Import a trade list: stress-test and certify results you already have
Most of AlgoCrucible drives a backtest engine: it re-runs your strategy hundreds of times to find out whether an edge survives data it never saw. But two of the questions traders ask most — how deep could the drawdown get? and would this pass a prop-firm evaluation? — don't need a re-runnable strategy at all. They need the trades. If you already have them, import the list and get those answers in seconds.
That covers cases the engine can't reach comfortably:
- A locked commercial strategy you can't optimize anyway — but whose Strategy Analyzer results you have.
- Range and Renko systems, where every engine backtest is built from tick data and takes minutes; a study of a hundred trials is an overnight job. The trade list is instant.
- Results from another platform. MultiCharts Performance Reports import today; a generic CSV with an exit time and a profit column works for most others.
What to export
NinjaTrader 8: run the backtest in the Strategy Analyzer, open the Trades tab, right-click → Export → CSV. Set the display unit to Currency first — the importer reads net profit per trade in account dollars. (Exports in points or percent are accepted but warned about loudly, because every downstream dollar figure would be wrong.)
MultiCharts: in the Export Backtest Report dialog click Uncheck All, tick only List of Trades, save as .xlsx. That sheet is the import; the other sections only make the file bigger.
Anything else: a CSV with at least an exit time and a net profit per trade. Common header names are recognised automatically (Exit time / Close time / Date, Profit / P&L / Net…). An entry time column, if present, unlocks intraday-ordered certification.
Import it
On the Strategies page, under Import a trade list, choose the file, optionally name the strategy, and click Import. You land on the imported record: trade count, net, profit factor, drawdown, durability figures, and any warnings the parser raised (a column it couldn't read, a unit it doubts).
The net figures are the file's own. The record states what happened, and the app does not re-price history it cannot re-run — so the cost model stands aside for imports, and the cost notes on the panels say so.
What you can do with it
- Drawdown risk (Monte Carlo) — reshuffles the trade order thousands of times to show the range of drawdowns this set of trades could have produced, not just the one path that happened.
- Certify — pick a prop-firm profile (or your own standards) and a contract size; the daily P&L is modelled through the firm's rules for a probability of passing, months to pass, and the bust risk. The recency check applies here too: a profitable record with a losing recent tail is stamped NOT CERTIFIED.
What you can't — and why it says so
Optimize, walk-forward, audit and perturbation are unavailable for imports. Those stages re-run the strategy on different data and different settings, which needs the strategy itself on a connected engine. An imported list can be analyzed, stress-tested and certified, but not anti-overfitting-validated — and every result derived from an import carries a banner saying exactly that, including on printed reports. A certificate that hid this distinction would dilute the one the full pipeline earns.
Next: Certification for reading the pass odds, or Durability metrics for what the record would have felt like to hold.