Cavport is being developed as an independent historical, quality-adjusted market-reference system based on verified arm's-length caviar transactions. During the pre-launch phase, the Public Retail Monitor uses separately labelled public-market observations and does not constitute a Cavport Verified Reference.
The methodology is versioned in two layers. Active public monitor control governs the public retail data shown today. Pre-launch verified reference control governs the future transaction-based reference. They share principles; they are not the same evidence hierarchy.
Scope: public retail asking prices for caviar and roe, observed on a defined date from public retailer pages. Only exact-pack observations with a resolved net weight, price and currency enter the active EUR analytical set. "From" prices, pack-size-not-stated rows, bundles and unresolved-currency rows are excluded and reason-coded.
No transaction inference is performed. The Public Retail Monitor describes observed asking prices only.
The verified reference uses verified historical arm's-length transactions, evidence tiers, clean-team processing, the arm's-length test, related-party treatment, product state, commercial lanes, normalization, sufficiency, suppression, model governance, corrections and auditability. Sensitive anti-gaming rules remain restricted.
Cavport's hedonic framework estimates the statistical relationship between observed transaction prices and defined product and commercial attributes, allowing quality-adjusted comparison within validated reference domains. An estimated association is not automatically treated as a causal premium or discount.
A participant-derived reference cell may publish only where the applicable sufficiency, contributor-diversity, aggregation, lag and legal-publication controls are satisfied. Cells that do not satisfy those controls are suppressed. Separately labelled research or demonstration estimates are governed independently and must never be presented as verified participant-derived references.
Under the production methodology, every Cavport Verified Reference must carry its observation window, methodology version, model version, eligibility state and integrity record, and must be reproducible from its frozen eligible observation set.
Cavport does not currently present itself as an EU-regulated financial benchmark and is not intended at launch for use in pricing financial instruments, financial contracts or investment funds. Such use is outside the standard service and triggers separate regulatory review.
Public Retail Monitor Methodology V1.0 — initial pre-launch public observation methodology. Verified Reference Methodology V1.5 — current pre-launch working methodology. Material methodology changes will not be applied retroactively without explicit restatement governance.