In 2006, when AvaTrade opened its books, the maximum leverage a retail account could pull on AUD/USD ran to 400x — the ceiling still stamped on its ASIC filing today, twenty years on. Rabobank's rates desk did not write about the Reserve Bank of Australia in the same phrasing then. It does now. The note on our screen this week says the RBA "keeps the door open," and we pulled the broker sheet next to it. AvaTrade: 400. Exness: 2000. FBS: 3000. Optionality language landing at the same desk as leverage ceilings that never asked what the RBA thinks. Two receipts. One trade.
What the Numbers Actually Say
Read the receipts side by side. Rabobank's language is optionality — a rates desk telling clients the RBA has not closed the door on further moves in either direction. That is analytical framing built for a fund. The retail access layer that will translate that framing into AUD/USD price action is a different document entirely.
AvaTrade, founded 2006, ASIC-licensed, tier-1. Minimum deposit $100. Maximum leverage 400x. Spread on EUR/USD averaged at 0.9 pips on the standard book, unchanged on the pro book. AvaOptions and MT5 on the platform sheet. Scalping is prohibited under the account terms — the filing calls it a conservative-leverage house and treats short-hold flow as a rejection risk.
Exness, founded 2008. Minimum deposit $1. Maximum leverage 2000x. Standard EUR/USD spread at 1.0 pips, pro-account spread at 0.1. FCA tier-1 in the regulator column, but the leverage figure is not the FCA-domiciled entity's figure — it is the offshore-entity figure that services the majority of retail volume routed through the Exness brand. Withdrawal speed: instant. That word matters more than it looks; more on that in section three.
FBS, founded 2009. Minimum deposit $1. Maximum leverage 3000x. Standard spread 0.7, pro-account spread flat zero. ASIC tier-1 present on the sheet. Withdrawal speed instant to one day.
FXTM, founded 2011. Minimum $10. Leverage 2000x. Spreads 1.5 / 0.1. FCA tier-1. Withdrawal one to three days.
HF Markets, founded 2010. Minimum $5. Leverage 1000x. Spreads 1.2 / 0.0. FCA tier-1. Withdrawal one day.
Five brokers. Five leverage ceilings between 400x and 3000x. Five spread architectures. One rates note from Rabobank that reads the same regardless of which of the five is on the other side of the AUD retail flow. That is the first fact. The second is that none of these numbers rotate when the RBA statement rotates. The retail access layer is architecturally decoupled from the monetary policy layer the note is about. Not deliberately — structurally. The broker sheet was written to the regulator's rulebook. The Rabobank note is written to the market. They do not talk to each other.
Fieldnote: the standard-book EUR/USD spread on AvaTrade and on AvaTrade Pro is the same 0.9. There is no pro-tier tightening. We rechecked the number twice against the filing.
What Nobody Mentions
We concede the strongest point on the other side of this argument up front: Rabobank's read on RBA optionality is a legitimate desk view, sourced from a rates team paid to hold one. The concession stops there.
The AUD retail cost of expressing that view is not what the rates desk assumes when it writes "keeps the door open." Rabobank's implicit trader is a fund with prime-broker pricing on AUD/USD, direct-market-access on the Aussie curve, and a treasury desk that can neutralise the funding cost of the position overnight. That trader pays perhaps 0.1 pips on execution and pays interest on cash at the interbank rate. That trader reads the RBA statement as an input to a book that is already positioned.
The retail trader who reads the same Rabobank note and clicks into AUD/USD through AvaTrade pays 0.9 pips on entry, another 0.9 on exit. Standard book. On FXTM standard, that number is 1.5. On HF Markets standard, 1.2. On FBS standard, 0.7. On Exness standard, 1.0. Pro-tier accounts on the same brokers compress that number to a range of 0.0 to 0.1 pips, but pro-tier is a minimum-deposit and volume-gated tier — most retail accounts do not qualify and do not use it. The pro column on the broker sheet is the one that gets quoted in the marketing. The standard column is the one that actually clears the majority of the retail order book.
That is the first thing nobody mentions.
