There is a contradiction sitting at the centre of on-chain private credit, and most of the industry has been working around it rather than solving it.

Public chains earned institutional attention because they make positions verifiable. An allocator does not have to take a manager's word for what is in the portfolio. The vault balance is right there, continuously, without waiting for a quarterly statement. That property is genuinely new, and it is the single strongest argument for moving credit on-chain at all.

But private credit is called private for a reason. The commercial terms of a facility, the identity of the borrower, the concentration of a counterparty book, the advance rate applied to a specific merchant, none of that is information a credit business can publish and still expect to win deals. In traditional structured finance, this data sits behind an NDA and a data room, disclosed to the parties entitled to see it and to no one else. Move the same facility onto a public ledger without further thought and you have not modernised it. You have leaked it.

That is the trade-off that has quietly capped institutional participation in on-chain credit: verifiability and confidentiality have been treated as opposites, and firms have been asked to pick one.

They are not opposites. Conflating them was always a limitation of the tooling, not a property of the asset class.

What actually leaks

It is worth being specific, because "privacy" in crypto is a word that has been stretched until it means very little.

For a receivables-financing book like ours, a fully transparent ledger exposes at least four things that have real commercial value to someone else:

Borrower-level economics. Which merchants are being financed, at what advance rate, on what settlement cycle. A competitor reading the chain can reverse-engineer the underwriting model and approach the same counterparties with a marginally better rate.

Counterparty structure. Which acquirers and payment processors sit in the flow, and how concentrated the book is across them. This is diligence material, appropriate for an LP under NDA and inappropriate as a public feed.

Merchant transaction volumes. Financing receivables means the merchant's underlying business activity becomes inferable from the financing pattern. Our borrowers did not consent to publishing their revenue.

LP position sizes. An institution allocating into a vault has a legitimate objection to any observer being able to size its exposure, time its entry, and watch for its exit. Treasury movements are strategy. Publishing them is a cost the allocator absorbs for no return.

None of these are edge cases. Together they explain most of the hesitation we encounter from institutional desks that are otherwise comfortable with the credit itself.

The wrong fix

The obvious response, move the sensitive parts off-chain and publish a summary, is worse than it looks.

It reintroduces exactly the trust assumption that on-chain credit was supposed to remove. If the collateral position is asserted in a monthly PDF rather than provable at the contract level, an allocator is back to underwriting the manager's reporting process rather than the portfolio. That is a legitimate way to run a credit fund. It is just not an improvement on the existing one, and it does not justify the operational cost of being on-chain.

The other common response is to hide everything behind a mixer or a shielded pool and treat opacity as the goal. This fails on the compliance side and, more practically, on the allocator side.

An institution cannot deploy into a position it cannot audit, and a regulated counterparty cannot transact with a pool it cannot screen.

Anonymity is not the requirement. Confidentiality with a disclosure path is.

Verifiability without visibility

The useful distinction is between proving a fact and publishing the data behind it.

An allocator does not actually need to see the merchant list.

▫️What they need is assurance that the pool is fully collateralised, that advance rates fall inside the stated policy band, that the maturity profile matches what was represented, and that no single counterparty exceeds a concentration limit.

Each of those is a statement about the data, not the data itself.. Each can be proven cryptographically while the underlying records stay confidential.

That is the shift zero-knowledge infrastructure makes available. A validity proof can attest that a balance is fully backed without revealing the composition of the backing. Encrypted balances can keep an LP's position size confidential from other observers while remaining fully accounted for at the protocol level.

And viewing keys allow selective disclosure. An auditor, a regulator, or a counterparty performing diligence can be granted visibility into exactly the slice of the record they are entitled to, without that access becoming a public broadcast.

The result is not less accountability. It is accountability aimed at the parties who are supposed to have it.

The infrastructure caught up

Until recently this was a paper argument. Building confidential settlement meant building the cryptography yourself, which is not a reasonable ask of a credit team, or migrating to a dedicated privacy chain and giving up composability with the assets and rails that matter.

That constraint has lifted over the past few months. Starknet's STRK20 framework applies confidential balances and private transfers to standard ERC-20 assets.

▫️Including the stablecoins that actually settle institutional flows

Without relocating them to a separate privacy environment, and with selective disclosure built into the design rather than bolted on afterwards. The relevant detail for a credit business is not the cryptography, which is well understood by now, but the fact that it arrives as a token-level capability rather than a bespoke integration project. Privacy stops being an architecture decision and becomes a property you can apply to an asset.

We are working through this problem alongside the @StarknetFndn as part of the first Proof of Privacy cohort, which is focused specifically on teams applying the framework to real financial use cases rather than to theoretical ones. Lending and payments were always going to be the demanding tests here, because they are where confidentiality obligations are legal rather than merely preferred.

What this changes for CREDI

Our model already carries a structural advantage that maps cleanly onto this.

CREDI finances short-duration VISA Ramp receivables, and settlement flows directly from the regulated merchant acquirer into the vault rather than through the borrower. The repayment path is a mechanical property of the structure, not a behavioural promise.

Which means it is the kind of thing that can be proven rather than asserted.

Applied to that structure, confidential settlement lets us hold two positions at once that have historically been in tension:

• Merchant-level data, counterparty terms, and borrower commercial information stay confidential, as they would in any private credit facility.

• Collateralisation, advance-rate policy, duration profile, and vault accounting remain continuously verifiable to allocators, with disclosure to auditors and regulated counterparties available through controlled access rather than public exposure.

For an LP, the practical version of this is straightforward: your position size stops being public information, and your ability to verify the book does not decrease.

Transparency was never the point. Verification was.

On-chain credit spent its first cycle conflating the two because the tooling left no alternative. That constraint is gone, and the asset class should stop paying for it.

CREDI operates an audited Enzyme permissioned vault on Ethereum mainnet financing short-duration VISA Ramp receivables. Target net yield is fixed at 15%, stepping to 18% over 12 months, with a three-month soft lock and a one-week cooldown. Zero realised losses to date.

This material is for informational purposes only. It is not investment advice, a recommendation, or an offer to sell or a solicitation to buy any security. Past performance is not indicative of future results.