In 2008, when Exness launched with a $1 minimum deposit and MetaTrader 4 was the only retail platform that mattered, the bridge technology connecting a broker's trade server to its liquidity providers was a competitive market. Multiple vendors built and sold bridge solutions independently of MetaQuotes. Pricing was negotiable, and a new broker setting up shop in Cyprus or the Seychelles could pit three or four bridge vendors against each other and walk away with infrastructure costs that amounted to a fraction of a pip per lot. The spread a retail client paid was mostly about the broker's own margin decision — not about the toll the platform vendor extracted upstream. The architecture was, by the standards of this industry, relatively open.
That is not how it works now. MetaQuotes — the company behind both MT4 and MT5 — has spent the better part of a decade tightening control over the ecosystem, from white-label licensing terms to bridge API access to the conditions under which third-party technology can plug into its servers. The result is a pricing structure that almost nobody in the retail forex press covers with any specificity. This week, the same structural question surfaced in the growing conversation around prediction markets: who profits from the platform, and is the game actually fair for the participant who pays the spread? We have read enough coverage of both topics — across trading blogs, fintech newsletters, and YouTube breakdowns — to identify what nearly all of it gets wrong.
What They All Get Wrong
The standard piece on prediction market fairness asks whether the odds are accurate. It asks whether the crowd is wise. It asks whether a given contract was correctly priced at 62% or should have been 58%, and whether the final settlement validated the market's collective judgment. This is the wrong question — or rather, it is a question that serves the platform's narrative rather than the participant's interest.
The right question is structural: who provides the liquidity on that contract, what bid-ask spread do they capture, and what happens to a retail participant who takes the opposite side of a position held by a market maker with superior information flow and a direct contractual relationship with the platform? That question does not get asked, because answering it would require the kind of cost decomposition that prediction market operators — like forex infrastructure vendors — have no incentive to publish.
The same misdirection operates in nearly every article about MetaQuotes and MT5 bridge technology. The coverage focuses on platform features — the depth-of-market window, the hedging mode, the 21-timeframe chart layout, the ability to run Expert Advisors in a multi-threaded environment. It does not focus on the licensing structure that determines what the broker pays MetaQuotes for access to that environment, what the bridge provider charges for connectivity on top, and how those layered costs arrive at the retail trader's terminal disguised as a "standard account spread."
*MetaQuotes does not publish its licensing fees on its website. This is verifiable — the pricing page, if it exists, is not publicly indexed.*
Consider the spread data across five brokers that all operate on the MT5 ecosystem. Exness offers EUR/USD at an average of 1.0 pip on its standard account and 0.1 pip on its pro account. FXTM lists 1.5 pips standard, 0.1 pip pro. HF Markets runs 1.2 pips standard, 0.0 pips pro. FBS publishes 0.7 pips standard, 0.0 pips pro. AvaTrade charges 0.9 pips on both tiers — the same spread regardless of account type. The gap between standard and pro is not a loyalty reward. It is a rough map of where the broker's infrastructure costs — including bridge licensing, MetaQuotes platform fees, and server hosting — get absorbed into the price. On the pro account, the trader pays something close to the raw liquidity spread. On the standard account, they pay that same spread plus everything the broker needs to cover upstream.
The error across conventional coverage is treating this markup as a purely discretionary broker decision, as though the broker simply chose to charge more and could choose not to. A portion is discretionary. But a material share of the gap is structural — it reflects the pricing power of a platform vendor that controls the ecosystem without meaningful competition at the retail scale. A broker offering MT5 pays whatever MetaQuotes charges, because MT5 is where the retail order flow is.
The same logic applies to prediction markets. The platform operator captures rent from every contract traded. The question is not whether the odds are fair. The question is whether the toll is disclosed.
What Is Almost Always Missing
What is missing from both conversations — prediction market fairness and MT5 infrastructure pricing — is the full incentive chain traced from top to bottom, with numbers attached at every link.
*The SEBI helpline is open 09:30–17:00 IST. It does not field questions about offshore bridge technology costs or prediction market fee structures.*
Start with an Indian retail trader running a ₹50,000 account on MT5 through Exness. At approximately 84 INR per dollar, that account holds about $595. Exness offers leverage up to 1:2000, but for EUR/USD on a standard account, a practical position size — assuming even minimal risk discipline — is roughly 2 to 3 standard lots per trade. Take 3 lots. At a standard-account spread of 1.0 pip on EUR/USD, each pip on a standard lot equals $10. Three lots at 1.0 pip spread means $30 in spread cost per round-turn trade. If this trader executes one trade per day across 22 trading days in a month, the total spread cost is $660 — approximately ₹55,440 at 84 INR/USD.
