About Kumasi Learn

Who runs this desk

Kumasi Learn is run by a small editorial desk that writes and checks gold (XAU/USD) material for readers in Ghana. The desk is not a company you can open an account with, and it holds no client money. Every page is written and reviewed by the same small group, and we name no individual because no single person is the product.

We write in plain English for people funding accounts in cedis, often through MTN Mobile Money, and we keep the instrument concrete: gold, 1 lot = 100 oz, one pip = 0.01, with a reference price around 4275.0 that is a dated marker, not a live quote.

How the desk is paid, and what payment cannot buy

Affiliate links are the only revenue this desk has. If you open an account with a broker through a link on this site, the broker may pay us a referral fee, and that costs you nothing extra. We take no payment from readers and sell no subscriptions.

Payment never buys a rating and never hides a weakness. A broker cannot pay to move up a ranking, to remove a paragraph about its funding route or its costs, or to have a caveat softened. If a broker's terms are awkward for a Ghanaian reader, we say so in the same words whether or not we earn a referral.

What we are not

Kumasi Learn is not a broker, not a bank and not a licensed financial adviser. We cannot open an account for you, cannot take a deposit, cannot place a trade and cannot recover money for you. Anyone claiming to act for this desk in those ways is not acting for us.

The brokers we write about are separate businesses with their own terms. FxPro is licensed by the FCA (UK), CySEC and FSCA; confirm any broker on the SEC Ghana licensed operators list. That check is yours to make before you send anyone money.

Why the Ghana entity question decides which numbers appear on a page

The entity that serves Ghana decides which numbers are relevant, which is why the desk asks the question before it writes a cost or a leverage sentence. A broker can operate through more than one licensed entity, and the cap, the account types and the documents a reader signs can differ by entity. The desk does not assume that a figure published for one entity applies to a Ghana account. It states that the cap which applies is the one attached to the entity serving this country, and it leaves the specific ratio to the broker's own account documents rather than printing a number it cannot tie to that entity.

This is also why the regulator caveat is repeated where it matters rather than buried on one page. FxPro is licensed by the FCA (UK), CySEC and FSCA, and that sentence is printed alongside the instruction to confirm any broker on the SEC Ghana licensed operators list. The two halves belong together: the first tells you where the firm holds licences, the second tells you where to check the local position. A reader who skips the second half has half the picture, and on a high-risk instrument half the picture is how accounts get opened with the wrong expectations.

The practical consequence for the desk is that some pages are shorter than they could be. Where the entity question cannot be resolved from the source list, the page describes the mechanism and stops: margin scales with position size, a 0.10-lot gold position needs about $85.50 margin at the applicable cap, and the cap itself is a ceiling. It does not fill the gap with a ratio borrowed from another country's entity. Readers who want the exact figure for their own account should take it from the account documents they are asked to sign, not from a page that cannot see those documents.

The correction log, the review cycle, and how an old page is kept honest

Every page carries an internal correction log, and the log is the reason an old page can still be trusted. When a sentence is changed, the log records the date, the old wording, the new wording and the challenge or source that caused it. The public page shows the current wording; the log shows that the desk knows it changed its mind. Pages are reviewed on a cycle rather than left to drift, and a page that names a broker's regulator or leverage treatment is reviewed sooner than a page that explains a formula, because the first kind goes stale faster.

The review is not a rewrite. The reviewer checks each number against the source list, checks that no adjective has crept in where a description of cost belongs, and checks that the risk framing still matches the instrument. For gold that means confirming the contract size of 100 oz and the pip of 0.01 are still stated wherever a position size is discussed, and that the reference price is still labelled a reference. A page that passes keeps its date. A page that fails is corrected before it is left up, not after.

The cycle exists because the alternative is a site that looks maintained and is not. A reader in Ghana funding through MTN Mobile Money does not need the desk to be fast; they need the page to be right when they read it. That is why the desk would rather publish six careful sections than sixty quick ones, and why a challenge that arrives months after publication is treated the same as one that arrives on the day. The log is the record that this happened, and it is kept whether or not anyone asks to see it.

How to challenge a published figure, and what happens to your challenge

To challenge something, send the page heading, the exact sentence, and the source you believe is correct. The desk accepts challenges through its contact route and does not require an account, a name or a reason beyond the correction. A challenge that says only that a number feels wrong cannot be acted on, because the desk has nothing to check it against. A challenge that names a regulator notice, a contract specification or a broker's own document can be acted on within the week, because it gives the reviewer something to compare with the source list.

