# Regulation Automated Trading

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/fr%3A2015-30533

## Record

- **Collection:** Federal Register
- **Document type:** Proposed Rule
- **Published:** December 17, 2015
- **Citation:** 80 FR 78824

## Text

COMMODITY FUTURES TRADING COMMISSION
17 CFR Parts 1, 38, 40, and 170
RIN 3038-AD52
Regulation Automated Trading

AGENCY:

Commodity Futures Trading Commission.

ACTION:

Notice of proposed rulemaking.

SUMMARY:

The Commodity Futures Trading Commission (“CFTC” or “Commission”) is proposing a series of risk controls, transparency measures, and other safeguards to enhance the regulatory regime for automated trading on U.S. designated contract markets (“DCMs”) (collectively, “Regulation AT”). The Commission's proposals build on efforts by numerous entities in recent years to promote best practices and regulatory standards for automated trading, including standards and best practices for algorithmic trading systems (“ATSs”), electronic trade matching engines, and new connectivity methods that characterize modern financial markets. In 2012 the Commission adopted rules requiring futures commission merchants (“FCMs”), swap dealers (“SDs”), and major swap participants (“MSPs”) to use automated means to screen orders for compliance with certain risk-based limits. It also adopted rules requiring certain financial risk control requirements for DCMs offering direct market access to their customers. In 2013 the Commission published an extensive Concept Release on Risk Controls and System Safeguards for Automated Trading Environments (“Concept Release”), compiling in one document a comprehensive discussion of industry practices, Commission regulations, and evolving concerns in automated trading.
1

Now, through this notice of proposed rulemaking (“NPRM”) for Regulation AT, the Commission seeks to update Commission rules in response to the evolution from pit trading to electronic trading. In particular, the Commission is proposing to adopt a comprehensive approach to reducing risk and increasing transparency in automated trading. Proposed Regulation AT is designed to consolidate previous work by industry participants, the Commission, and fellow regulators into a unified body of law addressing automation in order placement and execution in U.S. derivatives markets. The Commission welcomes all public comments.

1
Concept Release on Risk Controls and System Safeguards for Automated Trading Environments, 78 FR 56542 (Sept. 12, 2013).

DATES:

Comments must be received on or before March 16, 2016.

ADDRESSES:

You may submit comments, identified by RIN 3038-AD52, by any of the following methods:

•
CFTC Web site: http://comments.cftc.gov.
Follow the instructions for submitting comments through the Comments Online process on the Web site.

•
Mail:
Send to Christopher Kirkpatrick, Secretary of the Commission, Commodity Futures Trading Commission, Three Lafayette Centre, 1155 21st Street, NW., Washington, DC 20581.

•
Hand Delivery/Courier:
Same as Mail, above.

•
Federal eRulemaking Portal: http://www.regulations.gov.
Follow the instructions for submitting comments.

Please submit comments by only one method. All comments should be submitted in English or accompanied by an English translation. Comments will be posted as received to
http://www.cftc.gov.
You should submit only information that you wish to make available publicly. If you wish the Commission to consider information that may be exempt from disclosure under the Freedom of Information Act (“FOIA”), a petition for confidential treatment of the exempt information may be submitted according to the procedures established in 17 CFR 145.9. The Commission reserves the right, but shall have no obligation, to review, prescreen, filter, redact, refuse, or remove any or all of your submission from
http://www.cftc.gov
that it may deem to be inappropriate for publication, such as obscene language. All submissions that have been so treated that contain comments on the merits of the rulemaking will be retained in the public comment file and will be considered as required under the Administrative Procedure Act and other applicable laws, and may be accessible under FOIA.

FOR FURTHER INFORMATION CONTACT:

Sebastian Pujol Schott, Associate Director, Division of Market Oversight,
sps@cftc.gov
or 202-418-5641; Marilee Dahlman, Special Counsel, Division of Market Oversight,
mdahlman@cftc.gov
or 202-418-5264; Mark Schlegel, Special Counsel, Division of Market Oversight,
mschlegel@cftc.gov
or 202-418-5055; Michael Penick, Economist, Office of the Chief Economist,
mpenick@cftc.gov
or 202-418-5279; Richard Haynes, Economist, Office of the Chief Economist,
rhaynes@cftc.gov
or 202-418-5063; Andrew Ridenour, Senior Trial Attorney, Division of Enforcement,
aridenour@cftc.gov
or 202-418-5438; or John Dunfee, Assistant General Counsel, Office of General Counsel,
jdunfee@cftc.gov
or 202-418-5396.

