The business case for a chain that keeps its inputs
Public blockchains have a structural problem that no amount of throughput solves: they publish everything. Every order is a signal, every position is a disclosure, and every participant subsidises whoever is best equipped to read the ledger. The institutions with the most capital have looked at this trade-off and declined it.
Rand changes the trade-off. A program is public code. A call is a proof that the code ran on inputs only the caller holds, and the chain records the receipt and nothing else. The cases below examine what that single property does to market structure, who captures the value it releases, and what else has to be true before a veto becomes a purchase.
Every case on this page, by the industry that buys it
Twenty industries, each with the cases that apply to it. A link opens the case where it sits on the page. The first ten are where the capital and the incumbents are today; the second ten are industries the confidential model reaches that public chains never did.
Capital markets
- Tokenized stocks and onchain equity trading
- The programmable dark pool
- Sealed request-for-quote and OTC settlement
- Sealed-bid auctions
- A confidential prediction market
- Trade without becoming a public position
- Trade on information without revealing it
- Uniswap, Curve, Balancer
- CoW Swap, UniswapX, 1inch Fusion
- Polymarket and Azuro
- A dark pool clients can program
- A stop-loss nobody can hunt
Market infrastructure
Banking and credit
Payments and treasury
Regulation, identity and compliance
AI and data
Cybersecurity
Governance
Supply chain and manufacturing
Consumer and gaming
Healthcare and life sciences
Energy and commodities
Government and public sector
Advertising and media
Telecommunications
Trade finance and logistics
Professional services
Where the capital is, and why it has stayed away
A hedge fund will not trade on a venue that publishes its book. A company will not run payroll on rails that publish its compensation. A bank will not share a fraud model with a system that publishes its customers. Each is a veto, and in each case confidentiality removes the veto. It does not by itself create the venue, the liquidity or the legal wrapper, and every case below states what does.
Capital markets Tokenized stocks and onchain equity trading
Buyer: Hedge funds, asset managers, broker-dealers, tokenized-equity issuers
| Global listed equity market capitalisation | ≈ $152 trillion | World Federation of Exchanges, end-2025 (published 27 Feb 2026) |
|---|---|---|
| Equities tokenized on public chains today | ≈ $2.5 billion | rwa.xyz, July 2026; xStocks, Ondo, Dinari, Robinhood EU |
| Forecast for tokenized assets | ≈ $2 trillion by 2030 | McKinsey base case ($1–4 trillion, June 2024), excluding stablecoins. BCG and Ripple (April 2025): $18.9 trillion by 2033 including stablecoins and tokenized deposits |
| Hedge fund assets under management | ≈ $5 trillion | HFR, 2025. The buyers of the dark-pool case |
| Taken from Ethereum traders by sandwich bots | ≈ $60 million | in the twelve months to October 2025, most of it passed to block builders as fees (EigenPhi data via Cointelegraph Research, December 2025). The cost of trading in the open |
The status quo
Tokenized equities exist. Backed's xStocks trade on Solana and Kraken, Ondo and Dinari issue tokenized US shares, and Robinhood offers them to European clients. What they lack is institutional flow, and the gap between the market capitalisation of listed equities and the value tokenized so far is five orders of magnitude. The forecasts from McKinsey, BCG and Standard Chartered all assume that gap closes, and every one of them assumes institutions participate. They have not, and the reason is structural. A hedge fund's strategy is its product. On Ethereum or Solana every order it places is visible in the mempool before execution, every fill is a public state change, and its book can be reconstructed by any competitor from the explorer. A fund that trades in the open is publishing its alpha, and no compliance officer, no prime broker and no investment committee will sign off on that. The tokenized-equity market on transparent chains is therefore a retail market by construction, and retail flow alone does not produce the depth that would bring the issuers, the market makers or the forecasts' numbers.
The confidential alternative
A market for tokenized equities on Rand is a program: the matching rule is public code, the orders are private inputs, and the receipt records that a batch cleared. A fund's order is never in a mempool anyone can read, its fill never appears as a position on a wallet anyone can watch, and its strategy stays where it belongs. The objection that keeps hedge funds off Ethereum and Solana does not apply to a market inside Rand, because there is nothing to front-run. That is the difference between a venue that hosts tokenized stocks and one that can host the flow the forecasts are counting on. With institutional flow present, the market makers who currently decline to quote size arrive, depth follows, and the issuers gain a venue in which a tokenized share is more than a wrapper.
What it takes
Confidentiality removes the veto; it does not by itself deliver the market. Three things remain. Custody and legal: qualified-custodian rules, the status of a tokenized share under securities law in each jurisdiction, and prime-broker margin against a shielded position. Settlement: a fund's back office expects finality it can book against, and the venue must state what a receipt settles and when. Liquidity: nobody sources fifty million dollars of a large-cap name on chain today regardless of privacy, so the first products are the ones where onchain settlement already has an edge, such as after-hours trading, cross-border access and collateral mobility. The figures above are directional, drawn from the named sources, and are not a forecast of Rand's own share.
