What is Bitcoin market regime detection?

Bitcoin regime detection identifies hidden market states — such as Bull, Bear, or High Volatility — using quantitative models like Hidden Markov Models (HMM).

What is a Hidden Markov Model (HMM) in crypto trading?

An HMM is a probabilistic model that infers hidden market states from observable data such as returns and volatility. It detects regime shifts between Bull, Bear, and Transition states in real time.

How often does the Bitcoin regime update on BTCMonitors?

The BTCMonitors HMM regime dashboard updates every 30 seconds using the Viterbi algorithm applied to live Bitcoin price and volatility data.

Bitcoin HMM Regime Dashboard — Live Hidden Markov Model Market States

Current Regime updating…

HMM Decoded State (Viterbi)

Low Volatility Bull

Trending upward with compressed volatility.

Confidence

82%

Transition Matrix π(t→t+1)

States: 4 Obs: Returns+Vol Training: 3Y daily Baum-Welch

BTC/USD — Price with Regime Overlay

— 7D

Low Vol Bull

Low Vol Bear

High Vol Bear

Transition

Regime History (7D)

Realized Vol (24h)

—

Annualized

Implied Vol (ATM)

—

1-Month Tenor

Log-Likelihood

—

Model fit

Regime Duration

—

Current state

Expected Duration

—

1/(1−p_ii) days

State Entropy

—

Uncertainty

Rolling Volatility7D / 30D

Return DistributionPer Regime

Regime SignalsLive

Indicator	Value	Signal

Viterbi Decoded Path — Last 60 Periods Most Likely State Sequence

Each square = 1 daily observation · Left = oldest · Right = most recent

Forward Algorithm — Filtered State Probabilities (α) 7D Rolling

What is Bitcoin Regime Detection with Hidden Markov Models?

Hidden Markov Models (HMM) are probabilistic graphical models that identify latent (hidden) market states from observable price and volatility data. Unlike simple moving-average crossovers, HMMs model the full probability distribution of returns within each regime — capturing the clustering of volatility and directional drift that characterizes Bitcoin's distinct market phases.

This dashboard implements a 4-state Gaussian HMM trained on daily log-returns and realized volatility. States correspond to: Low Volatility Bull (trending upward, σ < 30%), Low Volatility Bear (slow downtrend), High Volatility Bear (crash/capitulation, σ > 80%), and Transition (regime-switching, elevated uncertainty).

Parameters are estimated via the Baum-Welch algorithm. Real-time state decoding uses the Viterbi algorithm for the most probable state sequence, and the forward algorithm for filtered posterior probabilities. Learn more about HMM theory →

View 90-Day History Set Up Regime Alerts API Access

⚠️ Disclaimer: This dashboard is for informational and educational purposes only. Regime signals are generated by a quantitative model and do not constitute financial advice. Past model performance does not guarantee future results. Cryptocurrency markets are highly volatile. Always conduct your own due diligence before making investment decisions.