SUPPLEMENTARY INFORMATION:

Table of Contents

I. Introduction

A. Overview—Development of Automated Trading Environment

B. Risks and Potential Benefits Associated With Automated Trading

C. The Proposed Regulations

1. Overview of NPRM

2. The Proposed Regulations Under Parts 1, 38, 40, and 170

II. Background on Regulatory Responses to Automated Trading

A. The Commission's Regulatory Response to Date

B. The Commission's 2013 Concept Release

C. Other Recent Regulatory Responses

1. SEC Regulatory Initiatives

2. FINRA Initiatives

3. European and Other Regulatory Initiatives

D. Industry and Regulatory Best Practices and Recommendations

1. NFA Compliance Rule 2-9: Supervision

2. FIA Reports on Automated Trading

3. IOSCO Reports on Electronic Trading

4. CFTC TAC Subcommittee

5. FIX Risk Management Working Group

6. Senior Supervisors Group (SSG) Briefing Note

7. Treasury Market Practices Group Best Practices

III. Recent Disruptive Events in Automated Trading Environments

IV. Overview of Regulation AT

A. Concept Release/Regulation AT Terminology

B. Commenter Preference for Principles-Based Regulations

C. Multi-Layered Approach to Pre-Trade Risk Controls and Other Measures

D. Codification of Defined Terms Used Throughout Regulation AT

1. “Algorithmic Trading”—§ 1.3(zzzz)

2. “Algorithmic Trading Compliance Issue”—§ 1.3(tttt)

3. “Algorithmic Trading Disruption”—§ 1.3(uuuu)

4. “Algorithmic Trading Event”—§ 1.3(vvvv)

5. “AT Order Message”—§ 1.3(wwww)

6. “AT Person”—§ 1.3(xxxx)

7. “Direct Electronic Access”—§ 1.3(yyyy)

E. Registration of Certain Persons Not Otherwise Registered With Commission—§ 1.3(x)

1. Concept Release Comments

2. Description of Regulation

3. Policy Discussion

4. Request for Comments

F. RFA Standards for Automated Trading and Algorithmic Trading Systems—§ 170.19

1. Policy Discussion

2. Description of Regulation

3. Request for Comments

G. AT Persons Must Become Members of an RFA—§ 170.18

1. Policy Discussion

2. Description of Regulation

3. Request for Comments

H. Pre-Trade and Other Risk Controls for AT Persons—§ 1.80

1. Concept Release Comments on Pre-Trade and Other Risk Controls

2. Description of Regulation

3. Policy Discussion

4. Request for Comments

I. Standards for Development, Testing, Monitoring, and Compliance of Algorithmic Trading Systems—§ 1.81

1. Concept Release Comments

2. Description of Regulation

3. Policy Discussion

4. Request for Comments

J. Risk Management by Clearing Member FCMs—§ 1.82

1. Concept Release Comments

2. Description of Regulation

3. Policy Discussion

4. Discussion of Persons Subject to Proposed §§ 1.80 and 1.82

5. Request for Comments

K. Compliance Reports Submitted by AT Persons and Clearing FCMs to DCMs; Related Recordkeeping Requirements—§ 1.83

1. Concept Release Comments

2. Description of Regulation

3. Policy Discussion

4. Request for Comments

L. Risk Controls for Trading: Direct Electronic Access Provided by DCMs—§ 38.255(b) and (c)

1. Concept Release Comments

2. Description of Regulation

3. Policy Discussion

4. Request for Comments

M. Disclosure and Transparency in DCM Trade Matching Systems—§ 38.401(a)

1. Concept Release Comments

2. Description of Regulation

3. Policy Discussion

4. Request for Comments

N. Pre-Trade and Other Risk Controls at DCMs—§ 40.20

1. Concept Release Comments

2. Description of Regulation

3. Policy Discussion

4. Request for Comments

O. DCM Test Environments for AT Persons—§ 40.21

1. Concept Release Comments

2. Description of Regulation

3. Request for Comments

P. DCM Review of Compliance Reports by AT Persons and Clearing FCMs; DCM Rules Requiring Certain Books and Records; and DCM Review of Such Books and Records as Necessary—§ 40.22

