There is a pattern we have watched on GBP/JPY for years now, and it shows up again whenever the cross drifts back toward a September low that was already tested, already defended, and already broken once before the chart decided to retest it. Readers write in asking which broker to use for the retest. The honest answer sits in the paperwork: of the firms a reader might pick today, the youngest in our citable set is HF Markets, founded in 2010, and the oldest is AvaTrade, founded in 2006. That gap matters more than the pip count on any single tick.
The Pattern We Keep Seeing Near Prior Lows
Here is the thing about a prior September low on a cross like GBP/JPY. It is not a level. It is a memory held by two different sets of participants who are not talking to each other, and the retest is the moment their silences collide. We have watched this on the tape across more cycles than we care to count, and the shape is always the same even when the macro backdrop is not.
The first set is the dealer book. The desk that sat the first break knows exactly where the stops cleared, where the next tranche of hedges got lifted, and where the pain in the inventory lives. The second set is the retail trader who read a thread last night and decided this is the level, this is the one, this is where the counter-trade finally works. Those two sets do not meet in the middle. They meet on the retest, and the retest decides which memory was the real one.
What we observe, repeatedly, is that sellers retain control for one unglamorous reason. The dealer book has been fed. The retail flow has not. When price returns to the September low after a bounce, the volume profile is almost always thinner on the upside leg than it was on the original break. Thinner volume on the recovery means the move up was positioning, not conviction. And positioning gets unwound against the next liquidity pocket, which — surprise — sits just under the prior low.
You do not need a signal service to see this. You need to look at the tape across three consecutive sessions and ask one question. Did the bounce reclaim the level with volume, or did it reclaim it on fumes? If the answer is fumes, sellers still own the chart. The September low is a floor only in retrospect; in real time it is a magnet, and the magnet is unfinished business.
Listen, I'll say it the way I'd say it to a friend who texted me at 2 AM asking what to do. The retest is not the trade. The retest is the question. The trade is what the tape answers in the forty minutes after the question gets asked. Everything else is noise, and most of what passes for GBP/JPY forecasting is noise dressed in a suit.
The Operator Memory Problem — Why Broker Age Decides Who You Argue With
There is a second pattern we see at these levels, and it has nothing to do with the chart. It is a pattern in who the retail trader is arguing with when the position goes wrong, and the answer depends on how old their broker is. This is the piece of the trade almost nobody models before entering.
Of the brokers in our citable set, the founding dates run across a tight nine-year window — AvaTrade in 2006, Exness in 2008, FBS in 2009, HF Markets in 2010, FXTM in 2011. That looks like a small spread. It is not. Those nine years span the pre-crisis, crisis, and post-crisis reconfiguration of the retail FX industry, and a broker's operational memory is shaped by which of those three environments it opened in.
A broker that opened in 2006, under AvaTrade's incorporation window, had to build its risk model before the 2008 cascade. Its reflexes around leverage, hedging policy, and client-fund segregation are shaped by having survived something that cost competitors their licenses. That is why you see AvaTrade capped at 400x max leverage and sitting under five regulators including ASIC as its tier-1 anchor. The paperwork reflects a firm that got its scars early and priced them in permanently.
A broker that opened in 2010, under HF Markets' timeline, built its risk model inside a different world. The crisis was already a case study, not a live threat. The playbook was already written. The firm optimized for a market in which retail flow had become the dominant customer, which is why HF Markets sits at 1000x max leverage and lists 1200-plus instruments across five regulators including the FCA. Same regulatory tier at the top, but the product shape underneath is a different animal.
Why does this matter at the September low? Because when the retest fails and your stop gets hit two pips wide of where it should have been hit, the firm you are arguing with has a culture. The 2006 firm argues back through a legal team that has been there since the pre-crisis days. The 2010 firm argues back through a client-support channel that was designed in an era of mass retail onboarding. Neither is better or worse in isolation. But the trader who does not know which one they picked is going to be very surprised by the texture of that conversation.
The quietest edge in retail FX is reading the broker's founding year the way an analyst reads a 10-K. It tells you what the firm learned by force, and what it only learned by case study.
The broker you argue with at 3 AM is the broker that was built by the market conditions of the year on its founding certificate — read the year, not the brochure.
The Cross-Rate Spread Illusion on GBP/JPY
Here is a pattern that trips up almost every reader who writes in asking about GBP/JPY execution. They look at the EUR/USD spread quoted by their broker, assume it transfers across pairs, and build their position-sizing on an illusion that falls apart the moment they trade the cross.
