https://github.com/fractal-solutions/paed-framework
https://github.com/fractal-solutions/paed-framework
Last synced: about 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/fractal-solutions/paed-framework
- Owner: fractal-solutions
- Created: 2024-12-06T09:10:40.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2024-12-12T20:09:28.000Z (5 months ago)
- Last Synced: 2025-02-07T01:45:33.435Z (4 months ago)
- Language: JavaScript
- Size: 2.92 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Price Action Energy Dynamics (PAED) Framework
The PAED framework is a first-principles method for efficient trading based on price action, inspired by the Hopfield neural network’s energy-minimization dynamics. This framework views price behavior as a system governed by energy functions, with stable market patterns as attractor states. Here's the concise yet complete documentation for reference.
---
## 1. Define the Problem
Efficient trading involves recognizing patterns in price action data, predicting the next likely movement, and acting on it profitably. The PAED framework reduces this to **pattern recognition**, **energy modeling**, and **decision-making**.
---
## 2. First Principles of Price Action
1. **Price Movement**: Driven by supply-demand imbalances, representing collective shifts in market perception.
2. **Market Structure**: Price action respects key zones (e.g., support/resistance) and trends due to herd psychology.
3. **Efficiency vs. Inefficiency**: Markets are inefficient in the short term (opportunity) but efficient in the long term (randomness).
---
## 3. Energy-Based Modeling of Price Behavior
### Energy Function
The price dynamics are described by the energy function:
\[
E = \text{Trend Energy} + \text{Volatility Energy} - \text{Liquidity Flow}.
\]
1. **Trend Energy**: Captures directional movement (momentum):
\[
\text{Trend Energy} = -\alpha \sum_t (P_t - P_{t-1}),
\]
2. **Volatility Energy**: Models random fluctuations:
\[
\text{Volatility Energy} = \beta \sum_t (P_t - \mu)^2,
\]
3. **Liquidity Flow**: Captures the effect of large orders/news:
\[
\text{Liquidity Flow} = -\gamma \sum_t \text{Volume}_t \cdot \Delta P_t,
\]
### State Updates
Define price states and update them to minimize energy:
\[
P_{t+1} = P_t + \eta \frac{\partial E}{\partial P_t},
\]
---
## 4. Recognizing and Storing Market Patterns
### Stable States as Patterns
Repeatable price patterns (e.g., trends, reversals) are encoded as local energy minima.
The weights of price levels are defined using a Hebbian-like learning rule:
\[
w_{ij} = \frac{1}{N} \sum_\mu (\Delta P_i^\mu \Delta P_j^\mu),
\]
### Attractor Dynamics
When price approaches a stored pattern (e.g., double top), it converges to that attractor.
Spurious states (random noise) are minimized through appropriate parameter tuning.
---
## 5. Decision-Making and Risk Management
### Actionable Trading Rules
1. **Entry Signal**: Enter when price approaches a stable state and Signal Strength exceeds a noise threshold:
\[
\text{Signal Strength} = \sum_i w_{ij} \Delta P_i - \text{Noise Threshold}.
\]
2. **Exit Signal**: Exit when:
\[
\Delta E \approx 0 \quad \text{or} \quad |\text{Volatility Energy}| > |\text{Trend Energy}|.
\]
3. **Stop-Loss Level**: Defined as:
\[
\text{Stop-Loss Level} = P_t \pm \sqrt{\text{Volatility Energy}}.
\]
---
## 6. Convergence Guarantees
1. The energy function E is bounded and decreases with each price update, ensuring convergence to a stable pattern.
2. The symmetric weight matrix w_{ij} prevents oscillations and ensures the system settles into attractors.
---
## 7. Practical Implementation
### Steps to Build the PAED Framework
1. **Data Preparation**: Use historical price and volume data to compute energy components.
2. **Pattern Encoding**: Train the system using a Hebbian-like rule to store market patterns as weights.
3. **Energy Minimization**: Use gradient descent to model price evolution and identify attractors.
4. **Backtesting**: Test the framework across datasets to validate profitability and refine thresholds.
---
## Summary
The PAED framework models price action as a system minimizing energy, with trends, volatility, and liquidity influencing dynamics. Patterns are encoded as attractor states, and actionable signals are derived by interpreting system behavior.