7 Use Cases for NLP in Large Hedge Funds

Interested in learning more?

Natural language processing (NLP) is capable of producing broadly aggregated and objective scoring and tagging on text assets with speed, accuracy, and transparency. Investors can use Symphony’s NLP to access comprehensive, real-time insights into equities and other asset classes — well ahead of when this information is furnished in a company’s reported information or revealed by the market. Symphony delivers these insights via dashboards or APIs integrated into internal systems, with the option for customization. Here are 7 examples of how large hedge funds are currently using Symphony’s NLP solutions.

Key Drivers Analysis (Earnings Call Transcripts)

Symphony’s NLP parses earnings call transcripts to reveal sentiment around the most important fundamental key drivers, including Margin Forecasts, Market Position, CapEx, Capital Returns, Guidance, Wages, and others. This language modeling and scoring system assesses each driver from the perspective of the equity and provides transparency to the underlying extractions. Asset managers and sell-side institutions such as Bank of America, Evercore, and QMA have used our sentiment scores for backtesting.

Weather-Related Analysis (News)

Our custom weather model allows you to create a long-short portfolio strategy that sorts the cross-section of commodity futures contracts according to a “hazard fear” signal. The weather model is based on 149 event types, and runs across a large set of news sources (via LexisNexis). The analysis on these event types leverages active attention to weather, disease, geopolitical, or economic threats. Fear of Hazards in Commodity Futures Market (SSRN).

Environmental, Social, Governance (ESG) and Sustainability Scoring (News)

Symphony provides systematic scoring on a highly configurable ESG framework applied to news articles. Our systematic method delivers transparent, granular, and human-level insights at scale and in real time. One model is deployed to analyze all companies, so you can understand how ESG profiles vary across peer groups. Additional features include Materiality and Greenwashing Analysis, which measures the difference in perception between self-reported and third-party reported information.

Business Classifications (Filings)

The description of operations in company filings can shed light on the differences in performance among publicly traded companies. Business model effects (asset types and rights) explain performance heterogeneity better than industry classifications, and are a powerful addition to risk and exposure. Symphony’s NLP trains the model to parse certain sections (such as the Business Description), auto-categorize those sections, and tag a company to the appropriate classifications. The data can also be viewed in a time series, which reveals how company business models shift over time. Do Business Models Matter? (MIT)

Earnings KPI Extraction (Press Releases)

Press Releases include key disclosures of financial performance and guidance. Symphony’s NLP processes and databases the valuable information contained within these documents (KPIs, forecast) quickly and accurately, and provides the data to analysts, PMs, and systematic teams for rapid analysis and response. With our news API, press releases are processed across a broad equity universe. An advanced NLP model is applied to deliver KPIs with normalized values, magnitudes, and time periods in under two minutes.

Research Tagging (Third Party Content)

Third-party research content is insufficiently and inconsistently tagged, leading to challenges in the discovery and timely consumption of content. Symphony provides a robust tagging framework that specializes in classifying thematic and topical content, as well as offering a more detailed targeting of asset classes and other entities than typical tagging platforms.

Buy-Side Research Analysis (Proprietary Content)

Covering analysts and portfolio managers write and update theses on investments, including recommendations and targets. However, the unstructured aspects of their communications contain additional signals that can help substantially with determining true analyst conviction, sentiment, and implicit timing/sizing recommendations. Symphony can securely analyze in-house research notes, and deliver a per-document score that provides this signal to the firm.

Interested in learning more?

You may also like