Foundation · Module F7
You can understand gold completely and still lose money. The structure you trade through matters as much as the analysis you bring to it.
What you'll learn
From the market to the mechanism
The first six modules were about the market. What gold is, how it's priced, what moves it, and when. That knowledge is necessary. It is not sufficient. You can read the four forces correctly, time the calendar perfectly, and still end the year down — because between your analysis and your account sits a structure that has its own economics, its own incentives, and its own way of taking a cut of every move you make.
This module is the uncomfortable one. Not because retail trading is a scam — it mostly isn't — but because the structure is genuinely stacked in ways most beginners never have explained to them. The goal here isn't to scare you off. It's to make the invisible parts visible, so that when you choose how to trade gold in the final two modules, you're choosing with your eyes open.
Everything that follows is sourced from regulators, published broker disclosures, and academic research. Where a number comes from, it's named. That's the standard this site is built on.
The thesis in one frame
Every trade you place travels through a structure before it reaches the market — if it reaches the market at all. The two paths gold traders take, spot and futures, handle that journey in fundamentally different ways. This is the whole module in a single diagram.
Same gold. Different structure. On the spot side, your counterparty may be the broker itself. On the futures side, the exchange matches you against the rest of the market. Neither path is free — but only one has a built-in conflict over whether you win.
The disclosed reality
Start with the statistic the industry is required to publish about itself. In the EU and UK, regulators force CFD brokers to display, on their own marketing, the share of their retail accounts that lose money. You've seen the line — it's the small print on every broker advert: a figure, then "of retail investor accounts lose money when trading CFDs with this provider."
Across the European market, the regulator ESMA found that between 74% and 89% of retail CFD accounts lose money, with average losses running from the low thousands into the tens of thousands of euros per client. The exact number varies by broker and by who's measuring — Interactive Brokers has cited around 69%, the UK's FCA has put the figure near 80%, and the 89% number appears on UK warnings often enough to have become the canonical example. Independent surveys of brokers land in the same band, roughly 62% to 82%.
Sit with the framing for a moment. This isn't a hostile statistic dug up by critics. It's the broker's own disclosure, mandated by its regulator, printed on its own page. The most generous reading available — the industry's own number — is that the large majority of retail accounts lose. Any honest account of retail trading has to start there.
Why this matters
Notice what that loss rate does not depend on. It's roughly the same whether the broker A-books or B-books, whether you trade spot or futures, whichever jurisdiction you're in. If the broker's conflict of interest were the real engine of losing, A-book traders would win — and they don't. So the broker isn't what beats most people. What's constant across every path is the game itself: leverage, frequent trading, and certain costs working against uncertain gains. That difficulty is inherent to the instrument, not inflicted by a counterparty — which is also why blaming the broker, the spread, or "the system" is the tell of a trader who hasn't yet looked at the real opponent. The hostile baseline is the same for everyone. What separates the minority who succeed isn't a better broker; it's the risk management, discipline, and edge to beat a game that's stacked against all participants equally.
Spot economics · Path A
When you trade XAU/USD through an FX or CFD broker, you are not buying gold and you are not trading on a central exchange. You're entering a contract with the broker that tracks the gold price. What happens to that contract behind the scenes is decided by the broker's business model. There are two pure models and one hybrid.
The A-book (agency) model. The broker passes your order through to its own liquidity sources and acts as an intermediary. For a retail broker that liquidity rarely comes straight from a tier-1 bank — the chain usually runs through a prime-of-prime provider, which in turn aggregates pricing from prime brokers and the tier-1 banks at the top. The broker makes money the same way regardless of whether you win or lose: from the spread it adds, or a commission on volume. Your outcome and the broker's revenue are not in conflict. If you're profitable, the broker is fine with it; it still got paid.
The B-book (market-maker) model. The broker keeps your trade in-house and takes the other side itself. It does not hedge your position with anyone. Now the relationship is inverted: when you lose, the money you lose is the broker's gain. When you win, the broker pays out of its own pocket. The broker's profit and loss is directly, mechanically opposed to yours.
The hybrid model. In practice almost every retail broker runs a hybrid — and this is the part that matters. Brokers profile their clients. The flow they expect to lose is internalised (B-booked), because losing flow is the most profitable thing a broker can hold. The smaller pool of clients who prove consistently profitable gets passed through to the real market (A-booked), so the broker isn't on the hook for their winnings. The decision of which book you're in is made by the broker, about you, using your trading history — and you're never formally told the answer, though execution quality leaves clues: clean, fast fills versus relentless slippage and requotes is often the tell of which book you're sitting in.
None of this is necessarily wrongdoing. A hybrid book is a legitimate, regulated business model, and internalising flow is how brokers manage risk without hedging every ticket externally. The problem isn't that B-booking exists. The problem is the incentive it creates: when most clients lose, and losing clients are the most profitable to keep in-house, the broker's commercial interest and your trading success point in opposite directions. That conflict is structural. It exists whether or not any individual broker ever acts on it.
The information gap
Here's the part that turns an abstract conflict into a real one: you cannot fully verify, from the outside, how your flow is handled. The labels brokers market themselves with describe intent, not a guarantee — and they're easy to wear.
NDD (No Dealing Desk) is the umbrella term: it means the broker claims not to route your orders through a desk that can intervene in your fills. Underneath that umbrella sit two execution styles.
