Secrets to Sourcing Corporate Event Data — Including History!

Exploring Corporate Event Data and Volatility (excerpt)

Research efforts and reports focused on event data have one overarching finding in common. Corporate events, especially revisions to previously scheduled events, can cause volatility. Understanding the types of events, event changes and other event-related behaviors can identify previously unknown or under-appreciated sources of alpha and risk.

Data Quality Considerations – The ‘Must-Have’ List

When an academic researcher or quant analyst is developing a systematic review of the near/mid/long-term financial implications that corporate events may have on a company, the data regarding the event must be highly accurate.

For example, an investigation of point-in-time data first requires defining what point-in-time means. There is not one set answer, which is often assumed, and that is a discussion that must be had with the data provider. Then the data you procure must be compared within the context of other market datasets and conditions. Essentially, you are creating a bespoke dataset during your investigation and each component must be highly accurate and well understood.

Following are brief descriptions of the characteristics to consider when sourcing and evaluating corporate event data.

Corporate Event Data Must be Primary and Publicly Sourced

Because publicly traded companies are disclosing information in an increasing number of ways, it is important that you subscribe to primary sourced, publicly available data that is 100% compliant.

• Need ongoing access to the data analysts and engineers who created and maintain the data. They should be available for content and technical questions.
• Researched with a repeatable, well-documented, fact-based methodology.
• Delivered with transparency as to how, when and from where the data is sourced.
• Formatted and delivered so that it is machine readable with specifications and content explanations.

Rigorous Curation of Archived Event History

Using historical data for model backtesting and strategy development is critical for quantitative firms and academic researchers alike.

• The data should be sourced directly from the forward-looking research efforts, keeping history as it was when originally published.
• The historical data is never revisited, tweaked or edited in any way.
• If event type requires updating the history, a separate change-log history detailing those revisits/updates must be maintained.
• Structured into a machine readable format, with detailed specification guides providing delivery, content, structure and key definitions of all terms.
• The data includes all relevant event types. Whether in response to the diminished role of broker research or changes in the regulatory industry, publicly traded companies are finding more ways to divulge information directly to investors. For example, within investor related conferences, there can be several subcategories including analyst days/capital markets days, business updates, company travel, conferences, forums, R&D days, seminars, summits, tradeshows and workshops.