Path B · Futures GC/MGC · Module B2
B1 ended on a single word: novation. The clearing house steps between you and a stranger and becomes the other side of your trade. That sentence is easy to say and easy to underestimate. This module takes it apart — because once you see what stands behind that word, every cost, every margin number, and every "what if my broker fails" question in Path B answers itself.
What this module does
The hinge, opened
B1 watched the clearing house interpose itself at the moment your order matched — buyer to the seller, seller to you, the stranger gone from your side. That's novation. But "your counterparty is the clearing house" is a claim that only reassures once you know what's standing behind it. On its face it sounds worse, not better: in spot you at least knew who you were facing. Here you face a faceless institution. Why is that the safer arrangement?
Because the institution is built so the question of who stops mattering. The clearing house — for COMEX gold, CME Clearing — does not take a market view. It is novated into every trade, long and short, in equal and opposite measure, so its own net position in gold is always zero. It does not win when you lose. It has no directional interest in your trade at all. That alone is the entire broker conflict, dissolved: a counterparty that is structurally flat cannot be a counterparty that profits from your loss.
What replaces the directional conflict is a different job: making sure the party on the other side of you can pay. The clearing house guarantees performance. If the trader who took the opposite side of your position defaults, the clearing house still settles with you — it has to, because as far as you're concerned, it is the other side. It can make that guarantee because it doesn't rely on goodwill. It relies on a system of collateral collected in advance and settled every single day. That system is the rest of this module.
How the guarantee is funded
Here is the mechanism that lets a flat institution stand behind every trade without ever taking a loss of its own: it never lets a loss accumulate. In spot, an open position's profit or loss floats unrealised until you close — your broker carries the running tab. In futures, the clearing house refuses to carry a tab. Every position is marked to the settlement price once a day, and the day's gain or loss moves in actual cash, that day, between the two sides.
This is mark-to-market, and it's not an accounting nicety — it's the spine of the whole guarantee. If gold falls and your long loses $40 today, that $40 leaves your account tonight and lands in the account of whoever was short. Tomorrow starts fresh. Because losses are collected daily rather than allowed to pile up, the most the clearing house can ever be exposed to is roughly one day's move — and that's exactly the amount it has already collected up front as margin. The guarantee isn't a promise to cover losses. It's an architecture that never lets a loss get bigger than the collateral already in hand.
And if a participant can't meet a daily call, the clearing house doesn't absorb it alone. There's an ordered sequence — the defaulter's own margin first, then their clearing member's contribution, then a mutualised guarantee fund the whole membership pays into, and the clearing house's own capital layered in. You will never see this machinery and never touch it. But it's why "your counterparty is the clearing house" is a stronger statement than any broker's balance sheet could ever be. You're not trusting a firm. You're trusting a waterfall designed so the firm never has to be trusted.
What your broker actually is
B1 called the firm you hold your account with a Futures Commission Merchant, and said it was a conduit, not a counterparty. Now we can say exactly what that means in the place it matters most: where your money sits.
Here is spot's problem. In spot, you deposit funds with the broker, and the broker is also the party you trade against. Your money and your counterparty are the same entity — so if that broker fails, your funds and your trades fail together. That's not a flaw in any particular broker; it's the structure. There is nowhere else for the money to be.
In futures, the structure pulls those apart. By US rule, an FCM must hold customer money segregated from its own — in accounts titled for the benefit of customers, which the FCM cannot use to fund its own operations or trades, and which sit ahead of the firm's other creditors if it goes bankrupt. The FCM holds your account and routes your orders; the clearing house holds the risk. The firm carrying your money is structurally not the party on the other side of your trade. If your FCM went under tomorrow, your positions and segregated balance are designed to be portable — transferred to another FCM — rather than caught in the wreckage. That is the precise inversion of spot: the entity that holds your funds cannot be the entity that profits from your loss, because the rules forbid it from being either at once.
