News Aggreagation

Reading, evaluating, and assigning news

The news summarization module reads various national and international news sources. These include stock market, economic, and political news. The news is extracted from online data streams. Each news item is then processed by an AI, specifically a large language model (LLM). The AI evaluates each news item and its impact on each individual security. The resulting trend assessment indicates whether the news is expected to have a positive, negative, or neutral effect on the price of the security. News with a neutral impact is not further processed.

If a piece of news is assigned to a security and a trend has been determined, the model also evaluates whether the trend is short-term, medium-term, or long-term, as well as the relevance of the news. These KPIs are used to generate a data record. A single news item may be assigned to one or more markets, indices, or stocks. Price forecasts are based on news items stored in the database. These forecasts are available in the form of a database and are used to calculate limits for buy and sell decisions.

News Aggregation Flow

Collaboration with Katastrophenmelder

The professional platform katastrophenmelder.de continuously analyzes news sources. Various sources from different domains are used to detect and evaluate anomalies. A catastrophe scale is generated, drawing on topics such as weather, politics, economy, and military conflicts. The news and sources are continuously monitored. Fortunately, the system only rarely triggers an alarm. It is rare that a threshold is exceeded in such a way that the news qualifies as catastrophic. This makes the system particularly robust.

STS Technology

Anomaly Detection

The module for detecting price anomalies

The anomaly detection module analyzes price behavior in order to identify sudden spikes—either upward or downward. Over a longer time period, it determines the typical range in which a security moves. Significant deviations above or below this range are detected and trigger an alert. Alerts from Katastrophenmelder and the anomaly detection module are manually reviewed and incorporated into the buy decision process.

High Anomaly Detection Low Anomaly Detection

News Processing System

Our advanced news processing system categorizes and analyzes information from multiple sources to provide actionable trading insights.

📰

News Collection

Gathering data from diverse global sources

🔍

Analysis

AI-powered evaluation of news impact

📊

Trend Assessment

Determining positive, negative, or neutral effects

⏱️

Time Classification

Categorizing as short, medium, or long-term impacts