Backtest an event-driven strategy.
8/18/2025 06:01am
To set up and run an event-driven back-test, I need to know exactly what should constitute an “event” and how trades should be opened and closed when that event occurs.
1. Event definition
• What specific signal or occurrence are we using?
– Examples: a scheduled macro release (e.g., CPI day), a news headline containing a keyword, a technical crossover, an earnings surprise, etc.
• Should we limit events to positive, negative, or all surprises?
2. Trading rules
• Entry: Do we open a position immediately on the event day’s open, the previous close, or some other timing?
• Exit: Fixed holding period, opposite signal, stop-loss / take-profit, or a combination?
• Position side: long only, short only, or both depending on event direction?
3. Back-test scope
• Ticker universe: I see “MSTR.O” listed – is this the only symbol to test?
• Historical window (start and end dates). If you’re unsure, I can default to the last five years.
• Price type: open or close when calculating P&L.
4. Risk controls (optional)
• Stop-loss %, take-profit %, maximum holding days, maximum drawdown, etc.
If you can outline (or approximate) the above, I’ll build the data-retrieval plan and run the back-test for you.