The honest limit
Segregation is structural protection — but this site's rule still holds: understand a thing, don't worship it. In 2011 a major FCM, MF Global, failed, and for the first time in the industry's history segregated customer funds were found missing — roughly $1.6 billion — because the firm had improperly dipped into client money to cover its own failing proprietary bets. Most US customers were eventually made substantially whole through bankruptcy, but the lesson is exact: segregation defends you against your FCM's insolvency; it does not defend you against your FCM's fraud. The structural protection is real and far stronger than spot's. It is not a force field. Choosing a properly regulated FCM still matters — for the same reason choosing a constrained broker matters in spot.
So the residual question in futures is not "will my counterparty trade against me" — that one is closed. It's the narrower, more honest question worth asking of any firm: is the one holding my money well-regulated and well-run? The conflict is gone. Ordinary operational trust is not. That's a far better place to stand than spot ever offered, and it's worth being precise about why.
Where the money inverts
In spot, your leverage — 20:1 on gold across Tier-1 regulators — is a number your broker offers you inside a regulatory ceiling. Different broker, different number. Futures invert this too. The collateral you must post isn't set by your FCM and isn't a leverage ratio it markets to you. It's set by the exchange's clearing house, by a risk model, and it's the same baseline for everyone trading that contract.
The model is called SPAN (now running as SPAN 2, a value-at-risk version of the same idea). You don't need its internals. You need its logic: rather than a fixed ratio, SPAN asks a question — across a realistic one-day range of gold prices, what's the most this position could plausibly lose? — and sets the required margin to cover that. Margin becomes a measure of the position's risk, not a number a salesperson picked. When gold gets more volatile, the plausible one-day loss grows, and the margin requirement rises with it, for everyone, automatically.
Two terms you'll meet the moment you open a position. Initial margin is what you must have to open the trade. Maintenance margin is the lower line your account must stay above to keep it open. If a day's mark-to-market loss drags your balance below maintenance, you get a margin call — top up, or the position is closed. Notice this isn't your broker's discretion deciding to liquidate you, the way a spot dealer manages its own book. It's the same daily-settlement machinery from Job 2, applied to your account: the system collects what it's owed every day, and the call is just that system speaking.
The honest limit
Don't mistake "set by the exchange" for "fixed" or "neutral in your favour." The exchange can and does change margin — sometimes sharply, mid-position — when it judges risk has risen. Gold's run through 2025 was volatile enough that in January 2026 CME switched precious-metals margining to a percentage of contract value (gold initial margin around 5%) precisely so requirements would track price and volatility more directly. A margin hike can force you to post more on a position you already hold, or close it. The difference from spot isn't that discretion disappears — it's that the discretion belongs to a neutral, flat clearing house applying one public rule to everyone, not to a broker managing a book it profits from. Exchange authority is real authority. It's just not pointed at you specifically.
The inversion, paid off
Stand the two structures side by side. Spot hands you an uncomfortable case, and the honest move is not to flinch from it: your spot broker is your counterparty, the conflict is real, and the right response is not to panic but to choose a broker constrained enough to be trusted with it. The victim framing is wrong — traders don't lose because of the b-book; difficulty is inherent to the instrument — but the conflict is not pretend. It exists. In spot, you manage it.
In futures, you don't manage it, because it isn't there to manage. The clearing house is flat by construction. The FCM is forbidden from being your counterparty and from touching your segregated money. Margin is a public risk model, not a private offer. Every lever a conflicted spot broker could pull — pricing against you, last-look rejections, internalising your loss as its gain — requires being on the other side of your trade, and in futures no one you deal with is. The conflict didn't get regulated into submission. It got engineered out of the structure.
That is the whole payoff of the inversion B1 promised. And it comes with its own honest price, which the rest of Path B will count: futures don't remove difficulty, they relocate it. There's no spread, but there's a commission and an exchange fee. There's no swap, but contracts expire and must be rolled. There's no broker liquidating you on a whim, but there's a margin model that can call you in a volatile session. The instrument is more honest about who stands where. It is not easier. The next module follows the money the way B1 followed the order — into what futures actually cost.
Carrying forward
B1 inverted the counterparty. B2 showed you the institution that the inversion installs in its place, and what that institution actually does — settle daily, guarantee performance, hold your money apart, price risk in the open. With the structure understood, Path B turns to the bill. Futures trade their conflicts for costs, and the next module puts a number on every one of them.