Capital markets The programmable dark pool
Buyer: Hedge funds, market makers, OTC desks
The status quo
The reason institutional capital has stayed off public blockchains is not custody or settlement. It is information leakage. A transparent mempool is a real-time feed of intent, a public ledger is a position report published every block, and any competitor can reconstruct a fund's book and measure how quickly its edge decays. Traditional markets solved this problem decades ago with dark pools, because block trades cannot survive being seen. Every onchain venue asks the institution to give that protection back.
The confidential alternative
On Rand the matching engine is a program and the orders are private inputs. The batch is matched inside a proof, and the receipt discloses only what the venue chooses to publish, typically the net change. More important, the client is not limited to the venue's order types. A fund can deploy its own execution logic and run it against pooled liquidity without disclosing the strategy, the size or the fill. That is not a private exchange. It is a dark pool with programmable order types, which traditional finance has never been able to offer.
What it takes
Three design decisions determine whether the product is credible. Minimum batch sizes, because a batch of one discloses everything. A solvency proof that liquidity providers can verify without a viewing key, because capital does not arrive on trust. And an explicit answer to who sees plaintext order flow at the moment of execution, because that party is the one positioned to front-run.
Capital markets Sealed request-for-quote and OTC settlement
Buyer: OTC desks, dealers, corporate treasuries
The status quo
Request-for-quote is the workhorse of institutional trading, and it lives offchain because it has to. On a transparent ledger every quote is visible the moment it is submitted, so the last dealer to respond wins by a tick, and the client's size is broadcast to the market before the trade is complete. The result is worse pricing for the client and a race to respond last for the dealers.
The confidential alternative
The quote-selection rule becomes a program. Each dealer submits a quote as a private input, the program selects the winner and generates a settlement instruction, and the receipt records that a trade occurred at a price only the two counterparties can open. Losing quotes are never revealed, which restores the incentive to quote aggressively.
What it takes
A commitment deadline, so no quote can be withdrawn once the others are in. And settlement of the value leg in the same bundle, so a winning dealer cannot observe the outcome and decline to honour it.
Capital markets Sealed-bid auctions
Buyer: Lending protocols, issuers, sequencers
The status quo
A public liquidation is an auction in which every bidder can see every other bid, and the outcome is decided by latency rather than valuation. Collateral clears at whatever the fastest bot will pay. Token sales on a transparent ledger have the same structure: the order book is visible before it closes, so the final bidder holds all the information and the first bidder holds none. Both are value transfers from the seller to the best-equipped participant.
The confidential alternative
The auction rule is a program. Bids are committed as private inputs before the close, the program computes a clearing price from all of them simultaneously, and the receipt publishes the winner and the price. No bidder learns another's bid, and there is no advantage to speed, because nothing is visible until the auction is over. The economics are those of a sealed-bid auction, which every auction theorist knows and every public chain has been unable to run.
What it takes
Bid escrow, so a losing bid cannot be an unfunded one, and a batch window or commit-reveal structure that closes before the program runs, so late bidders cannot react to early ones.
Banking and credit Private credit and lending
Buyer: Lending protocols, prime brokers, large borrowers
The status quo
Every position on Aave, Compound or Morpho is public down to its liquidation price. A large borrower near the threshold becomes a target: the market knows exactly what price triggers the liquidation, and it is profitable to move the price there. Liquidation bots then race one another for the bonus, and the borrower funds the race. Sophisticated borrowers respond by not borrowing, which caps the size of the market.
The confidential alternative
Positions become private state of a lending program. A liquidation is a proof that a position is undercollateralised at the oracle price, produced by a keeper who never sees the position itself. Keepers are compensated for proving rather than for seeing, and the borrower's threshold is no longer a public number the market can lean on. The credit discipline is unchanged; the attack surface is gone.
What it takes
A price oracle the program can read, and a mechanism for keepers to identify liquidatable positions without being told which ones they are. Both are open design questions, and a lending product built on Rand will have to answer them rather than inherit answers.
Market infrastructure Oracles under confidential compute
Buyer: Lending and derivatives venues, oracle networks, data vendors
The status quo
On a transparent chain the oracle update is the starting gun. The moment a price lands, everyone can see which positions it liquidates, and the race to act on it is what the industry calls oracle extractable value. Chainlink and API3 built auctions to sell that race back to the protocols it was taken from. Manipulation attacks such as Mango worked because the attacker knew exactly which thresholds a moved price would cross. And the data itself is a problem: a vendor with a proprietary feed cannot sell it into a public oracle without publishing it, which is why institutional-grade market data has stayed off chain.
The confidential alternative
Three things change. First, the update stops revealing targets. With positions private, a price tick discloses nothing about who is liquidatable, so the race shifts from who sees the positions to who proves fastest, and a venue can run that as a fair auction among keepers rather than a latency war. Second, targeted manipulation loses its target: moving a price to hunt a threshold you cannot see is a bet, not a strategy. Third, the oracle itself can go confidential. An aggregation program takes each provider's quote as a private input and publishes the median, so providers cannot copy or signal to one another, and a vendor can license proprietary data into a feed without exposing it. With the source's signature verified inside the guest, the price becomes a proof of provenance rather than a committee's word.
