt.” There is no such exemption. Payment services, capital-markets activity and financial advice are named regimes. Marketing language is not a regime.
• “The DAO is the issuer.” Someone is still writing the front end, holding an admin key, or spending the treasury. MAS has not said that a group chat is a licence holder.
• “It is decentralised, so nobody is responsible.” Responsibility, in a consumer case, has a way of finding a person with a Singapore presence, a domain name, or a bank account. Decentralisation is a design claim. It is not a cloak.

Licensing attaches to services: dealing in digital payment tokens, facilitating their exchange, transferring them, providing a custodial wallet. The noun “blockchain” is not a get-out.

What the technology cannot do for you

It cannot make a hollow claim valuable, replace a complaint process, or create deposit insurance. The Singapore Deposit Insurance Corporation covers eligible deposits at participating banks and finance companies, not a token balance. It cannot make a mistyped address come back — finality is the feature people advertise until they need a recall — and it cannot tell you whether a founder is telling the truth about reserves or a bank partnership. Those are disclosure problems. They are older than bitcoin.

A Singapore way to hear the word

When someone in this city says “we put it on the blockchain”, a useful follow-up is the same four questions that work for any book of record. Who writes? Who reads? What if two lines conflict? Who do you call when a line is wrong?

If the answers are “a known operator, a known auditor, a known court”, you may have a conventional system with a new interface. If they are “whoever can sign, whoever runs a node, an auction, and nobody”, you have a public-chain design with public-chain failure modes. Either can be fine, for people who wanted those facts.

What is not fine is the third answer, still the most common on the conference circuit: a shrug, a logo, and the implication that the legal structure took care of itself. It did not. In Singapore the structure is still a company, a licence, a trust, or a person. The chain is a ledger. The rest of the beat is about what people put on it, and what they claim while they do.