SocGen pushes data, analytics use cases for SG Markets

French investment bank Societe Generale is adding market data and analytics to SG Markets, its online B2B services marketplace. SG Markets Data and Analytics is currently available to some of SocGen’s buy-side clients, who can access market data across various asset classes.

Philippe Dufaÿ, head of data and analytics within SocGen’s Global Markets division, says the origins of what eventually became SG Markets can be traced back to when the team decided to modernize the methods through which they delivered data to customers.

“If you go back, maybe 15 years ago, we started to share data using many databases on an Excel file and sharing internally on a shared drive,” he says. “Then we would have a couple of people starting to send emails to clients to share the data with them. So at the end of the week, you’d go on the shared drive, you’d extract such data, and then you’d send over an email to the client.”

Ten years ago, nobody was thinking that data was of value in the bank, but now we realize that we are manipulating a lot of value

Philippe Dufaÿ, Societe Generale

Deeming this process too manual, Dufaÿ decided it needed automating, and thus occasionally sharing datasets with interested customers led to the creation of a fully digital platform, now SG Markets. Users can buy datasets curated by SocGen across a range of asset classes, and can use them to create and test models.

After Dufaÿ’s team gathered all the relevant datasets that clients wanted on a shared drive, they began to download data internally, so the data could be prepared and sent via email. Eventually, the email model fell out of favor as long-term clients moved companies and changed addresses, so the team built several websites to showcase their various data offerings, which were then further consolidated into a single web interface.

Once the datasets were available in a central location, SocGen’s data and analytics team decided that having valuable data on a shared drive was too risky and built out a data lake within the bank. Then Dufaÿ’s team plugged the data lake into a public cloud through an API on the website.

While SocGen now offers data to buy-side clients, Dufaÿ says this is not due to frustration with market data vendors—or their fees—or a belief that the vendor space would benefit from further competition.

“We are not here to compete with any market data vendor or to replicate data already available on some other framework. We are more focusing on data directly derived from our work and our traders,” he says.

Market data, but make it a service

Directly offering data and, to a lesser extent, analytics models is a relatively new venture for SocGen—the bank launched SG Markets Data and Analytics in November 2022—but it follows similar moves by some of its banking peers. Goldman Sachs has its Marquee platform, which lets clients access the firm’s analytics and trading capabilities, while JP Morgan’s Markets platform provides historical datasets going back to the 1960s.

SocGen adding a data and analytics offering “makes perfect sense,” says Stephen Kimsey, founder of market data consulting business Kimsey Consulting.

“This has been a developing situation for many years, with banks becoming data providers and solution providers,” he says. “I describe them as a sort of service. SocGen does security servicing and asset servicing, so adding data and analytics and building that out as a service makes perfect sense when lots of [banks] are doing that.”

SG Markets’ Data and Analytics section currently operates as a two-tiered system, through which all the data is available to a select group of SocGen’s hedge fund and asset manager clients, while a far more limited subsection of data is available to non-premium members.

SocGen’s Dufaÿ says this is primarily for security reasons, as data has become more valuable since the bank began its data modernization efforts.

“Some of the datasets are very unique, and we want to protect our IP,” he says. “Ten years ago, nobody was thinking that data was of value in the bank, but now we realize that we are manipulating a lot of value, and we can share that value easily.”

Premium or not?

Pointing to a specific model on his screen, Dufaÿ explains that for certain parts of the Data and Analytics unit, the difference between premium and non-premium is significant.

“[For non-premium], it’s very limited,” he says. “You have here only one week [of data], right? You can have up to 15 years. [Non-premium users] just want to quickly check the market and have some ideas to trade,” he says.

Developers who use SG Markets Data and Analytics benefit from the fact that part of the coding required to develop models is programmed in-house, and quants at buy-side firms can copy and paste the code into their own Python notebooks.

This has been a developing situation for many years, with banks becoming data providers and solution providers

Stephen Kimsey, Kimsey Consulting

“They get access to the data from a programmatic point of view,” Dufaÿ says. “So [within Data and Analytics], you have a couple of lines—and it’s not plain English, it’s more code lines—and you get a table with a value, and then you can use this value to build all the models you want and generate revenue.”

One of SocGen’s proprietary models available on Data and Analytics for users is CofBox, a trading tool that allows users to visualize the financing cost levels on equity indices.

CofBox is an example of a model that allows a much larger amount of data to be viewed by SocGen premium users than non-premium users. Kimsey says a business model where premium subscribers can access more datasets than non-subscribers is a product of the highly profitable and concentrated market data environment, where different services and datasets are often bundled.

“The vendors are no different, are they?” he says. “If you want 50 of something as opposed to two of something, you’re probably going to get those 50 cheaper than if you wanted the two if you’re an existing customer.”

Kimsey says providing data via the SG Markets service could lengthen the bank’s client list in the future.

“Banks have been providing data and services for years, almost as long as technologies like this existed, as it’s an add-on that helps retain customers by providing them with a useful service,” he says. “From the customer’s point of view, if they have a relationship with SocGen [through SG Markets], [data] may well be cheaper than going through an actual vendor.”