Nija

High Leverage Improvements Guide

Overview

This guide covers three critical improvements to NIJA that provide high-leverage value:

CodeQL + Security Hardening - Automated security scanning and best practices
5-Year Multi-Regime Backtesting - Comprehensive historical validation
30-Day Live Paper Trading - Real-world strategy validation before deployment

1️⃣ CodeQL + Security Hardening

What It Does

Automated security scanning integrated into the CI/CD pipeline to detect vulnerabilities before they reach production.

Components

CodeQL Analysis (.github/workflows/codeql.yml)
- Scans Python and JavaScript code
- Runs on every push and PR
- Weekly scheduled scans
- Detects security vulnerabilities and code smells
Security Scanning (.github/workflows/security-scan.yml)
- Safety: Python dependency vulnerability scanner
- Bandit: Python security linting
- TruffleHog: Secret scanning in git history

Setup

Security scanning is automatically enabled once the workflows are merged to main.

View Security Results:

GitHub Repository → Security Tab → Code scanning alerts

Manual Security Scan:

# Install security tools
pip install safety bandit

# Run dependency scan
safety check

# Run Bandit security linting
bandit -r . --exclude ./archive,./venv,./mobile,./frontend

# Generate reports
bandit -r . -f json -o bandit-report.json
safety check --json --output safety-report.json

Security Best Practices

See SECURITY_HARDENING_GUIDE.md for complete details:

✅ Never commit API keys
✅ Use environment variables for secrets
✅ Validate all webhook inputs
✅ Implement rate limiting
✅ Use HTTPS for all external calls
✅ Encrypt sensitive data at rest
✅ Implement proper error handling
✅ Log security events
✅ Regular dependency updates

Schedule

Daily: Review security scan results in PRs
Weekly: Check CodeQL alerts
Monthly: Update dependencies with security patches
Quarterly: Rotate API keys
Annually: Full security audit

2️⃣ 5-Year Multi-Regime Backtesting

What It Does

Comprehensive backtesting across 5 years of historical data with analysis broken down by market regime (bull/bear/ranging/volatile).

Why It Matters

Statistical Significance: 5 years provides sufficient sample size
Regime Analysis: Understand how strategy performs in different market conditions
Investor Confidence: Professional-grade performance validation
Risk Assessment: Identify maximum drawdowns across market cycles

Usage

# Run 5-year backtest on BTC-USD
python run_5year_backtest.py \
  --symbol BTC-USD \
  --years 5 \
  --initial-balance 10000 \
  --strategy APEX_V71 \
  --output results/5year_backtest_btc.json

# Run on ETH-USD
python run_5year_backtest.py \
  --symbol ETH-USD \
  --years 5 \
  --output results/5year_backtest_eth.json

# Custom parameters
python run_5year_backtest.py \
  --symbol BTC-USD \
  --years 3 \
  --initial-balance 50000 \
  --commission 0.0006 \
  --slippage 0.0003 \
  --output results/custom_backtest.json

Output

The script generates a comprehensive JSON report with:

Overall Performance:

Total return %
Sharpe ratio
Sortino ratio
Profit factor
Win rate
Maximum drawdown
Trade statistics

Regime Analysis: For each market regime (bull/bear/ranging/volatile):

Number of trades
Win rate
Profit factor
Return %
Max drawdown
Sharpe ratio
Average win/loss

Monte Carlo Simulation:

1,000 random resamplings of trades
Expected return distribution
Confidence intervals
Worst-case scenarios
Risk estimates

Statistical Significance:

T-test results
P-value
Sample size validation
Conclusion on edge

Example Output

================================================================================
5-YEAR MULTI-REGIME BACKTEST SUMMARY
================================================================================

📊 BACKTEST DETAILS
Symbol: BTC-USD
Strategy: APEX_V71
Period: 2021-01-28 to 2026-01-28 (5 years)
Initial Balance: $10,000.00

💰 OVERALL PERFORMANCE
Final Balance: $45,230.00
Total Return: 352.30%
Total Trades: 1,247
Win Rate: 61.2%
Profit Factor: 2.35
Sharpe Ratio: 1.92
Max Drawdown: 8.45%

