PAPER ONLY · PLAN BEFORE PERFORMANCE

Strategy backtest lab

Test a strategy, inspect every simulated trade, and see where it follows your plan.

Anonymous practice uses the example plan below. Runs are temporary. Sign in with Google and save your own plan to grade against your rules.

Grading plan: Example paper plan

Setup: Practice a closed-candle momentum or breakout hypothesis.

Entries: Confirm the selected signal at the close; enter at the next open with a stop.

Max leverage 3× · risk 2% · drawdown 15% · daily loss 3%

Symbols: BTC-USDT, ETH-USDT, SOL-USDT · sides: long, short · venues: blofin-demo

Written setup and entry text need your review. The grade checks the coded plan rules; it does not interpret prose.

1. Candle history

Data source

Symbol

Timeframe

Simulated venue (sample/import)

Sample data is synthetic, not market performance. Use USD/USDT-quoted history; the model assumes 1 USDT = 1 USD. For sample/import the venue is your declaration, not verified provenance. Demo history always uses Blofin demo and supports BTC-USDT, ETH-USDT and SOL-USDT, up to 300 recent bars. Market errors never substitute sample prices.

Import history (3–1000 candles)

OHLC JSON

Each candle needs ts (UTC milliseconds), open, high, low and close as numbers. Use continuous intervals matching the timeframe. Imported provenance is unverified. Unfinished bars are excluded.

2. Strategy to test

Strategy

Direction

Breakout lookback (bars) Starting equity (USD) Strategy leverage Target risk per trade (%) Stop distance (%) Fee per fill (basis points) Slippage per fill (basis points)

Strategy settings may differ from your plan so you can inspect violations. Risk sizing is capped by the strategy leverage budget. Risk % measures entry-to-stop loss before costs or gaps. One basis point = 0.01%.

How the replay and grade work

Momentum: enter after a close moves in the selected direction. Breakout: enter after a close exceeds the prior lookback highs (long) or lows (short). Signals use only completed candles; entries fill at the next open. Both strategies exit at the next open after an opposite close move, on a stop, or at the final close.

Fees and adverse slippage apply to every fill. Stops fill at the stop or a worse gap-open price. Drawdown starts at the initial balance and includes costs, exits and bar-close equity; intrabar peaks are not modeled. Daily loss uses realized balance changes per UTC day. An exit retains the grade and risk context of its entry decision.

This isolated replay grades rule violations and continues through them to show their consequences. It stops opening positions if equity is exhausted. It cannot submit orders and never changes execution caps. Funding, liquidation, liquidity and borrowing costs are not modeled. Counterfactual descriptions are rule-based illustrations, not a second historical replay.

Profit is separate from discipline. A high grade is not evidence of an edge. No holdout evaluation or automatic parameter optimization is performed.