QuantPerks

Advanced Volatility Models

Enhanced volatility modeling and portfolio risk analysis

✅ No Synthetic Data 🛡️ No Look-Ahead Bias 🎯 Enhanced Features
HAR Portfolio Analysis - Mobile Dark

HAR Portfolio Analysis

Simple 3-Variable Model: Yesterday, Last Week, Last Month

✅ Simple & Interpretable 🛡️ No Overfitting 🎯 Industry Standard
What is HAR?

HAR Model: RVtoday = α + β₁×RVyesterday + β₂×RVlast week + β₃×RVlast month

Daily: Yesterday's volatility impact
Weekly: Last week's average impact
Monthly: Last month's average impact
Portfolio Configuration
Leave empty for equal weights
Max 10 days
HAR Components Analysis
How much does each time period matter for volatility?
Pending Yesterday Impact
Pending Weekly Impact
Pending Monthly Impact

HAR Components Analysis

Shows relative importance of yesterday, last week, and last month
Daily (β₁): Yesterday's volatility impact
Weekly (β₂): Last week's average impact
Monthly (β₃): Last month's average impact

Component Contributions

Percentage breakdown of each time period's influence
Visual representation of which time horizon dominates your portfolio's volatility behavior
Volatility Forecasts
Portfolio volatility predictions for next few days
Pending Tomorrow
Pending 5-Day Average
Pending Trend

Volatility Forecasts

Pure HAR model forecasts - no synthetic data
HAR Forecasts: Real model predictions only
No Synthetic Data: Pure statistical forecasts
Trend Analysis: Model-based volatility direction
Model Summary
HAR equation, fit quality, and technical details
Pending R-Squared
Pending Persistence
Pending Observations
HAR Model Equation
RVt = α + β₁×RVdaily + β₂×RVweekly + β₃×RVmonthly
Coefficients will be shown after analysis

HAR Model Diagnostics

Statistical tests and model quality metrics will be displayed here
R-Squared: Model explanatory power
Persistence: Volatility clustering strength
Observations: Data points used in estimation
Business Insights
What these results mean for your investment strategy
Volatility Persistence

Volatility persistence measures how long market shocks last in your portfolio.

Higher persistence (>0.9) means volatility shocks take longer to fade away.

Dominant Driver

The time period that has the strongest impact on future volatility patterns.

Daily, weekly, or monthly components drive different trading strategies.

Model Quality

R-squared shows how well the HAR model explains volatility patterns.

Values above 30% are considered good for volatility modeling.

Investment Implications

HAR insights guide portfolio rebalancing frequency and risk management.

Different dominant periods suggest different optimal strategies.

EWMA Portfolio Risk - Mobile Dark

EWMA Portfolio Risk Analysis

Industry Standard Risk Management (RiskMetrics Methodology)

✅ Industry Standard 🛡️ No Overfitting ⚡ Fast Computation
What is EWMA?

EWMA Model: σ²ₜ = λ × σ²ₜ₋₁ + (1-λ) × r²ₜ₋₁

Lambda (λ): Decay factor (0.94 = RiskMetrics)
Exponential: Recent data matters more
Adaptive: Quickly responds to market changes
EWMA Portfolio Configuration
Comma-separated
Leave empty for equal weights
0.94 = RiskMetrics
No Overfitting
Industry Standard
Fast Computation
Portfolio Risk Overview
Pending Portfolio Vol
Pending VaR 95%
Pending VaR 99%
Pending Assets

Portfolio Risk Overview

EWMA-based risk metrics and portfolio analysis
Portfolio Vol: Annual volatility estimate
VaR 95%/99%: Value at Risk thresholds
EWMA λ=0.94: Industry standard RiskMetrics methodology
Risk Decomposition
Pending Top Asset
Pending Risk Contrib
Pending Diversification

Risk Attribution by Asset

EWMA-based risk decomposition analysis
Shows each asset's contribution to total portfolio risk using exponentially weighted correlations
EWMA Correlations
Pending Matrix Size
Pending Lambda Used
Pending EWMA Period

Asset Correlation Heatmap

Exponentially weighted correlation matrix
Dynamic correlations that give more weight to recent market relationships between assets
Risk Metrics Summary
Pending Daily VaR
Pending Annual Vol
Pending Lambda (λ)

Portfolio Risk Metrics

Comprehensive risk measurement and attribution
Daily VaR: Maximum expected daily loss
Annual Vol: Annualized portfolio volatility
Lambda (λ): EWMA decay parameter used