Advertisement IC Markets

The wick that only existed on one broker

Discuss 1-minute to 15-minute price action setups, fading intraday momentum, key support/resistance zones, and proven short-term trading methodologies.
Post Reply
Shadow Trader
Posts: 5594
Joined: Wed Sep 09, 2026 5:19 pm

The wick that only existed on one broker

Post by Shadow Trader »

A trader keeps a demo account on one broker and a live account on another, running the same strategy on USD/CAD. One afternoon the live chart prints a nasty 18 pip wick down through a support level, triggering a stop, and then recovers completely within the same minute. The demo chart shows nothing unusual. Just a calm candle that held the level.

The first instinct is to blame the live broker for hunting stops. That is a heavy accusation, and the evidence rarely supports it. Checking a few things tells a different story.

First, the wick was visible in the bid series only, and the ask series had a much smaller dip. That points to a quote problem on one side, rather than a real sale at a lower price.

Second, other pairs with CAD on the same live feed showed a small blip at the same second, while the USD side pairs were calm. A single liquidity source feeding CAD quotes had a momentary fault, or one provider briefly sent a bad price.

Third, the tick count in that minute was ordinary. A real 18 pip fall normally comes with a burst of updates.

Here comes the twist. When the broker's support looked into it, the explanation was not a stop hunt or a fat finger. A price from one provider outside normal tolerance had been accepted for a fraction of a second before the filter caught it. Outcomes of such events differ by broker: some cancel or adjust trades done on obviously erroneous prices, and some do not, depending on their terms.

The lesson for the trader was not the broker choice but the setup. The stop was a plain market stop sitting just under a visible level on a thin-ish minute. Moving it a little away from the obvious level, or using a time-based confirmation before exiting, would have avoided the fill.

Does it help to keep a second feed running for comparison, and has anyone successfully gotten a trade adjusted after a one-feed spike?
Recommended broker for automated trading & scalping IC Markets
FXS Prop Desk
Posts: 1689
Joined: Wed Oct 07, 2026 9:52 am

Re: The wick that only existed on one broker

Post by FXS Prop Desk »

Shadow Trader wrote: First, the wick was visible in the bid series only, and the ask series had a much smaller dip.
The bid-only wick is the telling detail. When the ask barely dips and only one feed prints the spike, a quote error from a thin CAD liquidity source fits better than a stop hunt. Sending the broker the timestamp, both bid and ask prices, and asking whether the fill can be reviewed is a reasonable next step.
FXS Official Team | The Idea Lab & Strategy
PTScalper
Site Admin
Posts: 7821
Joined: Mon Jul 20, 2026 1:28 pm

Re: The wick that only existed on one broker

Post by PTScalper »

Shadow Trader wrote: Sat Oct 31, 2026 11:47 am A trader keeps a demo account on one broker and a live account on another, running the same strategy on USD/CAD. One afternoon the live chart prints a nasty 18 pip wick down through a support level, triggering a stop, and then recovers completely within the same minute. The demo chart shows nothing unusual. Just a calm candle that held the level.

The first instinct is to blame the live broker for hunting stops. That is a heavy accusation, and the evidence rarely supports it. Checking a few things tells a different story.

First, the wick was visible in the bid series only, and the ask series had a much smaller dip. That points to a quote problem on one side, rather than a real sale at a lower price.

Second, other pairs with CAD on the same live feed showed a small blip at the same second, while the USD side pairs were calm. A single liquidity source feeding CAD quotes had a momentary fault, or one provider briefly sent a bad price.

Third, the tick count in that minute was ordinary. A real 18 pip fall normally comes with a burst of updates.

Here comes the twist. When the broker's support looked into it, the explanation was not a stop hunt or a fat finger. A price from one provider outside normal tolerance had been accepted for a fraction of a second before the filter caught it. Outcomes of such events differ by broker: some cancel or adjust trades done on obviously erroneous prices, and some do not, depending on their terms.

The lesson for the trader was not the broker choice but the setup. The stop was a plain market stop sitting just under a visible level on a thin-ish minute. Moving it a little away from the obvious level, or using a time-based confirmation before exiting, would have avoided the fill.

Does it help to keep a second feed running for comparison, and has anyone successfully gotten a trade adjusted after a one-feed spike?
Hi Shadow Trader,

Great forensic breakdown. Nine times out of ten, when a trader sees an isolated spike like this, the thread immediately turns into a "broker stop-hunting" rant without anyone checking the actual tick mechanics. Checking Bid vs. Ask separation, cross-pair triangulation (CAD crosses vs. USD crosses), and tick density is the exact checklist an execution desk uses to separate a bad tick from genuine order flow.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
PTScalper
Site Admin
Posts: 7821
Joined: Mon Jul 20, 2026 1:28 pm

Re: The wick that only existed on one broker

Post by PTScalper »

A few deeper layers on what happened under the hood—and direct answers to both of your questions:

Why Demo vs. Live Diverged (and What the Aggregator Missed)

Demo feeds are rarely true liquidity mirrors: Even on the same broker, demo servers typically run on smoothed composite feeds or isolated bridge plugins rather than live Tier-1/Non-Bank LP aggregation pools. Comparing a demo chart to a live ECN feed will almost always show cleaner wicks on demo because demo bridges filter out micro-spikes aggressively.

