40 Brokers Ranked by Commission With $40 Each: What the Methodology Actually Measures

Let us concede the obvious point first. A $40 deposit at each of forty brokers, a single round-trip per venue, a spreadsheet sorted by commission paid — yes, that produces a ranking. The ranking is real. The arithmetic is sound. The methodology, as a piece of consumer journalism, is reproducible. We have read the write-ups, watched the videos, and we accept what the numbers say at face value. What we want to argue, across three composite scenarios drawn from the execution-layer record of Refco, MF Global, and FXCM, is that the variable being ranked is not the variable that determined whether the depositor got the $40 back.

The argument is not that commission does not matter. It does. The argument is that across three different trader profiles — each one a composite illustration, not a person we interviewed — the commission line on the spreadsheet was the smallest input into whether the account ended the test with more dollars than it started with. The execution-layer record, the part the $40 methodology does not measure, decided the rest. Let us walk through three hypothetical traders and watch what the spreadsheet missed.

Scenario 1: The Weekend Scalper With a $40 Float Per Venue

Imagine a trader — call her the weekend scalper — who runs the $40 test across forty venues over a single Saturday morning. She is not a beginner. She knows the methodology she is reproducing: deposit, open one EUR/USD round-trip at a defined size, close it, withdraw, log the commission line. The whole point of her test is to isolate the cost-per-trade variable. She has built a Google Sheet with forty rows.

On paper, the broker that wins her ranking is the one with the tightest pro-account spread. Inside the grounding we have for this piece, that is Exness, with a EUR/USD pro spread of 0.1 pips, against a standard account of 1.0 pips. FBS is even tighter on its zero-spread tier at 0.0 pips, with a standard of 0.7. HF Markets and FXTM also offer 0.0 and 0.1 pips on their pro tiers, against 1.2 and 1.5 on standard. AvaTrade, with 0.9 pips on both tiers, ranks middle of the pack on the commission column.

She runs the round-trips. The spreadsheet ranks Exness pro and FBS zero at the top. The methodology is intact.

Here is where the execution layer enters. The historical record we work from — the published postmortems of Refco's 2005 unwind and the FXCM January 2015 negative-balance aftermath — describes a class of risk that does not show up in a single round-trip. It shows up in the reconciliation cycle that runs after the close. Refco's collapse was not a trading failure. It was a reconciliation failure: receivables that had been parked off the consolidated balance sheet for years through a related-party loan. The depositors who lost money in October 2005 did not lose it because of spreads. They lost it because the entity holding their cash was not the entity their statement said was holding it.

The weekend scalper's $40 spreadsheet has no column for which legal entity took her wire. She deposits, she trades, she withdraws. The column she ranks on is commission. The column that decides whether the withdrawal clears is which subsidiary, in which jurisdiction, under which segregation rule, holds the float between Saturday's deposit and Monday's settlement. The tier-1 regulator column in the grounding — FCA for Exness, FXTM and HF Markets; ASIC for AvaTrade and FBS — tells us part of that story. It does not tell us all of it. Withdrawal speed does. Exness instant, FBS instant-to-one-day, HF Markets one day, AvaTrade and FXTM one-to-three days. The spread ranking and the withdrawal-speed ranking do not produce the same order.

Her test, run on the commission column alone, gives her the wrong answer to the question she thought she was asking.

Scenario 2: The Swing Trader Who Holds Through a Wire Cutoff

Let us picture a second trader. He runs the same $40 test, but his profile is different. He is a swing trader. He opens a position, leaves it on for forty-eight hours, then closes. His commission exposure per trade is lower as a percentage of P&L than the scalper's, because his moves are larger. He cares less about the 0.1-versus-0.9 pip spread distinction and more about funding cost and withdrawal mechanics.

His ranking, sorted by the commission column, looks different. The spreads still favor Exness pro and FBS zero, but his weighting shrinks the gap. Where the methodology really starts to fail him is on the side that the spreadsheet does not contain a column for: the wire cutoff.

The execution-layer record from MF Global's October 2011 failure is the relevant grounding here. The trail that investigators reconstructed showed segregated customer funds being moved through the firm's accounts in the final days, and the question of whether each transfer counted as a permitted use turned on the timing of wire settlement, not on commission terms. The depositors whose withdrawals were caught on the wrong side of the cutoff lost access for months. The depositors whose withdrawals had cleared the day before did not. The commission column on their statements was identical.

