FinWiz Financial Libraries Strategy

When implementing financial calculations in FinWiz, follow this clear separation between standard calculations (use libraries) and business logic (custom code).

Library vs Custom Code Decision Matrix

Implementation Patterns

✅ CORRECT: Library + Custom Scoring

import talib
from empyrical import sharpe_ratio

# Library calculates the standard metric
rsi = talib.RSI(close_prices, timeperiod=14)
sharpe = sharpe_ratio(returns, risk_free=0.02)

# Custom code applies FinWiz business logic
if 40 <= rsi <= 60 and sharpe >= 1.5:
    recommendation = "BUY"  # FinWiz scoring logic
elif rsi > 70 or sharpe < 0.5:
    recommendation = "SELL"  # FinWiz scoring logic
else:
    recommendation = "HOLD"  # FinWiz scoring logic

❌ WRONG: Reimplementing Standard Calculations

# DON'T DO THIS - TA-Lib already provides RSI
def calculate_rsi(prices, period=14):
    # Complex RSI implementation...
    # Use talib.RSI() instead!

Key Libraries in FinWiz

TA-Lib (Technical Analysis)

• Use for: RSI, MACD, Bollinger Bands, moving averages
• Import: import talib
• Benefits: C-optimized (100x faster), battle-tested

Empyrical-Reloaded (Risk Metrics)

• Use for: Sharpe ratio, max drawdown, volatility, alpha/beta
• Import: from empyrical import sharpe_ratio, max_drawdown
• Benefits: Standard financial formulas, Quantopian-tested

Backtrader (Backtesting)

• Use for: Strategy backtesting, portfolio simulation
• Import: import backtrader as bt
• Benefits: Industry standard, comprehensive framework

Decision Framework

When implementing new financial calculations:

1. Is this a standard calculation?

   • YES → Use library (TA-Lib, Empyrical, Backtrader)
   • NO → Custom code

2. Is this business logic/scoring?

   • YES → Custom code (competitive advantage)
   • NO → Check if library exists

3. Does a library exist?

   • YES → Use the library
   • NO → Implement with proper documentation

Current FinWiz Architecture

What Uses Libraries (Calculations)

• Technical indicators: TA-Lib
• Risk metrics: Empyrical-Reloaded
• Backtesting: Backtrader

What Uses Custom Code (Business Logic)

• Scoring thresholds: "RSI 40-60 = score 1.0"
• Grading system: "Score >0.90 = A+"
• Asset-specific rules: ROE/debt thresholds

Migration Strategy

Current Custom Implementations to Migrate

Consider migrating these to Empyrical-Reloaded:

• Volatility calculation
• Maximum drawdown calculation
• Beta calculation

# CURRENT (custom)
volatility = np.std(returns) * np.sqrt(252)

# FUTURE (Empyrical for calculation)
from empyrical import annual_volatility
volatility = annual_volatility(returns, period='daily')

# KEEP (custom scoring threshold)
if volatility <= 0.10:
    vol_score = 1.0  # FinWiz business logic

Benefits of This Approach

Libraries Provide

• Performance: C-optimized implementations
• Correctness: Battle-tested by thousands of users
• Maintenance: Automatic updates and bug fixes
• Standards: Industry-accepted formulas

Custom Code Provides

• Competitive Advantage: Unique scoring methodology
• Flexibility: Easy threshold adjustments
• Business Logic: Domain-specific rules
• Customization: Per-asset-class logic

Remember: Libraries = Standard calculations, Custom Code = Business logic and scoring thresholds.