INVITE-ONLY SCRIPT
JINN: A Multi-Paradigm Quantitative Trading and Execution Engine
I. Core Philosophy: A Substitute for Static Analysis
JINN (Joint Investment Neural and Network) represents a paradigm shift from static indicators to a living, adaptive analytical ecosystem. Traditional tools provide a fixed snapshot of the market. JINN operates on a fundamentally different premise: it treats the market as a dynamic, regime-driven system. It processes market data through a hierarchical suite of advanced, interacting models, arbitrates their outputs through a rules-based engine, and adapts its own logic in real-time.
It is designed as a complete framework for traders who think in terms of statistical edge, market regimes, probabilistic outcomes, and adaptive risk management.
II. The JINN Branded Architecture: Your Command and Control Centre
JINN’s power emerges from the synergy of its proprietary, branded architectural components. You do not simply "use" JINN; you command its engines.
1. JINN Signal Arbitration (JSA) Engine
The heart of JINN. The JSA is your configurable arbitration desk for weighing evidence from all internal models. As the Head Strategist, you define the entire arbitration philosophy:
• Priority and Weighting: Define a "chain of command". Specify which model's opinion must be considered first and assign custom weights to their outputs, directly controlling the hierarchy of your analytical flow.
• Arbitration Modes:
First Wins: For high-conviction, rapid signal deployment based on your most trusted leading model.
Highest Score: A "best evidence" approach that runs a full analysis and selects the signal with the highest weighted probabilistic backing.
Consensus: An ultra-conservative, "all-clear" mode that requires a unanimous pass from all active models, ensuring maximum confluence.
2. JINN Threshold Fusion (JTF) Engine
Static entry thresholds can be limiting in a dynamic market. The JTF engine replaces them with a robust, adaptive "breathing" channel.
• Kalman Filter Core: A noise-reducing, parametric filter that provides a smooth, responsive centre for the entry bands.
• Exponentially Weighted Quantile (EWQ): A non-parametric, robust measure of the signal's recent distribution, resistant to outliers.
• Dynamic Fusion: The JTF engine intelligently fuses these two methodologies. In stable conditions, it can blend them; in volatile conditions, it can be configured to use the "Minimum Width" of the two, ensuring your entry criteria are always the most statistically relevant.
3. JINN Pattern Veto (JPV) with Dynamic Time Warping
The definitive filter for behavioural edge and pattern recognition. The JPV moves beyond value-based analysis to analyse the shape of market dynamics.
• Dynamic Time Warping (DTW): A powerful algorithm from computer science that compares the similarity of time series.
• Pattern Veto: Define a "toxic" price action template—a pattern that has historically preceded failed signals. If the JPV detects this pattern, it will veto an otherwise valid trade, providing a sophisticated layer of qualitative, shape-based filtering.
4. JINN Flow VWAP
This is not a standard VWAP. The JINN Flow VWAP is an institutionally-aware variant that analyses volume dynamics to create a "liquidity pressure" band. It helps visualise and gate trades based on the probable activity of larger market participants, offering a nuanced view of where significant flow is occurring.
III. The Advanced Model Suite: Your Pre-Built Quantitative Toolkit
JINN provides you with a turnkey suite of institutional-grade models, saving you thousands of hours of research and development.
1. Auto-Tuning Hyperparameters Engine (Online Meta-Learning)
Markets evolve. A static strategy is an incomplete strategy. JINN’s Auto-Tuning engine is a meta-learning layer inspired by the Hedge (EWA) algorithm, designed to combat alpha decay.
• Portfolio of Experts: It treats a curated set of internal strategic presets as a portfolio of "experts".
• Adaptive Weighting: It runs an online learning algorithm that continuously measures the risk-adjusted performance of each expert (using a sophisticated reward function blending Expected Value and Brier Score).
• Dynamic Adaptation: The engine dynamically allocates more influence to the expert strategy that is performing best in the current market regime, allowing JINN’s core logic to adapt without manual intervention.
