StockPulse, a German-based data analytics company, is specialized on emotional data intelligence, harvested from social media and news outlets, with particular focus on financial markets.
Emotional data is the next level of data processing and data driven decision making. While raw and unprocessed data, for example collected from social media, provides only isolated and limited pieces of information, emotional data matches and connects the dots and - most important - measures emotions that are inherent to the information. Like André Kostolany we believe that facts make up only 10% of the stock market - everything else is driven by psychology. Emotional data measures this psychology.
Resulting data-sets are refined by quantification, classification, interpretation and connection.
Emotional data intelligence can be used in various fields for more efficient decision making and one successful example are trading models. Different financial institutions use StockPulse data to optimize their decision making processes.
StockPulse collects and evaluates millions of opinions and news articles about more than 40.000 companies every day and shows at a glance, how much and in what kind of mood people discuss topics related to financial markets. The crawler engines collect data for regions all over the world, North America, Europe, and Asia with a large coverage on Chinese companies. Using latest Big Data and machine learning technologies, communication on stocks, indices, commodities, forex and major market events is monitored and evaluated in real-time around the clock. Text analysis is done in three languages: English, German and Chinese. Moreover, the software automatically generates Long / Short-signals for different asset classes. Currently, the StockPulse database contains more than 1,5 billion historical data points for more than eight years, which enables comprehensive and significant back testing.
Subscription Details
The following options are available: Free trial up to 2 months Annual subscription with payment option monthly, quarterly or annual.
Update Granularity
The data is generally updated in real-time. Depending on the use case data can be aggregated on different time periods. We provide 3 default periods: 10 minutes, hourly, daily.
StockPulse refines emotional data which is described as structured meta data based on raw data collected from the internet. The collection includes news about financial entities and activities, companies, including ESG (environmental, social and governance factors). Crawlers collect the data from publicly available web sources and save the data in a structured format to databases where the data is further processed. Also a large number of information on non-listed companies are completing our data set. From a collection point of view, we are constantly extend our sources and also track niche online communities as well as Twitter & Co.
1 TB full history, more than 1 Billion historic messages, several Billions of meta data and more than 45 Million users/authors. We process several Millions of messages per day of which about 1.5 Million are relevant to be included in our data products.
Regional Coverage
North America, Europe (Germany, UK), China, Hong-Kong, Singapore etc.
History
For US and EU stocks we have more than 7 years of historical data.
For Asian stocks we have 3 years of historical data.
Cadence
Update frequency of the datasets is real-time. Depending on the needs of customer or on the particular use case data can be aggregated on different periods, minutes, hours, days, weeks, or months.
Themes Covered
All discussions topics in financial markets in general and enterprise news in particular. For example FOMC meetings, EZB conferences, merger & acquisition, bankruptcy, earnings, etc.
Known Biases
Data about US stocks is based mostly on Twitter and Stocktwits
Data on EU and Asian companies is mostly from specific message boards, blogs, chats, comments to news articles, etc.