Companies' financial filings (Form-10K) have grown over time and are now equivalent in length to a 240-page novel. Even for analysts with modest coverage universes, parsing 10-Ks for relevant information can be time consuming, if not intractable. S&P Global Market Intelligence's Machine Readable Filings enables the process to be readily automated for Natural Language Processing (NLP). Two of the largest sections, the Management Discussion & Analysis (MD&A) and the Risk Factors sections, are information rich. This presentation will walk investors and analysts through the steps required to systematically extract value from the textual content within filings.
- How are different 10-K and 10-Q formatting conventions cleaned, reconciled, and standardized for a machine readable state?
- Leveraging a technique from linguistics and NLP called Cosine Similarity, significant changes in each section can be identified which has demonstrated historical outperformance.
- Comparing the incremental information content of 10-K sections.
- How Quality and Trend Following (Momentum) relate to 10-K changes.
- Evaluating the impact of small versus large textual revisions on stock returns and volatility.
Please join Joe Gits, President of Social Market Analytics and Frank Zhao, Quantamental Research, S&P Global Market Intelligence, for a wide-ranging discussion on the topic as we continue our webinar series presented by the Quantamental Research Group. This series grants you direct access to the team behind our Quantamental Research as they share actionable intelligence on how to enhance your modeling process leveraging the innovative data suite available via Xpressfeed™ and Snowflake.