What Makes Money Work on a Ledger?
This chapter asks what stablecoins must preserve to function as money rather than merely as transferable investment claims. The answer requires more than a stable exchange-rate target: the reader needs to distinguish a common unit of account from singleness, understand how liquidity and settlement are supplied, and see why a programmable ledger can improve payments while still weakening interoperability or financial integrity. Those distinctions provide the benchmark for judging the banking and macroeconomic effects that follow.
The institutional properties behind “one dollar”
Money is often introduced through three familiar roles: unit of account, medium of exchange and store of value. The more demanding test is whether a claim can settle an obligation with no questions asked. A shopkeeper accepts a payment because they trust not only the payer but also the monetary claim itself. A firm writes a contract in a currency because the unit will remain a common language for prices and balance sheets. A creditor accepts repayment because the received claim is final and usable by the next person. These conveniences depend on institutions that are easy to overlook precisely when they work.
The first foundation is a common unit of account. Prices, wages, debts and financial statements need a shared numeraire—a standard in which values are expressed. A dollar stablecoin usually borrows this unit from the dollar monetary domain. Its token may be labelled one dollar, and users may choose it because they want dollar exposure. But denomination alone does not make every claim in that unit interchangeable. A theatre ticket priced at $100 and a $100 bank deposit are both denominated in dollars, yet their usefulness as settlement assets depends on different rights and institutions.
The second foundation is singleness of money. In practical terms, claims that promise dollars should be redeemable at par for central-bank money, with finality. One dollar received should be accepted as one dollar without the recipient having to investigate which institution issued it, which network recorded it or whether a queue will form at redemption. Singleness is not the same as a token’s market price being close to one dollar most of the time. A small discount can look harmless in calm markets and become a coordination problem when users start asking whether the next holder will accept the claim at face value.
The existing two-tier monetary system supports this property by dividing responsibilities. The central bank issues the ultimate settlement asset for the currency domain, operates or oversees payment infrastructure, and can supply intraday credit, ordinary facilities or emergency liquidity to eligible solvent institutions. Commercial banks and other regulated intermediaries then provide the payment claims used by households and firms, extend credit, perform customer checks and absorb credit risk. Deposit insurance and resolution regimes help sustain confidence, although they do not make every deposit riskless under every legal or political circumstance.
This arrangement also supplies elasticity. Payment needs change during tax dates, quarter-ends, market stress or ordinary shifts in spending. If the supply of settlement balances were rigid, a temporary shortage of liquidity could prevent payments and turn an otherwise manageable problem into a solvency crisis. Central banks can expand or contract their balance sheets in response to monetary and financial conditions. Commercial banks replicate some of that flexibility one layer down: credit lines and new loans create deposits when households and firms need the means to pay, subject to capital, liquidity and supervisory constraints.
Elasticity is therefore not the same as unlimited money creation. It means that the system has a mechanism for meeting payment demand and containing liquidity shocks, while regulation limits the risks created by private intermediation. A stablecoin issuer can create tokens against incoming funds, but that cash-in-advance process is different from a bank’s capacity to create deposits against new loans. Its ability to meet redemptions depends on the liquidity and market depth of reserve assets, its legal obligations, the concentration of holders and any backstop available in stress.
A fourth condition is interoperability. Money gains value from network effects: sellers accept deposits because they know other sellers will, while buyers hold them because they can spend them widely. If payment instruments and platforms can communicate under common technical and institutional standards, each additional user expands the set of counterparties that can be paid at par. If value becomes trapped inside separate networks, each system develops a smaller network, liquidity must pass through extra intermediaries, and users face more questions about how a claim will move.
The fifth condition is financial integrity. A trusted payment system must limit money laundering, terrorism financing, fraud and other illicit uses. Regulated intermediaries perform know-your-customer checks, monitor transactions and file suspicious-activity reports. These controls concern the parties to a transaction; singleness concerns the acceptability of the settlement claim. The two functions are complementary. A payment can be technically final and still undermine trust if the system is widely used for illicit activity or if no accountable institution can investigate, stop or resolve abuse.
The point is not that traditional finance has eliminated every failure. It is that the current architecture places responsibility for settlement, liquidity, identity and risk management in identifiable institutions with defined mandates. Any proposed money-like instrument must either connect to those arrangements or provide credible substitutes. A stablecoin’s digital form does not answer that institutional question by itself.
What programmable ledgers improve—and what they leave unresolved
DLT, or distributed ledger technology, can address real frictions in financial infrastructure. When tokenisation represents an asset or claim on a programmable ledger, the record of ownership, the instructions for transfer and the settlement event can be brought closer together. A shared, tamper-evident record can reduce repeated reconciliation between institutions. Smart contracts can automate actions when specified conditions are met. Atomic settlement can exchange two legs of a transaction simultaneously, reducing the risk that one party delivers while the other does not. These functions could support fractional ownership, peer-to-peer transfers, around-the-clock operations and more efficient cross-border workflows.
