robust-rolling-regime-detection

Explainable Regime-Aware Portfolio Optimization using Robust Rolling Regime Detection (R2-RD) - Academic paper for International Review of Financial Analysis

View on GitHub

Information

Property	Value
Language	TeX
Stars	0
Forks	0
Watchers	0
Open Issues	4
License	No License
Created	2025-12-25
Last Updated	2026-02-19
Last Push	2026-01-11
Contributors	1
Default Branch	main
Visibility	private

Datasets

This repository includes 14 dataset(s):

Dataset	Format	Size

| sentence_review.json | .json | 74.91 KB |

| literature_results.json | .json | 128.09 KB |

| hallucination_report.json | .json | 0.67 KB |

| bic_selection.json | .json | 7.0 KB |

| hamilton_regime.json | .json | 6.63 KB |

| key_papers.json | .json | 1.79 KB |

| label_switching.json | .json | 6.82 KB |

| ms_garch.json | .json | 6.82 KB |

| mvo.json | .json | 6.85 KB |

| oos_performance.json | .json | 6.71 KB |

| portfolio_regimes.json | .json | 6.93 KB |

| regime_switching.json | .json | 7.01 KB |

| transaction_costs.json | .json | 6.73 KB |

| xai_finance.json | .json | 6.82 KB |

Reproducibility

No specific reproducibility files found.

Status

Issues: Enabled
Wiki: Enabled
Pages: Enabled

README

Robust Rolling Regime Detection (R2-RD)

Explainable Regime-Aware Portfolio Optimization using Robust Rolling Regime Detection

Authors

Amine Boukardagha (Citi Research, Citigroup Global Markets Inc.)
Alex Saunders (Citi Research, Citigroup Global Markets Inc.)

Target Journal

International Review of Financial Analysis (Elsevier)

Abstract

We propose an explainable framework for cross-asset portfolio optimization under time-varying market regimes. Our approach leverages a robust rolling regime detection model (R2-RD) to identify latent market states with interpretable mean and covariance structures, and a K-Nearest Neighbors (KNN) model as a nonparametric benchmark. Both methods are embedded within a regime-aware mean-variance optimization (MVO) framework and evaluated via a strictly causal expanding-window backtest on monthly data.

Repository Structure

robust-rolling-regime-detection/
├── manuscript/
│   ├── main.tex              # Main Elsevier article
│   ├── main.bib              # Bibliography
│   └── elsarticle.cls        # Elsevier template
├── figures/
│   ├── 01_hmm_pnl/           # R2-RD cumulative PnL
│   ├── 02_knn_pnl/           # KNN cumulative PnL
│   ├── 03_regime_timeline/   # Regime labels over time
│   └── 04_returns_by_regime/ # Asset returns by regime
└── theory/
    └── proofs.tex            # Supplementary proofs

Build

cd manuscript
pdflatex main.tex
bibtex main
pdflatex main.tex
pdflatex main.tex

Key Contributions

Robust Rolling Regime Detection (R2-RD): Rolling HMM estimation with BIC-based regime count selection and regime emergence policy
Label Matching Mechanism: Linear assignment problem to ensure temporal consistency of regime labels
Theoretical Foundations: Convergence properties, regime emergence bounds, and portfolio optimality conditions
Empirical Comparison: R2-RD vs KNN for cross-asset allocation