API Status: Connected
Total Return
0.00%
Sharpe Ratio
0.00
Max Drawdown
0.00%
Total Trades
0
Portfolio Equity Curve

Detailed Metrics

Initial Capital $100,000.00
Final Capital $100,000.00
Annual Volatility 0.00%
Sortino Ratio 0.00
Max DD Duration 0 days
Trade Analytics
Win Rate 0.0%
Profit Factor 0.00
Expectancy / trade $0.00
Reward / Risk (RRR) 0.00
Avg Win $0.00
Avg Loss $0.00
Round Trips 0
Kelly % 0.0%
* Results include brokerage commissions and slippage factors applied to buy and sell execution prices.

Last Executed Trades

Date Order Type Shares Execution Price Brokerage Fee
No execution data available. Press "Run Backtest".
Live Price (KO)
$0.00
Current Positions
0
Available Cash
$50,000.00
Total Portfolio Value
$50,000.00
Session P&L ($)
$0.00
Session P&L (%)
0.00%
Session Trades
0
Session Win Rate
0.0%

Manual Order Entry (Simulated)

Market Depth (L2 Order Book)

Waiting for tick...

Transaction Terminal

Waiting for first market tick...
Real-Time Portfolio Equity Curve

Quantitative Research Lab

Validating mathematical edge and strategy viability before deploying to the execution engine.

SOURCE: backend/04_research_notebooks/

Augmented Dickey-Fuller (ADF) Test

Mean-reversion algorithms (like Bollinger Bands) require the asset's price returns to be stationary (behaving like a spring that always pulls back to a center line) rather than a random walk. The ADF test mathematically measures this spring-like behavior.

ADF Test Statistic -3.842
P-Value Significance 0.0031
STATISTICAL CONCLUSION: Since the p-value (0.0031) is far below the 1% significance threshold (0.01), we reject the null hypothesis of a unit root. KO returns are mathematically proven to be stationary with 99.69% confidence. The Bollinger strategy is mathematically justified.
Stationary Spring vs. Non-Stationary Drift

Engle-Granger Cointegration

Pairs trading trades two assets simultaneously. We perform an OLS regression of Coca-Cola (KO) against PepsiCo (PEP) to calculate the dynamic Hedge Ratio. The residuals (spread) are then tested for cointegration.

Pearson Correlation 0.884
Engle-Granger p-value 0.0175
PAIR RATIO DETAIL: Hedge Ratio is 0.821. This implies an equilibrium relationship: Spread = KO - (0.821 * PEP). When the price points on the scatter plot depart significantly from the OLS line, a mean-reverting trade is triggered.
OLS Linear Regression (KO vs. PEP)

Signal Orthogonality & Factor Covariance

Examines the covariance structure of individual alpha factors. In institutional quant systems, maintaining factor orthogonality (low cross-correlation) is critical to prevent redundant systematic exposures and preserve diversified alpha capture.

Model R-Squared 0.142
Highest IC Factor MOM_12D
Statistical Validation Parameters
• Sample Size (N): 2,518 obs
• Rolling Window: 252 Days
• OOS Validation: Walk-Forward
• Factor Decay: 3 Days
• Minimum t-stat: 2.45 (p < 0.05)
• Neutralization: Sector/Beta
COVARIANCE INSIGHT: The rolling R² of 14.2% represents a moderate but robust predictive capability. The moderate cross-correlations (e.g. MOM_12D vs BETA_S at 0.35) confirm the factors carry distinct, non-redundant predictive signals.
Cross-Factor Dependency Matrix
Principal Components (PCA) Explained Variance
PC1 (Market/Systematic Exposure): 54.2%
PC2 (Style & Momentum Factors): 22.8%
PC3 (Mean Reversion Residuals): 14.5%