At the Finance Magnates London Summit (FMLS:25), while discussing the topic Art of the Dealer Risk Management and Industry Education, panelists painted a picture of the modern FX/CFD dealer as part quant, part firefighter and part educator, operating in an environment where poor risk discipline and weak internal understanding can still sink firms.

The discussion was moderated by Elina Pedersen, the Co-Founder and CEO of Your Bourse, and brought together panelists: Antois Patinios, Dealing Analyst at TMGM, Lee Goldfarb, the Executive Sales Trader at B2Prime, and Chariton Christou, the Co-Founder and CEO at Boltzam Research.

They stressed that while automation and analytics are advancing, the core edge remains a disciplined grasp of risk, market structure and client behavior — and the ability to communicate that across entire brokerages.​

• FMLS:25 Kicks Off With Networking Blitz at The Folly in London
• The FM Events App Is Now Live!
• Finance Magnates London Summit 2025: Global Fintech & Trading Leaders Unite

What Makes a Top-Tier Dealer

“It's a lot less about instinct and a lot more about understanding your underlying flows, who's trading, why they're trading, what they're looking to trade and what kind of risk you can actually manage,” Goldfarb said.

“So, you need to know your numbers very well. What is the net open position that you as a firm is comfortable in holding? What is the actual trade flows, how they behave, and how this, how as a broker, how you manage them.”

Patinios put adaptability and resilience under pressure at the top of the list, noting that around 80% of his day is structured around routine reporting, exposure checks and system monitoring, with the remaining 20% reserved for handling breaks, outages and sudden market moves.​

“I think adaptability and not folding under pressure. Because as a CFD dealer, you have routine daily tasks, but you also have monitoring system health, going through reports, analysis of client flow, how to deal with flow if you have to take defensive measures.”

On the question of tools, Christou mentioned that retail dealers increasingly share tools and concepts once reserved for institutions, such as spread decay and markouts, but still face distinct challenges around “toxic” or signal-driven flow.

The panel noted that dealers sit at the intersection of pricing, risk and technology, making decisions on whether to internalise or STP client trades while keeping spreads competitive and ensuring platform health in real time.​

Retail vs Institutional: An Information Gap

A recurring theme was the asymmetry of information between retail brokers and institutional liquidity providers. Goldfarb explained that on the institutional side, price-makers can see covariance across flow — for example, identifying clusters of gold EA traders and understanding when a signal is driving the market — giving them a visibility advantage over retail brokers who often see only headline markout metrics.

“Institutional space is a kind of different and you have to be sure that the liquidity is there, like kind of every single time you have to understand the flow from a kind of different perspective, Christou agreed.

“You don't want actually to increase the spreads. You don't want to have unhappy brokers. While on the CFD retail space, you do care about it, but you also care about consistency of the price.”

The panelists agreed that prime-of-prime providers sit “in the middle” of these worlds and need to translate between two sets of metrics and priorities to keep relationships sustainable.​

Stress Testing, Concentration Risk and Smaller Brokers

The discussion turned sharper when the moderator raised a comment from one partner that the industry should focus more on stress testing.

More from FMLS:25: “We Have Entered an Era Where Data Is Abundant, Accessible, and Tailored,” FMLS:25

In an environment where gold can move 5% in a day, that blind spot can be fatal if a skewed book goes the wrong way. According to Goldfarb, “that can be very, very damaging if you have a significant level of B book exposure that goes one way.”

“So, it's looking at your book, looking at whether or not the book itself is skewed and looking at how that affects payouts and net open positions. I think a lot of smaller brokers don't look at that.”

Education and Opening Up the “Black Box”

The second half of the session shifted to industry education, with Patinios offering a personal case study. He joined TMGM with no dealing background and spent roughly three months shadowing senior dealers before placing his first hedge, supported by internal training and regular sessions with teams in Australia.

“And I think that's the next step going forward to also not just allow dealers, but let's say people in marketing to understand what they're selling. It's very useful for the future.”

Goldfarb described the industry as operating like a “medieval guild”, where aspiring dealers must first be “brought in” by a firm and often cannot obtain platform-specific certificates without an existing license.

“I think our industry kind of operates a bit of a black box. It's like a medieval guild. You have to be bought in by a firm, you have to shadow that firm. And then you can go even if you wanted, for example, if you had a MetaTrader license, you can do the MetaTrader dealer certificate.”

Who Should Be Educated — and How

On the question of academic backgrounds, Christou favored quantitative disciplines such as mathematics, physics, engineering, computer science and finance, which equip dealers to reason about trade distributions and exposure in a structured way.

Keep reading: Pricing Can Make or Break a CFDs Broker: How Is It Done Right?

Asked how to close the gap between dealing and risk departments and the rest of the brokerage, Goldfarb argued that splitting them too rigidly can create conflicts: sales-driven dealers might court high-volume clients whose flow is nearly impossible to place profitably on the back end.

“The risk and the dealing kind of need to sit together for the company to run well and for it to be profitable. So that conversation has to be very, very tight. Separating them too much, so you get a conflict between.”

Market Reality vs Retail Expectations

Patinios put real-time exposure dashboards and internal controls at the top of the list, including clear NOP limits, dynamic margin settings ahead of news and continuous monitoring of server and bridge health rather than relying on periodic checks.

Responding to a question from the audience towards the end of the discussion, Christou explained that AI plays an everyday operational role, noting that his team was among the early adopters.

It includes having own in-house quantitative team to develop advanced machine learning models trained on real GPUs to predict market movements.