Inside SGX’s AI-powered digital journey

By Nurfilzah Rohaidi and Ng Yi Ming

Interview with Sigrid Rouam, Head of Data Science at the Singapore Exchange.

Artificial intelligence is transforming every sector and industry in ways we might not fully understand. This change is particularly acute in financial services, where every line of data could be worth billions.

The Singapore Exchange (SGX) has set out plans to use data and AI to help traders to trade faster and more effectively. Sigrid Rouam, Head of Data Science at SGX, has a vision to make SGX a global pioneer in this space.

GovInsider caught up with Rouam at the recent ConnectGov Leaders Summit in Edinburgh, Scotland, to find out how SGX is using Artificial Intelligence to stay at the cutting edge.

Superfast connections

SGX is exploring the use of AI in three areas, says Rouam. First is ‘latency anomaly detection’, or “how fast people can connect to the trading exchange”. Latency refers to the delay between when a human requests a trade, and when it actually begins. It can be just fractions of a second, but prices fluctuate frequently.

AI has allowed Singapore to analyse and understand latencies. “How traders are connecting to the trading engine is very important,” Rouam explains. SGX closely monitors latency problems that customers face, and recently implemented an anomaly detection algorithm to pick up on any delays and make trades as fast as possible. In the United States, traders have successfully used AI and superfast connections to make trades in fractions of a second, reacting before others can.

Second, the exchange is using AI to help customers better pull together trading data. SGX is building an engine that translates vast amounts of text from corporate announcements and reports, she says. With this new tool, users can glean more commercial insights from these documents. “We are also building a data dictionary, so that people know how to use the data”, Rouam continues.

Third, this requires automation to make the data clean and usable. SGX is automating PDF document classification and document auto-tagging, which also help to increase efficiency, she adds.

Understanding customer needs

SGX is also using AI to better understand the investor’s journey, Rouam explains. “The first touch point is our website for example, followed by opening an account with CDP (Central Depository),” she says. CDP accounts act like safes for an investor’s stocks. “They then open a trading account with the broker before they start trading.”

Customers have differing trading preferences, and SGX is looking to better understand the various types out there. Rouam gives some examples: “For example, we have customers who open a CDP account with us but never start trading, or there are customers who prefer to buy and hold. Then we have customers who are trading much more often.”

SGX is “starting to move more into machine learning and AI” to understand customer actions, Rouam says. “Extremely fast” data retrievals are crucial for such data analytics, she explains, where fractions of a second can make all the difference.

All this is so that SGX can relate to customers better, and build services that create the most value for them, she says. “We really have a big potential for increasing the use of data science and AI to digitalise our company,” she remarks.

SGX is a fast-growing, multi-asset exchange in the region. From DBS Bank to Singapore Airlines, big company stocks are traded on there every day. And with this vision for a tech-enabled future, SGX is poised to grow.

Rouam was speaking at the ConnectGov Leaders Summit in Edinburgh, Scotland on 8-10 July, organised by CIO Academy Asia.