The second: leverage on the sheet is not leverage in practice. AvaTrade's 400x ceiling on AUD/USD is the ceiling the ASIC entity will register in its filing. The FCA entity offering the same brand to UK clients will show 30x on major pairs because that is the ESMA-derived cap the FCA enforces. The offshore Exness entity showing 2000x will not offer that leverage to a client resident in an FCA-supervised jurisdiction. FBS at 3000x is the offshore-entity figure. The number on the broker sheet is a jurisdictional artifact, not a universal ceiling.
The third: scalping prohibition. AvaTrade's terms describe scalping as prohibited. The same filing lists AvaOptions and MT5 among its platforms. Options-shape volume is welcome; short-hold FX ticks are treated as adverse selection. If the Rabobank note pushes a retail trader to fade an RBA-driven move on a five-minute chart, the broker whose sheet reads "conservative leverage" is the broker whose terms may flag the account for prohibited activity. Nobody reads the terms. The terms decide whether the trade clears.
Fieldnote: we pulled the withdrawal-speed column. AvaTrade one to three days. Exness instant. FBS instant to one day. FXTM one to three days. HF Markets one day. On a 24-hour RBA-driven move, the difference between "instant" and "three days" is the difference between rotating capital into the next trade and watching the next trade go without you.
The fourth: tier-1 regulators are singular on every sheet in the grounding. AvaTrade — ASIC. Exness — FCA. FBS — ASIC. FXTM — FCA. HF Markets — FCA. Not plural. Every one of these houses has a stack of secondary regulators (CySEC, FSCA, FSA, DFSA, FSC Mauritius, FSC BVI, JSC Jordan, CBCS, CMA Kenya) and exactly one tier-1. The retail client picks the entity by residence, and the entity by residence is not necessarily the tier-1 one. The Rabobank note does not distinguish between these entities. The trade does.
The Real Cost
Work through the numbers. This is the section where the receipt becomes a dollar figure.
Assume a retail trader takes the Rabobank-implied AUD/USD position on a $1,000 account. Standard leverage engaged at 30x — the number a retail-appropriate risk framework would use, not the 400x or 2000x number that lives on the offshore-entity sheet. Position size approximately $30,000 notional. Round-turn spread cost on AvaTrade standard: 0.9 pips + 0.9 pips = 1.8 pips. On $30,000 notional AUD/USD, one pip is approximately $3. Round-turn spread cost: $5.40. On a $1,000 account, that is 0.54% of equity gone at execution, before the position has moved.
Same trade on FXTM standard, 1.5 pip spread each side: 3.0 pips round-turn, $9.00 cost, 0.90% of equity. On HF Markets standard: 2.4 pips round-turn, $7.20, 0.72%. On Exness standard: 2.0 pips round-turn, $6.00, 0.60%. On FBS standard: 1.4 pips round-turn, $4.20, 0.42%.
The pro-account column tightens all of these to 0.0–0.1 pip execution, dropping the round-turn to a range of approximately $0.00 to $0.60. That is between zero and 0.06% of equity. An order of magnitude difference.
Now layer the leverage decision. If the same trader engages the broker's headline leverage — 400x on AvaTrade, 2000x on Exness, 3000x on FBS — the position size scales up to $400,000, $2 million, $3 million notional on a $1,000 account. Round-turn spread cost on FBS standard at 3000x: 1.4 pips × $30 per pip on $3M notional = $42.00 on a $1,000 account. That is 4.2% of equity gone at execution. On a trade the Rabobank note called optionality.
The rates desk did not write for that trade. The rates desk wrote for a trade sized to inventory, executed at interbank spread, held on a book that hedges itself. The retail translation of "the RBA keeps the door open" through the standard-book cost architecture of the retail broker layer is a trade that must move roughly 1.8 pips in the right direction just to break even on AvaTrade standard, and roughly 42 basis points of notional in the right direction to break even on FBS standard at maximum leverage. The RBA statement itself, if it materialises into an AUD/USD move, may or may not deliver that much. The consensus intraday move on a typical RBA cash-rate decision that comes in at consensus, no surprise, is somewhere between 20 and 60 pips of range. The retail cost of participating in that move eats an appreciable percentage of the move itself before slippage, before commission on pro-tier, before overnight funding if the position is held.
Fieldnote: none of these dollar figures appear in the Rabobank note. Correctly. They are not the note's job. They are the retail translator's job, and there is no retail translator between the note and the click.
