Inspecting Freqtrade Signals, Backtests, and Trade Statistics
Summary
This guide outlines a notebook workflow for debugging and analyzing a Freqtrade strategy. It loads historical candles for a selected pair and timeframe, runs the strategy to inspect generated entry signals, and explains that signal counts do not equal executed trades: startup data, trade-slot availability, and repeated conditions can all affect actual entries.
It also shows how to examine saved backtest statistics and trades, group exits by pair, chart daily cumulative profit, and review live trades loaded from a database. Further analysis includes estimating simultaneous open positions to inform the max-open-trades setting, plotting candles with indicators and trades, and viewing the distribution of trade returns. These are analysis tools rather than evidence that a strategy is profitable. Results depend on the selected data, configuration, and strategy, and signal-level inspection alone does not reproduce full backtest behavior.
Key ideas
- Inspect recent candle rows after strategy analysis because indicators may need startup data.
- Generated entry signals do not necessarily become trades because trade slots and repeated conditions affect execution.
- Backtest statistics and trade records support pair, exit-reason, drawdown, and equity analysis.
- Trade-parallelism analysis can help assess a suitable limit on simultaneous positions.
- Charts and return distributions aid diagnosis but do not establish future profitability.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.