Confidential · Investor brief · September 2026
What the number is, who signs it, and why a new class of contracts will need it
Notes, funds, tokens, perps, and prediction contracts need an agreed risk state at expiry — how jumpy, how skewed, how far the tail. FloatNet is that state, signed by several named firms, so two wrappers that claim the same risk are the same object. On expiry the options that make the strip open in an auction; signers apply the published recipe; the registry holds the print the wrapper pays against. Venues keep matching. New assets reuse the same booklet. This round buys the publisher. Five named firms sign by 23 November.
The VIX succeeded because it converted a statistic into something a futures contract could pay against. That is the whole trick. Before 2004 you could quote implied volatility; you could not easily list a future on it. After Cboe listed VIX futures, the quote had a second life as a settlement object. The teaching text The VIX Forms puts it this way: any successor has to inherit three jobs at once — a public coordinate (a number people can read), a tradable settlement object (a number a contract can expire against), and a legal pointer (a sentence a licence or a court can cite). Miss one job, and the market keeps the VIX and treats the new number as research.
There is a fourth fact, which is the commercial problem. The VIX’s calculation agent is one firm, and settlement happens on an exchange that firm owns. A desk that wants the print is, in practice, a customer of that venue. The same pattern holds for later crypto vol indexes: one agent, one licensed venue. A sponsor that owns no exchange has had nothing to buy.
It does not mean an anonymous token vote, and it does not mean the trade itself moves off an exchange. It means this:
On expiry a Genus I strip opens in an auction — a live market in the options that enter it. Named signers apply the booklet. Supermajority of the five signs; cash sits behind each name. The registry holds the print. A Genus II wrapper — a perp — funds its mark to signed F2 on a published interval. ETH and Canton desks already settle that freeze. OMEGA-1 is a separate 1D ticket: readers recompute G and L from the public tape; signers seal Φ_Ω; GSL is the sleeve versus hold-only. The next listing uses the same quorum. F2 is listed now. First Genus I money line is tail mass, February 2027.
Three markets, none of which scales on a private well. Buyers and sellers need one risk state so two wrappers that claim the same risk are the same object. Matching needs a number both sides already accept, so a venue can list many wrappers without owning the recipe. Settlement needs an agreed risk state at expiry — not last-trade, and not a number that lives only in the listing house. Consensus is several named firms running the same public recipe, signing the same number, cash behind the signature as SLA.
Two facts sit next to each other. The first is easy to see. The second is why there is a company.
More contracts that need a print. Any contract that pays on volatility — or on a tail, a term, or a crash — needs an agreed number when it expires. That set has grown in two directions at once. On listed U.S. paper: VIX futures (2004), options on the index (2006), options on those futures (2024), and the notes and funds that roll them. On crypto and adjacent books: perpetual futures, listed options, prediction-market contracts, tokenized notes, on-chain funds, margin that reads an oracle. Each of those is a wrapper. A wrapper does not invent its own truth. It expires against a number. Hyperliquid is the tape we study because a lot of those wrappers already trade there. Prediction markets are the first listing because that class has the product and not the underlier. A newly licensed VASP that already quotes institutions is not a new customer. It is a book that already needs a number.
Almost no prints that several identified firms will sign. A live quote is common. A settlement print that a second named firm will put its identity and a cash bond behind is not. Cboe computes the VIX; settlement is Cboe’s opening auction; a third party cannot buy that calculation-agent seat. Volmex computes BVIV; venues license it; the agent is still one firm. A desk mark is not a contract. An oracle feed is often a price, not a licence with a fallback and a bond. So the market has many contracts and two kinds of number: a house number (one agent, one venue) and a private number (one desk). What it does not sell is a print several unrelated licensed firms have already signed, in dollars, that another venue can settle to without asking the house.
| What exists in size | What does not exist as a product you can buy |
|---|---|
| Futures, options, perps, notes, funds, prediction contracts — all of them expire against some volatility or tail number. | A settlement print several named firms will sign, with a cash bond, a frozen rule, and a licence a second venue can file. |
What “already visible” means. It does not mean twenty licences in hand. It means the first hour’s buyers are firms that already quote, already make markets, or already incubate wrappers. They do not need to be taught to want a number. They need a number they can take off the house’s venue. Papering the Read against those books this month is that demand becoming a paid feed on 5 October. The February product is that feed becoming a contract that pays. If five unrelated firms will not post a bond by 23 November, the demand was real and the supply of named desks was not — listed-open remains the architecture.
| Order | What | When it pays |
|---|---|---|
| 1 | The listing / methodology licence ($48k/year, primary SKU), the hourly Read ($12k/year, same price to everyone), and a royalty option on listed notional ($25k minimum or 0.5 bp). Freeze licence is residual integrity. | F2 is listed now. Read: 5 October 2026. First product: 8 February 2027. Year-3 base on the listed-open plan: $4.8M recurring. |
| 2 | Integrity: 4-of-5 bonded quorum as a public good a regulator can ask for — not a customer cohort. USD bond floor. Readers stay free. | Lab keys on testnet now. Named desks replace simulated seats. Method PV on the plan of record: $20.8M (band $11.3–29.4M). |
| 3 | The same booklet, more underliers. | P3 first money line in February. Curve, term, shortfall reuse the seats. Incremental cost is a new rule, not a new company. |
The published vector Φ, in ordinary language: wing mass (P3 — first money line); curve (P4); term at 9 / 30 / 90 (P6); expected shortfall (P7). Those four are functionals of q, each with a piecewise-linear strip, auction-settled. P1a (premium vs realised) mixes the surface with a physical path — desk only, with P1b, P2, and P5. High-order slices on a truncated grid are not equally listable. The Read advises. Sponsors pay signers directly ($30k per product per signer). The publisher does not take a cut of the vote.
W, V19 NetCo, the v0.17 SAFE, and DFY as a fee or bond rail are refused or parked. DFY is the unit of listed wrappers. USD is the floor. The network is not the product and not the investable object. Anyone who wants those rows reads the plan of record — they are not restated here.
This preferred stock buys the publisher of the booklet, a ConfigSecretAI subsidiary.
Business Plan v0.20, adopted 11 September 2026, is the plan of record. Priced preferred in the index publisher: $2.0 million at $16 million pre-money. After the round: investors 11%, option pool 10%, parent (ConfigSecretAI) 79%. Method present value $20.8 million; band $11.3–29.4 million. Listed-open architecture: listings primary. Proceeds run the publisher — people, hosting, data rights, counsel — through conversion of a live funding mark into paid licences. The plan does not need the money to survive; it needs it so a slow conversion does not also stop hiring. Open the plan of record.
Demand is the imbalance in §3. The SKUs, refused levers, and three-year cases are in the plan of record, not in this brief.
Drawn from The VIX Forms and CSAI Business Plan v0.20 (11 September 2026). Working figures in USD. ConfigSecretAI, Inc. Counsel has not opined. Not an offer to sell or a solicitation to buy any security or token.