Estimate Methodology

PTC estimates are modelled from public evidence and proprietary methodologies. They are not company-reported figures unless explicitly identified as such, and they are not audited industry totals.

1. 01Tracked Market
   Firms PTC actively researches and maintains as active.
2. 02Modelled universe
   The subset with sufficient evidence and a covered universe assignment.
3. 03Evidence & calibration
   Firm disclosures, observable signals and calibrated model parameters.
4. 04Estimate + range
   A central modelled value with a low–high band per firm.
5. 05Publication gate
   Classification decides whether a firm-level value is shown, shown range-first, or withheld.
6. 06Aggregates & baseline
   Universe and combined totals, locked into a dated industry baseline.

Firm disclosures

Figures a firm has itself published — trader or customer counts, funded populations, payout totals. Classified as observed evidence, with explicit semantics for what each figure counts (persons, accounts, currency, transactions) and for which period, so a payout amount is never read as a headcount.

Observable signals

Search demand, search visibility, public review activity, offer and platform structure, and operating history — recorded from public sources.

Calibration parameters

Model coefficients and conversion assumptions fitted by PTC and stored with the model version. These are model parameters, never presented as observed industry averages.

Operational status

Firms confirmed closed are excluded from current estimates and current aggregates rather than carried forward.

Not used

Private client data, firm-supplied unpublished figures and paid data that PTC cannot cite are not inputs to published estimates.

A firm whose individual estimate is withheld still contributes to the aggregate for its universe. Aggregates are computed across the full modelled universe; the publication gate governs firm-level display only. This is why the number of firms listed on an estimate page can be smaller than the number of firms behind its total. The modelled universe is expected to grow as coverage expands, and each page reports its own current counts rather than a fixed figure.

• Each firm-level estimate carries a central value and a low–high range produced by its model.
• Ranges are model uncertainty bands, not statistical confidence intervals. PTC does not publish a sampling-based confidence interval for any estimate.
• Where evidence is thin, the range is the primary figure and the central value is de-emphasised.
• Aggregate totals are published as central sums. PTC does not publish an industry-level uncertainty band, because summing firm bounds assumes independence that the models do not claim.
• Derived figures such as an annualized run-rate or a funded-to-active ratio are arithmetic on published values and are labelled Derived.
• Snapshot estimates are not a time series. Growth rates and forecasts are not inferred from a single snapshot.

Confidence

Each estimate carries a confidence band — High, Medium or Low — reflecting evidence breadth, calibration quality and how far the value depends on fallback assumptions.

Basis label

Every published estimate states its basis: an evidence-led estimate is anchored or bounded by firm-specific disclosure; a modelled estimate is produced from calibrated relationships; a scenario estimate is range-first by construction.

Publication gate

A classification is applied per firm and per metric: publishable as a value, publishable range-first, or withheld from firm-level display. Insufficient or unusable values are never shown with a substituted figure.

Withheld ≠ zero

A withheld firm-level estimate means PTC will not publish a firm-specific figure at that evidence level. It is not a statement that the firm is small.

Version filtering

Public pages read only explicitly approved model versions. An estimate produced under an unapproved or superseded version is never published, and superseded firm estimates are retained rather than overwritten.

• Estimated Active Traders™ is the upstream model. Its output can serve as an input to the Funded Trader and Monthly Payout models where firm-specific evidence is unavailable.
• Estimated Monthly Payouts™ prefers direct firm-specific payout evidence. Where none exists, it applies a calibrated universe coefficient to the upstream active-trader estimate.
• Estimated Funded Traders™ likewise uses the upstream active-trader estimate as a scale reference where firm-specific funded evidence is absent.
• A value carried down from an upstream model is never treated as independent corroborating evidence for the downstream estimate, and confidence for the downstream estimate cannot exceed what its own evidence supports.
• Firm-disclosed evidence must declare what it counts and for what period before it may bound a downstream estimate, so a payout amount or a cumulative population can never be read as a current headcount.
• Ranking indexes and estimate models are separate systems. An index score is not an input substitute for an estimate, and an estimate is not published as a ranking.

What a baseline is

A dated set of aggregate estimate values for a stated period, stored with the model versions and universe membership that produced it.

Locking

Once a period is locked or published, its stored values are not silently rewritten by a later model run. New work produces a new period, not an edit to the old one.

Restatement

A prior period can be explicitly restated under a newer methodology. Replacing or restating a baseline is a recorded, deliberate act — it is not the same as changing historical published values in place.

Comparability

Comparison between periods is only presented where the methodology versions permit it. A major methodology change breaks direct comparison until the earlier period is restated, and a baseline-to-quarter comparison is labelled as such rather than shown as a clean quarter-over-quarter movement.

• Forex and Futures are modelled and calibrated separately, because payout structure and account economics differ materially between them.
• Crypto-focused, multi-asset and equities-focused firms are outside current public estimate scope. No crypto estimate is published.
• Uncovered segments are not imputed, grossed up or extrapolated. An absent segment is absent from the total, not modelled into it.
• Industry revenue is not currently published as an estimate; the available inputs do not yet support it.
• Published totals are therefore totals for the PTC modelled universe, not for the entire prop trading industry.

• Estimates are modelled figures, not audited industry totals and not a census of the market.
• The modelled universe is smaller than the tracked universe, and both are smaller than the industry.
• Evidence quality varies by firm; some estimates rest on firm disclosure and others almost entirely on calibrated relationships.
• Published ranges express model uncertainty, not statistical sampling error.
• Individual firm estimates may be withheld while still contributing to aggregate totals.
• Uncovered segments are not automatically imputed, so aggregates understate the full market.
• Estimates must not be read as company-reported figures unless explicitly identified as such, and are not investment advice.

Related governance: the Confidence Framework, Data Sources and the Update Schedule. PTC EstimateModelled, not reported