Programmability is not exclusive to DLT; some automation can be built on existing systems. The important contribution of tokenised ledgers is the possibility of combining a shared record with executable rules and coordinated settlement. But those benefits depend on reliable money being available on the ledger. A programmable asset that cannot be delivered for final payment is still constrained by the older settlement chain it was meant to improve.
The network environment matters. Public permissionless blockchains are open to users and, in principle, allow any participant to validate transactions. This openness can support access and competition, but it also brings dispersed or pseudonymous validators, incentive rents, congestion, privacy problems and governance questions. Private permissioned networks restrict participation to approved entities and identifiable validators. They fit more naturally with regulated finance because identity, compliance and operational rules can be built into the system, but they can become walled gardens that limit competition and innovation.
Stablecoins have mostly circulated on public permissionless networks, even though their issuers are often centralised organisations. That combination creates a mismatch between control of the token and control of the rail. An issuer may manage supply and freeze particular addresses, while the underlying network determines how transfers are validated and what happens when a smart contract, bridge or oracle fails. The issuer may be accountable for reserves but unable to provide a common clearing system across every chain on which the token appears.
Fragmentation makes this concrete. A single issuer may create versions of a stablecoin on Ethereum, Solana or another blockchain. The versions can carry the same name, but they are recorded on separate ledgers that do not natively communicate. A user moving value from one chain to another may need a bridge, an exchange or another coordination mechanism. If that mechanism is compromised, a transfer can be delayed, misdirected or permanently lost. Even when no attack occurs, different fees, congestion and confirmation rules make the claims less interchangeable than their shared brand suggests.
The annual report describes similar fragmentation across layer-1 networks and layer-2 systems built on top of them. Layer-2 networks can reduce congestion on a base chain, but their proliferation can also split liquidity and make it harder for one user to reach another. Validator incentives and fees can rise as activity grows. Public visibility may support transparency but conflict with the privacy expected in financial markets. These are not merely user-interface inconveniences. They affect whether payment claims can circulate at scale, whether liquidity can move to where it is needed and whether a problem can be corrected without a central authority.
The result is a continuum rather than a binary label. At one end, a money-like stablecoin would aim to make payments at par in all relevant conditions. That requires high-quality liquid reserves, clear redemption rights, credible liquidity support and safeguards that keep a run from causing fire sales. At the other end, an investment-like stablecoin may accept secondary-market price movements, redemption fees or gates and a broader reserve portfolio. Such an instrument may be useful as a store of value or a source of foreign-currency exposure, but users must ask questions about price and redemption. It is closer to a security or fund share than to a settlement asset accepted automatically.
Reserve composition, redemption conditions, central-bank access and regulation determine where a design sits on this continuum. A coin backed by short-term government securities may have a liquid reserve portfolio but still expose users to market-sale conditions during a run. A coin backed partly by bank deposits may keep funds within the banking system but make banks more dependent on concentrated wholesale funding. A coin linked to central-bank reserves may be closer to the public settlement asset, but access rules and backstops still need to be defined. There is no single reserve label that resolves every risk.
The same is true of financial integrity. Public ledgers can create auditable trails, and issuers or analytics firms may help law enforcement trace or freeze particular balances. Yet routine compliance requires more than the ability to act after a high-profile incident. Pseudonymous accounts, unhosted wallets, mixers and cross-chain bridges can make it harder to identify parties and follow flows. A large-scale payment system needs accountable governance, identity standards, privacy protection and procedures for errors and disputes. Openness and programmability may be valuable, but they do not remove those requirements.
The comparison shows why a dollar peg is only one part of the analysis. The central-bank-anchored system joins a unit of account to final settlement, regulated intermediaries and elastic liquidity. The stablecoin system joins a private issuer to reserve assets and a token on one or more public rails. It may reproduce some functions, but par redemption, interoperability and integrity remain conditions to be designed and verified, not properties guaranteed by the word “stable.”
This distinction also changes how to think about innovation. Stablecoins can create competition for banks, improve access to certain dollar instruments and make programmable payments easier. They can also fragment liquidity or move risk into entities with less access to prudential safeguards. The relevant question is not whether one architecture is old and another is new. It is whether the new arrangement preserves the institutional properties that allow money to coordinate activity under both normal conditions and stress.
With that benchmark in place, the next problem becomes concrete: when households choose a stablecoin as an alternative liquid asset, how does that choice travel through bank funding, loan pricing and firm investment?