FINRA’s Guidance for Securities Traders Using Algorithmic Trading Strategies: A Perspective in 2016

Topic  - Regulatory Initiatives Regarding Algorithmic Trading

• June 2014 - SEC Chairman Mary Joe White announced that the SEC staff was preparing two rulemaking proposals addressing algorithmic trading: “the first, a rule to clarify the status of unregistered active proprietary traders to subject them to our rules as dealers; and second, a rule eliminating an exception from FINRA membership requirements for dealers that trade in off-exchange venues.”
  • The SEC has not yet issued a proposal regarding “unregistered active proprietary traders” -- a category of traders which conceivably could include institutional traders (such as hedge funds) which account for a sizable percentage of algorithmic trading activity.   The industry is anxiously awaiting this proposal.
  • March 2015 - the SEC proposed rule amendments that would require broker-dealers which currently are excepted from FINRA membership to become FINRA members if they engage in “significant” proprietary trading in the off-exchange markets.  These broker-dealers account for approximately 50% of all orders sent directly to alternative trading systems such as dark pools.
• September 2014 - in reaction to Chairman White’s announcement, FINRA’s Board authorized FINRA to issue a proposal establishing a registration requirement for associated persons who are (1) primarily responsible for the design, development or for directing the significant modification of an algorithmic strategy; or (2) responsible for supervising such functions.   March 2015 - FINRA issued the registration proposal. 
  • FINRA’s Board also authorized FINRA to publish a Regulatory Notice reminding firms of their existing supervisory obligations regarding the development and employment of algorithmic trading strategies.  March 2015 - FINRA issued the Guidance that forms the main subject matter of this webinar. 
• The pending SEC and FINRA regulatory initiatives could materially alter the regulatory landscape for algorithmic traders.   

Topic – FINRA’s Guidance Regarding Supervision and Control Practices for Algorithmic Trading Strategies

• Based upon its review of systems related issues at member firms,  FINRA has been focusing on whether and to what extent member firms have appropriate supervisory controls and procedures related to the creation, modification, usage and testing of trading algorithms.
• In Regulatory Notice 15-09, FINRA published formal guidance on effective supervision and controls practices for firms engaging in algorithmic trading strategies.
• FINRA’s guidance focuses on the following broad areas: 
  • Software / Code Development and Implementation.
  • Software Testing and System Validation.
  • Trading Systems and Implementation.
  • Compliance Monitoring.
• FINRA guidance does not purport to impose new substantive requirements on member firms. However, FINRA expects members firms to reevaluate their current practices in light of this guidance.
• Key to FINRA guidance is that firms that use algorithmic trading strategies must supervise every phase of the software development process, rather than just focusing on the end results.
• FINRA guidance raises question of whether FINRA may sanction firms for having deficient software development processes, even if such deficiencies do not cause more specific rule violations.

Topic - Recent Enforcement Actions/Examination Priorities Regarding Algorithmic Trading

• The SEC’s September 30, 2015 settlement with high frequency trading firm Latour Trading LLC illustrates many of the issues that FINRA is concerned about, the litigation and regulatory risks to firms and how those risks might be reduced.
• Latour’s violations of Reg. NMS over a multi-year period arose from computer coding errors, errors of logic, failure to conduct proper surveillance, and failure to have sufficient controls regarding modifications to trading system code.  
• In a number of prior cases, as in Latour, coding errors in algorithmic trading systems led to violations of securities regulations and sanctions.  An examination of the particular errors made in these cases, and the reasons the errors remained undetected, may assist firms in avoiding similar errors in the future. 
• In other enforcement matters, the SEC found high-frequency trading algorithms to be abusive because they manipulated the markets through layering, spoofing, wash sales, marking the close or other means.  These cases show how regulators identify abusive algorithms, and the bases for a finding of market manipulation in the context of algorithmic trading.  
• FINRA’s 2015 Examination Priorities Letter describes generally what examiners expect to review in connection with controls over trading technology and abusive algorithms.  The lessons learned from the enforcement cases discussed above may help firms to improve their controls in these critical areas, their preparation for future examinations and their overall compliance with FINRA guidance on algorithmic trading.