Methodology & Disclosures
Transparent research methodology, performance calculations, and systematic approach to stock analysis
System Overview
Our platform uses a multi-agent AI architecture that orchestrates specialized analysis agents to provide comprehensive stock research. Rather than relying on a single AI model, we deploy a team of expert agents covering different analytical domains.
Multi-Agent Analysis
Each query is processed by specialized AI agents working in parallel, ensuring comprehensive coverage from multiple analytical perspectives.
Systematic Approach
Our methodology applies structured filters and backtesting to identify opportunities beyond speculation or market sentiment.
Hot Ideas Selection Method
Every Monday at 3 PM New York time, our system scans hundreds of assets to identify up to 10 with the highest short-term potential. This weekly process helps both long-term investors and active traders apply systematic strategies.
Four Core Filters
Fresh News Catalyst
Finds stocks likely to move after company or sector announcements
Earnings Surprises
Targets companies with strong chances of beating forecasts
Price-to-Return Anomalies
Detects momentum and reversal patterns in price action
Web Search Candidates
Identifies companies trending online with potential for abnormal returns
Specialized AI Agents
Our system employs over 20 AI agents across 15 domains, each focused on specific aspects of financial analysis:
Fundamental Analysis
Financial statements, earnings reports, key ratios (P/E, debt levels)
Technical Analysis
Price dynamics, trading volumes, moving averages, support/resistance
News & Sentiment
Media monitoring, sentiment analysis, event significance
Earnings Analysis
Quarterly reports, management statements, forecast tracking
Macro & Industry
Economic indicators, sector trends, interest rates, inflation
Alternative Data
Web analytics, search trends, social sentiment, geospatial data
Performance Calculation Methodology
Portfolio Performance Metrics
Total Return
Cumulative percentage return from the start of the tracking period. Calculated as:(End Value - Start Value) / Start Value × 100
Average Return
Mean weekly return across all tracked periods. Provides insight into consistent performance patterns.
Win Rate
Percentage of positive weeks (positive weeks / total weeks × 100). Measures consistency of positive performance.
Maximum Drawdown
Largest peak-to-trough decline during the tracking period. Measures the worst-case scenario for risk assessment.
Calculation Assumptions
Backtesting & Validation
Our methodology is validated through extensive backtesting using historical market data. The system has demonstrated consistent outperformance against benchmarks like the S&P 500.
Historical Performance
Backtested results show the system often outperforms actively managed mutual funds and index-based ETFs, with improved selection quality when expanding the candidate pool from 100 to 300 assets.
Transparency
All performance data is tracked in real-time with daily updates, allowing for transparent verification of results and methodology effectiveness.
Important Disclosures
Not Investment Advice
All content is provided for informational and educational purposes only. This is not investment advice, and you should conduct your own research or consult a financial advisor before making any investment decisions.
Past Performance: Past performance is not indicative of future results. All investments carry risk, and you may lose some or all of your capital.
Data Accuracy: While we strive for accuracy, we make no warranties about the completeness, reliability, or accuracy of information provided.
Research Purpose: This is non-personalized research for general circulation. It does not take into account your objectives, financial situation, or needs.
Last updated: December 12, 2025