Among other things, the guidance seeks to address the results of a compliance review, which found that while most firms have integrated pre-trade controls into their risk management frameworks, the implementation of these controls isn’t consistent and firms’ governance of these controls isn’t always robust.
ESMA said added guidance was needed to address these shortcomings, given that weak pre-trade controls that allow for erroneous orders (aka “fat-finger” trades) pose risks that extend beyond the firm that enters the order, and can also threaten the stability of financial markets.
As a result, the regulator said it believes that it’s “necessary to issue guidance on this matter to ensure best practices are adopted by investment firms in the implementation of [pre-trade controls].”
The guidance also targets divergences in firms’ governance arrangements, stress testing and the outsourcing of algorithmic trading systems — and, it seeks to help regulators address the emerging risks posed by the growing use of AI for automated trading.
While the existing trading rules don’t specifically reference the use of AI, the guidance calls on both regulators and investment firms to recognize the impact of AI on algo trading.
In particular, it calls on firms to manage the risk that a series of minor model recalibrations could accumulate unchecked over time, leading to large, unexpected changes in model outputs.
To that end, it highlights existing expectations for investment firms to test and validate their trading systems, trading algorithms and their trading strategies.
It also highlights the obligation for firms to ensure that their compliance staff has “at least a general understanding” of how their automated trading systems and trading algorithms operate — and, continuous contact with the firm’s personnel that have detailed technical knowledge of those systems and algorithms.
“… on the one hand, the algorithmic trading systems and algorithms should be explainable and on the other hand, it is the investment firm’s responsibility to ensure they can adequately explain how AI impacts their algorithms’ decision-making,” it said.
The new guidance is being adopted to support the day-to-day work of trading supervision. ESMA also said that it’ll continue to monitor ongoing market and technological developments, which may result in revised guidance in the future.