That is 110% of the original account balance. Per month. In spread drag alone.
Now run the same arithmetic on the pro account at 0.1 pip: 3 lots multiplied by $1 per lot per trade, over 22 trading days, totals $66 per month — approximately ₹5,544. The difference between the two accounts is ₹49,896 per month. That gap — call it ₹50,000 in round numbers — represents the cost of the infrastructure stack sitting between the raw liquidity price and the retail trader's execution screen. It includes the broker's own margin, the bridge technology licensing fee, the MetaQuotes platform fee, and the data centre hosting costs. The coverage never traces this chain to its components. It never asks: of this ₹50,000 monthly cost difference, how much flows to MetaQuotes, how much to the bridge vendor, and how much does the broker retain?
And that is the critical omission. Without decomposition, the retail trader has no way to evaluate whether the spread they pay is competitive relative to the broker's actual infrastructure cost — or whether it reflects monopoly pricing upstream. The same information deficit operates in prediction markets. The platform extracts a fee. The market maker captures a spread. The retail participant pays both. Nobody publishes the breakdown, because every participant in the chain benefits from keeping it aggregated.
*AvaTrade charges the same 0.9-pip EUR/USD spread on its standard and professional tiers. The spread does not change between account types. The infrastructure cost is baked in identically. This is a structurally different pricing model from the standard/pro split — and nobody in the conventional coverage compares the two approaches or asks what each one reveals about the upstream cost structure.*
The missing analysis is not another broker comparison. It is structural cost accounting applied to an industry that has never been asked to provide it.
What I Would Say Instead
The frame should be this: both prediction markets and MT5 bridge pricing are toll collection systems, and the only honest analysis is one that names the toll collector and estimates the toll.
For prediction markets, the toll has two components. First, the bid-ask spread maintained by designated market makers — participants with contractual access to the platform's order flow data and the capital to warehouse inventory. A prediction contract trading at 62 cents "yes" and 38 cents "no" appears to sum to 100 cents — fair pricing, zero extraction. But the actual bid-ask available to a retail participant might be 61 to 63 cents, meaning the market maker captures 2 cents on every dollar of notional flow through the book. On a contract with $10 million in cumulative volume, that is $200,000 in market-maker revenue before the platform takes its transaction fee. The retail participant buying "yes" at 63 cents needs the contract to settle at 100 to realise a 58.7% return. The market maker needs the flow to continue. Their incentives are structurally different — and only one of them depends on being right about the outcome.
For MT5 bridge technology, the toll operates through the spread stack. Consider the four brokers in our data set that differentiate between standard and pro pricing — Exness, FXTM, HF Markets, and FBS. The average spread gap between standard and pro EUR/USD accounts across these four is 1.05 pips. That 1.05-pip gap, on one standard lot, represents $10.50 per trade in structural cost passed to the standard-account client. Multiply that by the total standard-account lot volume across a broker's entire client base in a given month, and you arrive at the broker's gross infrastructure subsidy extracted from retail flow. The fraction of that subsidy flowing upstream to MetaQuotes for platform licensing — and to the bridge provider for connectivity and execution routing — is disclosed by none of the brokers we examined.
*Exness lists "instant" withdrawal speed. FBS lists "instant to 1 day." The withdrawal is fast. The spread is permanent.*
This is not an argument for regulation. It is an argument for decomposition. When an Indian retail trader running a ₹50,000–₹1,00,000 MT5 account through Exness, FXTM, or HF Markets pays 1.0 to 1.5 pips on EUR/USD, they should have access to — even approximately — what fraction of that spread represents raw liquidity cost, what fraction represents the bridge technology toll, and what fraction is the broker's retained margin. The identical decomposition applies to prediction market transaction fees. The identical decomposition is absent from both industries' public discourse. The reason is not complicated: every participant in the chain profits from opacity, and no participant has a commercial incentive to break it.
This piece does not address the regulatory status of prediction markets under Indian law — SEBI's position on event-based trading platforms remains unresolved and is a separate, heavily contested argument. It does not address MetaQuotes' internal cost structure, because that data is not public and no broker we reviewed discloses the breakdown of their upstream licensing obligations. And it does not address whether the bridge technology toll is justified by the engineering complexity of the product — that is a question that requires access to vendor contracts, and we do not have them.