CFTC Approves Proposed Rule on Automated Trading
The CFTC approved proposed rules known collectively as "Regulation Automated Trading," or "Regulation AT." Regulation AT is intended to mitigate the risks of degradation in a person's algorithmic trading or of such trading's disruption of another person's ability to trade. The vote was unanimous. The proposal will remain open for a 90-day public comment period.
The following requirements are contained in Regulation AT:
Requirements for trading firms that fall within the proposed definition of "AT Person."
- Risk controls. Commission registrants, including a new class of persons who must be registered as floor traders ("AT Persons"), will be required to implement pre-trade and other risk controls to address the risks of algorithmic trading. These must include pre-trade risk controls (maximum order message and execution frequency per unit time, order price and maximum order size parameters), and order cancellation systems. The proposed rules provide AT Persons with flexibility regarding the design and calibration of required pre- trade risk controls.
-
An "AT Person" is defined as any person registered or required to be registered as a futures commission merchant ("FCM"), floor broker, swap dealer, major swap participant, commodity pool operator, commodity trading advisor or introducing broker that engages in "Algorithmic Trading" on or subject to the rules of a designated contract market ("DCM"). As noted above, the term also includes a new class of persons required to be registered as floor traders.
-
Development, testing, and monitoring. AT Persons must implement standards for the development, testing and monitoring of algorithmic trading systems ("ATSs"), and for the designation and training of algorithmic trading staff. These standards must include: keeping the development environment separate from the production environment; testing prior to implementation; a source code repository; real-time monitoring of such systems; and standards to ensure that systems comply with the Commodity Exchange Act and Commission regulations.
-
Compliance reports. AT Persons must submit annual compliance reports to DCMs regarding their risk controls, as well as copies of written policies and procedures developed to comply with testing and other requirements. AT Persons also must keep books and records regarding their Algorithmic Trading procedures to allow for inspection by DCMs.
Requirements for AT Clearing Members
- Risk controls. Clearing member FCMs must implement risk controls for algorithmic trading orders originating with AT Persons. For direct electronic access ("DEA") orders, FCMs must implement DCM-provided risk controls. For non-DEA orders, FCMs must establish the controls themselves. As with AT Persons, the proposed rules provide clearing member FCMs with flexibility regarding the design and calibration of required pre-trade risk controls.
- Compliance reports. Clearing member FCMs must submit compliance reports to DCMs that describe their program for establishing and maintaining the required pre-trade risk controls for their AT Person customers (in the aggregate). The reports must include (a) a description of the clearing member FCM's program for establishing and maintaining the pre-trade risk controls for its AT Persons at the DCM; and (b) a certification by the chief executive officer or chief compliance officer of the FCM that, to the best of his or her knowledge and reasonable belief, the information contained in the report is accurate and complete. FCMs also must keep, and provide to a DCM upon request, books and records regarding their risk controls for algorithmic trading orders for inspection by DCMs. AT Persons are required to provide more detailed reports than are clearing member FCMs.
Requirements for AT DCMs
- Risk controls. DCMs must implement risk controls for orders submitted through algorithmic trading. These must include pre-trade risk controls (maximum order message and execution frequency per unit time, order price and maximum order size parameters), and order cancellation systems. DCMs also must implement parallel controls for orders that do not originate from algorithmic trading (i.e., orders that are submitted manually). The proposed rules provide DCMs with flexibility regarding the design and calibration of required pre-trade risk controls.
- Compliance reports. DCMs must require risk control compliance reports from AT Persons and their clearing member FCMs. DCMs must review the compliance reports periodically, identify outliers and provide instructions for remediation. DCMs also must review, as necessary, books and records of AT Persons and clearing member FCMs regarding Algorithmic Trading procedures.
- Test environments. DCMs must provide test environments in which AT Persons may test algorithmic trading systems. The test environments must enable AT Persons to test compliance with proposed § 1.80 (risk controls, order cancellation systems) and § 1.81 (development and testing requirements).
- Risk controls for DEA orders. DCMs must establish risk controls for algorithmic orders submitted to DCMs by AT Persons using DEA, and require clearing member FCMs to use the risk controls for such DEA orders.
Use of Self-Trade Prevention Tools
-
DCMs must establish self-trade prevention tools, and either apply such tools or provide them to market participants and require their use. "Self-trading" is proposed to be defined as the matching of orders for accounts with common beneficial ownership or under common control. DCMs may either determine which accounts will be prohibited from trading with each other, or require market participants to identify such accounts. As an exception, DCMs may allow matching of orders for accounts with common beneficial ownership when initiated by independent decisionmakers. DCMs must publish quarterly statistics disclosing approved self-trading.
