Markov: Transition Matrix [Daily Timeframe]

16시간 전

Description

This indicator computes a 3-state Markov chain from price action and visualizes the transition probabilities between daily states:

• Up: daily % change > threshold
• Down: daily % change < -threshold
• Sideways: |daily % change| ≤ threshold

From those states, it builds transition matrices:

• Today → Tomorrow (1 day ahead)
• Today → In 2 days
• Today → In 3 days

Each matrix cell shows:

P(next state | current state)
Rows are the current state (today), columns are the future state (tomorrow / +2 / +3).
Each row sums to 100% (when there is sufficient sample size).
________________________________________
How to read it (trader workflow)

1. Identify the current regime (the most recent confirmed daily state).

2. Look at the row matching that regime:

• The ★ marks the highest probability outcome for that row (most likely next state).
• Heatmap intensity increases as probability increases.
• Each row shows its own sample size (n=...) so you can judge statistical support.

3. Use Quick-read:

• “Now” = current regime
• “Best” = top conditional outcome + probability
• “2nd” = second-best outcome + probability

4. Use Universe (N):

• Shows the marginal distribution: how often days are Up/Down/Sideways across the whole dataset.
________________________________________
Settings

Core logic

• Sideways threshold: controls how strict “Sideways” is.
Example: 0.001 = ±0.10% daily move is considered Sideways.

Display

• Toggle 1D / 2D / 3D matrices.
• Highlight best probability per row (★).
• Show n per row (row transition count).
• Focus: current state row only to reduce noise and speed decision-making.
• Quick-read row for the current regime.

Theme (fully customizable)

All colors can be customized:

• Up / Down / Sideways base colors
• Header background + header text
• Values text
• Quick-read neutral background

This makes it suitable for both light and dark chart themes.
________________________________________
Notes / Limitations

• The indicator is designed for daily sessions. It uses daily close-to-close returns to classify states and update the Markov chain once per day.
• On very volatile assets, a very small threshold can make Sideways rare. If you want a more frequent Sideways regime, increase the threshold.
• This is a statistical visualization tool, not a trading system.
________________________________________
Disclaimer (TradingView-friendly)

This script is provided for educational and informational purposes only and does not constitute financial advice. Trading involves risk. Past probabilities do not guarantee future results. Use at your own discretion and always apply proper risk management.

16시간 전

릴리즈 노트

Update: Live “Now” / Focus (optional) + confirmed stats preserved

• Added a new option: “Use LIVE daily state for ‘Now’ & Focus (may repaint)”.
• When enabled, the Now label and Focus row follow the current (in-progress) daily state (Up/Down/Sideways) using the live daily close, so the panel can reflect today’s direction before the daily bar is confirmed.
• The Markov matrices and probabilities remain based on confirmed daily closes only (non-repainting), to preserve statistical integrity.
• This resolves cases where extended sessions or intraday viewing could show Now = Down even when the current day is Up (close-to-close).

16시간 전

릴리즈 노트

Update: Added +4D and +5D transition matrices (optional)
• Added two new horizons:
o Today → In 4 days
o Today → In 5 days
• You can enable/disable them from the settings (Show 4 days / Show 5 days) to keep the table compact when needed.
• The script now supports up to 5 Markov transition matrices (1D–5D) with the same heatmap, best-cell ★ highlighting, sample size per row (n), and Quick-read workflow.
• Stat integrity preserved: the Markov chain and transition counts still update only on confirmed daily closes (non-repainting). The optional LIVE “Now” & Focus mode remains available for real-time regime display.