Most broker comparison sites do not tell you where their facts came from. This page does, including the parts that are unflattering to us.

It is worth reading for one reason above all: it tells you what we cannot tell you, and why. A site that only lists what it knows is easy to write and impossible to check.

On this page

What We Publish, and What We Will Not Claim

We separate two things that most of this category runs together.

A verified fact is something we read in a primary source, with the source and the date recorded. That a broker holds a specific licence number on a named regulator’s register is a fact. So is a clause in its client agreement, a figure on its own pricing page, or the text of a country’s foreign exchange regulations.

An opinion is our reasoning on top of those facts. That a broker is “safe”, or “a good choice for a first account”, or rated 4.2 — those are judgements. They are ours, they are argued, and they can be wrong.

We write them differently on purpose. A fact is stated flatly with its basis attached: this is the case, because this instrument says so, verified on this date. An opinion is framed as a conclusion and carries its reasoning where you can disagree with it.

Warning

What we will never write: that your money is guaranteed, that a broker ensures anything, or that you will not lose. No broker can promise that and no comparison site can promise it on their behalf. If you see that phrasing anywhere on this site, it is an error and we want to know.

brokersafrica.com

Where Our Facts Come From

Primary sources, read directly. The instrument itself — the Act, the regulation, the licensee register, the client agreement, the broker’s own published pricing page. Not a summary, not another comparison site, not an AI answer.

We go to the instrument, not to the search box. A regulator’s site search is not the regulator’s corpus. If a statute has to be read to answer a question, it gets read — and when we report that a term does not appear in a law, that means the text was searched, not that a search engine returned nothing.

Two authorities per country, not one. Almost every market in Africa splits this activity between a securities regulator and a central bank, and they answer different questions. The securities regulator decides whether a broker may be licensed. The central bank decides whether your money may leave the country at all — which, in several of our markets, turns out to be the rule that actually affects you. Checking only the first produces a confident article that misses the binding constraint.

Everything carries a date. Rules change, and a correct figure from fourteen months ago is a wrong figure today. Every verified fact in our records has a source URL and a verification date, and the articles are re-checked against the records when a record moves.

What We Do When a Fact Is Not Available

This is the part most methodology pages skip, and it is where the real differences are.

We do not guess, and we do not fill gaps from memory. If a figure is not in our record with a source, it does not go on the page — even where the omission is awkward and a plausible number would read better.

We do not defeat a login wall, a CAPTCHA or bot detection to get a fact. Several brokers have moved their deposit limits, withdrawal fees and processing times behind an account login. That is their decision. Ours is that we report the withholding rather than work around it — because a site willing to circumvent access controls for a number has a different relationship with rules than we want to have.

We treat “we could not reach it” and “it is not published” as different findings, and we say which. A page we could not load from where we were researching tells you about us. A figure a broker has chosen not to publish tells you about the broker. Collapsing the two would let our own limitations read as evidence against a company.

We do not publish from archived snapshots of pages that no longer exist. They are useful as leads for finding a live source. They are not a source.

Tip

And here is the one we would most like you to hold us to. Where a fact matters enough that you would act on it, and we cannot establish it, we say so on the page rather than quietly leaving it out. Where it is minor and unestablished, we leave it out without ceremony, because a paragraph about a trivial absence costs you attention the page needs elsewhere.

The distinction is deliberate, and the way it could fail is worth naming: the temptation would be to classify an inconvenient fact as a minor one. The test we hold ourselves to is not how hard the fact was to get. It is whether a reader would act on it.

How We Read an Absence of Evidence

A broker’s absence from a regulator’s list means completely different things in different countries, and almost nobody accounts for this.

Some regulators publish a blacklist — a named, dated list of firms operating without authorisation. Ghana’s SEC names its entities. Where such a list exists, a broker’s absence from it is real information: a regulator that names offenders and has not named this one has told you something.

Other regulators publish nothing of the kind. Kenya, Tanzania and Uganda have no standing warning list for this activity at all. There, a broker’s absence from the list is not reassurance — there is no list to be absent from. A comparison site that writes “not on any warning list” in both cases is saying something true in one country and meaningless in the other.

We record which shape each market has, and we write the sentence that is true for that market.

The same applies to enforcement. “No enforcement action found” means none has been published in a form we can read. It does not mean none happened.

Which Company Is Actually Serving You

A licence belongs to a named legal entity, and a large broker is usually a group of separate companies. The name on the website is a brand. The company on your client agreement is who actually holds the other side of your account — and those are often not the same, particularly for clients outside Europe.

So the regulatory question we ask is never “is this broker regulated”. It is: which entity will contract with you, in your country, and what does that entity’s regulator actually cover? A group-level licence that does not extend to the company taking your deposit is the most misleading thing a comparison site can repeat, and it is repeated constantly.

