Verifying the Number Billions Price Off Every Morning
Every business day around 4 p.m. Eastern, a fund administrator somewhere runs a ritual most investors never think about. Pricing feeds close, holdings get struck against end-of-day marks, accruals post, expenses hit, and a reconciliation runs against the custodian's records to make sure the books agree with the vault. Out of that process comes a single number, the net asset value (NAV) per share, against which billions of dollars in subscriptions and redemptions will price the next morning.
The arithmetic itself is simple enough that a junior analyst could handle it. What makes the ritual load-bearing is decades of accounting and regulatory infrastructure underneath: custodial segregation that separates fund assets from everything else, reporting requirements that make deviations expensive, pricing methodologies constrained by the rules that govern the product, auditor relationships that put real names on the quarterly opinion.
Tokenized money market funds (MMF) import the wrapper into an entirely different distribution environment without importing any of the apparatus that made it trustworthy.
Distribution at blockchain speed, verification at audit cadence
A tokenized MMF issues a token representing a share of the underlying fund, which holds the same basket as always: short-duration Treasury bills, repo, agency paper, sometimes a sliver of cash. Economics reach the holder through one of two mechanics: distributed as separate payouts against a stable token price, or accumulated into the token balance itself. The administrator still strikes a NAV each business day and the auditor still reviews the books on the same cadence as any other fund. What changes is the wrapper, which now settles in seconds, moves across permitted wallets without intermediaries, and gets posted as collateral in venues the original fund share could never reach.
This is where the mismatch opens. Tokens propagate across the network at onchain settlement speed, while the NAV behind them continues to update on a schedule designed for T+1 settlement and quarterly attestation. A holder who wants to check whether the token is actually backed by what the issuer claims has the same tools available as a holder of the untokenized share: read the disclosures, trust the auditor, wait for the next report.
Why price feeds and attestations can't close it
Two approaches have emerged to narrow the gap, and both leave the core problem in place.
Oracle networks do real work on price discovery for assets that trade in deep, continuous, multi-venue markets, aggregating independent sources into reliable feeds for stablecoin pegs, spot prices, and yield curves. A fund's NAV sits in a different category entirely, because it is not a market observation but an internal accounting result, produced by applying the fund's regulated pricing methodology to its actual holdings at a specific cutoff.
Attestations attack the gap from the opposite direction. The issuer publishes a NAV onchain and commissions third-party attestations against the underlying books on some periodic cadence. Several issuers also publish dashboards, daily reserve reports, or reserve-proof integrations. All of this adds real transparency on top of an already rigorous traditional foundation of regulated fund administration, SEC reporting, and annual audit. What the onchain holder still lacks is a native way to check the number independently, at the moment they are pricing off it, without trusting the party who produced it.
A proof, not a feed or an attestation
What neither approach produces is the primitive actually needed: the ability to compute NAV over the fund's real holdings data and return a cryptographic proof that the computation was performed correctly against that data. A proof establishes correctness directly, rather than sourcing it from external markets or delegating it to a trusted reviewer.
Space and Time is built to produce exactly that primitive. Proof of SQL lets a fund administrator, an issuer, or a third party run the NAV computation as a SQL query over holdings data and return the result alongside a cryptographic proof of correct execution against the committed data. Anyone with an internet connection can verify the proof independently, without re-running the computation. Verifiable compute does not replace the custody, audit, and reporting stack that establishes what the holdings are; it closes the last mile between those records and what a holder onchain can see and check.
The same primitive extends beyond money market funds. Tokenized credit funds run into an analogous gap on loan-level data, and real-world asset platforms face it on the receivables or commodity reserves backing their tokens.
The institutional standard has to travel onchain
The wrapper, distribution, and settlement layers of tokenized MMFs already work, and the product is growing faster than the verification layer underneath. What still runs on a traditional-finance cadence is the question of whether the token is actually worth what the issuer says.
The issuers that close this gap first will be the ones able to credibly sell tokenized funds to institutional allocators who expect the same standard of assurance they get from traditional custodians and administrators, delivered at the speed the onchain format promises. Anyone still treating the wrapper as the whole product will have to explain why the token moves in real time but the proof that anything backs it does not.
