Trade Journal

Methodology

Every closed deal from the autopilot system shows up here, fetched live from the bot's deals database. No filtering by outcome, no manual selection, no "since this date" cutoffs designed to flatter the numbers.

What "Status" means:

• TP / TP2 / Trail — closed at take-profit or trailing-stop. Profitable exit.
• SL — stop-loss fired. Realized loss.
• Time — time-based exit (max age reached). Profit can be either direction.
• Manual / Reset — operator-initiated or end-of-day cleanup.

Limitations: Paper-trading numbers don't capture full live execution friction — slippage, partial fills, exchange downtime. Expect 30–50% degradation from these numbers when the same strategies run on live capital. The methodology behind this expectation is in the nine-gate framework article.

What I do with this data

Every losing trade gets reviewed for pattern: was it strategy-level (entry signal weak) or execution-level (slippage or timing) or environmental (sudden volatility regime)? The losing trades teach more than the winners — pattern recognition on losses drives the next gate addition or configuration tweak.

Strategy-level losses with consistent patterns become candidates for backtest auto-disable (Gate 5). Coordinated losses across multiple bots flag cascade risk (Gate 1). Time-of-day clustering on losses feeds pattern memory (Gate 6).