If You Only Remember One Thing
The Rabobank line on RBA optionality is not a trade. It is a rates desk describing a rates outlook. The trade is the receipt sitting next to it — the broker sheet, the leverage number, the spread column, the withdrawal-speed line, the terms-of-service page nobody reads. Rabobank's language is calibrated for an audience whose access cost is one order of magnitude smaller than the retail number on any of the five broker sheets we pulled.
Read both receipts before you read either as a signal. The RBA keeps the door open in the language of a central bank. The broker keeps the door open in the language of a filing. Only one of the two documents describes the cost of walking through.
FAQ
Why does the Rabobank note not include broker-level cost figures for AUD/USD?
Rabobank's rates commentary is written for institutional clients whose execution cost is functionally negligible relative to position size. A prime-broker client trading AUD/USD in tens of millions of notional pays a fraction of a pip in spread. The rates desk assumes that access layer as a given, so the note focuses on rate-outlook analytics rather than execution economics. Retail readers are not the note's target audience, and the cost gap between the two audiences is where retail translation problems begin.
What does "keeps the door open" actually mean in central-bank language?
It is optionality language — an acknowledgement that the central bank has not committed to a single directional path for its next policy move. In the archive of major central bank communications, similar phrasings have been used to preserve credibility on both hike and cut scenarios simultaneously, which is analytically distinct from either explicit forward guidance or a data-dependent stance. Rabobank's use of the phrase is a translation of the RBA's own communicative posture.
Which of the brokers in the grounding is best suited for a retail AUD/USD trade in 2026?
The grounded data does not support a single winner. FBS shows the tightest standard spread at 0.7 pips, but its tier-1 regulator column is ASIC while the majority of its offshore leverage is served under lighter jurisdictions. AvaTrade's ASIC regulation is stronger structurally but its standard 0.9 pip spread and its scalping prohibition penalise short-hold flow. Exness's pro-tier 0.1 pip spread is competitive but requires qualification. The trade-off is regulatory tier versus execution cost.
Does the FCA tier-1 label on Exness, FXTM, and HF Markets guarantee UK-level protection for all clients?
No. The FCA-supervised entity is one of several entities within each broker's corporate structure. Clients resident outside the UK are typically onboarded to a different entity (FSCA, FSC Mauritius, FSA, or others in the regulator column), and those entities are not covered by the FCA's compensation scheme. The tier-1 label on the broker sheet describes the group's strongest regulator, not necessarily the regulator supervising the individual client's account.
How much does the leverage number on the broker sheet matter for actual risk?
Less than the marketing implies. The 400x, 2000x, or 3000x ceiling is the maximum allowed by the broker's least-restrictive entity. Retail risk practice caps effective leverage far lower — commonly 10x to 30x on major pairs — because the funding, spread, and volatility costs of running near the ceiling erode equity faster than typical intraday moves compensate for. The ceiling is a legal parameter, not a recommended operating point.
Why is withdrawal speed relevant to a rates-note-driven trade?
Because the ability to rotate capital into a follow-on position depends on liquidity availability, not on account balance. Instant withdrawal from Exness and near-instant from FBS mean that closed-position equity is redeployable within hours. Three-day withdrawal from AvaTrade or FXTM means the same equity is functionally frozen for the window in which the next RBA-driven move may occur. Over a hiking cycle, the compounding difference is material.
Do the pro-account spreads on the grounded brokers apply to all clients who deposit?
No. Pro-tier access is gated by minimum deposit or trading volume thresholds specific to each broker. The 0.1 pip pro spread on Exness, the 0.0 pip pro spread on FBS and HF Markets, and the 0.1 pip pro spread on FXTM are the numbers quoted in marketing material but are not the numbers the majority of retail accounts actually receive. The standard column — 0.7 to 1.5 pips depending on broker — is the realistic execution cost for a typical retail deposit.
If a rates desk note conflicts with a retail broker's account terms, which one governs the trade?
The account terms. The Rabobank note is analytical input; the broker's terms of service are a legal contract. When AvaTrade's terms prohibit scalping, a retail trader attempting to fade an RBA-driven move on short timeframes may find positions liquidated or the account restricted regardless of the analytical merit of the trade. The rates note describes a market opportunity. The terms describe the boundary of what the retail infrastructure will actually clear.