Measuring Historical Broker Spreads and Their Intraday Variation
Summary
SpreadAudit analyzes the spread recorded for each one-minute historical bar over a selected lookback period. It reports median, 90th-percentile, and 99th-percentile spreads in price units and pips, normalizes them against hourly ATR for comparison across instruments, and can export hourly average and maximum spreads. The example highlights that spreads can widen around daily rollover, so a strategy’s actual trading times may face costs well above typical levels.
The document explains how spread estimates can change the interpretation of a backtest: a signal’s average expected move may be smaller than the trading cost, or costs may overwhelm an otherwise real effect. It suggests timing filters and ATR-relative spread thresholds. A key limitation is that the platform’s historical spread field records the minimum spread within each bar, making the calculated costs an optimistic lower bound. Some brokers may provide empty or zero spread data, which also limits the analysis.
Key ideas
- The tool summarizes historical one-minute spreads using percentiles and hourly statistics.
- It scales spread measures by hourly ATR to support comparisons across instruments.
- Spreads can rise sharply around rollover and session openings, affecting strategy costs.
- Historical bar spreads are minimum values, so the estimates understate actual costs.
- Missing or zero spread history can prevent meaningful analysis.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.