Changing timeframe changes the bars and historical data the strategy calculates from, so signals, simulated fills, trade count, and performance can change. Compare backtests only on like-for-like settings.
What you may be seeing
- A strategy is profitable on one timeframe and unprofitable on another
- Trade count changes when you switch chart interval
- Entries or exits appear on different bars after changing timeframe
Most likely causes
- The strategy processes different chart bars on each timeframe
- Signal conditions can occur at different times with different OHLC data
- The amount and starting point of available historical data can differ
- Historical fill assumptions depend on the available chart data
Try this first
- Choose the timeframe the strategy is actually designed to trade
- Record the symbol, timeframe, date range, strategy inputs, commission, and slippage
- Compare results only after keeping those variables fixed
- Treat another timeframe as a separate robustness test rather than expecting identical results
If that didn’t fix it
- Use Bar Magnifier or higher historical detail when intrabar fill precision is important and your account supports it
- Export Strategy Report results before changing the timeframe so you keep a comparable baseline
Verify the fix
Repeated tests on the same symbol, timeframe, settings, costs, and date range reproduce the same intended baseline closely enough for comparison.
Why this happens
A TradingView strategy calculates from the chart data available to it, so changing the chart timeframe changes the bars the strategy processes. A 5-minute backtest and a 1-hour backtest are therefore not the same dataset sampled at different visual zoom levels.
The broker emulator also uses the available chart data to simulate historical fills by default. Changing the timeframe can change signal timing, OHLC values used by the strategy, the amount of available history, and the assumptions used for fills. Compare results only after keeping the symbol, timeframe, strategy settings, costs, and testing range consistent.