📈 REGIME ANALYSIS

BULL Market:
  Duration: 450 days
  Trades: 387
  Win Rate: 64.3%
  Return: 145.20%
  Profit Factor: 2.68
  Max DD: 5.20%

BEAR Market:
  Duration: 380 days
  Trades: 312
  Win Rate: 55.8%
  Return: 42.10%
  Profit Factor: 1.95
  Max DD: 8.45%

RANGING Market:
  Duration: 620 days
  Trades: 428
  Win Rate: 62.4%
  Return: 118.60%
  Profit Factor: 2.42
  Max DD: 6.30%

VOLATILE Market:
  Duration: 380 days
  Trades: 120
  Win Rate: 58.3%
  Return: 46.40%
  Profit Factor: 2.15
  Max DD: 7.80%

🎲 MONTE CARLO SIMULATION (1000 runs)
Expected Return: 352.30% ± 45.20%
95% Confidence: [275.40%, 428.60%]
Expected Sharpe: 1.92

📊 STATISTICAL SIGNIFICANCE
Sample Size: 1,247 trades
P-Value: 0.0001
Conclusion: Strategy shows statistically significant edge

================================================================================

Data Requirements

The script expects historical OHLCV data in data/ directory:

Format: CSV with columns: timestamp,open,high,low,close,volume
Filename: {SYMBOL}_historical_5y.csv (e.g., BTC-USD_historical_5y.csv)

If data file not found, the script automatically generates synthetic data for demonstration.

To use real data:

Download historical data from exchange or data provider
Save as CSV in data/ directory
Ensure proper column names and datetime format

3️⃣ 30-Day Live Paper Trading

What It Does

Live paper trading system that runs for 30 days to validate strategy performance in real market conditions without risking capital.

Why It Matters

Real-World Validation: Tests strategy with live data and execution
Performance Comparison: Compare vs backtest expectations
Risk Detection: Identify issues before deploying real money
Confidence Building: 30-day track record for scaling capital

Usage

Day 1-30: Record Daily Metrics

Run this once per day (set up as cron job):

# Record today's performance
python run_30day_paper_trading.py --record-daily

# Specify custom data directory
python run_30day_paper_trading.py --record-daily --data-dir data/paper_jan2026

Weekly Reports

# Generate week 1 report
python run_30day_paper_trading.py --weekly-report 1

# Generate week 2 report
python run_30day_paper_trading.py --weekly-report 2

# And so on...

Final 30-Day Report

# Generate comprehensive 30-day report
python run_30day_paper_trading.py --final-report

# Compare to backtest
python run_30day_paper_trading.py \
  --compare-backtest results/5year_backtest_btc.json \
  --final-report

Daily Output

============================================================
DAILY PAPER TRADING REPORT - 2026-01-28
============================================================

💰 BALANCE
Starting: $10,000.00
Ending:   $10,234.50
P&L:      +$234.50 (+2.35%)

📊 TRADING
Trades:   8
Wins:     5 (62.5%)
Losses:   3
Best:     +$125.30
Worst:    -$45.20

📍 POSITIONS
Open:     3
Exposure: $3,045.20 (29.7% of balance)
============================================================

Weekly Summary

📊 WEEK 1 REPORT
Period: 2026-01-21 to 2026-01-27
Return: +12.45%
Trades: 47
Win Rate: 59.6%
Profit Factor: 2.28
Sharpe: 1.85
Max DD: 3.20%

30-Day Final Report

================================================================================
30-DAY PAPER TRADING FINAL REPORT
================================================================================

📅 Period: 2026-01-01 to 2026-01-30
💰 Return: +48.50%
📊 Trades: 187
✅ Win Rate: 61.5%
📈 Sharpe: 1.94
📉 Max DD: 6.80%
================================================================================

Automated Monitoring

Set up daily cron job:

# Edit crontab
crontab -e

# Add daily execution at market close (5pm EST = 22:00 UTC)
0 22 * * * cd /path/to/Nija && python run_30day_paper_trading.py --record-daily >> logs/paper_trading.log 2>&1

# Weekly report on Sundays at 9am
0 9 * * 0 cd /path/to/Nija && python run_30day_paper_trading.py --weekly-report $(($(date +\%W) % 4 + 1)) >> logs/paper_weekly.log 2>&1

Performance Alerts

The system automatically generates alerts for:

Daily Loss > 5%: High severity
Win Rate < 40% (with 5+ trades): Medium severity
Drawdown > 12%: High severity
Exposure > 80%: Medium severity

Alerts are saved in data/paper_trading_30day/alerts.json and logged.