2. Lorentzian Classification and PCA-Lite EigenTrend
• Lorentzian Engine: A powerful probabilistic classifier that generates a continuous probability (0-1) of market state. Its adaptive, volatility-scaled distribution is specifically designed to handle the "fat tails" and non-Gaussian nature of financial returns.
• PCA-Lite EigenTrend: A Principal Component Analysis engine. It reduces the complex, multi-dimensional data from the Technical and Order-Flow ensembles into a single, maximally descriptive "EigenTrend". This factor represents the dominant, underlying character of the market, providing a pure, decorrelated input for the Lorentzian engine and other modules.
3. Adaptive Markov Chain Model
A forward-looking, state-based model that calculates the probability of the market transitioning between Uptrend, Downtrend, and Sideways states. Our implementation is academically robust, using an EMA-based adaptive transition matrix and Laplace Smoothing to ensure stability and prevent model failure in sparse data environments.
IV. The Execution Layer: JINN Execution Latch Options
A good signal is worthless without intelligent execution. The JINN Execution Latch is a suite of micro-rules and safety mechanisms that govern the "last mile" of a trade, ensuring signals are executed only under optimal, low-risk conditions. This is your final pre-flight check.
• Execution Latch and Dynamic Cool-Down: A core safety feature that enforces a dynamic cool-down period after each trade to prevent over-trading in choppy, whipsaw markets. The latch duration intelligently adapts, using shorter periods in low-volatility and longer periods in high-volatility environments.
• Volatility-Scaled Real-Time Threshold: A sophisticated gate for real-time entries. It dynamically raises the entry threshold during sudden spikes in volatility, effectively filtering out noise and preventing entries based on erratic, unsustainable price jerks.
• Noise Debounce: In market conditions identified as "noisy" by the Shannon Entropy module, this feature requires a real-time signal to persist for an extra tick before it is considered valid. This is a simple but powerful heuristic to filter out fleeting, insignificant price flickers.
• Liquidity Pressure Confirmation: An institutional-grade check. This gate requires a minimum threshold of "Liquidity Pressure" (a measure of volume-driven momentum) to be present before validating a real-time signal, ensuring you are entering with market participation on your side.
• Time-of-Day (ToD) Weighting: A practical filter that recognises not all hours of the trading day are equal. It can be configured to automatically raise entry thresholds during historically low-volume, low-liquidity sessions (e.g., lunch hours), reducing the risk of entering trades on "fake" moves.
• Adaptive Expectancy Gate: A self-regulating feedback mechanism. This gate monitors the strategy's recent, realised performance (its Expected Value). If the rolling expectancy drops below a user-defined threshold, the system automatically tightens its entry criteria, becoming more selective until performance recovers.
• Bar-Close Quantile Confirmation: A final layer of confirmation for bar-close signals. It requires the signal's final score to be in the top percentile (e.g., 85th percentile) of all signal scores over a lookback period, ensuring only the highest conviction signals are taken.
V. The Contextual and Ensemble Frameworks
1. Multi-Factor Ensembles and Bayesian Fusion
JINN is built on the principle of diversification. Its signals are derived from two comprehensive, fully customizable ensembles:
• Technical Ensemble: A weighted combination of over a dozen technical features, from cyclical analysis (MAMA, Hilbert Transforms) and momentum (Fisher Transform) to trend efficiency (KAMA, Fractal Efficiency Ratio).
• Order-Flow Ensemble: A deep dive into market microstructure, incorporating Volume Delta, Absorption, Imbalance, and Delta Divergence to decode institutional footprints.
• Bayesian Fusion: Move beyond simple AND/OR logic. JINN’s Bayesian engine allows you to probabilistically combine evidence from trend and order-flow filters, weighing each according to its perceived reliability to derive a final posterior probability.
2. Context-Aware Framework and Entropy Engine
JINN understands that a successful strategy requires not just a good entry, but an intelligent exit and a dynamic approach to risk.
• Shannon Entropy Filter: A direct application of information theory. JINN quantifies market randomness and allows you to set a precise entropy ceiling to automatically halt trading in unpredictable, high-entropy conditions.
