How the Framework Is Built
This page explains the structure behind every score in full: the criteria, the weighting, how a score is calculated, where the data comes from, and how often it's updated. The internal process an individual rating goes through is our product's methodology and stays undisclosed — everything that determines how the framework itself is built and checked is below.
The Evaluation Framework
Every broker in a live market is scored against 54 fixed criteria, grouped into 8 categories covering regulation, costs, trading conditions, platforms, investor protection, and customer service. Each criterion has a public, written definition that applies identically to every broker — the same question is asked of all of them, every time.
Each criterion is scored on a 1–5 scale. Criterion scores roll up into a category subscore (the average of every criterion in that category), and category subscores roll up into one overall score out of 100, using the weighting below. That aggregation is arithmetic, and it's the same formula for every broker — it's disclosed here in full.
Category Weights
Sample Calculation
A simplified, illustrative walk-through of the arithmetic — not real data for a specific broker.
Every category subscore (Trust, Regulation, Costs, Platform, Security, Transparency, Products, Customer Support) is multiplied by its weight from the chart above and summed — the same formula shown there, applied to all 8 categories, produces the single overall score you see on a broker's page.
Source Traceability
Every individual rating links back to where it came from — the broker's own pricing page, a regulator's public register or filing, or an independently verifiable comparison source. Nothing is scored from an unlabelled or private source, and nothing is scored from the broker's own marketing claims alone.
Update Cycle
Every broker in a live market is re-verified on a weekly cycle against current public sources. A change on a broker's side — a fee adjustment, a new regulatory filing, a platform update — is picked up at the next weekly run, not the moment it happens. The date each rating was last verified is shown alongside it, so you always know exactly how current it is.
Limitations
What this framework deliberately does not claim to be.