Comparison to Backtest

python run_30day_paper_trading.py \
  --compare-backtest results/5year_backtest_btc.json

Output:

================================================================================
PAPER vs BACKTEST COMPARISON
================================================================================

WIN_RATE:
  Backtest: 0.612
  Paper:    0.615
  Diff:     +0.003

SHARPE_RATIO:
  Backtest: 1.920
  Paper:    1.940
  Diff:     +0.020

MAX_DRAWDOWN:
  Backtest: 8.450
  Paper:    6.800
  Diff:     -1.650

✅ EXCELLENT - Paper trading matches backtest expectations
================================================================================

Integration Workflow

Recommended Implementation Order

Week 1: Security Hardening

Merge security workflows to main
Review and fix any security alerts
Implement security best practices
Set up alert notifications

Week 2: Run 5-Year Backtests

Download historical data for top trading pairs
Run comprehensive backtests
Analyze results by regime
Document findings and expected performance

Week 3-6: 30-Day Paper Trading

Start paper trading with live bot
Record daily metrics (automated via cron)
Generate weekly reports
Compare to backtest expectations
Adjust strategy if needed

Week 7: Go Live Decision

Review 30-day paper trading results
Compare vs backtest
Assess risk metrics
Make go/no-go decision
If go: Start with small capital
If no-go: Refine strategy and repeat

Metrics to Track

Security Metrics

Number of vulnerabilities detected
Time to fix critical issues
Dependency update frequency
Secret scanning false positive rate

Backtest Metrics

Sharpe ratio by regime
Maximum drawdown by regime
Win rate consistency
Profit factor stability
Statistical significance

Paper Trading Metrics

Daily P&L
Win rate vs backtest
Sharpe ratio vs backtest
Drawdown vs backtest
Alert frequency
Performance degradation signals

Success Criteria

Security

✅ Zero high/critical security vulnerabilities
✅ All dependencies up to date
✅ No secrets in codebase
✅ Security scans running automatically

Backtesting

✅ Sharpe ratio > 1.5 across all regimes
✅ Profit factor > 2.0
✅ Win rate > 55%
✅ Max drawdown < 15%
✅ Statistically significant edge (p < 0.05)

Paper Trading

✅ Performance within 10% of backtest metrics
✅ Max drawdown < backtest max drawdown
✅ No critical alerts
✅ Sharpe ratio > 1.5
✅ Consistent profitability week-over-week

Troubleshooting

Security Scans Failing

Review GitHub Actions logs
Check if dependencies are compatible
Update outdated packages
Address reported vulnerabilities

Backtest Issues

Verify data format and quality
Check for missing data points
Ensure sufficient historical data (5 years)
Validate strategy parameters

Paper Trading Issues

Verify paper account has sufficient balance
Check that bot is running continuously
Review trade logs for errors
Ensure market data is being received

Next Steps After Validation

Once all three components are validated:

Scale Testing
- Run backtests on additional symbols
- Extend paper trading to more trading pairs
- Test with larger position sizes
Live Deployment
- Start with minimum capital
- Monitor closely for first week
- Gradually increase capital as confidence builds
- Implement circuit breakers
Continuous Improvement
- Keep security scans running
- Quarterly backtest updates
- Monthly paper trading validation
- Regular performance reviews

Resources

Security: SECURITY_HARDENING_GUIDE.md
Backtesting: LIVE_EXECUTION_BACKTESTING_GUIDE.md
Paper Trading: bot/paper_trading.py
Strategy Docs: APEX_V71_DOCUMENTATION.md

Support

For issues or questions:

Check relevant documentation files
Review error logs
Create GitHub issue with details
Tag with appropriate label (security/backtest/paper-trading)