• Adaptive Exits and Regime Awareness : The script uses its entropy-derived regime awareness to dynamically scale your Take Profit and Trailing Stop parameters. It can be configured to automatically take smaller profits in choppy markets and let winners run in strong trends, hard-coding adaptive risk management into your system.
VI. The Dashboard: Your Mission Control
JINN features a dynamic, dual-mode dashboard that provides a comprehensive, real-time overview of the entire system's state.
Mode 1: Signal Gate Metrics Dashboard
This dashboard is your pre-flight checklist. It displays the real-time Pass/Fail/Off status of every single gating and filtering component within JINN, including:
• Core Ensembles: Technical and Order-Flow Ensemble status.
• Trend Filters: VWAP, VWMA, ADX, ATR Slope, and Linear Regression Angle gates.
• Advanced Models: Dual-Lorentzian Consensus, Markov Probability, and JPV Veto status.
• Regime and Safety: Shannon Entropy, Execution Latch, and Expectancy Gate status.
• Final Confirmation: A master "All Hard Filters" status, giving you an at-a-glance confirmation of system readiness.
Mode 2: Quantitative Metrics Dashboard
This dashboard provides a high-level, institutional-style data readout of the current market state, as seen through JINN's analytical lens. It includes over 60 key metrics for both Signal Gate and Quantitative Metrics, such as:
• Ensemble and Confidence Scores: The raw numerical output of the Technical, Order-Flow, and Lorentzian models.
• Volatility and Volume Analysis: Realised Volatility (%), Relative Volume, Volume Sigma Score, and ATR Z-Score.
• Momentum and Market Position: ADX, RSI Z-Score, VWAP Distance (%), and Distance from 252-Bar High/Low.
• Regime Metrics: The numerical value of the Shannon Entropy score and the Model Confidence score.
VII. The User as the Head Strategist
With over 178 meticulously designed user inputs, JINN is the ultimate "glass box" engine. The internal code is proprietary, but the control surface is transparent and grants you architectural-level command.
• Prototype Sophisticated Strategies: Test complex, multi-model theses at your own pace that would otherwise take weeks of coding. Want to test a strategy that uses a Lorentzian classifier driven by the EigenTrend, arbitrated by JSA in "highest score" mode, and filtered by a strict Markov trend gate? These can be configured and unified.
• Tune the Engine to Any Market: The inputs provide the control surface to optimise JINN's behaviour for specific assets and timeframes, from crypto scalping to swing trading indices.
• Build Trust Through Configuration: The granular controls allow you to align the script's behaviour precisely with your own market view, building trust in your own deployment of the tool.
JINN is a commitment. It is a tool for the serious analyst who seeks to move from discretionary trading to a systematic, quantitative, and adaptive approach. If this aligns with your philosophy, we invite you to apply for access.
Disclaimer
This script is for informational and educational purposes only. It does not constitute financial, investment, or trading advice, nor is it a recommendation to buy or sell any asset.
All trading and investment decisions are the sole responsibility of the user. It is strongly recommended to thoroughly test any strategy on a paper trading account for at least one week before risking real capital.
Trading financial markets involves a high risk of loss, and you may lose more than your initial investment. Past performance is not indicative of future results. The developer is not responsible for any losses incurred from the use of this script.
JINN (Joint Investment Neural and Network) represents a paradigm shift from static indicators to a living, adaptive analytical ecosystem. Traditional tools provide a fixed snapshot of the market. JINN operates on a fundamentally different premise: it treats the market as a dynamic, regime-driven system. It processes market data through a hierarchical suite of advanced, interacting models, arbitrates their outputs through a rules-based engine, and adapts its own logic in real-time.
It is designed as a complete framework for traders who think in terms of statistical edge, market regimes, probabilistic outcomes, and adaptive risk management.
II. The JINN Branded Architecture: Your Command and Control Centre
JINN’s power emerges from the synergy of its proprietary, branded architectural components. You do not simply "use" JINN; you command its engines.