What it takes
Discipline about what remains public. A burst of receipts after a downward tick is still a liquidation wave, so venues must batch or pad calls that are conditioned on prices. Staleness must be closed off by design: the program pins the oracle epoch in its public output and the venue rejects receipts older than the current one, or confidentiality makes manipulation easier rather than harder. And latency: proof time plus oracle cadence sets the minimum quote age, so the front-running problem is replaced by a stale-quote problem unless the feed updates faster than proofs settle. On Rand today a program cannot read the chain, so the working design is the signed price as a private input, verified in the guest, with the price and epoch as public outputs.
Payments and treasury Confidential payroll and corporate treasury
Buyer: Companies paying in stablecoins, DAOs, finance teams
The status quo
A payroll run in USDC is a permanent, public spreadsheet: every employee's compensation, every vendor's terms, every advisor's retainer. A treasury on a public ledger discloses its runway, its counterparties and its negotiating position to anyone who cares to look. Compensation and supplier terms are the two categories of information a company will never voluntarily publish, so the rails go unused regardless of their cost advantage.
The confidential alternative
Rand already carries USDT and USDC as shielded notes. A payroll program takes the pay table as a private input, generates the transfers, and the receipt records that payroll ran. Who was paid what is visible to the employer, the employee and the holder of the viewing key, and to no one else. The cost advantage of stablecoin settlement becomes available to companies that could not previously accept its disclosure cost.
What it takes
A viewing key held by the company's accountant and auditor, so books close and taxes are filed. In this use case the compliance design is not a constraint on the product. It is the product.
Regulation, identity and compliance Private to the market, transparent to the regulator
Buyer: Broker-dealers, funds with AML obligations, chief compliance officers
The status quo
Regulators do not accept privacy systems whose disclosure depends on the cooperation of the person being investigated. Tornado Cash was sanctioned not because privacy is illegal but because the design offered no compliance surface at all. A fund routing flow through a venue whose position is that money laundering is the counterparty's problem inherits that problem, because the fund carries its own AML liability. Any venue with an identifiable operator becomes the pressure point regardless of what its documentation claims.
The confidential alternative
Disclosure is a design feature rather than a hope. The regulated venue through which a fund onboards holds a viewing key, so an auditor can open the relevant batch under subpoena without the user's cooperation. Entry to the pool requires proof of having passed KYC and of absence from sanctions lists, without revealing identity. And disclosure is scoped: an auditor's key reveals counterparty and amount but not the strategy that generated the order. The positioning becomes private to the market and transparent to the regulator by construction, which is a considerably easier conversation with a general counsel than private from everyone unless compelled.
What it takes
A venue operator willing to hold keys and stand behind the compliance programme. That is a commercial arrangement rather than a protocol feature. The protocol's contribution is to make the scoped viewing key a first-class object so the arrangement is possible.
Banking and credit Cross-institution analytics
Buyer: Banks, payment networks, insurers
The status quo
Fraud moves between institutions precisely because no institution can see the others' records. Sharing customer lists is prohibited in most jurisdictions and commercially unthinkable in the rest, so a pattern that would be obvious across two ledgers remains invisible within each one.
The confidential alternative
The join is a program. Each institution supplies its records as a private input, the program computes the intersection or aggregate statistic, and the receipt publishes only the result both parties agreed to release. No record from either input leaves the proof, which is a materially different regulatory posture from a shared database.
What it takes
Agreement in advance on exactly what the output discloses, because a poorly chosen statistic leaks its inputs. Bounding that leakage is a question of program design, not protocol design.
Governance Governance without a bribe market
Buyer: DAOs, protocol foundations, token holders
The status quo
Governance tokens trade in bribe markets because votes are public. The Curve wars and platforms like Votium exist because a payer can verify that a voter delivered. Large holders telegraph their positions, delegates are lobbied in the open, and outcomes are frequently settled before the vote opens. This is a governance failure created entirely by transparency.
The confidential alternative
The tally is a program. Ballots are private inputs weighted by a private proof of holdings, and the receipt publishes the totals. Because a voter cannot prove to a payer how they voted, the vote cannot be sold. The tally remains publicly verifiable, so auditability is preserved.
What it takes
A snapshot of holdings the program can verify against, and a coercion-resistance design under which a voter cannot produce a receipt of their own ballot. Without the second, the bribe market returns.
Cybersecurity Vulnerability disclosure as a market
Buyer: Security teams, bounty programmes, auditors
The status quo
A researcher who discovers a critical vulnerability must trust the vendor with the details before being paid, and the vendor must trust the researcher's claim before paying. Because the evidence of severity is the exploit itself, bounties are slow, disputed and frequently underpaid, and a grey market exists to fill the gap.
The confidential alternative
The target's bytecode is the program's public code, the exploit is a private input, and the program executes it to a state that demonstrates the break. The receipt proves an exploit exists without containing it. Payment settles on the receipt, and the exploit is delivered afterwards under a key the vendor holds. Both sides' counterparty risk is eliminated.