For our hypothetical swing trader, the question is not which broker has the tightest pro-spread. It is which broker's withdrawal speed survives the worst case. The grounding gives us a workable proxy. Exness publishes instant. FBS publishes instant-to-one-day. HF Markets, one day. AvaTrade and FXTM, one-to-three days. Those windows are the published averages, not the stressed-condition figures. The MF Global lesson is that the stressed figure is what matters, and the published average is not its predictor.

He cannot run a stressed-condition test for $40. Nobody can. What he can do is read the published one-to-three-day figure as a floor, not a typical. AvaTrade's conservative leverage cap of 400 and FXTM's similar profile do not, in isolation, make either broker safer than the higher-leverage venues — but the slower withdrawal cycle does add operational lag that, in a stressed event, sits on the wrong side of the depositor's interests. The $40 methodology gives both venues a middling rank on commission. It gives neither venue a rank at all on the variable that would have mattered in October 2011.

Scenario 3: The Carry Trader Who Never Sees the Commission Line

Now imagine a third trader. She is running a carry strategy. She opens a position and holds it for weeks. Her commission exposure is genuinely negligible — one round-trip per multi-week cycle, divided across a position size that the spread cost rounds to noise against. The $40 methodology, applied to her profile, is asking a question she does not need answered.

What she actually pays is the overnight swap. The $40 test does not measure swap. It cannot — a single round-trip closed inside one session never crosses a rollover. Her real cost is the daily debit or credit that the broker books against her position at the end of each trading day, calculated against the interest-rate differential and a markup that varies by venue and by currency pair. The methodology she is reading about online is not designed to surface this number.

The grounding gives us a partial map. Islamic-account availability — true for AvaTrade, Exness, FBS, FXTM and HF Markets in the data we have — is a swap-substitute structure for traders who cannot pay or receive interest. The existence of the Islamic account, by itself, tells us the broker has built the machinery to apply a different daily-carry mechanic. For our carry trader, that machinery is the variable. Whether she uses the Islamic account or the standard rollover, the broker's pricing of the daily carry — not its EUR/USD pro spread — is the line item that compounds across her holding period.

There is a second layer here that the FXCM January 2015 aftermath surfaced and that we want to name carefully. When EUR/CHF gapped, the positions that destroyed FXCM's US capital base were not the small-size scalp trades. They were the leveraged carry-style holdings that had accumulated against a peg widely treated as a hard floor. The negative-balance event that followed was not about commission. It was about the gap between the published leverage cap and the venue's ability to close positions inside the price chain that actually printed. Exness's published 1:2000 max leverage and FBS's 1:3000 are facts; their behavior under a printing-gap event is a separate fact, and the $40 test does not measure it.

For her, the spreadsheet column that mattered was never on the spreadsheet.

What All Three Share

Across all three composite cases, the variable the $40 methodology measures most precisely is the variable that mattered least to the outcome. The scalper's ranking was correct on its own terms; the term was the wrong one. The swing trader's commission ranking captured a fraction of his real cost; his real cost was the operational lag the methodology did not measure. The carry trader's profile sat almost entirely outside what the test was designed to capture.

What unites them is not the conclusion that commission does not matter. It does. What unites them is that the cost of being wrong about commission, in the range the grounding shows us between 0.0 pips on a zero-spread tier and 1.5 pips on a wider standard tier, is small relative to the cost of being wrong about the execution layer. Refco's depositors did not lose money on the commission column. MF Global's did not. FXCM's January 2015 negative-balance customers did not.

The execution-layer record we work from has a specific texture. It is about which entity holds the cash. It is about which jurisdiction the segregation rule was written in. It is about the wire cutoff on the day the bad print arrived. It is about whether the published leverage cap survives contact with a gap that exceeds the broker's stop-out distance. None of these variables are in the $40 spreadsheet. All of them have been, in the historical record, the variables that decided whether a customer's withdrawal cleared.

This is not an argument that the $40 test is useless. It is an argument that the test is well-calibrated for the question it asks and badly calibrated for the question the customer thinks it is answering. The two questions are not the same.

Which Scenario Is You

If you are the scalper — small float, short holding period, many round-trips, commission a meaningful share of P&L — then the test's ranking is closer to useful for you than for the other two. Read the commission column. Then, before you act on it, add a second column for which legal entity took your wire and a third for what the withdrawal-speed figure looked like in the worst published month, not the average.

If you are the swing trader, the ranking is not your map. Your map is the operational lag under stress, and the published withdrawal-speed window is the closest proxy you have without running a real stress test you cannot afford. Treat the slower end of the published range as the figure that matters, not the faster end.