1. Concept Release Comments

2. Description of Regulation

3. Policy Discussion

4. Request for Comments

Q. Self-Trade Prevention Tools—§ 40.23

1. Concept Release Comments

2. Commission Analysis of Amount of Self-Trading in the Marketplace

3. Description of Regulation

4. Policy Discussion

5. Request for Comments

R. DCM Market Maker and Trading Incentive Programs—§§ 40.25-40.28

1. Policy Discussion

2. Description of Regulations

3. Request for Comments

V. Related Matters

A. Calculation of Number of Persons Subject to Regulations

1. Request for Comments

B. Calculation of Hourly Wage Rates Used in Related Matters

C. Regulatory Flexibility Act

1. FCMs and DCMs

2. AT Persons

3. Request for Comments

D. Paperwork Reduction Act

1. Information Provided by Reporting Entities/Persons

a. § 1.3(x)(3)—Submissions by newly registered floor traders

b. § 1.83(a)—Compliance reports submitted by AT Persons to DCMs

c. § 1.83(b)—Compliance reports submitted by clearing member FCMs to DCMs

d. § 1.83(c)—AT Person retention and production of books and records

e. § 1.83(d)—Clearing member FCM retention and production of books and records

f. § 38.401(a) and (c)—Public dissemination of information by DCMs pertaining to electronic matching platforms

g. § 40.23—Information publicly disseminated by DCMs regarding self-trade prevention

h. § 40.25—Information in public rule filings provided by DCMs regarding Market Maker and Trading Incentive Programs

i. § 40.26—Information provided by DCMs to the Division of Market Oversight upon request regarding Market Maker and Trading Incentive Programs

2. Information Collection Comments

E. Cost Benefit Considerations

1. The Statutory Requirement for the Commission to Consider the Costs and Benefits of its Actions

2. Concept Release Comments Regarding Costs and Benefits

3. The Commission's Cost-Benefit Consideration of Regulation AT—Baseline Point

4. The Commission's Cost-Benefit Consideration of Regulation AT—Cross-Border Effects

5. General Request for Comment

6. The Commission's Cost-Benefit Consideration of Regulation AT—Proposed Definitions

7. Pre-Trade Risk Controls, Testing and Supervision of Automated Systems, Requirement to Submit Compliance Reports, and Other Related Algorithmic Trading Requirements

8. Requirements for Certain Entities to Register as Floor Traders

9. Transparency in Exchange Trade Matching Systems

10. Self-Trade Prevention

11. Market-Maker and Trading Incentive Programs

VI. Aggregate Estimated Cost of Regulation AT

VII. List of All Questions in the NPRM

I. Introduction

A. Overview—Development of Automated Trading Environment

U.S. derivatives markets have historically relied on manual processes for the origination of orders, transmission of information, and execution of trades. Trading decisions were typically initiated by natural persons, and transmitted through intermediaries via comparatively simple communications networks. Execution occurred in open-outcry trading pits operated by DCMs. Access to these pits was limited to brokers and traders granted trading privileges by the exchange. A range of other processing and risk management services were equally reliant on manual processes, and the complete trading system could move only as fast as its human decision-makers. Trading information was often recorded on paper order tickets and trading cards, and time-stamps were recorded only to the nearest minute. The physical element of trading was reflected in exchange or Commission rules governing diverse matters such as the types of trading permitted from the top step of a futures pit,
2

as well as requirements that certain orders for execution in a trading pit be recorded in “non-erasable ink.” This basic structure remained constant for decades, and produced a parallel regulatory framework also premised on natural persons and human decision-making speeds.

2
For example, press reports surrounding the initiation of CME's “top step” rule in the S&P 500 stock-index pit in 1987 indicated that brokers preferred the top step to “get a panoramic view of the trading activity and quickly grab customer order sheets being relayed by nearby clerks.” They described a trading pit where “[s]ome 400 traders are jammed shoulder-to-shoulder in the amphitheater-like pit, which accounts for three-fourths of the nation's stock-index futures trading.”
See
Jouzaitis, Carol, “Merc Launches `Top-step' Reform,” Chicago Tribune (June 22, 1987) available at
http://articles.chicagotribune.com/1987-06-22/business/8702160155_1_dual-trading-stock-index-futures-market-chicago-mercantile-exchange.

Today, derivatives markets have transitioned from the manual processes described above to highly automated trading and trade matching systems. Modern DCMs and DCM market participants, in particular, are characterized by a wide array of algorithmic and electronic systems for the generation, transmission, management, and execution of orders, as well as systems used to confirm transactions, communicate market data, and link markets and market participants through high-speed networks. Collectively, such DCM and market participant trading systems constitute the “automated trading

environment” at the center of Regulation AT. Automated trading environments often make use of automated systems for either the generation or the execution of orders (in many cases, both). Such automated systems are based on sets of rules or instructions (commonly referred to as algorithms) and related computer systems used to automate the execution of a trading strategy.
3

In futures markets, orders generated by automated trading systems are ultimately transmitted to DCMs that accept, manage and match orders by automated means.