What it takes
A faithful execution of the target inside the guest, which for a contract on another chain means an interpreter for that chain's virtual machine. Feasible for small targets, expensive for large ones.
Supply chain and manufacturing Supply chain assurance without supplier disclosure
Buyer: Manufacturers, importers, certification bodies
The status quo
A supply chain on a public ledger is a map of every supplier relationship, volume and price, which is precisely the information a competitor most wants. Provenance systems therefore remain private, and the sustainability and origin claims they support cannot be verified by anyone outside the company making them.
The confidential alternative
The chain of custody is a private input and the compliance rule is the program. The receipt states that the finished good satisfies the rule, with a proof any customer or regulator can check, while the network that produced it stays with the manufacturer.
What it takes
Signed attestations at the physical boundary, which no cryptography provides. The proof is exactly as strong as the records that went into it.
AI and data Model evaluation that cannot be gamed
Buyer: Model labs, benchmark maintainers, enterprise procurement
The status quo
Benchmark scores drive procurement decisions, and they are increasingly unreliable because public test sets leak into training data. The alternative, a private evaluator who runs the model, requires the lab to hand over its weights. Buyers are left choosing between contaminated scores and unverifiable ones.
The confidential alternative
The evaluation harness is the program, the test set is a private input held by the benchmark maintainer, and the model is a private input held by the lab. The receipt is a score bound to the harness's code hash and a commitment to the test set. No one learns the questions and no one learns the weights, and the score is verifiable by the buyer.
What it takes
The same scaling as private inference: a model that fits in the guest. Small classifiers and scoring functions are within reach today; frontier models are not.
Capital markets A confidential prediction market
Buyer: Prediction-market operators, event-contract exchanges, market makers, institutional hedgers
| Combined monthly volume, Kalshi and Polymarket | ≈ $53 billion | July 2026, up from under $5 billion in September 2025 (Pew Research Center, 23 September 2026) |
|---|---|---|
| Kalshi valuation | ≈ $22 billion | after a $1 billion round in March 2026 (Bloomberg, 19 March 2026); in talks at about $40 billion in August (CoinDesk, 13 August 2026) |
| Polymarket valuation | ≈ $21 billion | post-money, $1 billion round led by 1789 Capital (Bloomberg, 31 August 2026) |
| Share of Kalshi's volume in sports | ≈ 80% | 2026; 39% on Polymarket. Event markets are where retail money already is |
The status quo
On Polymarket every order, fill and position sits on a public chain. Whale trackers and copy-trading terminals follow the wallets with the best records and replicate their entries within seconds, so a trader with real information is paid for it only until the first copy fills. Kalshi keeps its book off chain, which hides positions from other traders but asks every participant to trust one operator with all of them, and gives the market no proof that it settled each contract by its rules. The informed participants a prediction market exists to attract face a choice between broadcasting their edge and handing it to a custodian.
The confidential alternative
On Rand the market is a program: the contract terms, the matching rule and the settlement rule are public code, and the orders are private inputs. Orders are batched and cleared together, and the receipt publishes the clearing price and the open interest, not who bought or how much. A position is a shielded balance its holder can prove to an auditor or a counterparty when they choose, and settlement pays each winner from the pool by a rule everyone can check, with no operator able to see the book or move it.
What it takes
Confidentiality removes the copy-trading leak; it does not by itself deliver a regulated venue. Three things remain. An oracle the program can read, so a contract resolves from a signed outcome rather than an operator's say-so; Rand programs cannot read one today. A scoped viewing key, so the exchange's surveillance team and its regulator can see the trades that insider-trading rules require them to see while the public cannot; the key is designed and not yet built. And the legal wrapper: event contracts are regulated as swaps in the United States and as gambling or financial products elsewhere, and the operator, not the chain, holds that licence. Batching matters as much as privacy: a single visible price move from a single hidden order is still a signal, so the market clears in intervals.
The retail participant, who pays for everyone else's edge
A salary, a trade, a diagnosis, a hand of cards, a preference. Each is something a person would disclose to one counterparty and to no one else. On a transparent chain the choice is to disclose it to everyone or to stay away, and the individuals who stay are the least equipped participants on the venue. These are the same programs as the institutional ones, invoked by one person at a time.
Capital markets Trade without becoming a public position
Buyer: Anyone who trades
The status quo
Every swap on a public chain is visible before it executes, and on the order of a billion dollars has been extracted from ordinary traders by bots that observe an order and trade around it. After execution, the wallet is a public portfolio anyone can follow. The individual trader is the least sophisticated participant on the venue and pays for everyone else's edge.
The confidential alternative
The order is a private input to the venue's program. No one sees it before it fills, and the fill does not appear as a position on a wallet anyone can watch.
What it takes
A venue running one of the matching programs on the institutional side of this page. The individual case is the same program with a smaller order, which is why institutional adoption comes first.
Payments and treasury Be paid on chain without publishing your compensation
Buyer: Employees, contractors, freelancers
The status quo
Being paid in USDC means your compensation sits on a public ledger with your identity effectively attached, along with every purchase made from that wallet afterwards. No employee accepts that, so employers cannot offer it.
