The second day of Neudata’s Winter Summit explored the unique data use cases that investors are exploring, including new ways of working with natural language processing technology and how investors are still finding alpha in transactional data.

Our first speaker was Robert Levy from Cboe Global Markets, who discussed how investors can use the option volatility surface for informed equities trading.

Next, Chafic Merhy from Ostrum Asset Management discussed his firm’s machine learning-led approach to evaluating extra-financial risks for sovereign bonds. He discussed how his approach integrates financial and economic indicators with ESG data, and the importance of creating interpretable ML models (i.e., models that aren’t black boxes).

Data vendors Thomas and YouGov then spoke to attendees. Hamid Benbrahim presented on how investors could use industrial sourcing data to develop a viewpoint on the health global industry, while Ryan Gross spoke to the power of using survey data to help investors make better investing decisions.

Our conversation transitioned into data privacy topics as Neudata’s Jose Elias Terriquez explained upcoming changes to Apple’s iOS, which could throw parts of the alt data ecosystem that rely on mobile app panels into disarray next year. Jose predicted that Apple’s decision to force users to opt-in to certain data collection policies could compel vendors to work with smaller panels, lose mobile data as one of their inputs, or cause dataset terminations among vendors who aren’t able to keep up with changes. Vendors that rely on third-party providers to source app install and usage metrics, mobile device and network penetration, mobile clickstream and geolocation data could see the highest risks, he explained.

After a quick networking break, Sebastian Owen from HIPSO discussed new benchmarks for web-scraping technology.

HIPSO’s CTO Andrii Pylypenko then joined Sanne De Boer from Voya Investment Management, Poorya Ferdowsi from Bristol Gate Capital Partners and Neudata’s Amy Dafnis for a discussion of natural language processing-derived datasets. The group explained that NLP is being used to help quants and fundamental investors ingest more data than ever before, while other investors are using it to identify risks within their models. Amy explained that NLP data vendors who offer some degree of customisation are usually more successful in working with investors.

Amy next tackled the thorny topic of dataset terminations in 2020. As we predicted at our London Summit in 2019, this year spelled the death of both well-known and obscure datasets. In conversations with Neudata, vendors cited several reasons for removing their dataset from the market, including a lack of commercial success, pausing their data sales efforts to the investment management vertical, and regulatory/compliance concerns. This year, location, web and app tracking, web-scraping and sentiment datasets were most likely to be terminated, Amy explained.

Our next two speakers from Yewno and Quant IP presented case studies on how their data can be used by investors. Yewno’s Bernardo Bravo spoke about the use of knowledge graph technology to power thematic investing, while Lucas von Reuss explained the importance of using patent analytics to track and measure innovation.

Our final panel of the day saw Neudata’s Ian Webster return to the virtual stage, in conversation with Max Chapados and Invesco’s Edward Leung. The trio discussed how to find alpha within transactional datasets, historically some of the most popular and expensive data products for investment managers. A greater focus on consumer privacy could spell trouble for some products, but the group advised listeners that any new transactional products coming out — particularly in less saturated markets — still held value for the right investor.

Photo by JR Korpa on Unsplash