3

See
IOSCO Report on Regulatory Issues Raised by Technological Changes,
infra
note 103 at 10.

While technologies have evolved, the underlying functions of derivatives markets remain the same, as do the Commission's responsibilities under the Commodity Exchange Act (the “CEA” or “Act”). Such markets, typically operated by DCMs, provide valuable risk mitigation and price discovery services for numerous financial and physical commodities businesses, including producers and consumers of energy, foodstuff, metals, and other raw materials, as well as natural person investors. The Commission is committed to the safety and integrity of U.S. markets as they continue their rapid technological change. Through proposed Regulation AT, the Commission is taking its next steps in ensuring that its regulatory standards and industry practices properly address current and foreseeable risks arising from automated trading, and promote responsible innovation and fair competition among markets and market participants.
4

4

See
CEA Section 3, “Findings and Purposes,” noting in Section 3(a) that transactions subject to the CEA are “affected with a national public interest” and in Section 3(b) that “[t]o foster these public interests, it is further the purpose of this Act to deter . . . any other disruptions to market integrity; to ensure the financial integrity of all transactions subject to the Act and the avoidance of systemic risk; . . . and to promote responsible innovation and fair competition among boards of trade . . . and market participants.”

Within U.S. derivatives markets, DCMs represent a significant catalyst in the transition to automated trading. From its beginnings with CME Globex in 1992, DCM on-exchange trading now occurs almost exclusively on electronic matching platforms, using internal algorithms to rapidly match incoming orders from an array of market participants.
5

Data available to Commission staff indicates that in an approximately two-year period through October 2014, over 95 percent of all on-exchange futures trading occurred on DCMs' electronic trade matching platforms.
6

In this regard, the Commission notes that CME Group, the largest U.S. exchange operator, announced in February 2015 its intention to close all but one of its open-outcry trading floors for futures.
7

IntercontinentalExchange, the second largest DCM operator, ended all futures open-outcry trading in March 2008, and ended all options open-outcry trading in October 2012. On-exchange trading on DCMs other than the CME Group exchanges and IntercontinentalExchange now occurs exclusively on electronic matching platforms. Concurrent with their transition to electronic trade matching platforms, DCMs have taken steps to increase the speed of trading in their markets. These include offering co-location and proximity hosting services to reduce latencies between the DCM and market participants, as well as measures taken by DCMs to reduce processing times within their electronic trade matching platform. The two largest DCMs, for example, have for several years indicated in their public materials average or median order entry round trip times of less than one millisecond.
8

5
Trading on CME Globex was initially limited to “after-hours” periods when the Exchange's open-outcry pits were closed. The first products offered on Globex in 1992 included German mark and Japanese yen futures and options on futures contracts, followed by other FX and currency products. In 1997, CME launched the E-mini S&P 500 futures contract, the first CME product available exclusively on Globex, including during regular (open-outcry) trading hours in other CME products. Globex monthly volume exceeded 100,000 contracts for the first time in 1997. In 1999, CME for the first time began offering “side-by-side” trading, allowing its Eurodollar contract to be traded both on Globex and in open-outcry during regular trading hours. Side-by-side trading was expanded in the ensuing years, including for example to FX products in 2001. Globex average daily volume exceeded 1,000,000 contracts for the first time in 2002. By 2004, Globex trading volume began exceeding open-outcry volume for the first time. Through agreements or mergers, CME began listing NYMEX products (2006) and CBOT products (2007) on Globex as well.
See
Aldinger, Lori, and Labuszewski, John W., “ELECTRONIC TRADING Twenty Years of CME Globex” (2012), available at
http://www.cmegroup.com/education/files/globex-retrospective-2012-06-12.pdf.

6
Haynes, Richard & Roberts, John S., “Automated Trading in Futures Markets,” CFTC Office of Chief Economist (Mar. 13, 2015), available at
http://www.cftc.gov/ucm/groups/public/@economicanalysis/documents/file/oce_automatedtrading.pdf.

