Skip to main content

The power of alternative data

Publish date: 30 October 2019
Author: SESAMm

Alternative data is increasingly being used by companies to unearth valuable information about their customers and products. SESAMm’s founder, Sylvain Forté, recently spoke with Efma’s Boris Plantier about alternative data, what it is exactly, and how companies are utilizing it.

What types of alternative data do you use?

SESAMm creates alternative data based on text through its platform TextReveal. We leverage Natural Language Processing to analyze assets, companies, brands, executives, economic concepts, macro indicators, events, etc. It is based on SESAMm's web database of more than 2M data sources (news, social media, blogs and forums), 8B articles and 10 years of history, one of the largest text data lakes in the world!

We can also leverage other types of data sources (credit card data, shipping data etc.) when building investment signals for our clients. We find it interesting to combine different data sets (including market data) to create actual trading strategies. This is done with our machine learning toolbox SignalReveal.

How is the data used to help trading?

Our data is used by both quantitative and fundamental teams. For quantitative teams, alternative data is a new factor that can be used in existing or new models (in many cases based on machine learning techniques). Fundamental analysts usually use this data in dashboards to better understand companies. We can for example track the specific sentiment of Chinese consumers (in Chinese language), regarding the launch of a new product by a listed company and identify the topics they mention.

How does the data determine if a company is doing well in terms of CSR?

Having access to millions of data sources allows us to track in real-time environmental, social, or governance events on companies. We create specific knowledge bases around these three key topics and link them to relevant companies. An article mentioning one of the company’s products in the context of pollution would be automatically detected and its tonality (sentiment) analyzed. This allows us to solve for the problem of timeliness with these types of indicators. Where current ESG indicators would take months to properly integrate these events, our indicators change the scoring in less than a day and generate alerts.

How can alternative data be used to help banks in the field of retail banking?

Alternative data is now used at a corporate level to make better strategic decisions. It could help analyze interactions between employees and clients, pre-fill CRM tools, identify potential tensions between teams in the organization, plan a product launch, track the company’s global e-reputation or provide early warnings for scandals. There are so many ways alternative data can help position a company towards its customers and competitors that the adoption of these new data sets by corporations is today’s most important topic in this field. The CEO of Nasdaq specifically described this challenge after acquiring a large alternative data distributor last year.

Banks collect a great deal of data on their customers - what are some ways in which they can monetize this information?

I believe that the first use of that data should be internal. Combining internal and external data sets gives companies a competitive edge and helps them better understand their own inner workings and challenges. Data monetization is a good opportunity, but the greater challenge (and ROI!) is probably in using that data to make decisions at the company level.