The confidential alternative
Pay arrives as a shielded note. You can prove to a landlord or lender that you received it without opening the rest of your history, and you hand a viewing key to your accountant at tax time.
What it takes
An employer running payroll on Rand, and a wallet that can produce a scoped disclosure of a single note rather than the whole account. The wallet experience decides whether this is usable.
Regulation, identity and compliance Prove eligibility without surrendering identity
Buyer: Anyone who must pass a check to use a service
The status quo
A service that needs to verify your age collects your address. An exchange that needs to confirm you are not sanctioned retains a copy of your passport indefinitely. Over-collection is the default because the alternative has not been available, and every retained record is a liability for the company holding it.
The confidential alternative
A credential signed by an issuer is a private input, and the program checks the single predicate the service requires. The receipt is the answer, bound to the issuer's key and nothing else. The same credential serves as the compliance gate on the institutional side of this page.
What it takes
Issuers who sign credentials in a form the guest can verify, and a revocation mechanism, because a credential that cannot be revoked will be sold.
Banking and credit Creditworthiness from your own records
Buyer: Borrowers, lenders
The status quo
A lender requests months of statements to establish a single fact about the aggregate. The borrower discloses every transaction, and the statements remain in the lender's systems long after the decision is made.
The confidential alternative
The scoring model is the lender's program, published so the borrower can read it. The statements are the private input. The receipt is the score, and the lender never sees an individual transaction. The lender's data liability falls to near zero.
What it takes
Statements signed by the issuing bank or exchange. Without a signature, the input is whatever the borrower asserts it to be.
Capital markets Trade on information without revealing it
Buyer: Forecasters, analysts and hedgers, on a venue that meets event-contract rules
The status quo
On Polymarket, an informed trader reveals their edge by acting on it. The order is public, the wallet is public, and the price moves against the trader before the position is filled. Informed participants stay out, which defeats the purpose of the market.
The confidential alternative
Orders are private inputs to the market's program. The price reflects the information without disclosing who brought it, and a position is not a public claim about what its holder knows.
What it takes
The same batching discipline as the dark pool. A single visible price update from a single hidden order is still a signal.
Consumer and gaming Matching without a platform that knows everything
Buyer: Dating, hiring, co-founder search
The status quo
Every matching product knows everyone's preferences because it has to compute the match. Preferences become the platform's asset, and a rejection is information the other party was never meant to receive.
The confidential alternative
The matching rule is a program with both parties' preferences as private inputs. The receipt says match or no match. A no reveals nothing about which side declined, and the platform retains nothing.
What it takes
A platform willing to hold no data, which is a business-model decision at least as much as a technical one.
AI and data Monetise data without giving it away
Buyer: Anyone with data worth buying
The status quo
Selling data means handing over a copy, and a copy is permanent. Once a buyer has the file there is nothing left to sell, so the seller prices the first transaction as if it were the only one, and most transactions never clear.
The confidential alternative
The buyer's computation is the program, the data is the seller's private input, and the receipt is the buyer's result. The buyer receives what was paid for and the seller retains the asset. The same data can be sold many times because it never leaves.
What it takes
A price for computation over data rather than for data itself, and buyers who accept a result they did not compute on their own infrastructure. Both are new commercial conventions.
Ten industries public chains never reached
Finance adopted public blockchains first because its assets were already digital and its participants could tolerate some disclosure. The industries below could not. Each runs on data that is regulated, competitively sensitive or personal, and each has an expensive institution whose only job is to be trusted with it. A program over private inputs is a substitute for that institution.
Healthcare and life sciences Clinical evidence without moving patient data
Buyer: Hospitals, pharmaceutical companies, contract research organisations, payers
Where value leaks
Patient records cannot leave the institution that holds them, so every study that spans institutions is negotiated site by site, and most of the evidence that could be produced never is. De-identification is both lossy and, in practice, reversible. A pharmaceutical company that wants to screen its compound library against a partner's target model must expose one asset or the other.
What changes
A trial endpoint or a cohort statistic is a program run across sites, with each site's records as a private input and the receipt as the result. The compound library and the target model are two private inputs to one screening program, and neither party sees the other's asset. The receipt also carries a proof that the pre-registered protocol was followed, which is the artefact a regulator actually wants.
Insurance Underwriting and claims on data the insured keeps
Buyer: Insurers, reinsurers, brokers, managing general agents
Where value leaks
Underwriting requires the applicant to disclose everything so the insurer can compute one number. Parametric products on public chains publish who holds which policy and how much exposure each carries. Reinsurance requires ceding a copy of the book. And fraud that spans insurers is invisible to each of them, for the same reason fraud that spans banks is.
What changes
The risk score is a program over the applicant's private records, and the insurer receives the score. A parametric payout is triggered by an oracle reading inside a proof, without publishing the policyholder or the exposure. A reinsurer prices a portfolio from proofs about it rather than from a copy of it, and cross-insurer fraud checks run as a join that neither party's data leaves.