Every challenge gets one of three outcomes, and the desk tells you which. If the source holds, the page stays and you are told why. If the source does not hold, the sentence is corrected or removed and the change is noted on the internal version with the date. If the challenge raises a question the desk cannot settle, the sentence is rewritten to say what the figure depends on rather than left as a claim. The third outcome is the most common, and it is not a fudge: it is the honest form of a statement the desk cannot verify to the standard it set.

Challenges about costs and leverage get particular attention because those are where invented numbers do the most damage. If you believe a spread, a swap or a minimum deposit has been implied rather than stated, say so; the desk will check whether an adjective slipped past the reviewer. The same applies to leverage: if any page appears to recommend the cap rather than describe it, that is a fault and the desk wants it flagged. Challenges are logged even when they are rejected, so a pattern of the same objection is treated as a signal that the page is unclear, not that the readers are wrong.

The two-person Kumasi desk and the weekly rule that settles disagreements

The desk is run by two people: a markets writer who drafts every explainer and a reviewer who checks each number against the source list before publication. Neither can publish alone. The writer produces the draft, the reviewer signs off, and the reviewer's name is recorded on the internal version of the page. This is why you will not find an author byline on the public page: accountability sits with the desk, not with a personality. Both work from Kumasi, and the desk does not employ sales staff, affiliate managers or anyone paid per account opened.

Decisions about what to cover are made once a week, on a fixed call, against a written list of reader questions. The writer brings the questions that arrived through the contact route; the reviewer brings the gaps found in existing pages. A topic enters the plan only if it can be answered with facts the desk can cite. Anything that would require inventing a spread, a swap or a minimum deposit is rejected at that meeting rather than softened later. The call is short by design, and the plan is frozen for the week so drafts are not rewritten to chase traffic.

Disagreements are settled by a single rule: the stricter reading wins until a source settles it. If the writer believes a cost can be described and the reviewer cannot find the number in the source list, the sentence is removed, not hedged. If both agree a figure exists but disagree on how it applies to a Ghana account, the page says what the figure depends on and names the entity question. This rule has cost the desk paragraphs it wanted to keep. It is still the rule, because a wrong number on a high-risk instrument is worse than a missing one.

How the desk is funded and the three things funding is not allowed to touch

The site is funded by the desk's own resources and by nothing that depends on a reader opening an account. There is no per-signup payment, no revenue share on spread and no payment for a mention of FxPro or any other firm. When a page names a broker it is because the page is answering a question about that broker's platform, entity or regulator, and the naming is part of the answer. The funding model is deliberately boring: the desk would rather write less than owe a sentence to a payer.

Funding is not allowed to touch three things. It cannot decide which instrument gets covered, so gold is covered because readers in Ghana ask about it, not because it pays. It cannot decide what a page says about risk, so the warnings about leverage and margin stay in the same place regardless of who is named. And it cannot buy the removal of a caveat, including the line that FxPro is licensed by the FCA (UK), CySEC and FSCA and that any broker should be confirmed on the SEC Ghana licensed operators list.

What funding does influence is volume. A self-funded desk publishes fewer pages than an advertising-funded one, and it updates them more slowly. That is the trade the desk has chosen. If a page looks thin on a topic, the likely reason is that the desk could not source enough to write it properly, not that a payer asked for silence. Readers who want more coverage on a specific question should send it through the challenge route; the weekly plan is set from those questions, and a well-argued request carries more weight than a request for a particular broker to be praised.

What counts as a number we will print, and the three tests it must pass

A number is printed only if it passes three tests: it appears in the desk's written source list, it is stated in the same unit a reader would use, and it can be traced to the instrument as traded. For gold XAU/USD that means the contract size of 100 oz per standard lot and the pip value of 0.01 are printed because they are fixed by the contract, not because they were observed on a screen. A reference price near 4275.0 is labelled a reference, never a quote, because it moves and the page cannot move with it.

Costs are described by what they consist of, never by an adjective. The desk will say a gold position carries a spread, a commission where the account type applies one, and a swap if held overnight, and it will say each of those depends on the account and the entity serving Ghana. It will not call any of them low, tight or competitive, because the desk was not given a number to compare and will not invent one. When a reader asks which account is cheapest, the honest answer is that the desk cannot rank what it has not measured.

Leverage is handled the same way. The page states that a cap applies to the entity serving Ghana and that the cap is a ceiling, not a target. The one worked figure the desk prints is that at the applicable cap a 0.10-lot gold position needs about $85.50 margin. That figure is there to show how margin scales with size, not to suggest anyone should use the cap. A number that cannot be shown to a reader with its unit, its instrument and its source does not go on the page, however useful it would be.