BRISMA

Bantleon Risk Model Analysis - Implied Risk Premia Framework for Portfolio Optimization

0 Python Tests

0 R Test Suites

0 Interactive Charts

0 Real Data Charts

0 Page Paper

0 Portfolio Assets

0 Risk Factors

0 Years History

Read Research Paper View Real Data Simulated Data

Core Methodology

Inverse Optimization

mu* = lambda * Q * w + r_f

Extract implied expected returns from observed portfolio weights using mean-variance framework. The key insight: if the portfolio is optimal, what returns must the investor expect?

View Implied Returns Chart 2

Lambda M1 (Market-Based)

lambda = ln((1+y_10Y)/(1+r_f)) / beta_ref

Calibrate risk aversion using the 10-year government bond yield spread. Anchors implied returns to market-observable term premium.

View Lambda M1 Time Series 3

Lambda M2 (Historical)

lambda = sum(w_t * f_t)

Exponentially weighted average of historical factor returns. Captures recent market dynamics with decay parameter controlling memory.

View Lambda M2 Time Series 4

Hybrid Blending

mu = R^2 * mu_M1 + (1-R^2) * mu_M2

R-squared regime switching: when factor model explains well, use market-based M1; when idiosyncratic risk dominates, rely on historical M2.

View Lambda Comparison

Dashboard Overview

Simulated Data Dashboard

19 Charts

Real Data Dashboard

19 Charts - NEW

Interactive Chart Gallery

Showing 50+ charts across simulated and real data

📈

View Chart

Simulated

Asset Returns Time Series

Daily returns for all assets

📈

View Chart

Real Data

Portfolio Time Series

31 assets, 2558 observations

📈

View Chart

Simulated

Cumulative Returns

Wealth evolution over time

📈

View Chart

Real Data

Cumulative Returns

10+ years wealth paths

📈

View Chart

Simulated

Factor Returns

Systematic factor performance

📈

View Chart

Real Data

Risk Model Time Series

61 factor returns

📊

View Chart

Simulated

Portfolio Weights

Asset allocation stacked area

📊

View Chart

Real Data

Portfolio Weights

31 asset weights over time

📊

View Chart

Simulated

Risk Contribution

Marginal risk per asset

📊

View Chart

Real Data

Risk Contribution

Real portfolio risk breakdown

📊

View Chart

Simulated

Implied Returns

mu* from inverse optimization

📊

View Chart

Real Data

Implied Returns

Real data implied expectations

📊

View Chart

Simulated

Rolling Implied Returns

Time-varying mu* estimates

📅

View Chart

Simulated

Yield Curve Data

Term structure for M1

📅

View Chart

Real Data

Yield Curve Data

Actual yield curve history

📅

View Chart

Simulated

Lambda M1 Time Series

Market-based risk aversion

📅

View Chart

Real Data

Lambda M1 Time Series

Real market-implied lambda

📅

View Chart

Simulated

Lambda M2 Time Series

Historical factor-based lambda

📅

View Chart

Real Data

Lambda M2 Time Series

Real historical lambda

📅

View Chart

Simulated

R-Squared Time Series

Factor model fit for regime

📅

View Chart

Real Data

R-Squared Time Series

Real regime indicator

📅

View Chart

Simulated

Lambda Comparison

M1 vs M2 analysis

📅

View Chart

Real Data

Lambda Comparison

Real M1 vs M2

📅

View Chart

Real Data

Time Weights

Exponential decay for M2

🛠

View Chart

Simulated

Factor Loadings

Asset betas to factors

🛠

View Chart

Real Data

Factor Loadings

Real factor exposures

🛠

View Chart

Simulated

Factor Premia

Implied factor returns

🛠

View Chart

Real Data

PCA Components

Principal components

🛠

View Chart

Simulated

Hansen-Jagannathan Bounds

SDF volatility bounds

📋

View Chart

Simulated

Correlation Heatmap

Asset correlation matrix

📋

View Chart

Real Data

Correlation Heatmap

31x31 correlation matrix

📋

View Chart

Simulated

Rolling Volatility

Time-varying vol estimates

📋

View Chart

Real Data

Rolling Volatility

10+ years vol history

📋

View Chart

Real Data

Covariance Comparison

Empirical vs shrinkage vs GARCH

🚀

View Chart

Simulated

Efficient Frontier

Mean-variance optimal

🚀

View Chart

Real Data

Efficient Frontier

Real data frontier

🚀

View Chart

Simulated

Summary Dashboard

Complete overview

🚀

View Chart

Real Data

Summary Dashboard

Real data overview

Documentation Hub

📚

Mathematical Wiki

Complete theory with all formulas and derivations (Sections 14-15)

🎓

Step-by-Step Tutorial

Hands-on guide with exercises and solutions

🛠

Architecture Diagrams

System design and data flow visualizations

📑

Introduction to Implied Premia

Comprehensive overview (17 sections)

📊

Fama-French Data Pipeline

Real data source and processing

📈

Factor Premia Documentation

Implied factor returns methodology

Project Milestones

Phase 1

Project Kickoff

Research objectives, methodology overview, team assignments. 10 slides covering inverse optimization fundamentals.

View Kickoff Presentation

Phase 2

Week 4 Review

Implementation progress, initial results, Lambda M1/M2 calibration. 10 interactive charts demonstrating methodology.

View Week 4 Review

Phase 3

BRISMA Framework Complete

Full Python/R implementation with 173 tests, 18 R test suites. Covariance, GARCH, factor models operational.

View 17-Step Walkthrough

Phase 4

Academic Primer Published

69-page research paper: "Implied Risk Premia for Factors: Theory, Estimation, and Applications" with real FF data.

Read Research Paper

Phase 5

Real Data Dashboard

19 interactive charts using actual portfolio data: 31 assets, 61 factors, 2558 observations, 10+ years history.

Explore Real Data

Key Formulas Reference

Implied Returns

mu* = lambda * Q * w + r_f

Extract expected returns from portfolio weights and covariance. Click to copy.

Lambda M1 (Market)

lambda = ln((1+y_10Y)/(1+r_f)) / beta_ref

Calibrate risk aversion from 10-year yield spread.

Lambda M2 (Historical)

lambda = sum(w_t * f_t)

Exponentially weighted average of factor returns.

Hybrid Blend

mu = R^2 * mu_M1 + (1-R^2) * mu_M2

Blend based on factor model R-squared regime.

Covariance Shrinkage

Q_shrink = alpha * Q_emp + (1-alpha) * Q_factor

Optimal shrinkage between sample and structured estimators.

GARCH(1,1)

sigma_t^2 = omega + alpha * r_{t-1}^2 + beta * sigma_{t-1}^2

Time-varying volatility with persistence and shock impact.