The citable spreads we have on EUR/USD read like this: FBS at 0.7 average, AvaTrade at 0.9, Exness at 1.0 standard and 0.1 on Pro, HFM at 1.2 standard and 0.0 on Pro, FXTM at 1.5 standard and 0.1 on Pro. Those numbers are the showroom. They are what the broker displays because EUR/USD is the pair every comparison engine indexes on, and the firms know that.
GBP/JPY is not EUR/USD. It is a cross, and the mechanics of a cross-rate spread are built from the two underlying legs — GBP/USD and USD/JPY — plus whatever inventory risk the broker is carrying on the composite. In practice this means the GBP/JPY spread on a standard account is routinely two to four times the EUR/USD spread, sometimes more depending on the hour. If you are trading the September low retest at 2 AM Tokyo on a standard account, you are paying not what the brochure quoted, but what the dealer's inventory anxiety says you should pay.
The Pro accounts tell a different story. Exness at 0.1 Pro, HFM and FBS at 0.0 Pro, FXTM at 0.1 Pro — these are commission-model accounts where the spread compresses because the economics shift to a per-lot fee. For a trader actually working a defined level like the September low, the Pro spread arithmetic almost always beats the standard arithmetic once position frequency crosses a modest threshold. Readers rarely do that math because the Pro accounts have minimum deposit tiers or activity requirements that filter out the dabblers. Which, to be clear, is the point.
What this means for the retest trade is simple and almost nobody internalizes it. The move from the current price to the September low is a fixed distance in pips. Your execution cost is a variable that depends entirely on which account tier you funded, which hour you trade, and whether your broker's inventory book is net-long or net-short the cross at that moment. The pip-count forecast is the easy part. The real forecast is on your own transaction cost, and that forecast lives in the account-type paperwork, not the price chart.
The Leverage Shelf Nobody Reads Twice
The fourth pattern we see, and this is the one that costs retail traders the most money at prior-low retests, is a misunderstanding of what the leverage number on a broker's homepage actually represents. The citable numbers in our set: AvaTrade at 400x, HF Markets at 1000x, Exness and FXTM at 2000x, FBS at 3000x. Readers see those numbers as a shelf of options they can pick from. They are not. They are regulatory ceilings, and the ceiling is not the trade.
Here is what we observe when a GBP/JPY retest goes sideways. The trader who had 400x available but used 10x survives the retest. The trader who had 3000x available and used 300x — thinking they were being conservative because they used a tenth of what was offered — gets wiped on the first stop run under the September low. Both traders made the same directional call. One is still in the game next week. One is reading a termination email from their broker's compliance desk.
The leverage ceiling tells you two things, and neither is "use this much." The first is what the regulator in the brokers' top-tier license permits. AvaTrade's 400x and FBS's 3000x are not an argument about which broker is better — they are an argument about which license is primary. The FCA, ASIC and CySEC caps are nothing like what the offshore entities permit, and the broker's homepage displays the highest number in its stack because marketing works that way.
The second thing the ceiling tells you is the firm's assumption about who its customer is. A 3000x ceiling is a product designed for a client whose average position life is measured in minutes, not sessions. A 400x ceiling is a product designed for a client whose position life is measured in sessions, not weeks. The ceiling is a filter, and when you pick the wrong ceiling for your actual holding period, you are picking a broker whose entire operational muscle is calibrated for a different kind of trader. The customer-service response times, the margin-call policies, the overnight-swap economics — all of it is downstream of that assumption.
The September low retest is a multi-session trade for most readers. Which means the brokerage with the eye-catching 3000x ceiling is almost certainly the wrong vehicle for it, not because 3000x is dangerous — you can use 5x on a 3000x account — but because the firm's internal plumbing is optimized for someone the reader is not. The leverage number on the brochure is a tell about who the brochure was written for. Read the shelf, and then shop at a different shelf.
So What Do You Actually Do
Stop forecasting the pip count. Start forecasting your own transaction stack. The September low is a level the market has already shown you how it treats — tested, defended, broken, retested. The directional read is almost never where the money is lost or made at these levels. The money is lost in the forty minutes after the retest when execution, spread, slippage, and margin policy all collide at once, and most readers do not know what any of those four variables are set to on their own account until it is too late.
Here is what I would do if a friend called me tonight about this trade. First, I would open the account statement and read, in order: the account tier (standard or Pro), the stated spread on GBP/JPY specifically (not EUR/USD), the max leverage, the overnight swap rate, and the regulator listed on the client agreement. If any one of those five items surprises them — if they cannot answer each from memory before looking — the trade does not happen tonight. Not because the trade is wrong, but because the vehicle is unknown, and no amount of chart work compensates for an unknown vehicle.