1. JINN Signal Arbitration (JSA) Engine
The heart of JINN. The JSA is your configurable arbitration desk for weighing evidence from all internal models. As the Head Strategist, you define the entire arbitration philosophy:
• Priority and Weighting: Define a "chain of command". Specify which model's opinion must be considered first and assign custom weights to their outputs, directly controlling the hierarchy of your analytical flow.
• Arbitration Modes:
First Wins: For high-conviction, rapid signal deployment based on your most trusted leading model.
Highest Score: A "best evidence" approach that runs a full analysis and selects the signal with the highest weighted probabilistic backing.
Consensus: An ultra-conservative, "all-clear" mode that requires a unanimous pass from all active models, ensuring maximum confluence.
2. JINN Threshold Fusion (JTF) Engine
Static entry thresholds can be limiting in a dynamic market. The JTF engine replaces them with a robust, adaptive "breathing" channel.
• Kalman Filter Core: A noise-reducing, parametric filter that provides a smooth, responsive centre for the entry bands.
• Exponentially Weighted Quantile (EWQ): A non-parametric, robust measure of the signal's recent distribution, resistant to outliers.
• Dynamic Fusion: The JTF engine intelligently fuses these two methodologies. In stable conditions, it can blend them; in volatile conditions, it can be configured to use the "Minimum Width" of the two, ensuring your entry criteria are always the most statistically relevant.
3. JINN Pattern Veto (JPV) with Dynamic Time Warping
The definitive filter for behavioural edge and pattern recognition. The JPV moves beyond value-based analysis to analyse the shape of market dynamics.
• Dynamic Time Warping (DTW): A powerful algorithm from computer science that compares the similarity of time series.
• Pattern Veto: Define a "toxic" price action template—a pattern that has historically preceded failed signals. If the JPV detects this pattern, it will veto an otherwise valid trade, providing a sophisticated layer of qualitative, shape-based filtering.
4. JINN Flow VWAP
This is not a standard VWAP. The JINN Flow VWAP is an institutionally-aware variant that analyses volume dynamics to create a "liquidity pressure" band. It helps visualise and gate trades based on the probable activity of larger market participants, offering a nuanced view of where significant flow is occurring.
III. The Advanced Model Suite: Your Pre-Built Quantitative Toolkit
JINN provides you with a turnkey suite of institutional-grade models, saving you thousands of hours of research and development.
1. Auto-Tuning Hyperparameters Engine (Online Meta-Learning)
Markets evolve. A static strategy is an incomplete strategy. JINN’s Auto-Tuning engine is a meta-learning layer inspired by the Hedge (EWA) algorithm, designed to combat alpha decay.
• Portfolio of Experts: It treats a curated set of internal strategic presets as a portfolio of "experts".
• Adaptive Weighting: It runs an online learning algorithm that continuously measures the risk-adjusted performance of each expert (using a sophisticated reward function blending Expected Value and Brier Score).
• Dynamic Adaptation: The engine dynamically allocates more influence to the expert strategy that is performing best in the current market regime, allowing JINN’s core logic to adapt without manual intervention.
2. Lorentzian Classification and PCA-Lite EigenTrend
• Lorentzian Engine: A powerful probabilistic classifier that generates a continuous probability (0-1) of market state. Its adaptive, volatility-scaled distribution is specifically designed to handle the "fat tails" and non-Gaussian nature of financial returns.
• PCA-Lite EigenTrend: A Principal Component Analysis engine. It reduces the complex, multi-dimensional data from the Technical and Order-Flow ensembles into a single, maximally descriptive "EigenTrend". This factor represents the dominant, underlying character of the market, providing a pure, decorrelated input for the Lorentzian engine and other modules.
3. Adaptive Markov Chain Model
A forward-looking, state-based model that calculates the probability of the market transitioning between Uptrend, Downtrend, and Sideways states. Our implementation is academically robust, using an EMA-based adaptive transition matrix and Laplace Smoothing to ensure stability and prevent model failure in sparse data environments.
IV. The Execution Layer: JINN Execution Latch Options
A good signal is worthless without intelligent execution. The JINN Execution Latch is a suite of micro-rules and safety mechanisms that govern the "last mile" of a trade, ensuring signals are executed only under optimal, low-risk conditions. This is your final pre-flight check.
