What do you want to test?
Start with a hypothesis, not a promised return.
Editable experiment draft
PRACTICE DATA · PRIVATE PYTHON ENGINE
Testing your strategy · demo
Your strategy results.
Full Strategy Report
Equity and comparison
Computed portfolio equity in USD; the report identifies its price source.
Drawdown from running peak
Calculated from marked portfolio equity, in percent.
Read sampled chart observations as a table
| Interval | Timestamp UTC | Equity $ | Comparison $ | Drawdown % |
|---|
Monthly net P&L
Grouped by exit date; reconciles to net P&L.
Trade outcomes
Breakeven trades are not counted as losses.
Risk ratios and their limitations
Sharpe and Sortino use daily marked returns, zero risk-free rate and √252 annualization. Small samples and repeated strategy searches can make these estimates unreliable.
Simulated fills use the selected price source. Costs model commission, slippage, spread and short borrow; they are not actual broker fills.
| # | Symbol / side | Entry → exit | Qty | Entry $ | Exit $ | Gross $ | Costs $ | Net P&L $ |
|---|
Configured hypothesis
This immutable snapshot belongs to the completed example. Editing the draft does not rewrite it.
| Setting | Value | Fixture role |
|---|
About this research run
What the engine evaluates
Completed-candle strategy signals, shared-capital positions, explicit entry and exit rules, sizing constraints and modeled costs. Practice prices are artificial. Historical mode requires independently configured Massive access and verified coverage.
Bounded optimization
The optimizer compares a small stop/target set using an earlier training period, then evaluates the selected settings on a later holdout. It does not search every setting or prove a globally optimal or profitable strategy.
Personal local candidate
Optimization and report downloads are available for local evaluation. No public accounts, subscription billing or broker execution are connected. A future paid plan is not implemented here.