There is a pattern this desk keeps seeing whenever AUD/USD prints above 0.7000 on a geopolitical risk-on headline — an Iran deal, a Fed pivot rumor, a tariff pause. The Reuters wire lights up, the pair gaps, and within ninety seconds a specific type of retail trader posts a screenshot: "0.0 pip spread on EUR/USD at Exness Pro, why isn't AUD/USD the same?" The answer sits four layers deep in the pricing stack. We are going to decompose one specific spread, on one specific pair, at one specific moment — the Sunday 22:00 GMT reopen following a risk-on Friday close — and show what the marketing number leaves out.
The Pattern: Headline Risk-On Moves Compress the Wrong Number
Let us concede something upfront. The marketing numbers are not lies. When Exness quotes a 0.1 pip average on EUR/USD Pro accounts, that figure is defensible under the sampling methodology brokers use — time-weighted, London session, one-lot notional, main venue liquidity intact. The advertised spread on AUD/USD is similarly defensible under the same conditions. We are not writing a broker-fraud piece. We are writing about what happens when the reader stops trading at the sampled hour and starts trading at a moment the sampling deliberately excludes.
Every time AUD/USD gaps through 0.7000 on a risk-sentiment headline, the same type of retail trader asks the same question in the same forum thread. Why did my stop skip? Why did the spread balloon to 4 pips when the site advertises 0.9? Why did the "instant" execution take 800 milliseconds? The answers are not fraud. They are the pricing model doing exactly what the pricing model is designed to do — reprice the risk of quoting into a market where the counterparty is asymmetric.
Here is the observation across cases. The word "spread" as retail traders use it is a single number. The word "spread" as an interbank prime-of-prime desk uses it is a stack of four numbers plus a time-of-day multiplier plus a headline flag. When a broker advertises 0.9 pips average on AUD/USD, they are averaging across the flat hours. When you actually trade the reopen after Friday's Reuters headline about a US-Iran de-escalation, you are trading a moment that is definitionally excluded from the average. The advertised number was accurate. It was also irrelevant to your fill.
The Cheap Spread Fallacy: What 0.0 Pips Actually Means at 22:00 GMT Sunday
OK so here's where it gets really interesting, and I love this detail, so let me walk through it slowly. The 0.0 pip and 0.1 pip pro-account marketing figures — Exness quotes 0.1 on EUR/USD Pro, FBS quotes 0.0 on their pro tier, HF Markets quotes 0.0, FXTM quotes 0.1 — are all describing the same accounting move. It is not a claim about the market. It is a claim about where the broker has decided to book the cost. The cost has not disappeared. It has been shifted to a different line item.
The historical version of this argument runs through the commission-plus-raw-spread model that ECN venues built in the mid-2000s. On that model the broker charges — say — $3.50 per side per lot as an explicit commission and passes through the raw interbank spread. On a lot of AUD/USD the raw interbank spread at London open on a quiet Tuesday might be 0.2 pips. Add the commission converted into pip-equivalent and you get a total cost of roughly 0.9 pips round-trip. The commission is visible on the trade blotter. The reader can reconcile it. The reader can also decide, at 22:00 GMT Sunday after an Iran headline, that they do not want to trade because the raw spread has widened to 3 pips and the commission is still $3.50.
Now flip to the zero-commission-spread-markup model that came to dominate retail marketing after roughly 2015. The advertised number goes to 0.0 or 0.1 pips because the broker is subsidizing the cost during the sampling window — buying attention with a headline figure. The cost is recouped in three places: swap markup on overnight positions, execution latency during volatile moments, and — most importantly — spread widening during the exact moments the sample deliberately excludes. The 0.1 pip average holds during the 87% of trading time that is boring. During the other 13% — which contains most of the moments the reader actually wants to trade — the spread reprices to whatever the venue liquidity permits.
The concession is real. Exness's 0.1 pip Pro spread on EUR/USD is genuinely the tightest headline figure in this comparison set — FBS matches it at 0.0 on their raw tier, HF Markets matches at 0.0, FXTM matches at 0.1. The teardown is that the headline number was chosen to win the comparison-shopping click. It was not chosen to describe the moment you are actually trading. When AUD/USD gaps through 0.7000 on a risk-on headline, the number that priced your fill was calculated four layers deep, and none of those layers appear on the broker's homepage.
