One denominator per setting
Allocation is a percentage of the total example balance. Trade size is a percentage of the allocated bot capital. Fees are added to the entry notional, so a 100% position cannot fit with a positive entry fee.
The original 50% allocation and 5% of total per trade are represented here as 50% allocation and 10% of the bot budget per trade.
A net outcome, not a promise
Buy fill = reference × (1 + adverse slippage). Quantity = entry notional ÷ buy fill. Sell fill = reference × (1 + target or stop move) × (1 − adverse slippage). Net outcome subtracts both entry and exit fees.
Targets and stops are trigger references, not guaranteed fills. A stop does not cap the full economic loss.
What the market can tell you
Spread is 10,000 × (ask − bid) ÷ midpoint. Loaded depth is the sum of price × quantity within ±50 basis points. One basis point is 0.01%. Book checksums cover the nearest 25 levels per side, not the full market.
“Mean true range” is a simple average of 14 consecutive closed candles, including gaps from the previous close. It is not Wilder ATR, a future volatility estimate or a recommended stop distance.
What the scenario actually does
The selected sequence supplies the outcome of each sequential position in one hypothetical session. Positive net profit can be routed to funding after recovering earlier losses. Fees and slippage remain fixed assumptions.
Before each new entry, the model checks available cash, account-risk cap, remaining cumulative realized-loss budget and bot-linked peak drawdown. Profits do not reset the session loss budget. Gaps can overshoot limits after admission.
Cash, profit and drawdown
Funding transfers do not create profit. Account equity = outside-bot cash + bot cash + funding. The drawdown gate uses the peak of bot cash plus funding, so an internal transfer does not create a false drawdown.
Concurrent slots are a capital and stop-exposure illustration. The sequential scenario does not model simultaneous open positions or intratrade unrealized losses.
What is not implemented
No exchange account, entry signal, strategy execution, authenticated save endpoint, order placement, funding transfer, exchange precision rules or real bot deployment is connected. Live observations only change market context.
The values are educational software examples, not a personalized trading recommendation. The preview uses floating-point display math; a real execution backend needs exchange-aware decimal accounting and independent risk checks.
Risk-linked editing & comparison
Fixed planned stop cost adjusts the position size, rounding down to the 0.05 percentage-point increment. Cash, account-risk and session-budget reservations can force a smaller size. No risk cap is raised automatically.
Draft A freezes a reference, account baseline and candle sample. Draft B is evaluated on that common basis; the comparison never estimates a win probability.
Behaviour replay is a different model
The event rehearsal uses fixed synthetic price points, not the supplied win/loss sequence. It admits concurrent positions, includes partial fills, reserves planned risk and leaves unfilled stop-limit positions exposed. The arithmetic Scenario Lab remains a separate conditional-outcome illustration.
Observation memory is a five-minute foreground sample. The demo history is labeled synthetic; public gaps remain unknown.
ACCESSIBLE CONTROLSUse the number fields or sliders for every setting. On chart handles, arrow keys move by 0.05 percentage points; Shift moves by 0.5. Home and End select bounds. Chart arrow keys inspect points. Motion follows your device’s reduced-motion preference.