Experimental system · Paper trading phase · Not financial advice · Past performance does not guarantee future results

How the Engine Works

A systematic approach to detecting anomaly signals in US equities, entering on confirmation, and compounding returns through multi-layer exit and re-entry.

The Core Insight

Markets move between quiet and active phases. The engine identifies statistically rare, anomalous conditions in a stock's own behavior that have historically preceded outsized moves. The direction may be uncertain, but the engine focuses on the conditions where the magnitude of the subsequent move tends to be large.

A.I.ccelerate monitors 141 US equities in real time, using proprietary composite indicators to detect when an engine signal fires. The engine doesn't predict direction — it identifies the conditions where outsized moves are statistically likely.

Four-Stage Autonomous Pipeline

1

Signal Detection

The engine continuously monitors the full 141-symbol universe. Multiple proprietary indicators are combined into a composite signal that identifies when a stock's behavior reaches anomalous levels relative to its own history.

Timeframe: 30-minute bars, real-time aggregation
2

Entry Signal

When the composite engine signal crosses a critical threshold, the engine generates an entry signal. This is fully systematic — no discretionary input, no emotional override, no second-guessing. The signal fires or it doesn't.

Multiple entry types across different timeframes (intraday + daily)
3

Compounding Re-entry

After an exit at target, the engine doesn't walk away. It calculates precise re-entry levels based on the move's magnitude. When price retraces to these levels, the engine re-enters — often at a lower cost basis, compounding share count trade over trade.

Capital recycling: exit proceeds fund re-entry at lower prices
4

Multi-Layer Exit

Capital protection comes from multiple exit layers operating at different timeframes:

  • Exit Layer 1: Short-term distribution detection
  • Exit Layer 2: Medium-term structural shift identification
  • Exit Layer 3: Deep-value re-entry after extended declines

What Makes This Different

📉

Not Trend Following

The engine enters on signal, not on momentum. It's contrarian by design — acting on anomaly conditions, not chasing trends.

🔄

Capital Recycling

Most systems buy and hold or buy and sell. This engine exits, re-enters lower, and compounds. Each cycle builds on the last.

🛡

Layered Protection

No single exit point. Multiple independent exit layers protect capital at different time horizons, reducing the risk of any single failure mode.

Honest Disclosure

  • Backtesting uses historical data with perfect hindsight on bar-level OHLCV. Real-world execution will include slippage, latency, and partial fills.
  • Not all symbols perform equally. Some lose money. The engine's edge comes from aggregate compounding across the universe, not from every individual trade.
  • The engine is currently in paper trading validation. No real capital is at risk. Results will be tracked transparently.
  • This is not financial advice. The engine is an experimental research system.

Follow the engine's journey from backtest to live trading.

Follow @AIccelerateEng