The Liquidity Premium Nobody Prices Into an Iran-Headline Gap
Here is the second pattern, and it is the one that most cleanly separates traders who have looked at this from traders who have not. The Sunday 22:00 GMT reopen — the moment the interbank tape starts printing again after the New York close on Friday and the Asia-Pacific desks come online — is the single most mispriced hour of the retail trading week. Not mispriced in the market's sense. Mispriced in the reader's mental model.
Interbank liquidity providers do not reset instantaneously at 22:00 GMT Sunday. The tier-1 banks that quote into ECN venues run their weekend risk books through a specific unwind sequence. Sydney desks open first, Tokyo desks about an hour later, and Singapore adds a third quote stream around 23:30 GMT. Until all three are live and matching each other's quotes within a tolerance band, the venue's effective spread is set by whichever desk is quoting widest — which is almost always the one carrying the largest weekend gap-risk position. AUD/USD is particularly exposed to this because the pair's natural liquidity center is the Sydney desk, which is also the desk pricing the weekend Iran-headline risk into its Monday book.
The premium that gets stacked into the spread during this reopen has four components. Raw interbank spread expands from its Tuesday-London floor because fewer venues are quoting. Broker markup on the zero-commission model expands because the broker's own risk desk is hedging against the possibility that the retail flow it is warehousing turns out to be informed rather than noise. Liquidity provider premium expands because the LP is quoting into a market where the next tick could be a headline update from Reuters or Bloomberg that repositions the pair by 40 pips before the LP can pull the quote. And volatility premium — the implied-vol input that the LP's pricing engine uses to widen the quote — expands mechanically because the option market opened Sunday with elevated implied vol on AUD/JPY and AUD/USD cross rates.
Add those four components and the 0.9 pip advertised spread on a standard AUD/USD account becomes, at 22:03 GMT Sunday, something between 2.5 and 5 pips depending on how many LPs are actually quoting and how the broker's smart-routing engine has been configured for the session. This is not a broker choosing to gouge the reader. This is the pricing model reflecting what the market will actually pay to warehouse the risk of quoting into a headline-driven gap.
The advertised spread is what the broker charges when the market is boring. The real spread is what the broker charges when you actually want to trade.
The Commission Model That Made This Legible — And Why Retail Abandoned It
The third pattern is historical, and it is the one this desk finds genuinely interesting because it explains why the current retail marketing environment is what it is. Before roughly 2015, the dominant model for spread advertising among ECN-style brokers was transparent commission plus raw spread. The trader saw two numbers on the blotter — a commission line and a spread line — and could reconcile them against the venue's own tape. Pepperstone's standard commission-model account and IC Markets's standard commission-model account are the two survivors of that era in this comparison set. XM and Exness both offer zero-commission variants that quote the tighter marketing figure.
The advantage of the commission-plus-raw model was legibility. When AUD/USD's raw spread went from 0.2 pips to 3 pips during a Sunday reopen, the trader saw the widening on the tape. The commission line stayed at $3.50. The trader could decide — in real time, with a visible price — whether the moment was worth trading. The disadvantage, from the broker's marketing perspective, was that "0.9 pips total including commission" reads worse in a comparison-shopping table than "0.1 pips average" — even though the two describe substantially the same cost structure at moderate volumes.
The pivot to the zero-commission-spread-markup model was not driven by traders demanding it. It was driven by affiliate-comparison sites ranking brokers by the headline spread number, and by the mechanics of Google search rewarding pages that could truthfully claim "lowest spread." The commission model lost the marketing war for the same reason airlines that advertise the ticket price separately from the baggage fee lose to airlines that bundle both into an opaque "total" — the bundled number wins the comparison click even when the unbundled price is objectively better at the shopper's actual usage pattern.
At retail volume — a few lots per day — the difference between the two models is genuinely small. The commission is not a hidden tax; it is roughly what the spread markup would have been, unbundled. At institutional volume the transparent commission wins by a wide margin because the commission scales linearly while the spread markup compounds with size and volatility. This is why the commission model survived at the institutional tier — where the reader is negotiating volume rebates against a visible commission schedule — and died at the retail tier, where the reader is comparing headline numbers on a listicle.
So What Do You Actually Do
The specific question this piece opened with — why is the advertised spread not the fill you got on the AUD/USD gap through 0.7000 — has a specific answer. You did not get the advertised spread because the advertised spread was calculated during hours that structurally exclude the moment you traded. This is not a claim about broker misconduct. It is a claim about what the advertised number is describing. If you want your fill to match your expectation, you need to price the moment, not the average.