There is a second layer we check where it applies. In some markets the local licence is an agency licence, meaning the locally regulated company is not the counterparty to your trades at all. The licence is real; the protection it implies is not what most readers assume.

Where we know this, we say it, and we say which entity the answer depends on.

What Counts as a Finding

A fact is not a finding because it exists. It is a finding because it differs from what everyone else does.

Every retail broker’s contract contains clauses that sound alarming read cold: that prices may move, that trades may be voided on manifest error, that accounts may be closed, that dormant accounts may be charged. These are in essentially every contract in the industry, including those of brokers we rate highly. Quoting them as though we had uncovered something proves diligence and tells you nothing.

Our editorial guidelines state this as a rule. A clause reaches our pages as a criticism only when it departs from what the rest of the industry does — an unusual discretion, a missing time limit, a condition nobody else imposes. Then we say so, and we say what the normal version looks like.

The same applies to complaints, and this is where the category is at its worst. Every broker with a large client base generates withdrawal complaints. At scale, they are statistically certain. A complaint count without a denominator is not evidence, and we do not publish one.

What is worth reporting is a pattern that differs: not “some users report slow withdrawals”, which is true of everyone, but a specific, repeated behaviour that is not the industry norm. Where we can establish the pattern but not its base rate, we report what we saw and draw no conclusion from it.

Warning

The single test behind this whole section, and we apply it in both directions:

Would we write this same sentence about our highest-rated broker, if the fact were identical? If not, we are not reporting — we are building a case.

And the reverse: would we leave this out of a review of a broker we liked? If a flaw is worth mentioning in a critical review, it goes in a favourable one too. Selective generosity and selective scrutiny are the same decision.

brokersafrica.com

Where Industry Facts Belong

How client money is held, what a licence category permits, what a regulator does and does not do — these are true, important, and not facts about any particular broker. They belong in our guides, and that is where we put them.

A review is for what distinguishes that broker. If a sentence could be written identically about every firm in the category, it is not a review sentence, however true it is.

How Our Ratings Work

This section is deliberately incomplete while we rebuild our rating model, and we would rather say that than describe a method we are in the middle of changing. It will be completed before any rating appears on this site.

What we can state now, because it will not change: a rating is an opinion built on verified facts, never a fact itself. Every rating on this site will show what it is made of, and any score we publish will be accompanied by what it measures — and, where it matters, by what it does not.

What Our Testing Consists Of

Reading a broker’s documents tells you what it promises. It does not tell you what happens when you try to use it, so we do that part ourselves.

We open an account with every broker we cover and work through registration and verification as a reader in that country would. We record which identity documents are actually demanded, how long verification takes, and where the process stops and asks for something else.

We go as far as the deposit screen in each country we cover, and record what is really there. This matters more than it sounds: a broker’s payment page often lists methods that do not appear once you are inside, and the country you registered from changes what you are offered. What the deposit screen shows is a fact about that market, not a marketing claim.

We use the platform on the kind of phone our readers have — a mid-range Android on mobile data, not a desktop on fibre. We record what the demo account shows about spreads, order types, the instruments actually available to that account, and whether the thing is usable with one hand on a slow connection.

We read every document the broker publishes, not the page that summarises them: the client agreement in full, the fee schedule, the bonus terms and the parts written in smaller type than the rest. Most of what we find that nobody else prints comes from there.

What we publish about deposits, withdrawals, costs and platforms comes from those two places — the account we opened, and the documents themselves. Where a figure exists in neither, we say so rather than fill the gap.

How We Make Money, and What It Cannot Buy

We earn commission from some — not all — of the brokers on this site, through affiliate arrangements. The disclosures page sets them out in full. If you open an account through one of our links, the broker may pay us. It costs you nothing and changes nothing about your account.

Here is the part that matters. The people and processes that produce our verdicts do not have access to the commercial terms. Our editorial records contain no commission rates, no payout models and no commercial priorities — that information is kept deliberately separate, so that a rating cannot be influenced by a number nobody writing has seen.

We cover brokers we cannot earn anything from, and we rate them by the same standard. A broker that pays us will be ranked below a better one that does not, and the reason will be stated in the paying broker’s own review.

That is either true or the whole site is worthless, so we have built the separation into how we work rather than asking you to take it on trust.

When We Get It Wrong

We will. Rules change, brokers change their terms, and we make mistakes of our own.

Our records are re-checked on a schedule, more often for markets where the rules are moving. When a source changes, every article standing on it is flagged for review rather than left to age quietly.

When we correct something substantive, we correct it visibly. A page that silently loses a claim teaches nobody, including us.

If you find something on this site that is wrong, out of date, or that we have stated more confidently than the evidence supports, tell us. A specific correction with a source is the most useful thing anyone can send us.