Disclosure Regarding DCM Trade Matching Systems
-
DCMs must provide public disclosure regarding certain elements of their electronic trade matching systems. Disclosure must include a description of the rules, specifications or known attributes of the trade matching platform that materially affect the time, priority, price or quantity of the execution of market participant orders; the ability to cancel or modify orders; and the transmission of market data and order or trade confirmations to market participants.
Market-Maker and Trading Incentive Programs
-
DCMs must provide certain disclosures regarding their exchange market-maker and trading incentive programs, and implement other controls. Such programs must be submitted as "Rule" filings pursuant to Commission Regulations 40.5 or 40.6. Filings must describe eligibility criteria as well as payments and other benefits. Payments are prohibited for trades between accounts with common ownership.
Role of Registered Futures Associations
-
Registered futures associations ("RFAs") must adopt certain membership rules - as deemed appropriate by each RFA - relevant to algorithmic trading for each category of member in the RFA. RFAs have discretion regarding the rules they issue and the categories of members to which their rules apply. Furthermore, to ensure that all AT Persons are subject to rules of an RFA regarding algorithmic trading, AT Persons must become members of at least one RFA. The proposed rules would allow RFAs to supplement elements of Regulation AT as markets and trading technologies evolve over time.
Chair Timothy M. Massad said this about the rules: "We have tried to be principles-based. We have set forth requirements for certain controls, but we have avoided prescribing the parameters or levels at which they should be set."
Commissioner Sharon Y. Bowen supported the rules, but recommended more certification and testing of employees hired to develop algorithms at automated trading firms. Additionally, she emphasized the principles-based nature of the rule. "The rule before us today should not substantially change how many firms utilize algorithms," she remarked.
Although Commissioner J. Christopher Giancarlo voted for the rule proposal for comments to be published, he indicated that he was doing so largely to review the comments. He stated that he has "serious doubts" about the proposed rulemaking and expressed his concern over the following areas: (i) some of the requirements appear to be window dressing, (ii) the requirement that registrants hold their proprietary source code in data repositories available for inspection by the CFTC or the U.S. Department of Justice is burdensome, and (iii) there could be "regulatory inconsistencies" regarding market participants that must comply with this rule.
Commentary
This release is over 500 pages, including all of the questions and discussions. Regarding the rule's fairly substantial requirements, CFTC Chair Massad said (at p. 501) that for "a firm that already uses risk management strategies, has various protections against malfunctions in place, and retains the services of talented attorneys, this new regulation will not create significant new burdens for that firm." Does anyone believe that a 500-page release will not impose significant compliance burdens? (For those who may be skeptical, please note the Chair's carve-out: in order for this rule not to be burdensome, the firm must "retain the services of talented attorney." Contact information is readily available through the Cabinet.)
Unfortunately for those firms that do not retain talented attorneys (you know who you are), this rule has the potential to impose (i) very substantial compliance burdens and (ii) completely open-ended obligations that may result in tremendous liability on the basis of after-the-fact assessments of fault.
As for Chair Massad's statement that Rule AT will be principles-based, this observation cuts two ways. Either (i) it may be difficult for the CFTC to prove that a firm violated the rule or (ii) it may be difficult for a firm to prove that it complied with the rule. When something goes wrong with a firm's algorithmic trading, the odds shift toward the latter being the case. To use just one of the requirements from the Rule as an example: a firm that is subject to the rule must have in place "[a] plan of internal coordination and communication between compliance staff of the AT Person and staff of the AT Person responsible for Algorithmic Trading regarding Algorithmic Trading design, changes, testing, and controls, which plan should be designed to detect and prevent Algorithmic Trading Compliance Issues."
What does that mean? How can a firm demonstrate that its plan of internal communication was sufficient if something has gone wrong? (Certainly those firms that do not have a sufficiently "talented attorney" on retainer are likely to fall into a deep hole as they begin their defense.) If the firm's internal coordination was insufficient, then the firm will be liable if an event originating with it disrupts, or materially degrades, (i) the algorithmic trading of itself, (ii) the operation of the designated contract market on which it trades or (iii) the ability of other market participants to trade on that market.
Commentary
In fact, the "race to the bottom" theory is widely discredited in modern economic circles. Economic studies suggest that, rather than engaging in such a "race to the bottom," actors in a competitive marketplace become involved in a race to find an equilibrium, which takes a number of factors into account, including the quality of regulation. Those who adhere to the older theory believe that "unhealthy" competition among firms needs to be stifled - and replaced with regulation - in order to prevent a "race to the bottom." However, the problem is twofold: First, the theory underlying the commission's policy has no empirical support. See, e.g., Daniel R. Fischel, "The Race to the Bottom Revisited: Reflections on Recent Developments in Delaware's Corporation Law," 76 Northwestern University Law Review 913 (1982). Second, the theory conflicts with antitrust concepts designed to promote competition, and with the CFTC's obligation under CEA Sec. 15 to take the public interest in being protected by such laws into consideration.