Central Bank publishes review of firms undertaking algorithmic trading

11 May 2021 Press Release

Central Bank of Ireland 

  • Thematic review assessed how firms undertaking algorithmic trading have incorporated MiFID II requirements into their risk management and control frameworks.
  • Some positive practices identified but risk and control frameworks require continued improvement. Engagement ongoing with firms where specific concerns were identified.
  • All firms engaging in algorithmic trading should review and take all necessary remedial action.

The Central Bank of Ireland has today (11 May 2021) written to CEOs of firms undertaking algorithmic trading following a thematic review which assessed how such firms have complied with risk management and control framework requirements as required by regulatory technical standards for investment firms.

This letter outlines the Central Bank’s expectations of firms in relation to the governance, testing and controls surrounding algorithmic trading, and has also been issued to all firms undertaking algorithmic trading, who must now take appropriate action to address the issues identified.

Algorithmic trading gives rise to significant risks stemming from potential failures of algorithms, IT systems and related processes. In recent years, a number of significant algorithmic trading failures have resulted in substantial losses, fines and reputational damage for firms globally.

Firms undertaking algorithmic trading must comply with specific regulatory requirements. The framework was introduced in 2018 and is key to protecting investors, and making the financial market resilient and transparent.

While the reviewed found certain positive practices, a number of key concerns were identified, including: 

  • Governance - Firms failed to evidence effective oversight of controls and practices related to algorithmic trading. This included minimal Board involvement, both in setting or challenging key controls, and in the oversight of the development of trading algorithms resulting in a lack of regular reporting to the Board and an absence of formalised governance documentation.
  • Development and Testing - Significant disparities identified between firms with respect to the level of detail within documentation on development, testing and deployment processes.
  • Risk and Measurement Control - Weaknesses identified around annual self-assessments, and lack of clarity with regard to the third line of defence and the role of Internal Audit.
  • Trade Lifecycle Management – Lack of appropriate documentation with respect to pre- and post-trade controls.

As a result of the review, the Central Bank has engaged with investment firms where specific concerns have been identified, issuing specific time-bound actions called risk mitigation programmes to address issues.

Mary-Elizabeth McMunn, Director of Credit Institutions Supervision, said:

“Some of the findings of this review do not align with our expectations, given the nature, scale and complexity of firms within the scope of this review. We expect all firms to ensure risk management and control frameworks in respect of algorithmic trading are appropriately embedded and are operating to a high standard. The regulations provide a framework to mitigate these, and other risks, through the requirement to maintain effective systems, procedures, arrangements and controls.

"We did note many positive practices, including the presence of experienced, competent professionals across the first and second lines of defence, supported by a comprehensive suite of controls in terms of monitoring, development, testing and deployment of trading algorithms. But continued vigilance is necessary to ensure the risk management frameworks and implementation are to a high standard.

"The Central Bank will continue to drive higher standards of governance and oversight of the use of technology in financial services firms. All firms engaging in algorithmic trading should review the content of this letter. We will continue to assess whether firms have taken sufficient steps to reduce risks arising from algorithmic trading through our ongoing intrusive supervisory work.”

Notes: