Business Model

Demostatics is an investment company built on top of an information platform. It sells the intelligence, advises on it, manages capital with it, and trades its own book with it.

Founder decision, recorded 2026-08-01

This supersedes the earlier positioning. The platform was originally documented as a data-and-analytics product only, with a standing recommendation never to touch client capital. That recommendation is overridden by founder decision: Demostatics will operate as an investment company across all four lines below.

The consequences are not small, and they are set out on Regulatory Posture. Nothing on that page is legal advice, and every item on it needs counsel before it becomes a plan.

The four lines

#	Line	What is sold	To whom	Status
1	Data and intelligence	Structured global data, risk views, reports, API access	Investors, analysts, research institutions, policy teams	Planned — the platform surfaces exist, nothing is sold
2	Investment advisory	Recommendations, research, model portfolios, signals	Advised clients	Planned — licensed activity, nothing built
3	Discretionary management	Capital managed on the client’s behalf	Managed clients, fund investors	Planned — licensed activity, nothing built
4	Proprietary trading	Nothing — the firm trades its own capital	The firm itself	Planned — nothing built

Line 1 is the original business and remains the only one with any code behind it. Lines 2, 3 and 4 have no implementation of any kind in any repository.

What changes architecturally

The platform stops being only a product. It becomes both a product and internal infrastructure — the same pipeline that feeds paying subscribers also feeds the firm’s own advisory calls, its managed portfolios and its own trading book.

That single fact drives most of what follows.

The platform is now a system of record, not just a display tier

If an advisory call or a managed-account trade is made on the strength of a Demostatics figure, that figure has to be reconstructible years later: what the value was at the moment of the decision, where it came from, how it was derived, and whether it was subsequently revised. That is a record-keeping obligation, not a feature.

Concretely, the schema needs things it does not have:

Requirement	Why	Exists today
Source attribution per observation	”Where did this number come from”	No
Method and method version	A revised model changes the number; you must know which produced it	No
Measured-at and ingested-at	The lag between reality and your knowledge of it	Partially — one timestamp only
Revision history, append-only	A corrected value must not erase the value you acted on	No
Point-in-time reconstruction	”What did we know at 14:02 on the day of the trade”	No

See Data Pipeline and Worker Tier.

This is the expensive one to defer

Retro-fitting provenance means backfilling every observation you have ever published or acted on. Adding it before the tables exist costs a column set. The roadmap’s Phase 2 is the moment it is cheap; after that it is not.

Selling the data you trade on

Lines 1 and 4 sit in direct tension. If Demostatics publishes a figure to subscribers and also trades on it, subscribers will eventually ask the obvious question: did you trade first?

There is no way to answer that with policy alone. It has to be answerable from the system: publication timestamps, the firm’s own order timestamps, and a rule about which comes first that is enforced rather than promised. Deciding that rule is a founder decision, and the options are genuinely different products:

• Subscribers first. The firm’s own book cannot act until after publication. Cleanest to defend, and the most restrictive on line 4.
• Simultaneous release. Everyone sees it at once, including the prop desk. Requires the publication path and the internal path to be the same path.
• Tiered timing. Higher tiers see it earlier. Legitimate and common, but it must be disclosed and it makes the conflict question sharper, not softer.
• Firm first. Defensible only if the data is never sold to anyone who could be disadvantaged, which effectively means abandoning line 1.

Whichever is chosen, it needs to be recorded here and enforced in code, because the question will be asked by a client, an auditor or a regulator rather than by you.

Advising and managing at the same time

Lines 2 and 3 are separate regulated activities in most jurisdictions, and running them together introduces its own obligations — suitability, best execution, allocation fairness across accounts, and treatment of clients who receive advice against those whose money you manage. These are covered on Regulatory Posture.

What this does to the roadmap

The previous plan assumed the firm never touched capital, which let it defer everything about money to a conditional final phase. That assumption is gone.

Was	Now
Phase 9 (money movement) conditional on a contracted partner, probably never	Mandatory, and gated on licensing rather than on engineering
”No money schema goes in the repo”	Superseded — but see the caveat below
Licensing not on the critical path	On the critical path, with the longest lead time of anything
Provenance a display detail in Phase 4	A record-keeping requirement in Phase 2

One piece of the old advice survives intact

The reason for “no money schema before a partner is under contract” was never that custody is wrong — it was that a half-designed money schema is a liability the day it merges rather than the day it is used. That reasoning still holds. Design the money layer against a real licence and a real custodian, not against a guess about what they will require.

The roadmap page carries the phase-by-phase detail, and the source file it mirrors — demostatics-mobile_application/docs/roadmap.md — has been updated to match. See Roadmap.

What has not changed

The data business is still the foundation, and it is still the only line with code behind it. Nothing about lines 2, 3 and 4 makes the platform less necessary — an investment company whose edge is proprietary information is exactly the case where the pipeline, the provenance and the analytics matter most.

The engineering critical path is unchanged: numbers with history, from a real source. Everything in lines 2 through 4 is built on top of that, and none of it is possible without it.

Next

• Regulatory Posture — the licensing and conflicts questions
• Roadmap — sequencing
• Open Questions — what is still undecided