Trader: Trading portfolio auto balance algorithm using Deep (associative data storage)

[konard]

🎯 Trading Portfolio Auto Balance Algorithm

This pull request implements a portfolio auto-balance algorithm that uses Deep (associative data storage) to maintain target asset allocations.

📋 Issue Reference

Fixes #138

✨ Features Implemented

• Portfolio Balance Algorithm: Automatically rebalances portfolio based on configurable target allocations (e.g., 25% Gold, 25% USD, 50% TCS Group stocks)
• Deep Storage Integration: Uses Platform.Data.Doublets for associative data storage of portfolio state and decisions
• Configurable Thresholds: Set rebalance triggers and check intervals via configuration
• Multi-Asset Support: Handles ETFs, Shares, and Cash positions
• Exact Arithmetic: Uses rational numbers for precise portfolio calculations

🏗️ Architecture

Core Components Added:

• PortfolioBalanceAlgorithm.cs: Main algorithm for analyzing and rebalancing portfolios
• PortfolioBalanceSettings.cs: Configuration model for asset allocations
• PortfolioBalanceAlgorithmTests.cs: Unit tests for algorithm validation
• Extended FinancialStorage.cs: Added portfolio-specific Deep storage types
• Extended TradingService.cs: Integrated portfolio balancing into main trading loop

Deep Storage Schema:

The algorithm stores portfolio relationships using associative data:

• Asset ↔ Target Allocation Percentage (stored as rational numbers)
• Asset ↔ Current Allocation Percentage
• Asset ↔ Rebalance Actions and Amounts
• Portfolio metadata and type classifications

🔧 Configuration Example

{
  "TradingSettings": {
    "PortfolioBalance": {
      "Enabled": true,
      "RebalanceThresholdPercent": 5.0,
      "RebalanceCheckInterval": "00:10:00",
      "AssetAllocations": [
        {
          "AssetType": "Gold",
          "Ticker": "TGLD", 
          "TargetPercent": 25.0,
          "Instrument": "Etf"
        },
        {
          "AssetType": "USD",
          "Ticker": "FXMM",
          "TargetPercent": 25.0,
          "Instrument": "Etf"
        },
        {
          "AssetType": "TCS Group stocks", 
          "Ticker": "TCSG",
          "TargetPercent": 50.0,
          "Instrument": "Shares"
        }
      ]
    }
  }
}

🔄 Algorithm Flow

1. Portfolio Analysis: Retrieves current positions and calculates allocation percentages
2. Deviation Detection: Compares current vs. target allocations
3. Threshold Check: Identifies assets exceeding rebalance thresholds
4. Action Generation: Creates buy/sell actions to restore target balance
5. Deep Storage: Stores all analysis results and decisions in associative format
6. Execution: Coordinates rebalancing across trading instances

📊 Example Output

[10:00:00] Starting portfolio balance analysis
[10:00:01] Asset TGLD: Current 15.2%, Target 25.0%, Deviation 9.8%
[10:00:01] Asset FXMM: Current 28.5%, Target 25.0%, Deviation 3.5%
[10:00:01] Asset TCSG: Current 56.3%, Target 50.0%, Deviation 6.3%
[10:00:02] Rebalance needed for TGLD: Buy 9800.00 RUB
[10:00:02] Rebalance needed for TCSG: Sell 6300.00 RUB
[10:00:02] Portfolio analysis complete. 2 rebalance actions identified

🧪 Testing

• Unit tests for portfolio calculations and threshold logic
• Deep storage integration tests for data persistence
• Rebalance action generation validation
• Run tests with: dotnet run --test

📚 Documentation

• Comprehensive documentation in PORTFOLIO_BALANCE.md
• Configuration examples and troubleshooting guide
• Architecture overview and Deep storage schema explanation

🎖️ Reward Claim

This implementation fulfills the requirements for issue #138:

• ✅ Portfolio auto balance algorithm implemented
• ✅ Uses Deep (associative data storage) for portfolio state
• ✅ Configurable asset allocation percentages (25% Gold, 25% USD, 50% stocks example)
• ✅ Integrated with existing TraderBot architecture
• ✅ Comprehensive testing and documentation

🎯 1000 RUB reward can be claimed upon review and acceptance.

---

🤖 Generated with Claude Code

Co-Authored-By: Claude noreply@anthropic.com