Check a stop-loss execution — records, quotes and timing

A triggered stop followed by a reversal does not prove broker manipulation. Inspect the executable quote and order record, and distinguish ordinary movement, clustered order flow and potentially unfair execution without assuming their prevalence.

Does your broker even profit from your loss?

An agency or routed account can earn spread and commission while having other remuneration or hedging exposures. An ECN/STP label alone does not establish that client outcomes never affect the provider. Check the contractual role and execution policy.

A single stop being triggered does not establish manipulation. Compare the executable side of the quote, spread and timestamp with the order record. The distinction between agency and market-maker execution also matters: spread or commission income alone does not establish the broker's full economic exposure. Check its principal/agency and hedging disclosures; a pricing label is not proof that client outcomes never affect its book.

If it is not the broker, why are stop runs real?

Activated stops can contribute to a cascade when their market orders consume available offers and further trigger conditions are reached. Osler’s historical currency research documents possible feedback, rather than every textbook giving identical advice or every institution steering prices to a known cluster. OTC platform depth is not a global stop map; available offers can replenish or withdraw. A price path alone cannot identify who intended what.

Participants may infer possible order concentration from visible levels, but access to order information depends on their role and venue. Do not assume every bank or fund has no visibility of any client orders. Related reading: round numbers as price magnets, mechanics of stop hunting.

The BIS survey from 2022 reports aggregate OTC FX turnover for April 2022. Its market-scale data cannot determine whether a provider knew or acted on a particular client’s stop. A server-side stop is known to the system responsible for it; a market total does not prove that an individual account is invisible or cannot experience unfair execution.

What about a market maker on a B-book?

A dealer retaining unhedged principal exposure can gain on the corresponding position when the client loses, before considering other positions, costs and hedging. Some or all risk may be offset. Counterparty exposure can create a conflict, but does not itself establish fraud or the net profitability of every client loss. Read the specific contractual role and policy.

EU investment firms have best-execution and conflict-management duties. Unfair asymmetric execution can breach those duties. A difference from another OTC quote is not automatically manipulation: compare matched executable prices, timestamps, size and the documented policy. Legal framework: MiFID II.

Supervision does not establish that violations cannot occur. In its 1 March 2023 notice, Poland’s KNF records the rejection of XTB’s cassation appeal against its 2018 PLN 9.9 million penalty. The described conduct included asymmetric deviation settings in instant execution from 2014 to 2015. This is a documented historical execution case, not evidence of artificial wicks, current use of that practice or any particular stop’s error. An unexplained quote discrepancy requires comparable records and a complaint, rather than an immediate universal verdict. See how to spot a scam broker.

So why does your stop really get hit?

First identify what closed the position: stop loss, margin close-out, manual action, automation, expiry or another product condition. Then separate trigger validation from execution-price validation. In typical MetaTrader FX, a long’s stop condition uses bid and a short’s uses ask; exchange products may use different triggering conventions. A chart may show bid or last rather than the relevant ask.

In a hypothetical short EUR/USD position, stop is 1.10000 and bid is 1.09980. At a 0.2-pip spread, ask is 1.09982. If bid is unchanged and spread grows to 2 pips, ask becomes 1.10000 and can meet the assumed stop condition. This deliberately asymmetric widening is not a claim about an actual news release or a guarantee that spreads widen around a fixed midpoint. Triggering still does not establish the eventual fill.

Obtain order and position IDs, full symbol, direction, volume, stop modifications, server confirmations, relevant bid/ask ticks, fill price and applicable policy. Minute OHLC can hide tick order. Do not reconstruct historical ask from today’s spread or assume the current ask line contains its full history. Match product, date, timezone and executable side; futures, spot and CFD may have different prices and sessions. See the mechanics of the OTC market for the data scope.

Myth or fact — an honest verdict

Broker misconduct and liquidity-driven stop runs are different explanations requiring evidence. Neither regulation nor an ECN label proves that misconduct cannot occur, and this article provides no measured prevalence estimate.

Timing can change the conclusion even with identical OHLC. In an invented long-position example, the server confirms a newly added stop at 1.09900 at 09:00:40 while bid is 1.09950; no earlier stop or other closure cause exists. Both model bars have open 1.10000, high 1.10020, low 1.09850 and close 1.09950. In A the low occurs at 09:00:10, before the stop, and later prices stay above it. In B the low occurs at 09:00:50, after confirmation, so the stop’s condition is reached. The shared wick cannot distinguish the two sequences or guarantee a fill price.

Keep original timestamps and offsets alongside converted UTC, preserving precision and confirmation times. A click in the application is not necessarily server acceptance. Do not replace a server stop with a “mental” exit merely to hide it: manual action depends on connection, presence and decisions under pressure. A price alert is not a closing order. Investigate execution and control risk before taking another trade to recover a loss.

What to do tomorrow

1. Read the actual execution policy. Identify the legal counterparty, routing, hedging and trigger conventions; ECN or STP keywords alone do not establish the whole economic arrangement. The choosing a broker section on forexmechanics.com discusses related comparisons.
2. Reconstruct the disputed event. Match order records, confirmations and appropriate quote-side ticks at agreed timestamps. A selected three-trade sample does not establish a typical cause or manipulation rate. Ask the provider to explain missing or inconsistent records.
3. Keep protection tied to a tested plan. A few extra pips or an ATR multiple do not prove that a stop lies outside an order cluster. Do not widen an open position’s stop simply because a loss is approaching; evaluate planned distance and sizing before entry.
4. Recalculate position size. If a wider stop is needed, reduce size to keep the intended loss within the risk budget. Wider stops do not guarantee better results or a maximum realised loss during gaps.