AI is going to transform the role played by buy-side analysts. AI makes searches easier. Patterns can be detected earlier. Such tools inform decision-making which is still the analysts’ ultimate duty.
Juan-Carlos Rhodes writes about the brave new world facing buy-side analysts in efinancialcareers.com:
As a sell-side trader with long years of experience on the buy-side, I think the big changes will come to order management and execution management systems, which may yet be combined into one. The amount of data these systems can process is about to increase exponentially.
A.I. is great for finding behavioral exceptions, outliers and anomalies. When you have a basket to trade, you will soon be able to use A.I to run a pre trade and identify the large weighted potential issues. A.I will allow you to look at the illiquid tails, and you will be able to use this to create a focus list, depending upon market conditions. The same basket on a different day is a totally different trade. Understanding the characteristics of the type of trades and experience in trading them is the driver of performance. Protecting your data going forward becomes of greater importance. The ability to be anonymous to the market will become a bigger part of how you reduce your impact.
Back-testing and understanding of your flow will become easier as your access to data improves. Real time realized and unrealized expected impact models will make it quicker to make intra-trade implementation adjustments. Accordingly, real-time actionable transaction cost analysis will be adopted across all buy side traders in the future. This will make trader efficiency far more quantifiable. The good buy-side traders will stand out from the bad.
Buy-side traders will still be required. However, they will need to act in combination with these systems. AI is about identifying issues early on and providing traders with probable outcomes which they can act upon. Tomorrow’s buy-side traders will have to be flexible to increase their efficiency. – And it will be very evident if they are not.