One-sided quote leakage in the bridge: In a multi-LP aggregator (PrimeXM, oneZero, Integral, etc.), the engine builds a composite Top-of-Book from several liquidity providers. When a single LP’s pricing engine hiccups on a currency matrix—say, an internal hedge calculation on CAD goes stale or throws a bad Bid for 50–150 milliseconds—the aggregator’s spike filter is supposed to quarantine that quote before publishing it. When one slips through inside the tolerance threshold for a fraction of a second, the Bid drops 18 pips while the Ask stays anchored, creating a momentary synthetic spread blowout.

Why your stop got filled: A standard broker-side Stop-Loss on a long position is simply a conditional market order triggered by Bid <= Stop_Price. The microsecond that rogue Bid hit the book, the matching engine converted your stop into a market sell and matched it against the only Bid available—the bad one—before the filter clamped the feed back to normal.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
PTScalper
Site Admin
Posts: 7821
Joined: Mon Jul 20, 2026 1:28 pm

Re: The wick that only existed on one broker

Post by PTScalper »

1. Does it help to keep a second feed running for comparison?

Yes—not for real-time manual intervention, but as indispensable audit evidence.

You cannot manually react to an 80-millisecond spike, and a second chart on your screen won't stop a server-side market stop from firing. Where a second feed pays for itself is post-trade verification and formal disputes:

Live-to-Live only: Never use a demo account as your reference feed. Keep a small live account open at a reputable, institutional-grade ECN venue (or a broker known for clean raw feeds like LMAX, Pepperstone, or IC Markets) running on a VPS with UTC-synced time.

Log raw ticks, not just candles: M1 OHLC bars hide the timing and spread geometry. Running a lightweight background tick-logger script/EA on your secondary live account gives you timestamped Bid / Ask records down to the millisecond. When you can show a support desk that 4 other Tier-1 feeds were quoting 1.3850 / 1.3851 at the exact millisecond their feed printed 1.3832 / 1.3851, the conversation changes immediately.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
PTScalper
Site Admin
Posts: 7821
Joined: Mon Jul 20, 2026 1:28 pm

Re: The wick that only existed on one broker

Post by PTScalper »

2. Has anyone successfully gotten a trade adjusted after a one-feed spike?

Yes, quite regularly—provided the spike is demonstrably off-market and you present clean data.

How smoothly it goes depends heavily on the broker's execution model and regulatory jurisdiction:

True STP / Agency ECN Brokers: When an LP sends a manifestly erroneous quote outside the composite market range, the broker doesn't want to keep that bad fill on their books either—their prime broker or LP will usually bust or adjust the leg under the LP's Off-Market / Manifest Error clause. If you submit a ticket promptly with the order ID, exact server timestamp, and comparative tick proof, reputable ECN desks will typically either reinstate the position at your original parameters or credit the exact pip difference back to your balance.

Market Makers / B-Book Desks: It is more of an uphill battle, though still possible with regulated brokers (FCA, ASIC, CySEC, FINMA). Most retail Terms of Business include a "Manifest Error" clause—usually written to protect the broker from paying out profits on bad ticks—but compliance teams generally apply it symmetrically if you keep the dispute strictly objective rather than accusatory.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
PTScalper
Site Admin
Posts: 7821
Joined: Mon Jul 20, 2026 1:28 pm

Re: The wick that only existed on one broker

Post by PTScalper »

What works best when filing a dispute ticket:

Skip words like "manipulation" or "stop hunting"—those get routed to generic tier-1 support templates.

Use execution-desk terminology: "Requesting an execution audit on Ticket #XXXX for an off-market / single-LP bad tick outside composite market pricing."

Attach three things: your broker's tick export showing the one-sided spread blowout, a secondary live ECN feed's tick/bar data at the exact same UTC timestamp, and the cross-pair comparison showing USD legs were flat.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
PTScalper
Site Admin
Posts: 7821
Joined: Mon Jul 20, 2026 1:28 pm

Re: The wick that only existed on one broker

Post by PTScalper »

Practical Execution Tweaks to Prevent It Next Time

Beyond moving stops slightly further away from obvious structural liquidity pools, two structural adjustments eliminate almost all vulnerability to sub-second feed glitches:

Spread-Guard Filter on Algorithmic / Soft Stops: If you manage exits via an EA or cBot, never trigger a market exit if Ask - Bid exceeds a maximum spread threshold (for example, if USD/CAD normally trades at 0.3–0.6 pips and suddenly widens beyond 2.5 pips, the EA pauses exit execution until the spread normalizes). That single line of logic makes your strategy 100% immune to one-sided Bid spikes.

Two-Tier Stop Architecture (Hard Catastrophe + Soft Tactical): Keep a hard broker-side SL 25–30 pips away purely as insurance against black-swan events, flash crashes, or VPS disconnects, while handling your actual technical exit via a spread-filtered, time-buffered soft stop (e.g., requiring price to hold beyond the level for 2–3 seconds or close an M1 candle) closer to structure.
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
Post Reply