How We Score
The Market
Our proprietary AI pipeline analyzes thousands of financial data points in real-time to bring you actionable intelligence before the market reacts.
1. Natural Language Processing
We ingest thousands of articles, SEC filings, press releases, and social media posts every minute. Our custom LLMs (Large Language Models) are trained specifically on financial terminology, allowing them to comprehend complex market dynamics and differentiate between noise and material news.
2. Sentiment Scoring (Bullish / Bearish)
Once an article is processed, our models determine its core sentiment relative to specific assets or the broader market. The outcome is categorized as Bullish (positive catalyst), Bearish (negative catalyst), or Neutral. This happens in milliseconds, giving you an edge.
3. Market Impact Algorithm
Not all news is created equal. A "Bullish" article about a CEO's new dog has a heavily reduced impact compared to a "Bullish" article about a massive earnings beat. Our algorithm scores the impact on a scale from 1 to 10 based on:
- Historical price action correlation to similar news events
- Source credibility and distribution velocity
- Magnitude of the reported event (e.g., M&A vs Routine updates)
4. Continuous Learning
Our models are constantly backtesting their predictions against real-time market data. When a scored event occurs, the system analyzes the subsequent price action of the affected assets and adjusts its future weights accordingly, ensuring our intelligence grows sharper every day.