Energy and commodities Power, carbon and commodity markets without disclosing the plant
Buyer: Utilities, energy traders, grid operators, carbon registries, commodity houses
Where value leaks
A utility's trading volumes reveal its plant availability and hedging position, so bilateral power contracts are negotiated in private and settled slowly. Plant-level emissions data is competitively sensitive, so reporting is aggregated and unverifiable. Capacity and balancing auctions are run by the operator on trust, and carbon credits carry provenance that a registry cannot prove without exposing the project's counterparties.
What changes
Capacity, balancing and bilateral markets run as sealed-bid programs with a verifiable award. A generator proves its emissions satisfy a rule without publishing plant-level data, and a registry proves a credit's provenance without disclosing the project's supply chain. Commodity trade finance follows the same construction as the trade finance case below.
Real estate Property, rent and mortgages without a public record of the buyer
Buyer: Real estate funds, mortgage lenders, title companies, property managers
Where value leaks
Tokenized property on a public chain discloses every holder and every distribution. Mortgage underwriting requires the borrower to hand over their finances to a lender who retains them. A fund assembling a position telegraphs it with every purchase, and the price of the remaining assets moves accordingly.
What changes
Rent and distributions flow as shielded notes with viewing keys for the auditor and the tax authority. Affordability is a program over private income, and the lender receives a decision rather than a file. A fund acquires through a program that does not broadcast its accumulation, which is the onchain equivalent of buying through nominees, with the disclosure to the regulator built in.
Government and public sector Procurement, tax and statistics with proofs instead of disclosures
Buyer: Treasuries, procurement agencies, tax authorities, statistical offices
Where value leaks
A sealed tender is verified by the agency that ran it, which is the party the losing bidders suspect. An open tender lets suppliers coordinate. Tax compliance requires full disclosure of income, and official statistics require a statistical office to collect and hold the raw returns of every respondent, which is a liability and a target.
What changes
Procurement runs as a sealed-bid program with a publicly verifiable award, so losing bidders can check the rule was applied without seeing the winning bid. A taxpayer proves that the tax computed on private income is correct and has been paid. A statistical office computes an aggregate from private returns and publishes the aggregate with a proof, never having held the returns. Elections belong with the governance cases above.
Advertising and media Ad auctions without the data leak, royalties without the audit fight
Buyer: Publishers, advertisers, ad exchanges, rights holders, streaming platforms
Where value leaks
Programmatic advertising is, by volume, the largest personal-data leak in the economy: every impression sends a profile to every bidder, and every bid is visible to the exchange. Privacy regulation is dismantling the model without offering a replacement. Separately, royalty statements from platforms to rights holders are unverifiable, and the audit disputes that follow are a cost of doing business.
What changes
The auction is a program over private bids and a private profile, and the receipt is a winner and a price. No profile leaves the user's side and no bid leaves the bidder's. Royalties are computed from private play counts with a proof the rights holder can verify, which replaces the audit with a check.
Telecommunications Inter-carrier settlement and spectrum auctions
Buyer: Carriers, tower and infrastructure companies, regulators
Where value leaks
Roaming and interconnect settlement depend on call and data records that each carrier keeps and neither will share, so disputes are settled by negotiation long after the traffic. Spectrum auctions are sealed by regulator fiat, and the award is taken on trust.
What changes
Settlement is a program over both carriers' private records, and the receipt is the net amount with a proof it was computed by the agreed rule. Spectrum and infrastructure auctions run as sealed-bid programs with a verifiable award. Network-sharing agreements, where each operator's traffic is commercially sensitive, follow the same construction.
Trade finance and logistics Invoice financing without double-pledging, letters of credit without revealing the chain
Buyer: Trade finance banks, invoice financiers, exporters, freight forwarders, insurers
Where value leaks
Double financing of receivables is a recurring multi-billion-dollar fraud because no lender can see whether another lender has already advanced against the same invoice. A letter of credit exposes the whole trade chain to every bank in it. Shipping documents are checked by hand against conditions, which is why trade finance settles in days.
What changes
A lender checks an invoice against a registry of private commitments with a proof that no prior claim exists, and no lender's book is revealed. Letter-of-credit conditions are proven met against private shipping documents, and the receipt releases payment. The construction is the cross-institution join from the banking cases applied to a market that has been waiting for it.
Professional services Due diligence from a sealed data room
Buyer: M&A advisers, private equity, law firms, audit firms
Where value leaks
Due diligence requires opening the data room to every bidder, including the ones who will not win and the ones who compete with the target. An audit opinion requires the auditor to hold every transaction. Both are disclosures made because there was no way to verify without seeing.
What changes
A bidder verifies revenue, churn, customer concentration and covenant compliance as proofs computed over the target's private books, and learns the answers rather than the ledger. Audit tests run as programs over the ledger with a proof that each test passed. The seller, not the bidder, decides what is disclosed, and the disclosure is a receipt rather than a copy.
Human capital Pay benchmarking and credential checks without publishing anyone's payroll
Buyer: HR functions, compensation consultants, recruiters, universities, regulators
Where value leaks
A salary benchmark exists only because a consultant collects every participant's payroll. A pay-gap report requires the regulator to hold the underlying data or to trust the company. Employment and degree verification hands a candidate's full history to whoever asks.
