Commit Graph

7 Commits

Author SHA1 Message Date
mindesbunister
7c4adff4e4 Implement risk checks: cooldown, hourly limit, and daily drawdown
Implemented 3 critical risk checks in /api/trading/check-risk:

1. Daily Drawdown Check
   - Blocks trades if today's P&L < maxDailyDrawdown
   - Prevents catastrophic daily losses
   - Currently: -0 limit (configurable via MAX_DAILY_DRAWDOWN)

2. Hourly Trade Limit
   - Blocks trades if tradesInLastHour >= maxTradesPerHour
   - Prevents overtrading / algorithm malfunction
   - Currently: 20 trades/hour (configurable via MAX_TRADES_PER_HOUR)

3. Cooldown Period
   - Blocks trades if timeSinceLastTrade < minTimeBetweenTrades
   - Enforces breathing room between trades
   - Uses minutes (not seconds) thanks to previous commit
   - Currently: 0 min = disabled (configurable via MIN_TIME_BETWEEN_TRADES)

Added database helper functions:
- getLastTradeTime() - Returns timestamp of most recent trade
- getTradesInLastHour() - Counts trades in last 60 minutes
- getTodayPnL() - Sums realized P&L since midnight

All checks include detailed logging with values and thresholds.
Risk check called by n8n workflow before every trade execution.
2025-10-30 10:50:08 +01:00
mindesbunister
25d31ff75a Fix: Save MAE/MFE values when trades exit
Bug: MAE/MFE was tracked in memory during trades but not saved to database on exit
Cause: updateTradeExit() wasn't receiving or saving MAE/MFE parameters

Changes:
- Added MAE/MFE fields to UpdateTradeExitParams interface
- Modified updateTradeExit() to save maxFavorableExcursion, maxAdverseExcursion, maxFavorablePrice, maxAdversePrice
- Updated both updateTradeExit() calls in Position Manager to pass MAE/MFE values
- Enhanced exit logging to show final MAE/MFE percentages

Impact: Future trades will now properly save MAE/MFE data for analytics
Note: Past 2 trades (from before this fix) don't have MAE/MFE saved
2025-10-30 07:37:10 +01:00
mindesbunister
e068c5f2e6 Phase 2: Market context capture at entry
- Added getFundingRate() method to DriftService
- Capture expectedEntryPrice from oracle before order execution
- Capture fundingRateAtEntry from Drift Protocol
- Save market context fields to database (expectedEntryPrice, fundingRateAtEntry)
- Calculate entry slippage percentage in createTrade()
- Fixed template literal syntax errors in execute endpoint

Database fields populated:
- expectedEntryPrice: Oracle price before order
- entrySlippagePct: Calculated from entrySlippage
- fundingRateAtEntry: Current funding rate from Drift

Next: Phase 3 (analytics API) or test market context on next trade
2025-10-29 20:51:46 +01:00
mindesbunister
65e6a8efed Phase 1: Add MAE/MFE tracking and analytics schema
- Added 20+ analytics fields to Trade model (MAE/MFE, fill tracking, timing, market context, slippage)
- Implemented real-time MAE/MFE tracking in Position Manager (updates every 5s)
- Enhanced database schema with comprehensive trade analytics
- Updated all API endpoints to initialize MAE/MFE fields
- Modified updateTradeState() to persist MAE/MFE in configSnapshot

Database changes:
- maxFavorableExcursion/maxAdverseExcursion track best/worst profit %
- maxFavorablePrice/maxAdversePrice track exact price levels
- Fill tracking: tp1Filled, tp2Filled, softSlFilled, hardSlFilled
- Timing metrics: timeToTp1, timeToTp2, timeToSl
- Market context: atrAtEntry, adxAtEntry, volumeAtEntry, fundingRateAtEntry, basisAtEntry
- Slippage tracking: expectedEntryPrice, entrySlippagePct, expectedExitPrice, exitSlippagePct

Position Manager changes:
- Track MAE/MFE on every price check (2s interval)
- Throttled database updates (5s interval) via updateTradeMetrics()
- Persist MAE/MFE in trade state snapshots for recovery

Next: Phase 2 (market context capture) or Phase 3 (analytics API)
2025-10-29 20:34:03 +01:00
mindesbunister
d3c04ea9c9 feat: Position Manager persistence + order cleanup + improved stop loss
- Add Position Manager state persistence to survive restarts
  - Auto-restore open trades from database on startup
  - Save state after TP1, SL adjustments, profit locks
  - Persist to configSnapshot JSON field

- Add automatic order cancellation
  - Cancel all TP/SL orders when position fully closed
  - New cancelAllOrders() function in drift/orders.ts
  - Prevents orphaned orders after manual closes

- Improve stop loss management
  - Move SL to +0.35% after TP1 (was +0.15%)
  - Gives more breathing room for retracements
  - Still locks in half of TP1 profit

- Add database sync when Position Manager closes trades
  - Auto-update Trade record with exit data
  - Save P&L, exit reason, hold time
  - Fix analytics showing stale data

- Add trade state management functions
  - updateTradeState() for Position Manager persistence
  - getOpenTrades() for startup restoration
  - getInitializedPositionManager() for async init

- Create n8n database analytics workflows
  - Daily report workflow (automated at midnight)
  - Pattern analysis (hourly/daily performance)
  - Stop loss effectiveness analysis
  - Database analytics query workflow
  - Complete setup guide (N8N_DATABASE_SETUP.md)
2025-10-27 10:39:05 +01:00
mindesbunister
8e5c592cac Fix database persistence and add analytics
- Fixed Prisma client not being available in Docker container
- Added isTestTrade flag to exclude test trades from analytics
- Created analytics views for net positions (matches Drift UI netting)
- Added API endpoints: /api/analytics/positions and /api/analytics/stats
- Added test trade endpoint: /api/trading/test-db
- Updated Dockerfile to properly copy Prisma client from builder stage
- Database now successfully stores all trades with full details
- Supports position netting calculations to match Drift perpetuals behavior
2025-10-27 09:35:01 +01:00
mindesbunister
d64f6d84c4 feat: implement dual stop system and database tracking
- Add PostgreSQL database with Prisma ORM
  - Trade model: tracks entry/exit, P&L, order signatures, config snapshots
  - PriceUpdate model: tracks price movements for drawdown analysis
  - SystemEvent model: logs errors and system events
  - DailyStats model: aggregated performance metrics

- Implement dual stop loss system (enabled by default)
  - Soft stop (TRIGGER_LIMIT) at -1.5% to avoid wicks
  - Hard stop (TRIGGER_MARKET) at -2.5% to guarantee exit
  - Configurable via USE_DUAL_STOPS, SOFT_STOP_PERCENT, HARD_STOP_PERCENT
  - Backward compatible with single stop modes

- Add database service layer (lib/database/trades.ts)
  - createTrade(): save new trades with all details
  - updateTradeExit(): close trades with P&L calculations
  - addPriceUpdate(): track price movements during trade
  - getTradeStats(): calculate win rate, profit factor, avg win/loss
  - logSystemEvent(): log errors and system events

- Update execute endpoint to use dual stops and save to database
  - Calculate dual stop prices when enabled
  - Pass dual stop parameters to placeExitOrders
  - Save complete trade record to database after execution

- Add test trade button to settings page
  - New /api/trading/test endpoint for executing test trades
  - Displays detailed results including dual stop prices
  - Confirmation dialog before execution
  - Shows entry price, position size, stops, and TX signature

- Generate Prisma client in Docker build
- Update DATABASE_URL for container networking
2025-10-26 21:29:27 +01:00