Sternshus publishes financial reports for humans and creates sustainable value indices for investors by reading publicly available data using natural language processing, constructing a graph, and generating an appraisal so that companies are comparable and market prices measurable.
Query, Filter, & Index
Websites, unlike financial information, are optimized for algorithmic search engines. The current state of the art uses natural language processing to understand the intent of long-tail keyword queries provided by their consumers. Search engines return three types of responses:
a list of relevant websites,
entity data from a knowledge base, distilled from webpage content
Sternshus first seeks to replicate the search engine model with publicly available financial information by
constructing & maintaining a knowledge base,
using a custom trained model to extract entities, documents, and classify and
querying with kb entities
The goal is to return an index of performant companies that match customer sustainability goals.
Hedge, Moat, & Markets
Where information is free and transparent, market prices reflect inherent value. The only option is to hedge one's investments against risk. Markets the world over, however, continue to suffer from inefficiencies of information flow. Professional investors maintain in-house research and analysis departments. Expertise is costly.
Retail investors lack access to the same information, though the data is public. Their most reasonably priced alternative is index funds. And although low-cost, these funds invest in companies that many eschew as consumers.
Sternshus estimates the intrinsic value of companies and models competitive advantages (moats) based on both the numbers and text of data using statistical learning and state of the art natural language processing. Both professional and individual investors benefit from timely analysis at a minimal cost.
Annotate, Curate, & Search
Index makers have little incentive to develop custom benchmarks. Financial Service Companies won't retail them. Even if they wish to invest following environmental, social, and good governance (ESG) recommendations, prohibitive fees keep investors trapped in index funds tied to generic benchmarks. Costly mutual funds use unregulated self-reporting by corporations to determine their ESG offerings.
Using transfer learning and domain adaptation, Sternshus annotates reports with sustainability comparisons and curates searchable indices for investors.
Occasionally we release publicly available information. Sign-up and get put on the waitlist for our limited release program.
© 2020 Sternshus.
All rights reserved.