Practically, three moves. If you trade Sunday reopens or headline-driven gaps regularly, use a commission-model account — Pepperstone standard or IC Markets standard from this comparison set — because the raw spread on the tape will show you the widening in real time and let you decide whether the trade is worth the cost. If you trade primarily during London and New York overlap hours on major pairs, the zero-commission Pro accounts (Exness at 0.1, FBS at 0.0, HF Markets at 0.0, FXTM at 0.1) genuinely deliver something close to the advertised figure. The two models are not better or worse. They are optimized for different sessions and different sizes.
Second — reconcile your actual fill data against the advertised average once a quarter. Every broker's blotter exports the spread at fill for every trade. Pull the CSV. Compute your own average. Compare it to the marketing figure. The gap between the two is the cost of trading the moments the sample excludes. If you find the gap is small, the marketing was honest for your usage pattern. If the gap is 3x or 5x, the marketing was describing a trader you are not.
Third — accept that the moment AUD/USD gaps through 0.7000 on a risk-on headline is a moment the broker has priced in a way you cannot see. Retail execution during high-headline hours is a different product than retail execution during the London morning, and the price difference is real. This piece does not cover swap-rate markup on carry-trade positions held through the Sunday rollover — which is a substantial secondary cost on AUD/USD specifically. It does not cover the mechanics of last-look rejection at the LP level, which is a separate and larger argument. And it does not cover the tax treatment of forex trading gains under any specific jurisdiction — that varies too much to responsibly generalize. Each of those is a separate teardown.
FAQ
Why did my AUD/USD stop-loss skip through my level during a risk-on gap?
Stop-loss orders are filled at the next available price when your level is touched, not at your level. During a Sunday reopen or a headline-driven gap, the next available price after 0.7000 might be 0.7008 — because there was no quote between them. The broker did not skip your stop. The venue did not print a price at your level. Commission-model accounts show this on the raw tape; zero-commission accounts absorb the gap into a wider spread but the mechanics are identical.
Is Exness Pro at 0.1 pips genuinely cheaper than Pepperstone standard at commission plus raw?
At retail volume — say, 5 lots per day — the two are within a fraction of a pip of each other on average. Exness Pro wins during the sampling hours the marketing figure describes. Pepperstone standard wins during the excluded hours because the commission does not widen with volatility — only the raw spread does, and it is visible. The honest comparison depends on when you trade, not on which headline number is smaller.
What is the raw interbank spread on AUD/USD outside marketing samples?
The raw interbank spread on AUD/USD floats between roughly 0.2 pips during peak London-New York overlap and 3-plus pips during Sunday reopen or immediately after a Reuters headline. The advertised average — Exness at 1.0 pips standard, AvaTrade at 0.9, HF Markets at 1.2, FXTM at 1.5 on standard accounts — is a time-weighted figure across the flat hours. It is accurate to its methodology and misleading to reader expectation.
Does the FCA or ASIC regulate how brokers sample the advertised spread?
The FCA and ASIC both require that advertised spreads be "typical" and disclose the sampling methodology if requested. They do not standardize the sampling window across brokers. Exness (FCA authorized), AvaTrade (ASIC authorized), FBS (ASIC authorized), and HF Markets (FCA authorized) each publish their own methodology. The regulator polices honest disclosure; it does not police whether the sampled hours reflect the hours the trader actually uses.
Should high-frequency traders use commission or zero-commission accounts?
At high frequency the commission model almost always wins. The zero-commission markup compounds across trades because each trade is subsidizing the marketing number, and volume amplifies the drag. IC Markets standard and Pepperstone standard were designed for this usage pattern and survived the industry pivot to zero-commission for precisely this reason. The zero-commission Pro tiers work best for readers holding positions for hours or days rather than minutes.
Why do Islamic accounts often show slightly wider spreads?
Islamic (swap-free) accounts — offered by AvaTrade, Exness, FBS, FXTM, and HF Markets in this comparison set — replace the overnight swap charge with either an administration fee or a marginally wider spread. This is a compliance mechanism, not a markup on Muslim traders. The broker cannot charge or pay interest on positions held overnight, so the interest-equivalent cost is repositioned into the spread or fee line. The total cost is comparable to a standard account for positions held less than a few days.
How much of the spread on a Sunday reopen is broker markup versus interbank widening?
Roughly speaking — and this varies by broker — during a Sunday reopen the raw interbank component might account for 40-60% of the total spread widening, and the broker's own risk markup accounts for the rest. The broker markup is not gouging; it is the broker's risk desk hedging against the possibility that early Sunday retail flow is informed by a weekend headline the broker has not yet priced. A commission-model account lets you see the interbank component directly on the tape.