What changes
The benchmark is a program across companies' private payrolls, and each participant receives the statistic without any payroll leaving. A pay-gap figure is proven to the regulator without publishing the data behind it. Degree and employment checks are selective disclosures signed by the issuer, which is the same credential construction as the identity case above.
The incumbents: what changes when an existing product's calls are confidential
The pattern across the largest applications on Ethereum is consistent. Wherever a product depends on an offchain solver, relayer, keeper or sequencer, transparency was patched with trust. The confidential version is strictly better for a business that already has users and revenue, which is why the incumbents are the first adopters rather than the last.
Capital markets Uniswap, Curve, Balancer
Where value leaks
Every swap is visible before it executes, and over a billion dollars has been extracted through sandwich attacks. Every liquidity position is a public statement of exposure. The venue's users are its weakest participants, and they pay for everyone else's advantage.
What changes
Orders are private, so there is nothing to sandwich. But the price of an automated market maker is its reserves, so hiding the trades means either batching them and publishing the net, or pricing from an oracle and using the curve only for residual imbalance. A confidential AMM is a batched or oracle-priced venue, and the product changes shape rather than simply gaining a feature.
Banking and credit Aave, Compound, Morpho
Where value leaks
Collateral, health factors and liquidation prices are public. Bots compete on latency for the liquidation bonus, and large borrowers are pushed toward the threshold by participants who can see exactly where it is.
What changes
Positions become private state. A liquidation is a proof of undercollateralisation at the oracle price, and keepers are compensated for proving rather than for seeing. The addressable market expands to borrowers who currently stay away.
Banking and credit MakerDAO and Sky
Where value leaks
Vault positions are public, and governance votes on rates are public, so large holders signal their intentions to the market before a vote closes.
What changes
Vaults receive the private-lending treatment, and the rate vote becomes a private ballot with a public tally.
Capital markets CoW Swap, UniswapX, 1inch Fusion
Where value leaks
These venues exist because the mempool is public. Users sign intents and solvers compete offchain to fill them, which means the user trusts the solver with the order flow, and the solver's execution is unverifiable.
What changes
Batch execution against sealed intents becomes a program the chain runs and proves. The business model survives; the trust assumption does not.
Market infrastructure Flashbots and MEV-Boost
Where value leaks
Proposer-builder separation is an elaborate workaround for a readable mempool. Searchers see, builders order, validators sell the right to order, and the value extracted comes from users.
What changes
A transfer or a call reaches Rand's mempool as a proof. The amounts, the parties and a call's inputs were never in it, so a builder has nothing to read and trade ahead of. The mempool itself is not encrypted, and the block's leader still chooses the order of what it holds: what changes is that the order is of proofs, not of readable intent. Shutter's encrypted mempool is the current attempt to get there on Ethereum.
Governance Snapshot, Tally, Compound Governor
Where value leaks
Every vote is public and verifiable, which is exactly the property a vote buyer needs. Votium and the Curve wars are the market that transparency created.
What changes
A private ballot with a public tally changes DAO governance more than any adjustment to token economics, because a voter can no longer prove how they voted and therefore cannot sell it.
Capital markets Polymarket and Azuro
Where value leaks
Order flow is public, so anyone with real information reveals it by acting on it, and the market moves against them before they are filled.
What changes
Private orders let informed participants trade without disclosing their edge, which is the condition under which a prediction market actually aggregates information.
Payments and treasury Safe
Where value leaks
Every signer, every threshold and every queued transaction is visible. Treasuries move slowly because every movement is announced in advance to the market and to adversaries.
What changes
The signing policy is a program, approvals are private inputs, and the receipt is an authorised transfer. This is the control environment a corporate treasurer expects from a bank.
Regulation, identity and compliance ENS
Where value leaks
Reverse resolution ties a wallet to a name, and through the name to every transaction that wallet has ever made.
What changes
A private lookup answers the query without publishing the association, which changes what a name discloses about its holder.
Payments and treasury Stablecoin payroll and payments
Where value leaks
Sablier streams, Request invoices and plain USDC transfers all publish the amount, the parties and the schedule. Compensation and supplier terms are the two categories of information no company will put on a public ledger.
What changes
Confidential transfers of a shielded stablecoin, with a scoped viewing key for the accountant. Rand already has the shielded stablecoin; the payroll program is the thin layer on top.
Consumer and gaming OpenSea, Blur, NFT lending
Where value leaks
Much of the auction dynamics on Blur are transparency exploits: bids are visible, so the final bidder wins and the first bidder is merely information.
What changes
Sealed bids and a program that selects the winner. Sellers capture more of the value and bidders compete on valuation rather than timing.
Market infrastructure Lido, EigenLayer, restaking
Where value leaks
Validator and operator economics are public, along with who delegates to whom and how much slashing risk each operator carries.
What changes
Delegation becomes a private input to a staking program. The change is real but modest, and it is listed so the reader can calibrate: not every product on a transparent chain is materially improved by confidentiality.
