Editorial Policy and Guidelines

This page covers how content gets produced at CompareForexBrokers: who writes it, how it’s checked, and how the commercial side of the business stays separate from editorial decisions.

Editorial principles

Independence

The editorial team doesn’t know which brokers have commercial agreements with CompareForexBrokers. Writers, researchers, and editors aren’t told what affiliate agreements pay, or which brokers we have one with. Management handles commercial relationships. Editorial handles content. The two sides don’t share files.

A commercial relationship doesn’t entitle a broker to editorial influence. Where a placement in a comparison table is commercially influenced (featured positions, the visibility of “Go to Site” buttons) we label it. None of those placements feed into a broker’s score, rating, or review content.

When a broker asks us to remove a finding from their review, we explain how we reached the finding and what evidence would change it. We don’t remove it.

Accuracy and data: How we tested brokers in 2026

All data on the site comes from primary sources: live funded accounts at each broker, official regulatory registers, and broker documentation. Nothing comes from broker marketing or third-party aggregators. Every data point is verified before publication.

Accuracy and data image

Broker scores are built from over 100 data points per broker. Our technical testing covers 67 brokers across spreads, execution speeds, commissions, and platform behaviour. If we can’t measure it, we don’t score it. Reputation and how long a broker has been a commercial partner aren’t inputs.

We also have a model that gives eight key trading factors a weighting (defined on our Methodology page), which then feeds into an overall score used to formulate our broker comparisons. The following screenshot gives a snapshot of the weighting we give to each trading factor when reviewing brokers.

Trading CostsTrading ExperienceTrustTrading PlatformsCustomer ServiceRange of MarketsEducationFunding
Weighting25%15%20%15%10%5%5%5%

When broker conditions change, the page changes. When errors are identified, the page is corrected with a dated note added.

Objectivity

Scores reflect the data. If testing produces a weak result, that result goes in the review. Commercial partners get assessed against the same methodology as everyone else. No softening for relationships.

Transparency

For full transparency of information, you can check out Our Methodology to find out how we compare brokers and what data we use to test. Commercial relationships are detailed at Advertiser Disclosure. Author credentials are on every piece of content.

Accountability

If there’s a factual error on this site, we want to know about it. If a reader or broker challenges a conclusion, we explain how we got there. If a broker tries to pressure us into changing editorial content, we refuse. The corrections process below explains how to flag something.

How content gets made

Who writes

Every piece of content is written by a named team member with trading experience and a financial industry background. The full team is on the Meet the Team page. Three of the people you’ll see most often:

  • Justin Grossbard (CEO, Co-Founder). Justin has been trading since 1998 and co-founded CompareForexBrokers in 2014. He’s published over 100 finance articles in Forbes, Kiplinger, and Finance Magnates. Justin oversees research direction and editorial standards across the site.
  • David Levy (Head of Content). David has been reviewing forex brokers since 2018. His IT background helps when broker platforms behave strangely, which they often do. David runs review quality and editorial consistency.
  • Ross Collins (Chief Technology Researcher). Ross spent 35 years across SWIFT, UBS, PwC, and IBM before moving full-time into broker testing. He runs purpose-built Expert Advisors against live broker accounts, with single test windows of 20 to 30 hours so the numbers don’t depend on what was happening for an hour on a Tuesday.

Who fact-checks

Fact-checking isn’t one person’s job. Justin, David, and Ross each verify in their own area. Trading conditions and fees are checked against live accounts and broker documents. Regulatory status is checked against the official registers at ASIC, FCA, CySEC, and NFA/CFTC. Data hits a shared spreadsheet before it reaches writers, who confirm the details again as part of their own write-up.

Sources we rely on

  • Live funded accounts at every broker we test
  • Official regulatory registers (ASIC, FCA, CySEC, NFA/CFTC)
  • Broker documentation and published fee schedules
  • Live spread testing using IceFX SpreadMonitor across six major currency pairs
  • Execution speed testing using ExTest_ForExpat and Broker Latency Tester
  • CFB commission and VPS benchmarking data
  • Public financial filings where they’re relevant

log in and scanning servers

spread data review seciton

When pages get updated

Spread data is reviewed monthly. Full broker reviews are formally refreshed in January each year. If something material happens outside that cycle (regulatory action, platform change, ownership change), the affected sections are updated immediately, with a “What Changed?” note at the top of the page.

Where AI fits in (and where it doesn’t)

Every broker review, score, ranking, and editorial opinion on this site is produced by a human writer. Spread testing, execution speed testing, and commission data come from live funded accounts using tools we built for the job. None of it is AI-generated.

Our writers use AI tools for two narrow tasks:

  • Summarising long regulatory documents (ASIC, FCA, CySEC, NFA filings) to identify the sections relevant to a specific review.
  • Final grammar and clarity checks on already-written drafts.

Rest assured, AI is not used to draft reviews, score brokers, generate testing data, rank brokers in comparison tables, write opinions, or fact-check claims. Every published piece is drafted by a human writer, reviewed by a second team member, and signed off before it goes live.

The tools currently approved for the two tasks above are Claude (Anthropic) and ChatGPT (OpenAI). Any change to this list is logged and the policy page is updated.

Corrections and updates

If you spot a factual error, tell us via our Contact page. Include the page URL and a description of what’s wrong.

Confirmed factual errors are fixed straight away, with a dated correction note added to the page. Challenges to a review conclusion get a reply explaining the reasoning. We update factual content when something is genuinely wrong. We don’t change scores or conclusions because a broker is unhappy with them.

How long we’ve been doing this

CompareForexBrokers has been comparing brokers since 2014, strenuously reviewing 67 brokers worldwide across spreads, execution, regulation, and platforms.

We test more brokers, more thoroughly, than most. Live accounts, named writers, published methodology, dated corrections: that’s the standard the site is built on.

For more on how we work, see About Us and Why Trust Us. Other related pages: Terms of Service, Risk Disclaimer, and our full Methodology.