Speaker: James Tromans, Technical Director, OCTO, Google
James Tromans: Hello. My name is James Tromans, and I am a technical director within Google Cloud's CTO office.
What does the future of capital markets look like with AI?
Trend #1
There will be increased levels of automation in the middle & back office, not just the front.
James Tromans: Well capital markets is a long and rich history of automation and quantitive of analysis within the front office, especially around market making and sales-related activities. So while I think that there is certainly a role for machine learning and AI to play in this space, some of the lower hanging fruit is actually in the middle and back office, for example, automation of some of the more manual activities around trade failures and settlements, or predicting when those things will actually happen, which can free up balance sheet for better use elsewhere within the firm.
Trend #2
The best AI/ML outcomes will still involve humans.
James Tromans: So one thing is for sure. In the case of capital markets, the best outcomes for machine learning and AI are still where a human is in the loop, which means those firms that are able to dissolve the boundary between the business trading and sales and technology that has historically supported those businesses so they can effectively build machine learning and AI models well together are going to lead to the best outcomes in the medium to long run.
Trend #3
Increased accessibility of AI/ML solutions will empower nontechnical business stakeholders.
James Tromans: Capital market firms have long benefited from having quantitative technical business people that are building automated systems for the last 20-30 years. However, modern machine learning and AI is now much more accessible, and when combined with cloud, is going to enable nontechnical business stakeholders to build state-of-the art baseline AI models that, if the human experts are not able to surpass, it enables them to focus on the areas where they can add value, they can add alpha, such as core differentiated market making activities, in addition to building a better white glove customer experience. But this means that AI will still find its way into other areas that were historically deemed too low value for human experts to be spending their time. So this is going to see more proliferation of machine learning in these other areas within capital markets.
