There is so much discussion about using AI for FX Sales Teams you could be forgiven for thinking that is the only way forward for organisations looking to be future focused. A recent article in South African Treasurer Magazine (see page 33) by Aphile Molefe head of EFICC Sales for Absa Bank (a Hedgebook customer) took a look at FX Hedging in the age of big data and AI and provided some interesting insights.
1. AI does have a place for FX Sales Teams – but currently it’s limited
In pointing out that many of the proposed AI use cases in finance – such as making lending and investment decisions or fraud detection – has always been evident and are still relevant. But according to Aphile Mofele if you were to be pedantic most of these are actually machine learning technologies versus AI.
So where does he see AI being applied:
“Banks may need AI for Support chat bots so clients can see certain information or even execute transactions in natural language! But these use cases may explain the main reason why AI technology has not taken off with FX functions at banks: regulation and market risk.”
The article goes on to explain how financial market regulations and the high value (often millions of dollars) of transactions exposes extreme risk should something go wrong. He asks: “If something does go wrong who would be held to account with any losses experienced? The algo? The developer?”
2. Enter, FX client specific insights
One area of technology that Aphile Mofele highlights in the article as being immediately deployable is around big data analytics in the FX sales world. He highlights how it is now possible to refer to a client’s specific requirement as risk or exposure analytics. He explains that in taking it a step further salespeople can simulate how certain market levels would impact that company’s profit and losses – including if they transact using proposed trade ideas.
“This sort of tool can transform how a treasurer interacts with his FX bank because they can better collaborate with the bank based on relevant and easier to understand informatics.”
3. Finally advanced FX trade analytics
Salespeople can now identify which of their clients are likely to trade and when. Aphile Mofele explains how using advanced trade analytics tools FX sales teams can build up to date reports on the fly and use machine learning modules to predict a whole range of client behaviours based on prior history.
This includes highlighting to the salesperson if a client does something unexpected – like a new tenor or currency pair or ticket size. Mofele explains:
“With market data included (i.e. historical tick data) it can also predict client behaviour in certain market conditions such as paid changes in prices or even activity around a specific level. This capability is becoming a reality for some banks – including Absa.”
In summary: benefits are clear
In summary he concludes that the benefits of big data analytics and machine learning in the dealing room are clear:
“Collaborating with a data-powered dealing room can ensure a more client centric service with proactive salespeople that also recommend bespoke and well informed strategies suited to meet the specific treasurer’s objectives.”
Hedgebook is proud to be part of building this data-driven power house with the Absa FX Sales Team and would gladly give a demo to those interested in seeing how this could work for their bank.