Second, I would ask them to describe the broker's founding year and what regulatory regime it opened under. If the answer is "I don't know," then the real work this week is not the GBP/JPY retest — it is the paperwork review. The firm you argue with when the stop runs through your level is the firm whose institutional memory was shaped by the market it was born into. A reader who skips this homework is not trading GBP/JPY. They are trading a position inside a counterparty they have not read, and that is a different game than they think they are playing.
Whether the September low holds on this retest or breaks for a third time is a question the tape will answer in its own time, probably without warning, probably at an hour when the reader is not watching. The question we cannot answer from the archive is whether retail GBP/JPY traders at prior-low retests do better when their broker is older than they are, or when their broker is their own age. The data to settle that lives inside the brokers' own books and will never see daylight. If anyone has run that study properly, write in.
FAQ
Why do GBP/JPY retests of a prior low usually favor sellers on the second visit?
Across the cycles we have watched on the tape, the recovery leg after an initial break of a September low tends to run on thinner volume than the break itself. Thinner volume on the way up signals positioning rather than conviction, which means the long side is holding inventory it has not fully distributed. On the retest, that inventory gets re-offered into the next liquidity pocket, which almost always sits beneath the prior low rather than above it.
Does broker founding year actually affect how a GBP/JPY retest trade plays out in practice?
Yes, more than readers assume. A broker founded in 2006 like AvaTrade built its risk model inside the pre-2008 environment and carries scars that shape its leverage caps and segregation policy. A broker founded in 2010 like HF Markets built its model after the crisis was already a case study. The founding year predicts the firm's reflexes around margin calls, overnight swaps, and client-support texture — all of which matter precisely when the retest goes against you.
Is the EUR/USD spread a reliable guide to what I will pay on GBP/JPY?
No, and this is one of the most expensive mistakes we see readers make. The EUR/USD figures our citable set discloses — ranging from 0.7 at FBS to 1.5 at FXTM on standard accounts — are the showroom pair. Cross-rate spreads on GBP/JPY are built from two underlying legs plus the broker's inventory risk, and they routinely run two to four times the headline EUR/USD number depending on the hour and the dealer book's position.
How much leverage should I actually use for a GBP/JPY retest, given that offerings range from 400x to 3000x?
The ceiling is not the trade. AvaTrade's 400x and FBS's 3000x are regulatory maxima, not recommendations. For a multi-session retest trade, the usable leverage is governed by your account equity, your stop distance from the September low, and the GBP/JPY spread at your execution hour — rarely more than 10x to 20x effective, regardless of what the shelf permits. The ceiling tells you what customer the firm is optimized for, not what you should deploy.
Which account tier matters more for GBP/JPY retests — standard or Pro?
For readers actively working a defined level, the Pro tier arithmetic typically wins once trade frequency crosses a modest threshold. Exness Pro at 0.1, FBS and HFM Pro at 0.0, FXTM Pro at 0.1 — these commission-model accounts compress the spread cost that matters most on cross-rates. Standard accounts can be fine for occasional swing positions, but readers trading retests should run their own cost-per-lot math across both tiers before committing.
What regulatory tier should I prioritize when choosing a broker for GBP/JPY exposure?
Our set shows every broker holds at least one tier-1 regulator — AvaTrade and FBS under ASIC, Exness, FXTM and HFM under the FCA. The practical difference lives in the stack behind the top tier. HFM lists five regulators including the DFSA; Exness lists nine including the FCA. More secondary regulators does not mean safer, but it does mean the firm's dispute resolution has more jurisdictions in play, which can matter when a retest-related complaint drags on past the standard review window.
Can I use an Islamic account to trade the GBP/JPY retest without swap costs?
All five brokers in our citable set — AvaTrade, Exness, FBS, FXTM and HF Markets — offer Islamic accounts, which remove overnight swap charges on compliant pairs. For a multi-day retest trade this materially changes the cost structure, since the swap on GBP/JPY can be significant depending on rate differentials. The caveat is that Islamic account mechanics vary by broker and typically carry administrative fees that replace the swap, so the net savings depend on holding period.
How fast will I actually see my money back if the GBP/JPY retest trade goes right and I withdraw?
Withdrawal speeds in our set range from instant at Exness to one to three days at AvaTrade and FXTM, with HFM at one day and FBS at instant to one day. The headline figure is the broker's internal processing; the wall-clock figure adds the payment rail — wire transfers add banking days, Skrill and Neteller clear faster, card refunds depend on the issuer. The first withdrawal from a new account is routinely slower than subsequent ones because of the initial compliance review.