Charting on one feed and trading on another
Posted: Mon Nov 02, 2026 9:01 am
Charting on one feed and trading on another
A level looks perfect on a charting website. Execution happens at a broker whose feed prints the same level half a pip off. The trade enters a bit late, exits a bit early, and the numbers never quite match the backtest.
Many brokers draw charts from their own bid feed. Charting services use a consolidated blend from several banks or a different aggregator. In quiet markets the gap is a tenth of a pip. In busy ones it can reach half a pip or more.
With targets near two pips, half a pip is a quarter of the whole trade.
Charting on the same feed that executes removes the issue. If a separate tool is needed, comparing its prices with the broker's at several points in the day shows how far apart they really drift.
A simple way to check is opening the same pair on the broker chart and the outside chart, then comparing the highs and lows of ten random one minute bars during a busy session. If the differences average more than a tenth of a pip, setups drawn from the outside chart need a buffer.
Has a level ever looked perfect on one chart and never quite worked at the broker?
A level looks perfect on a charting website. Execution happens at a broker whose feed prints the same level half a pip off. The trade enters a bit late, exits a bit early, and the numbers never quite match the backtest.
Many brokers draw charts from their own bid feed. Charting services use a consolidated blend from several banks or a different aggregator. In quiet markets the gap is a tenth of a pip. In busy ones it can reach half a pip or more.
With targets near two pips, half a pip is a quarter of the whole trade.
Charting on the same feed that executes removes the issue. If a separate tool is needed, comparing its prices with the broker's at several points in the day shows how far apart they really drift.
A simple way to check is opening the same pair on the broker chart and the outside chart, then comparing the highs and lows of ten random one minute bars during a busy session. If the differences average more than a tenth of a pip, setups drawn from the outside chart need a buffer.
Has a level ever looked perfect on one chart and never quite worked at the broker?