📊 Professional EDA Dashboard

📋 Dashboard Overview

• Filter Variable: Symbol (10 unique values)
• Numeric Metrics: 92 columns analyzed
• Categorical Variables: 1 columns identified

📊 STOCK PRICE & BASIC DATA

• Source: Direct feeds from stock exchanges (NASDAQ, NSE, BSE, NYSE) via market data providers like Bloomberg, Refinitiv Eikon, or FactSet
• Open, High, Low, Close, Volume: Raw tick data aggregated from the exchange order books
• Adj Close: Calculated and provided by data vendors (e.g., Yahoo Finance, Bloomberg) by adjusting for corporate actions like dividends and stock splits

📈 RETURNS & VOLATILITY

• Source: Calculated from the Exchange-Sourced Price Data
• Daily_Return: (Today's Close / Yesterday's Close) - 1
• Volatility: The standard deviation of Daily_Returns over a rolling window, annualized

📊 TECHNICAL INDICATORS

• Source: Calculated exclusively from Exchange-Sourced Price and Volume Data
• SMA, EMA, RSI, MACD, Bollinger Bands: Standard calculations applied to the Close price and Volume

🤖 ALGORITHM SCORES & SIGNALS

• AMSF_Score/Signal: Generated by the Multimodal Sentiment Fusion Engine processing Text, Audio, and Visuals
• Text: Transcripts from earnings calls (from Bloomberg/Refinitiv) and SEC filings (from EDGAR database), analyzed with FinBERT
• Audio: The audio track of earnings calls, processed with Whisper for speech-to-text and prosodic analysis (pitch, tempo, pauses) to detect stress
• Visuals: Slides from investor presentations, analyzed by CLIP to interpret charts, design tone, and layout cues
• FTCS_Score/Formula: Output of a separate, proprietary quantitative model based on market data
• SignalConfidence_Entropy: A direct measure from the Self-Assessing AI quantifying prediction uncertainty using techniques like Monte Carlo dropout

🔄 MARKET REGIMES & PATTERNS

• RABE_Regime: Identified by statistical models (HMM, change-point detection) analyzing Price Data, Macro Sentiment, and Volatility Regimes
• Hurst Exponent: Calculated from the price time series to measure trend persistence
• Fundamental_DecayFactor: Determined by the NeuroSync Memory Agent tracking how new earnings calls, news, and filings make old fundamental data obsolete

⚡ VOLATILITY & RISK MEASURES

• Source: Calculated by applying advanced statistical models to price data and options market data to create volatility-adjusted versions of standard indicators

🧠 ADVANCED ALGORITHM COMPONENTS

• ARFS_RobustFundamental: A score generated by analyzing company financials with NLP to identify accounting quality and resistance to manipulation
• MHRP_ReturnForecast: A prediction generated by machine learning models trained on historical prices, macro data, and fused multimodal sentiment scores
• FinalScore/Signal: The consensus output after synthesizing all other scores, including Network, Causal, and Multimodal inputs

🕸️ NETWORK & CLUSTER ANALYSIS

• Source: Built by calculating a dynamic correlation matrix of returns for all stocks within an index (e.g., NASDAQ-100, NIFTY 50)
• AHF_GNN_Score: A stock's rating based on its connections in this financial network
• AHF_ClusterID: The group of stocks it is most correlated with, validated against sector-specific news sentiment

🔍 CAUSAL ANALYSIS

• CAAE_CausalAlpha: The portion of returns attributed to genuine skill, isolated using causal inference models
• CAAE_Attribution: A breakdown of which specific factor caused a move (e.g., CEO's hesitant tone, sector-wide news negativity)

⚡ HIGH-FREQUENCY TRADING (HFT) DATA

• Source: Logs from a dedicated HFT system connected directly to exchange co-location servers (NASDAQ, NSE)
• SPRINT_ExecutionCost: Calculated from the actual bid-ask spreads and market impact observed on the exchange

📊 MULTI-RESOLUTION ANALYSIS

• Source: The result of running the same analysis on market data sampled at different timeframes
• MRS_Uncertainty: High uncertainty indicates conflicting signals across timeframes, key input for the Abstention Engine

🔒 PRIVACY & FEDERATED LEARNING

• Source: Outputs from a Federated Learning System where multiple institutions train models without sharing proprietary trade data

📉 RISK MANAGEMENT

• Source: Calculated from the portfolio's holdings and their historical returns from exchange data
• DIR_ConditionalVaR: A worst-case loss estimate derived from the tail of the portfolio's return distribution

🌊 MARKET ENVIRONMENT

• NEAD_MarketEntropy: A measure of market chaos calculated from disagreement between different data modalities

📐 FACTOR INVESTING

• Source: A multi-factor model using data from financial databases (MSCI Barra, Axioma)
• WFO_Factor1/2/3/4: Exposures to common risk factors like Value, Momentum, Size, and Quality

💰 TRADING COSTS

• Source: Output from a Transaction Cost Analysis (TCA) model using historical trade data and real-time exchange liquidity data

🛡️ ROBUSTNESS & PERSISTENCE

• TPS_RobustnessScore: Measures a strategy's performance stability across different market regimes
• EnhancedFinalScore: The FinalScore penalized by a low RobustnessScore, ensuring only durable signals are acted upon