• Execution Latch and Dynamic Cool-Down: A core safety feature that enforces a dynamic cool-down period after each trade to prevent over-trading in choppy, whipsaw markets. The latch duration intelligently adapts, using shorter periods in low-volatility and longer periods in high-volatility environments.
• Volatility-Scaled Real-Time Threshold: A sophisticated gate for real-time entries. It dynamically raises the entry threshold during sudden spikes in volatility, effectively filtering out noise and preventing entries based on erratic, unsustainable price jerks.
• Noise Debounce: In market conditions identified as "noisy" by the Shannon Entropy module, this feature requires a real-time signal to persist for an extra tick before it is considered valid. This is a simple but powerful heuristic to filter out fleeting, insignificant price flickers.
• Liquidity Pressure Confirmation: An institutional-grade check. This gate requires a minimum threshold of "Liquidity Pressure" (a measure of volume-driven momentum) to be present before validating a real-time signal, ensuring you are entering with market participation on your side.
• Time-of-Day (ToD) Weighting: A practical filter that recognises not all hours of the trading day are equal. It can be configured to automatically raise entry thresholds during historically low-volume, low-liquidity sessions (e.g., lunch hours), reducing the risk of entering trades on "fake" moves.
• Adaptive Expectancy Gate: A self-regulating feedback mechanism. This gate monitors the strategy's recent, realised performance (its Expected Value). If the rolling expectancy drops below a user-defined threshold, the system automatically tightens its entry criteria, becoming more selective until performance recovers.
• Bar-Close Quantile Confirmation: A final layer of confirmation for bar-close signals. It requires the signal's final score to be in the top percentile (e.g., 85th percentile) of all signal scores over a lookback period, ensuring only the highest conviction signals are taken.
V. The Contextual and Ensemble Frameworks
1. Multi-Factor Ensembles and Bayesian Fusion
JINN is built on the principle of diversification. Its signals are derived from two comprehensive, fully customizable ensembles:
• Technical Ensemble: A weighted combination of over a dozen technical features, from cyclical analysis (MAMA, Hilbert Transforms) and momentum (Fisher Transform) to trend efficiency (KAMA, Fractal Efficiency Ratio).
• Order-Flow Ensemble: A deep dive into market microstructure, incorporating Volume Delta, Absorption, Imbalance, and Delta Divergence to decode institutional footprints.
• Bayesian Fusion: Move beyond simple AND/OR logic. JINN’s Bayesian engine allows you to probabilistically combine evidence from trend and order-flow filters, weighing each according to its perceived reliability to derive a final posterior probability.
2. Context-Aware Framework and Entropy Engine
JINN understands that a successful strategy requires not just a good entry, but an intelligent exit and a dynamic approach to risk.
• Shannon Entropy Filter: A direct application of information theory. JINN quantifies market randomness and allows you to set a precise entropy ceiling to automatically halt trading in unpredictable, high-entropy conditions.
• Adaptive Exits and Regime Awareness : The script uses its entropy-derived regime awareness to dynamically scale your Take Profit and Trailing Stop parameters. It can be configured to automatically take smaller profits in choppy markets and let winners run in strong trends, hard-coding adaptive risk management into your system.
VI. The Dashboard: Your Mission Control
JINN features a dynamic, dual-mode dashboard that provides a comprehensive, real-time overview of the entire system's state.
Mode 1: Signal Gate Metrics Dashboard
This dashboard is your pre-flight checklist. It displays the real-time Pass/Fail/Off status of every single gating and filtering component within JINN, including:
• Core Ensembles: Technical and Order-Flow Ensemble status.
• Trend Filters: VWAP, VWMA, ADX, ATR Slope, and Linear Regression Angle gates.
• Advanced Models: Dual-Lorentzian Consensus, Markov Probability, and JPV Veto status.
• Regime and Safety: Shannon Entropy, Execution Latch, and Expectancy Gate status.
