Research standards

Research Methodology

There is no single test that can tell you which broker is right for everyone. Our job is to make each question narrow, each comparison fair, and each limitation visible.

Last updated: Version: Framework 1.0

1. Start with a specific research question

Research begins by defining what the page is trying to answer. “Which broker quoted the lowest spread during this test window?” is measurable. “Which broker is best for everyone?” is not.

The question determines the evidence, comparison set, exclusions, calculation, and limits. If the scope changes, the method must change with it.

2. Use the strongest suitable evidence

RinTrade separates evidence used to verify a broker fact from material used only to discover a topic or understand context. A source is judged by how directly it supports the claim, not by how confidently it is written.

Regulator and legal records
Preferred for licences, regulated entities, formal permissions, and public enforcement information.
Official broker material
Used for account terms, platform support, fee schedules, product details, and operational information, with the applicable entity and checked date attached where relevant.
First-hand observations
Used for measured platform conditions and workflow tests, with the account, server, platform, time window, and data-processing method documented.
Independent secondary material
Useful for context and discovery, but not automatically promoted into a verified broker fact or RinTrade score.

3. Make uncertainty visible

Reusable broker fields move through three statuses. The status describes the evidence state of the field, not the reputation of the broker.

needs_source
No verified source-backed value is ready for public use.
partial
Some research exists, but the evidence is not complete enough for a final display or score.
verified
The value has source IDs, a checked date, and has passed the applicable validation rules. It still needs rechecking when the underlying fact changes.

4. Build reviews from reusable evidence

A broker review is assembled from central evidence rather than a separate set of facts written only for that page. Regulation, account conditions, trading costs, platforms, funding, and country context can each have different evidence and freshness requirements.

Analysis and editorial conclusions are written around those records. If a key field remains unresolved, the review should explain the gap instead of smoothing it over.

5. Keep hands-on tests reproducible

A test result is useful only when readers can understand what was tested. RinTrade records the account category, platform, server or environment, instruments, timestamps, sample interval, cleaning decisions, and calculation used for the result.

Where publication is safe and practical, the cleaned observations, methodology, and checksums are made available so the result can be inspected. Raw private account details and internal system paths are excluded.

6. Rank only what the evidence can support

A data-study ranking follows its published calculation. The current spread study, for example, ranks a defined set of Standard and Standard-equivalent accounts using synchronized quoted-spread observations and an equal-weight average across seven major currency pairs.

Composite broker scores are different. RinTrade will not publish an overall, trust, fee, spread, or swap score until the page-specific weights, required inputs, missing-data rules, freshness gates, and tie handling are complete and every displayed input is verified.

Comparisons reuse the same verified broker fields. They do not create a new fact merely because two brokers appear side by side.

7. Recheck when the reason to trust a fact changes

Checked dates are part of the evidence. High-change fields need a new review when a source changes, a scheduled check becomes due, a broker updates its terms, a reader identifies credible contrary evidence, or the tested setup no longer matches the public description.

A correction updates the affected page and its structured source where applicable. Material changes should remain understandable from the page date or an explicit correction note rather than disappearing silently.

8. Keep compensation outside the calculation

Affiliate availability can affect whether RinTrade offers an account-opening link. It cannot determine which brokers enter a measurement set, change a recorded observation, override an exclusion rule, or supply a missing score.

Relevant pages disclose the commercial relationship near the recommendation or link. The fuller editorial rules explain how affiliate and sponsored material is separated from the research conclusion.

9. State what the method cannot show

Broker conditions vary by time, account, platform, entity, country, market conditions, and client classification. A quoted-spread study is not an execution test; a licence record is not a guarantee against loss; and a global comparison set does not prove that every broker accepts clients everywhere.

RinTrade therefore treats every result as an answer to a bounded question, not as a permanent promise about a broker.

10. Version meaningful method changes

When a method change could alter who is included, how a metric is calculated, or how a conclusion is interpreted, RinTrade updates the methodology date and explains the material change on the affected study or policy page.