If you are the carry trader, the methodology is the wrong instrument entirely. Your line item is swap, your structural question is daily-carry mechanics, and the $40 round-trip will not surface either. Build the test you actually need, or accept that the published rankings you are reading are answering somebody else's question.

Whether the aggregate effect of the $40 test is to make depositors better off — by giving them any framework at all — or worse off, by giving them a precise answer to the wrong question, is something the published execution-layer record does not settle. If you have run the test on more than one venue and watched what the reconciliation cycle did afterward, write.

FAQ

Does the $40-per-broker commission test actually rank brokers correctly on cost?

On the narrow variable it measures — commission paid on a single EUR/USD round-trip — yes, the arithmetic holds. Across the brokers in our grounding, that ranking would put Exness pro at 0.1 pips and FBS zero at 0.0 pips ahead of AvaTrade at 0.9, HF Markets pro at 0.0 against a 1.2 standard, and FXTM pro at 0.1 against a 1.5 standard. The ranking is real; what it is not is a ranking of total cost-to-outcome for any holding period beyond one session.

Why does withdrawal speed matter more than commission in the methodology's blind spots?

Because the execution-layer record from Refco 2005 and MF Global 2011 shows that depositor losses, when they happen, cluster around the reconciliation and wire-settlement cycle, not around per-trade cost. A broker can be cheapest on commission and still be the slowest to return funds when the operational machinery is stressed. Exness publishes instant withdrawal, FBS instant-to-one-day, HF Markets one day, AvaTrade and FXTM one-to-three days — those windows are the floor on the variable the test does not measure.

How does tier-1 regulation factor into this if every major broker advertises it?

The grounding shows tier-1 coverage for all five named brokers — FCA for Exness, FXTM and HF Markets, ASIC for AvaTrade and FBS — but tier-1 regulation governs the entity licensed in that jurisdiction, not necessarily the entity that takes a non-EU or non-Australian depositor's wire. The legal-entity question is the one the $40 test does not ask, and it is the one the Refco postmortem identified as decisive. Read the wire instructions, not the marketing page.

Is the swap line really invisible to a $40 round-trip test?

Yes, by construction. A round-trip opened and closed inside one trading session never crosses a rollover, so the swap charge or credit is never booked. For a carry trader holding for weeks, the swap line compounds against the position daily and rapidly exceeds any plausible commission savings. The presence of an Islamic account at AvaTrade, Exness, FBS, FXTM and HF Markets indicates each venue has built alternate daily-carry machinery, but the pricing of that machinery is a separate question the methodology does not surface.

What did the FXCM January 2015 event actually reveal about leverage caps?

That the published leverage maximum and the venue's behavior under a gapping print are two different facts. When EUR/CHF moved outside the range FXCM's stop-out machinery could close inside, the resulting customer deficits were not commission-related — they were structural, on the gap between the printed price and the executable price. Exness's 1:2000 and FBS's 1:3000 published caps describe normal-state behavior. Stressed-state behavior is not what the $40 test measures, and it is what the negative-balance aftermath taught the desk to weight.

Should retail traders run the $40 test at all?

The test has value if you treat it as a measurement of one variable, not a recommendation. Run it, log the commission column, and then add the columns the methodology omits: legal entity holding the float, jurisdiction of the segregation rule, worst-month withdrawal time rather than average, and whether your strategy crosses overnight rollovers. The test is a starting point; treating its ranking as the answer is what gets depositors hurt in the events the execution-layer record documents.

Does broker founding date or platform selection matter for this analysis?

Marginally. The grounding shows founding dates clustering between 2006 and 2011 across the five brokers — a similar vintage, all with MT4/MT5 coverage and proprietary apps layered on top. Platform selection matters for execution latency and order types, which is a real variable, but it sits below the operational-lag and legal-entity questions in the hierarchy of what has historically decided whether depositors got their money back.

What would a methodology that actually measured the right variable look like?

It would deposit, hold across a wire cutoff, attempt a withdrawal under a non-standard condition, and time the reconciliation cycle — not just the commission round-trip. It would record the receiving entity's name on the wire confirmation and check it against the regulator's licensed-entity register. It would, for carry profiles, hold the position long enough to book a week of swaps. It would cost more than $40 per venue, which is precisely why the cheaper methodology dominates the published comparisons and why the depositor reading them is being given a precise answer to the wrong question.