• Final Confirmation: A master "All Hard Filters" status, giving you an at-a-glance confirmation of system readiness.
Mode 2: Quantitative Metrics Dashboard
This dashboard provides a high-level, institutional-style data readout of the current market state, as seen through JINN's analytical lens. It includes over 60 key metrics for both Signal Gate and Quantitative Metrics, such as:
• Ensemble and Confidence Scores: The raw numerical output of the Technical, Order-Flow, and Lorentzian models.
• Volatility and Volume Analysis: Realised Volatility (%), Relative Volume, Volume Sigma Score, and ATR Z-Score.
• Momentum and Market Position: ADX, RSI Z-Score, VWAP Distance (%), and Distance from 252-Bar High/Low.
• Regime Metrics: The numerical value of the Shannon Entropy score and the Model Confidence score.
VII. The User as the Head Strategist
With over 178 meticulously designed user inputs, JINN is the ultimate "glass box" engine. The internal code is proprietary, but the control surface is transparent and grants you architectural-level command.
• Prototype Sophisticated Strategies: Test complex, multi-model theses at your own pace that would otherwise take weeks of coding. Want to test a strategy that uses a Lorentzian classifier driven by the EigenTrend, arbitrated by JSA in "highest score" mode, and filtered by a strict Markov trend gate? These can be configured and unified.
• Tune the Engine to Any Market: The inputs provide the control surface to optimise JINN's behaviour for specific assets and timeframes, from crypto scalping to swing trading indices.
• Build Trust Through Configuration: The granular controls allow you to align the script's behaviour precisely with your own market view, building trust in your own deployment of the tool.
JINN is a commitment. It is a tool for the serious analyst who seeks to move from discretionary trading to a systematic, quantitative, and adaptive approach. If this aligns with your philosophy, we invite you to apply for access.
Disclaimer
This script is for informational and educational purposes only. It does not constitute financial, investment, or trading advice, nor is it a recommendation to buy or sell any asset.
All trading and investment decisions are the sole responsibility of the user. It is strongly recommended to thoroughly test any strategy on a paper trading account for at least one week before risking real capital.
Trading financial markets involves a high risk of loss, and you may lose more than your initial investment. Past performance is not indicative of future results. The developer is not responsible for any losses incurred from the use of this script.
Script sob convite
Somente usuários aprovados pelo autor podem acessar este script. Você precisará solicitar e obter permissão para usá-lo. Normalmente, essa permissão é concedida após o pagamento. Para obter mais detalhes, siga as instruções do autor abaixo ou entre em contato diretamente com INVESTATI.
A TradingView NÃO recomenda pagar ou usar um script, a menos que você confie totalmente em seu autor e entenda como ele funciona. Você também pode encontrar alternativas gratuitas e de código aberto em nossos scripts da comunidade.
Instruções do autor
DO NOT request access in the comment's section. Please visit https://www.investati.com to gain access to JINN Quantitative Trading and Execution Engine.
Aviso legal
As informações e publicações não se destinam a ser, e não constituem, conselhos ou recomendações financeiras, de investimento, comerciais ou de outro tipo fornecidos ou endossados pela TradingView. Leia mais nos Termos de Uso.
Script sob convite
Somente usuários aprovados pelo autor podem acessar este script. Você precisará solicitar e obter permissão para usá-lo. Normalmente, essa permissão é concedida após o pagamento. Para obter mais detalhes, siga as instruções do autor abaixo ou entre em contato diretamente com INVESTATI.
A TradingView NÃO recomenda pagar ou usar um script, a menos que você confie totalmente em seu autor e entenda como ele funciona. Você também pode encontrar alternativas gratuitas e de código aberto em nossos scripts da comunidade.
Instruções do autor
DO NOT request access in the comment's section. Please visit https://www.investati.com to gain access to JINN Quantitative Trading and Execution Engine.
Aviso legal
As informações e publicações não se destinam a ser, e não constituem, conselhos ou recomendações financeiras, de investimento, comerciais ou de outro tipo fornecidos ou endossados pela TradingView. Leia mais nos Termos de Uso.