The Story of Systemic Collapse On March 14, 2008—six months before Lehman Brothers would collapse—our Hidden Markov Model assigned the financial sector (XLF) an 89.3% bearish confidence rating. The S&P 500 (SPY) showed only 50.4% sideways confidence at the same moment. The model had detected systemic contagion building …

The Story of Irrational Exuberance At the NASDAQ peak on March 10, 2000, our Hidden Markov Model revealed a striking divergence: Amazon showed 85.2% bearish confidence while Microsoft peaked at 98.3% bullish. This stark regime divergence between speculative dot-coms and profitable technology companies foreshadowed …

The Signal That Came Too Early On January 27, 2020, while Carnival Cruise Lines stock traded at 44.75 — still near its high — our 3-state Hidden Markov Model assigned it a 99.99% confidence to “Crisis ” Regime . QQQ, the Nasdaq-100 ETF, simultaneously received a 99.95% Bearish confidence rating at 210.72, well before …

Glossary A comprehensive reference for statistical, mathematical, financial, and trading terms used throughout Hidden Regime. Each entry includes definitions, formulas (where applicable), interpretation guidance, use cases, and cross-references to related concepts. Updated: October 2025 - Expanded with 51 new trading …

Master the production-ready Hidden Regime pipeline: one-line setup, automated regime classification, duration analysis, and model selection for real-world regime detection.

Learn how to apply HMM regime detection to real trading: regime-specific risk analysis, dynamic position sizing, technical indicator integration, and backtesting strategies.

HMM Basics: Understanding Regime Detection From Log Returns to Market Regimes Using Hidden Markov Models In the previous notebook, we proved that log returns are the correct transformation for financial time series. Now we’ll use those log returns to detect market regimes using Hidden Markov Models (HMMs) . What …

Why Log Returns? A Mathematical Foundation Understanding the Correct Transformation for Financial Time Series Before we can detect market regimes , we must answer a fundamental question: How should we transform price data for statistical analysis? This notebook provides the mathematical proof for why log returns …

HMM 101 An example of how we’ll be using our Hidden Markov Models in this project. HMM Workflow: Initialization (k-means): Cluster observations to get initial emission parameters Training (Baum-Welch ): Refine transition matrix & emission distributions Prediction (Viterbi ): Decode the most likely state sequence This …

Stochastic Modeling of Market Dynamics and the Efficient Market Hypothesis If the Efficient Market Hypothesis holds, then what hope do we have for using technical analysis? I. Introduction: The Dialectic of Efficiency and Predictability I.A. Contextualizing the Conflict Modern financial economics operates under a …