SIPPM Algorithm Documentation

Systematic Investment Plan for Precious Metals
Version: 1.2
Author: Swapnil Sharma
Last Updated: May 18, 2025

---

Objective

The SIPPM algorithm is designed to automatically monitor a user's precious metal portfolio (Gold, Silver, Platinum, Palladium) and rebalance it based on configurable thresholds, volatility factors, and market price changes—ensuring an optimized and diversified investment strategy.

---

Core Inputs

Parameter	Description
investment_amount	Total user investment (e.g., ₹100,000)
rates	Current metal rates (₹ per gram)
previous_rate	Rate used during last purchase or rebalance
allocation_percent	User's target allocation for each metal
volatility_factor	Adjustment factor based on historical price volatility
transaction_cost	Fixed % per buy/sell trade for that metal
min_trade_amount	Minimum trade amount to be executed (e.g., ₹200)
delta_threshold	Acceptable % deviation before triggering rebalance

---

Walkthrough: How It Works

Step 1: Load User Portfolio

• Retrieve user metals with fields: symbol, weight (grams), previous_rate, allocation_percentage.
• Calculate current_value = weight * current_rate.

Step 2: Market Rate Update

• Fetch real-time INR rate for all metals.
• Recalculate total current value of the portfolio.

Step 3: Delta Calculation

For each metal:

actualPercent = (currentValueOfMetal / totalPortfolioValue) * 100  
delta = actualPercent - allocationPercent  
adjustedDelta = delta * volatilityFactor  

• If abs(adjustedDelta) > deltaThreshold, the metal qualifies for rebalance.

Step 4: Weighted Dynamic Selling

• Metals above allocation → mark for SELL.
• Calculate how much to sell to move back to target allocation.
• Account for transactionCost and skip trades under min_trade_amount.

Step 5: Dynamic Buying

• Distribute freed cash from selling among under-allocated metals.
• Instead of equal buy, prioritize metals with:
  • Highest negative delta
  • Lower volatility (safer bets)
• Use weightage proportional to:

(abs(delta) * (1 - volatilityFactor))

Step 6: Transaction Simulation

• Simulate buy/sell based on cash available.
• Calculate expected new weights & project new portfolio allocation.
• Apply transactionCost to both buy and sell amounts.

Step 7: Persist and Log

• Record all rebalanced units and new weights.
• Log transactions with:
  • metal, units, action, price, value
  • date/time, user_id, delta breached

---

Features Implemented

Feature	Description
Capped Daily Rebalancing	Limit rebalance to once/day/user
Volatility-Based Delta	Adjusts sensitivity per metal
Transaction Cost Simulation	Includes trading fees in logic
Minimum Trade Amount	Skips micro trades
Threshold Tolerances	Prevents over-rebalancing
Weighted Dynamic Selling	Sell based on deviation magnitude
Dynamic Buying	Buy based on need, not equally
INR-Based Investment	All values calculated in ₹

---

Sample Snapshot (Before Rebalance)

[
  {
    "metal": "GOLD",
    "weight": 5.2,
    "previous_rate": 9215.72,
    "current_rate": 9250.0,
    "allocation_percent": 40.0,
    "actual_percent": 45.0,
    "delta": 5.0,
    "volatility_factor": 1.1,
    "transaction_cost": 0.01
  }
]

---

Smart Behaviors

• Cash pool is retained if not enough under-allocated metals are found.
• Trade batcher accumulates trades until they meet the min_trade_amount.
• Cap limit: Only X metals rebalanced per cycle (e.g., 3).

---

Future Enhancements

• Machine Learning-based prediction for weight adjustment.
• User-level volatility override.
• Threshold auto-adjustment based on market activity.