The two tiers carry fundamentally different credit risk profiles, and the asymmetry is what motivates the entire architecture of the wholesale settlement system. A unit of central bank money in domestic currency is the cleanest credit position a domestic counterparty can hold. A unit of commercial bank money is exposed to single-name risk to the issuing bank, mitigated by supervision, capital, liquidity rules, and an insurance overlay that has its own perimeter. Tokenisation does not change the underlying risk position, but it can shift where the legal title sits, and that shift sometimes matters for whether the insurance overlay still applies.
Why one tier is risk-free in domestic currency and the other is not
Central bank money in domestic currency is risk-free because the issuer cannot run out of the currency it issues. A commercial bank can fail, and depositors learn this every several years in some part of the world. A central bank, in its own currency, cannot fail in the same sense, because the currency is its liability and it determines the supply. This is not a comment on inflation, on monetary policy, or on whether central bank money holds purchasing power over time. It is a comment on nominal credit risk for a balance denominated in the issuer's own currency, which is what settlement systems care about.
Commercial bank money is exposed to single-name credit risk to the bank that issued the deposit. If the bank fails, the deposit is impaired. The depositor is in the queue with other unsecured creditors. This risk is real. It is also, for most depositors most of the time, mitigated by deposit insurance, supervision, and the broader safety net that surrounds a regulated bank.
The implication for wholesale flows is direct. A payment between two banks is final, in credit-risk terms, only when it settles in central bank money. A payment that settles in commercial bank money leaves the receiving bank holding a claim on the paying bank, which is a different and often unacceptable risk position for a flow of any size. This is why every major economy operates a wholesale settlement system that uses central bank reserves, and why the payment system architecture cleaves so cleanly along the wholesale-retail boundary, covered in the next part.
BIS research on the singleness of money, including the Garratt-Shin work, frames the two-tier structure as the institutional answer to the question of how a domestic monetary system maintains uniform value across heterogeneous private issuers. The "singleness of money" paper is the standard reference for the wholesale-settlement-in-central-bank-money argument when it is needed in policy discussions.
Deposit insurance overlay
Deposit insurance is the public-policy intervention that makes commercial bank money function like risk-free money for retail and small-business depositors, up to a defined cap. Each major jurisdiction operates its own scheme. The Federal Deposit Insurance Corporation (FDIC) in the United States. The Financial Services Compensation Scheme (FSCS) in the United Kingdom. The Deposit Protection Scheme administered by the Hong Kong Deposit Protection Board. The Singapore Deposit Insurance Corporation (SDIC) scheme. The Deposit Insurance Corporation of Japan (DICJ). The caps differ. The structure does not. Each scheme reimburses covered depositors when an insured bank fails, up to the cap, on a defined timeline.
What the schemes do not do. They do not insure non-deposit liabilities. They do not insure balances above the cap, which means they do not extend to corporate treasury balances at any meaningful scale. They do not extend in any general way to e-money, payment stablecoin reserves, or tokenised fund balances, even where the underlying instrument is held in a bank deposit, because the legal title to that deposit sits with the issuing entity rather than with the end holder.
The implication for tokenisation is direct. A tokenised deposit issued by an insured bank, where the token holder is the legal depositor on the bank's books, can in principle benefit from deposit insurance, subject to the scheme's eligibility rules and the operator's structuring. A tokenised deposit issued through a sub-custody or trust structure, where the token holder is one step removed from the legal deposit, often does not. Designers of tokenised deposit programmes pay close attention to which side of this line their structure lands on, because the answer determines retail eligibility and marketing language.
The same point applies in reverse for payment stablecoins. A stablecoin holder typically has a contractual claim on the issuer. The stablecoin issuer holds a deposit at a custodian bank as part of its reserve. That deposit may or may not be insured at the issuer level, depending on the scheme's rules and how the reserve account is titled. The stablecoin holder is several legal steps away from the insurance perimeter and usually outside it. This is one of the reasons the Stablecoin types taxonomy treats payment stablecoins as a third category sitting outside both money tiers, rather than as commercial bank money in disguise.
What this changes when you tokenise
The structural conclusion is that tokenisation does not, on its own, alter the credit risk attached to either tier. A tokenised reserve carries the same risk-free property as a non-tokenised reserve, because the issuer is unchanged. A tokenised deposit carries the same single-name credit risk as a non-tokenised deposit at the same bank, for the same reason. What tokenisation can do is shift where the legal title sits, by routing the on-chain unit through a wrapper, a trust, or a sub-custody structure. Each of those shifts can move the holder closer to or further from the deposit insurance perimeter, the wholesale settlement system, and the lender-of-last-resort framework. The token does not change the tier. The wrapper around the token can change which protections the holder is entitled to.