Start with a hypothesis
A strategy should begin with a reason to expect an edge — for example, "trends in XRP tend to persist for weeks after a breakout" — rather than with a search through indicator combinations. A clear hypothesis tells you what to test and what result would show you are wrong.
Make every rule precise
A complete strategy specifies, without ambiguity:
- Market and timeframe — which pair, which candles, which session boundaries.
- Entry — the exact condition, evaluated on closed candles.
- Exit — targets, trailing rules or time-based exits.
- Stop — where the idea is wrong and the position must close.
- Sizing — how much to risk per trade.
- Filters — conditions when the strategy should not trade, such as a volatile regime.
An educational example
The following is a template for learning the structure, not a recommendation: go long when the daily close is above the 50-day SMA and the 20-day SMA is above the 50-day; exit when the close falls below the 50-day SMA; place an initial stop two ATRs below entry; risk 1% of equity per trade. Each part is testable, and each part can fail.
Resist curve-fitting
Every additional rule and parameter makes it easier to fit the past and harder to succeed in the future. Prefer few, simple rules with a clear rationale. Decide your evaluation criteria — minimum number of trades, acceptable drawdown, benchmark comparison — before you look at results, so you cannot move the goalposts afterwards.
Version and invalidate
Treat a strategy like software: give each change a version and record why it was made. Also write down in advance what would make you stop using it, such as a drawdown beyond anything seen in testing. A strategy without an invalidation rule tends to be abandoned at the worst moment or kept long after it stopped working.
A testing sequence
Move an idea through increasingly demanding stages, stopping as soon as it fails one:
- Write the hypothesis and the precise rules.
- Backtest on in-sample data with fees and slippage.
- Validate on out-of-sample data you have not looked at.
- Replay history bar by bar to see how the signals feel in real time.
- Paper trade forward for a meaningful number of trades.
- Review against the criteria you set before starting.
Key takeaways
- Begin with a hypothesis, not an indicator search.
- Define entry, exit, stop, sizing and filters precisely.
- Fewer rules generalise better.
- Set evaluation criteria before seeing results.
- Version every change and define invalidation in advance.