ECN pricing. There's no such thing as a true retail "ECN broker" — what's actually on offer is ECN pricing. The broker receives raw, market-depth pricing from its liquidity sources and passes those quotes straight to your terminal: tight variable spreads, real depth, paid for with a visible commission rather than a marked-up spread. Crucially, ECN describes the price feed, not what happens to your order behind it. A broker can show you flawless ECN pricing and still B-book the execution — you see institutional-quality quotes while the broker quietly takes the other side. Good pricing is not proof of clean routing.
STP (Straight-Through Processing). Your order is passed to one or more liquidity providers, but which provider, at what markup, and with what fill quality is far less transparent. STP execution is more exposed to slippage and to a synthetic markup layered on top of the raw price — or some combination of both. It can be perfectly fair. It can also quietly cost you on every fill. From your side of the screen, clean and costly look almost identical.
And here's the catch that ties it all together: a broker can advertise "NDD," "STP," or "ECN pricing" and still run a B-book behind the label. The marketing word describes the quotes or the routing intention; it is not proof of how your specific trades are handled. To actually know, you'd need to see the broker's execution policy and order-routing records — documents the average retail trader never sees and couldn't fully audit if they did.
Consider what that means in practice. Genuine, fully A-book retail brokers — ones that pass all client flow to the market and never take the other side — are vanishingly rare. They're a low-margin, hard business to run, which is why almost no broker is purely A-book. Try to name one you could verify with certainty, and the difficulty of the exercise is the point. Even brokers who build their entire brand around execution transparency are hard to confirm from the outside. This isn't an accusation against any particular firm. It's a statement about what you, the retail trader, are actually able to know: less than you'd like.
The losing mindset
Everything above is real — and it's also the raw material for the most expensive habit in retail trading: blaming everything except your own trading. The reflexes are familiar. The broker slipped me on purpose. Market makers hunted my stop. The whole system is rigged against the little guy. Sometimes there's truth in them. But notice the pattern — every one of them puts the cause outside your control, which conveniently means there's nothing for you to fix. The serious trader asks the opposite questions. Was my stop sitting on an obvious level everyone could see? Was I holding through a tier-one release with no edge? Was my size too big for my account? The losing trader explains why the loss wasn't their fault. The improving trader assumes it was — because that's the only version of the story they can actually do something about.
Futures structure · Path B
Gold futures resolve the conflict from the other direction — not by trusting the broker, but by removing the broker as counterparty entirely. On the CME's COMEX exchange, every order is sent to a central order book that matches buyers and sellers by price and time of arrival. Your futures broker is a conduit to that exchange, not the party on the other side of your trade. There is no A-book/B-book decision to make, because the broker structurally cannot internalise your flow. The conflict simply doesn't exist in the same form.
The contract most retail traders use is Micro Gold (MGC). Each MGC contract represents 10 troy ounces of gold — one-tenth the size of the standard GC contract — and it moves in ticks of $0.10 per ounce, worth $1.00 per contract. Settlement and pricing run through the same regulated exchange institutions as the rest of the futures market. (A nice detail for this site: COMEX physical delivery requires gold of 995 fineness or better — the same standard 2nines5 takes its name from.)
But structural cleanliness is not the same as safety, and this is where futures traders fool themselves. Leverage on futures is severe. As of 2026, with gold near $4,200 an ounce, a single MGC contract carries a notional value of roughly $42,000 — and a broker may let you control it on intraday margin of well under $2,000. That's leverage on the order of 20-to-1 or more. The exchange removed the counterparty conflict; it did nothing to remove the leverage that does most of the damage.
And the outcomes confirm it. The CFTC's own research on retail futures traders found the familiar pattern: cohorts shrink over time as losers drop out, and the traders who remain are still, on average, losing money. A cleaner structure produces better-defined costs and no dealer betting against you — real advantages — but it does not turn a leveraged speculator into a winning one. The game still applies.
The arithmetic
Step back from brokers entirely and the deepest reason emerges. Most retail traders don't lose because they're bad at reading gold. They lose because the arithmetic of frequent, leveraged trading is hostile in a way that has nothing to do with skill.
The foundational research is Barber and Odean's work on retail investors, bluntly titled "Trading Is Hazardous to Your Wealth." Tracking thousands of accounts, they found that the investors who traded the most earned the worst returns after costs — the extra gains from extra trading were more than eaten up by the costs of doing it. The most active traders didn't just underperform; their net returns went negative. The most active quintile underperformed the market by around 6.5 percentage points a year.
The obvious objection is that this was the expensive 1990s, and trading is nearly free now. It doesn't hold. When researchers examined modern low-cost brokerage data, the same shape appeared: more active traders still generated higher returns before costs, but once costs were subtracted, the most active traders still slid into negative territory. Cheaper trading didn't repeal the math. It just lowered the threshold at which the math turns against you.
Now add leverage, which is the defining feature of retail gold trading on both paths. Leverage doesn't change your average outcome — it multiplies your variance, and variance is what kills accounts. A study of leveraged retail FX traders found active traders losing on the order of 9% of their account per month after fees. At that rate, even modest leverage compounds small, repeated costs and small, repeated mistakes into account-ending drawdowns. The researchers' conclusion was that capping leverage was the single most effective protection — precisely because leverage is the accelerant, not the analysis.
This is the honest core of the module. The spread, the swap, the markup, the leverage — each is a small, certain drag. Your winning trades are large and uncertain. Run that asymmetry across hundreds of trades and the certain costs grind down the uncertain gains. That's not pessimism, and it's not the broker's fault. It's arithmetic — the inherent math of the game — and it's why the disclosed loss rates look the way they do. The traders who beat it don't find a friendlier structure; they trade less, size smaller, and respect the arithmetic the majority ignore.
Carry this into F8