Advertiser Disclosure: we may earn commission

Our Review Methodology

We follow a rigorous, evidence-based process for our editorial reviews, publish the exact EngineForex Trust Score formula behind every review, and use a transparent scoring formula for card-tier broker profiles. All three are documented below in full.

Step 1

Multi-Market Investigation

We read each broker's live websites from several countries to see which legal entity serves which region, then map every entity, regulator, and licence number directly from official sources.

Step 2

Register Verification & Hands-On Checks

We confirm every claimed licence against the regulator's own public register and record exactly what we find. Where we hold live accounts with a broker, we also test funding, withdrawals, and support first-hand.

Step 3

Scoring & Rating

Verified data feeds our deterministic EngineForex Trust Score, while our analysts grade the broker across seven editorial pillars: trust, fees, platforms, research, education, support, and range of markets.

Step 4

Fact-Check & Compliance Review

Every factual claim is checked back against its original source, and the full draft passes a compliance review before it goes live, with no unverified numbers and no vague safety claims.

Step 5

Ongoing Updates

Reviews are re-checked and re-scored as regulation, Trustpilot data, and a broker's offering change, so the rating you see reflects the broker today.

EngineForex Trust Score

Every review-tier broker carries an EngineForex Trust Score, the 0-100 number shown at the top of its Trust & Regulation section. It's computed deterministically from verified, public facts: the same inputs always produce the same score, with no opinion mixed in. Here's exactly how the 100 points are weighted:

ComponentPointsHow it's scored
Regulator tier coverage40The single strongest trust signal. One confirmed tier-1 licence (FCA, ASIC, CFTC, FINMA, MAS, JFSA) saturates this axis; tier-2 (CySEC, BaFin, FSCA, etc.) scores lower; offshore-only lower still. Only licences we confirm on the official register count.
Number of regulators15Breadth of regulatory footprint, counting each distinct regulator once. Plateaus at four or more, where extra licences add diminishing signal.
Years of operation15Institutional stability. 10+ years scores fully; a brand-new broker scores lowest. Taken from the founding year on the register or corporate filings.
Trustpilot rating20The broker’s live public Trustpilot rating when available. Brokers with no Trustpilot data score zero on this axis rather than being assumed good.
Trustpilot review volume10How much confidence to place in that rating, based on the number of reviews behind it.

Grade bands

A 85+ · B 70-84 · C 55-69 · D 40-54 · F below 40.

What does and doesn't count

Only verified data earns points. A licence a broker claims but that we can't confirm on the official register earns no tier or count credit. EEA passport notifications and representative offices are excluded entirely, since they aren't standalone trading licences, so they never inflate the regulator count. The score recomputes whenever the underlying facts change, so it always reflects the broker's current standing.

Card-tier Provisional Score

Brokers in our directory fall into two tiers. Review-tier brokers have been opened, tested, and scored by our analyst team across seven editorial criteria, the rating you see on /reviews/[broker]. Card-tier brokers haven't been editorially reviewed yet; they're shown with an algorithmic Provisional Score labeled PROV wherever they appear, so you can never confuse the two.

The Provisional Score is a deterministic 0-10 number computed from public, structural facts about the broker, with no editorial judgment, no opinion, no interpretation. The formula is reproducible: same inputs always yield the same score. Here's exactly how it's weighted:

ComponentWeightHow it's scored
Regulation40%Tier-1 regulators (FCA, ASIC, CFTC, FINMA, SFC, MAS) count fully; tier-2 (CySEC, BaFin, CONSOB, CNMV, AMF, FSA, FMA, JFSA, FSCA) at 60%; everything else at 30%. Two tier-1 licenses saturate this axis. We never reward offshore-only registration.
Trustpilot20%User rating, weighted by review-count confidence (saturates at 500 reviews). Brokers with no Trustpilot data default to a neutral 0.5 contribution, with no penalty for being absent.
Longevity15%Years in business, log-shaped saturation at 20 years. A 20-year-old broker scores fully on this axis; a 1-year-old scores about 24%.
Minimum deposit10%Lower is better, an accessibility metric. $0–$10 is perfect, $1,000+ is zero, linear in between. Unknown defaults to neutral.
Platforms10%Number of supported trading platforms (MT4, MT5, cTrader, proprietary, etc.). Three or more saturates the axis.
Polish5%Whether we have a clean square brand icon. A small signal that the broker is a real business worth looking at, not a faceless skin.

What we deliberately do NOT score

  • Spreads, fees, swaps. Advertised spreads aren't live spreads, and judging cost competitively requires real trading. That's analyst work, and it lives on review-tier brokers only.
  • Platform quality, execution speed, charting depth. Listing which platforms a broker offers is fact; grading them requires using them.
  • Research, education, customer support quality. Presence is detectable; quality isn't. We'd rather show no opinion than a fake one.
  • Leverage. A neutral signal, since some markets prefer low leverage (stricter regulator) and others prefer high. Rewarding either direction would bias the formula.

Output range

The weighted sum of the six components above is mapped to 0-10 with a 4.0 floor, so visually empty cards never show 0, and the practical ceiling is 10.0. Scores round to one decimal.

Provisional Scores recompute on every save of the broker document, so any change in regulators, founded year, or Trustpilot data immediately reflects in the published number. The formula source (in TypeScript) is auditable in the open-source EF repo.