PTC Research Methodology

This document describes how PropTradeCenter researches, structures, reviews and publishes firm and industry intelligence. It covers research findings only. Modelled figures are governed separately by the Estimate Methodology.

Research scope

PTC researches proprietary trading firms and the structure of the market around them: firm identity and operating structure, markets and products offered, trading rules and account conditions, execution platforms, and dated company events. This research feeds firm dossiers, the PTC Risk Monitor™, company timelines and industry data publications.

Research finding

Evidence-backed conclusion

A statement supported by identifiable sources — a rule, a platform, a product, a dated event. Published as a finding, with its evidence recorded.

Modelled estimate

Model output

A number produced by a disclosed PTC model. Never a research finding, never presented as an observed fact, and always labelled as an estimate with a range.

Source framework

Sources are recorded with a tier. Tier informs how a finding may be published and how much confidence it carries; it does not by itself make a claim true.

Official: Firm-published documentation — rules, terms, FAQs, help centres, pricing pages, official announcements and firm-owned channels. Also regulators, corporate registries and government sources where the firm or its operating entity appears.
High quality: Credible financial, business and industry publications with editorial accountability, and platform or technology provider announcements naming the firm.
Secondary: Other third-party reporting and aggregators. Usable as a lead or corroboration, not as a sole basis for a material claim.
Community: Structured trader and community evidence, including reports submitted through the PTC report intake. Treated as signal, not proof: it can direct research and support a pattern, but is held to review and is not published as a verified fact on its own.

Source quality, recency and relevance all affect the confidence attached to a finding. A source is described as corroborated only where more than one independent source supports the same claim; single-source and unverified findings are recorded as such and are held to a higher review bar before publication.

Research workflow

Every public finding travels the same path. Nothing enters a public page directly from a research pass.

Step 1
Discovery
A firm, change or event is identified as worth researching.
Step 2
Structured research
Evidence is gathered against a fixed question set per firm.
Step 3
Normalization
Findings are imported into a structured record with source fields.
Step 4
Evidence review
Sources, dates, identity and materiality are checked.
Step 5
Editorial review
A decision is recorded: accept, edit, hold or reject.
Step 6
Publication
Only accepted records become publicly visible.

Evidence & provenance

PTC uses one four-class provenance vocabulary across dossiers, industry data and research publications. Where a figure or statement could be read as either fact or estimate, it carries its class explicitly.

Observed: Directly measured or recorded data.
PTC Estimate: Modelled values based on a disclosed PTC methodology.
Derived: Calculated from PTC datasets, rankings or published snapshots.
Research Finding: Evidence-backed findings from PTC research.

Observedrecorded or measured evidence·PTC Estimatemodel output·Derivedcalculated from published PTC data·Research Findingevidence-backed conclusion

Firm research & dossiers

A dossier is a compiled record, not a summary of one source. Each dimension is researched against its own evidence, so a firm can be well documented on platforms and thinly documented on operating structure at the same time.

Identity: Brand, operating entity where disclosed, jurisdiction of record, and the relationship between related brands so a group's history is not attributed to a sibling brand.
Markets & products: Asset classes offered, account and challenge types, and which programmes are actively sold versus retired.
Rules: Account conditions as published by the firm — drawdown, targets, time limits, payout and consistency conditions — recorded against the firm's own documentation.
Platforms: Execution platforms and technology providers named by the firm or by the provider.
History & events: Dated company events assembled into the company timeline (section 06).
Operating structure: Publicly documented structure, ownership and regulatory footprint where such records exist. Absent records are left empty rather than inferred.
Evidence conflicts: Where sources disagree, both are recorded. The dossier states the better-evidenced position, or presents the claim as unresolved (section 08).

Company events & timeline

A company timeline is an edited record, not a research dump. A candidate event becomes a public timeline entry only when it clears every test below.

Test

Correct attribution

The event is tied to a specific firm identity. Events that cannot be resolved to a firm are never shown on a firm dossier.

Test

Dateable

The event has a date at some usable precision. Undated events are not published.

Test

Materially relevant

Routine promotion and marketing activity is excluded. Structural, operational, regulatory and product milestones are in scope.

Test

Evidence-supported and reviewed

A source is recorded, and an editor has accepted the entry for publication.

Date precision

Exact date: Day, month and year are evidenced.
Month known: Month and year are evidenced; the day is not. Shown as a month.
Year only: Only the year is evidenced. Shown as a year, never as a fabricated date.

Precision is displayed as it is recorded. PTC does not sharpen a year-only event into a specific day to make a timeline look more complete.

Editorial review

Research at PTC may be AI-assisted or automated. Substantive public content is not automatically published: it passes editorial review first, and the decision is recorded against the record.

Accepted: Published as researched.
Edited: Published after correction of wording, attribution, date precision or scope.
Held: Kept unpublished pending further research or better evidence.
Rejected: Not published. The record and its reason are retained internally.

Records may also be superseded or retracted after publication if better evidence emerges. Retracted and superseded records stop being publicly displayed.

Confidence & conflicting evidence

Incomplete evidence is normal in this market. PTC's default is to publish less precisely rather than to publish a false precision.

Incomplete evidence: The field is left empty or marked as not publicly documented — not filled with an assumption.
Conflicting sources: The better-evidenced position is stated, the conflict is disclosed, or the claim is withheld.
Uncertain dates: Published at the precision the evidence supports (month or year).
Low confidence: Published as a range, with a qualification, or withheld from public view.

How reliability is graded and displayed is documented in the Confidence Framework.

Research updates

Update frequency varies by dataset and publication type. PTC does not operate a single research refresh cycle across all content.

Firm dossiers: Updated when a research pass is completed for that firm or when a material change is evidenced. There is no guaranteed per-firm refresh interval, and firms are not all researched at the same depth or at the same time.
Company events: Added as events are discovered and reviewed, including historical events found after the fact. A timeline can gain older entries.
Estimate datasets: Published as versioned baselines with prior versions retained as history, rather than continuously edited in place.

Dataset-level cadence is set out in the Update Schedule.

Limitations

Public information about this industry is incomplete and unevenly disclosed.
Firm rules, products and pricing can change quickly, including without announcement.
Some claims cannot be independently verified and are published as such or withheld.
Absence of evidence is not evidence of absence.
Research reflects the best available evidence as at the publication date shown.
Modelled estimates are governed separately by the Estimate Methodology.