Tokenization is a record. It is not a credit enhancement, a custody upgrade, or a substitute for knowing what sits behind the record. The first capital-loss risk in a tokenized book is not a smart-contract bug and it is not a thin order book. It is an opaque underlying: an asset, a pool, or a legal claim that cannot be inspected, valued, or enforced on the same clock as the token that represents it.
A token can be transferred in seconds. The thing it points to often cannot. That mismatch is the work. If the underlying is a short-duration government bill in a named custodian, with a published holdings file and a redeem path that a qualified holder can actually use, the token is a wrapper around something an investment committee can underwrite. If the underlying is a borrower schedule, a fund of funds, a “cash-equivalent” sleeve, or a note whose collateral is described but not independently confirmed, the token is a claim on a story. Loss arrives when the story and the cash diverge.
Opacity has a structure. It is not the same as privacy. A private-credit pool can be confidential to the market and still transparent to a holder: collateral schedule, borrower concentration, valuation policy, audit, and who can halt redeem. Opacity is the absence of those facts at the moment a decision is required. Marketing pages list categories. Underwriting files list CUSIPs, counterparties, and gates. The distance between the two is where capital is lost.
The legal form is part of the underlying. A token that is title to a security is not the same instrument as a token that is a note, a limited-partnership interest, a tracker, or a contractual claim on an issuer that happens to hold something similar. Recovery sits against the stack that issued the token, not against the asset the marketing deck photographed. If that stack is thinly capitalized, multi-jurisdictional, or able to pause mint and redeem at discretion, the “real-world asset” is, in a default, a real-world lawsuit. Committees that skip issuer, security interest, and enforcement agent are not being efficient. They are postponing the first question.
Valuation is the second face of the same risk. On-chain supply is easy to count. Off-chain NAV is easy to print. Neither is a mark if the inputs are stale, manager-provided, or circular — a token whose “backing” includes other tokens from the same platform, or a fund whose largest sleeve is another tokenized product with its own gates. When the market is calm, the print and the cash path agree. In stress they do not. The first loss is not a liquidation on a public book. It is a gap between the last published NAV and what a redeeming holder can collect after gates, fees, and ordinary settlement delay.
Tokenization can make some of this better. A public holdings file, a named custodian, an independent administrator, and a redeem window that is contractual rather than discretionary are improvements in the information set. They do not appear automatically because a balance was minted. The chain records the token. It does not, by itself, record the loan tape, the custody account, or the board resolution that authorizes a pause. Those remain off-chain facts. Treating on-chain visibility as a substitute for those facts is how an allocator buys opacity at a premium.
The practical test is short. Can a third party, without the issuer’s narrative, name the underlying, the legal claim, the valuation source, and the party that can stop an exit? If any of the four is missing, the token is not “the asset on-chain.” It is a pointer to a file the committee has not opened. That is not a reason to refuse every wrapper. It is a reason to refuse the ones whose first page is a logo and whose last page is a disclaimer.
Hedgen’s work starts there: the underlying, then the wrapper, then the market around the wrapper. Liquidity, composability, and secondary prints come later. They are not irrelevant. They are not first. An opaque book that trades is still an opaque book. Capital loss in tokenization is usually not mysterious. It is the moment the token’s speed meets an underlying that was never required to be slow — or honest — on the same terms.