New categories: products neither a transparent chain nor a trusted operator could build
The previous section is existing products made better. This one is different. Each of these was blocked in both worlds, on public chains because they conceal nothing and in traditional institutions because someone has to be trusted to see everything. A proof over private inputs is the first mechanism that is neither, and each of these is a market that did not have a supplier.
Capital markets A dark pool clients can program
Why the market never produced it
A dark pool is dark because a trusted operator keeps the book, and no operator runs a client's proprietary code inside that book. A public chain, which will run anyone's code, cannot keep a book dark. The two properties required different institutions, and no institution could offer both.
What changes the economics
A program on Rand is public code that executes on private inputs. A fund deploys its own matching logic, submits orders only that logic can see, and the chain proves the logic ran. The operator, and the trust the operator required, are removed from the value chain.
Regulation, identity and compliance A venue private to the market and transparent to the regulator
Why the market never produced it
Transparent chains are transparent to everyone, including competitors. Mixers are opaque to everyone, including regulators. The position in between, private to the market and open to an authorised auditor by design, requires a scoped disclosure primitive that neither category possessed.
What changes the economics
Viewing keys scoped to a batch, a counterparty or an amount, held by the regulated venue rather than the end user. An auditor opens what a subpoena names and nothing more, and the client's strategy remains sealed even under examination. Compliance becomes a feature the venue sells rather than a cost it absorbs.
AI and data Inference in which neither party sees the other's asset
Why the market never produced it
Running a model on data requires the model and the data to be in the same place, and whoever controls that place sees both. Every privacy-preserving alternative to date has been a research prototype or a bet on hardware.
What changes the economics
The weights and the query are two private inputs to one program, and the receipt proves the answer came from the model identified by its code hash. The place where they meet is a proof, and a proof has no operator. Markets that currently fail on negotiation cost can clear.
AI and data A benchmark that cannot be contaminated
Why the market never produced it
A public test set leaks into training data. A private test set requires trusting the evaluator with the model or trusting the lab with the score. Benchmarks have been either contaminated or unverifiable, and buyers have been choosing between the two.
What changes the economics
The harness is public code, the test set and the model are private inputs, and the receipt binds the score to the harness's code hash and a commitment to the test set. The score is verifiable and the questions are never published. Evaluation becomes a service with a defensible product.
Governance A vote that cannot be bought
Why the market never produced it
Onchain governance is verifiable because ballots are public, and a public ballot is a receipt a buyer can pay against. Every DAO has lived with this because the alternative appeared to require giving up verifiability.
What changes the economics
The tally is proven rather than the ballots. A voter cannot demonstrate to a buyer how they voted, so payment cannot be conditioned on delivery, so there is no market.
Cybersecurity A bug bounty paid on proof, before the details change hands
Why the market never produced it
The evidence of a vulnerability is the vulnerability itself. Disclosure has always required trust in one direction or the other, which is why bounties are slow and disputed.
What changes the economics
The target executes inside the guest, the finding is a private input, and the receipt proves the target reached a broken state. The vendor's bounty clears on the receipt and the details are delivered to that vendor afterwards, under its key. Counterparty risk is removed from both sides of a disclosure that stays between researcher and vendor.
Banking and credit A liquidation nobody can hunt
Why the market never produced it
A lending protocol needs someone to trigger liquidations, and on a public chain that someone needs to see the positions. Seeing the positions is indistinguishable from being able to hunt them.
What changes the economics
A keeper proves a position is underwater at the oracle price without learning the position. The trigger is separated from the sight, which no transparent system could accomplish, and the market opens to borrowers who currently decline to participate.
Capital markets A stop-loss nobody can hunt
Why the market never produced it
A stop-loss on a transparent chain is a published invitation: the trigger price is visible, so the market pushes to it and the order fills at the worst moment. Stop hunting is a feature of every onchain venue that has tried conditional orders, and the response has been to keep them offchain with a trusted operator, which is the thing the chain was meant to remove.
What changes the economics
The threshold is a private input, the price is a public oracle reading verified inside the guest, and the receipt is a legal order that fires only when the condition holds. Nobody sees the trigger before it executes and nobody can prove it existed if it never fires. Take-profits, trailing stops and any strategy that is a rule over public prices follow the same construction, which turns algorithmic execution from a service a broker sells into a program a client deploys.
Market infrastructure A mempool with no orders in it to front-run
Why the market never produced it
A transaction has to be visible to be executed, and once visible it can be traded around. Encrypted mempools on Ethereum are retrofits that depend on a committee to decrypt at the right moment.
What changes the economics
A call on Rand is a proof over private inputs. What reaches the mempool is already decided, and the inputs that decided it were never there. The mempool is not encrypted: a call's program and its eight output words are public, and the leader orders transactions as it likes. Ordering remains a fact, but it is a fact about proofs, not about orders anyone can read, so the extractive layer has far less to work with.
Today a Rand program is a small RISC-V guest with private inputs and eight public output words, and a call's outputs are data rather than effects: a program cannot move value by itself. The cases above describe what the model is designed to support. Several require larger guests, an oracle the guest can read, or a scoped viewing key that does not yet exist. The developer documentation states precisely what works on the current chain, and readers evaluating a build should start there.