Fast answer
Crypto signal backtesting needs assumptions, bias checks, and live-forward records.
Before accepting a crypto signal backtest, record the market, date range, data source, entry and exit rules, fee model, slippage assumption, liquidity filter, rule changes, out-of-sample period, and whether the same method has live-forward signal records.
If a backtest shows smooth historical returns without data source, costs, slippage, rule history, out-of-sample checks, and live-forward proof, do not treat it as verified signal evidence.
Backtest checks
What to inspect in crypto signal backtesting claims.
Data and period
The test should state exchange, symbols, date range, candle or tick source, excluded markets, and whether delisted or illiquid assets were removed.
Fees and slippage
Crypto execution costs can change the result. Check taker fees, maker fees, spreads, funding, price impact, and missed-fill assumptions.
Bias and curve fitting
Watch for optimized rules chosen after the fact, look-ahead data, survivorship bias, and tests that only show the most favorable market regime.
Live-forward link
A backtest is stronger when the same rules are preserved before live alerts and later compared against timestamped signal outcomes.
Source context
Backtesting can test rules, but it does not prove future execution.
Investopedia explains backtesting as applying a trading strategy to historical data and notes that useful tests should account for realistic market conditions and trading costs. For crypto signals, historical tests still need live-forward evidence before trust increases.
Review standard
A reviewable backtest preserves rules before results are known.
For CSR evidence review, crypto signal backtesting claims should preserve the original rule set, source data, test period, asset universe, cost model, slippage model, out-of-sample results, rule-change log, and live-forward comparison. Historical fit is not the same as provider proof.