7

See
CME Press Release, “CME Group to Close Most Open Outcry Futures Trading in Chicago and New York by July; Most Options Markets to Remain Open,” (Feb. 4, 2014) available at
http://cmegroup.mediaroom.com/2015-02-04-CME-Group-to-Close-Most-Open-Outcry-Futures-Trading-in-Chicago-and-New-York-by-July-Most-Options-Markets-to-Remain-Open?pagetemplate=article.

8

See
CME Group, “The World's Leading Electronic Platform: CME Globex,” (2014) at 3, available at
http://www.cmegroup.com/globex/files/globexbrochure.pdf
; IntercontinentalExchange, 2010 Annual Report, (2011) at 26, available at
http://ir.theice.com/~/media/Files/I/Ice-IR/annual-reports/2010/ice-2010ar.pdf.

The largely complete transition of DCMs to electronic trade matching platforms has occurred alongside an equally important shift in the technologies used by market participants to place and manage orders. Market participants have applied a range of sophisticated technological tools to their trading. For example, market participants are increasingly using ATSs, often coupled with high-speed communication networks. Market participants are also increasingly relying on electronic market and other data feeds to inform trading decisions, and on multiple computer algorithms to generate, manage, or route orders to DCMs. Market participants may also make use of direct electronic access and/or co-location services to minimize latencies between an ATS, market data systems, and a DCM's electronic trading matching platform.

Data available to the Commission highlights the importance of ATS trading on DCMs today. The Commission's analysis of data covering the same approximately two-year period addressed above (through October 2014) indicates that ATSs were present on at least one side in almost 80 percent of foreign exchange futures volume, 67 percent of interest rate futures volume, and 62 percent of equity futures volume analyzed. They were also present on at least one side in approximately 47 percent of metals and energy product volumes. Even in agricultural products, a category not typically associated with automation in recent years, ATSs were present in at least 38 percent of futures volume analyzed. Finally, in the aggregate, ATSs were present in over 60 percent of all futures volume traded across all products in the nearly two-year period that the Commission examined. In highly liquid product categories, ATSs represented both sides of the transaction over 50 percent of the time.
9

9

See
Haynes & Roberts,
supra
note 6 at 4.

Market participants using ATSs may transact on DCMs through registered intermediaries, including their clearing members. Such intermediaries themselves often rely on extensive automation, using ATSs for functions ranging from simple order routing to the generation of independent trading decisions. These registered intermediaries include FCMs, commodity pool operators (“CPOs”), commodity trading advisors (“CTAs”), introducing brokers (“IBs”), and floor brokers (“FBs”). In addition, Commission-registered SDs and MSPs

may use ATSs to conduct trading on DCMs. As discussed in more detail below, each of these categories of Commission registrants may be subject to Regulation AT in the event that they conduct algorithmic trading on a DCM.

B. Risks and Potential Benefits Associated With Automated Trading

Regulation AT proposes a series of pre-trade risk controls and other measures intended to address the risks related to automated trading on DCMs. The proposed rules primarily address operational risk issues, as well as related issues such as self-trading and market maker and trading incentive programs.

The potential risks of automated trading were recently described in a report discussing the events of October 15, 2014, when the market for U.S. Treasury securities, futures, and other closely related financial markets experienced an unusually high level of volatility and a very rapid round-trip in prices. On July 13, 2015, five regulatory agencies issued a joint staff report on the unusual market events of October 15, 2014 (the “October 15 Joint Staff Report”).
10

In addition to discussing the events of October 15, the report includes an Appendix C that summarizes many of the risks of automated trading. These risks include the following: Operational risks (ranging from malfunctioning and incorrectly deployed algorithms to algorithms reacting to inaccurate or unexpected data); market liquidity risks (arising from abrupt changes in trading strategies even when a firm executes its strategy perfectly); market integrity risks (automated trading can provide new tools to engage in unlawful conduct); transmission risks (shocks based on erroneous orders impacting multiple markets); clearing and settlement risks (as more firms gain access to trading platforms, trades may not be subject to sufficient settlement risk mitigation techniques); and risks to effective risk management (the speed of trade execution may make critical risk mitigation devices less effective).

10

See
Joint Staff Report: The U.S. Treasury Market on October 15, 2014 (July 13, 2015) [hereinafter “October 15 Joint Staff Report”], prepared by the U.S. Department of Treasury, Board of Governors of the Federal Reserve System, Federal Reserve Bank of New York, U.S. Securities and Exchange Commission, and U.S. Commodity Futures Trading Commission, available at
http://cftc.wss/OCE/conceptrelease/documentlibrary/Regulation%20AT/Reg%20AT%20--%20DRAFT%20PREAMBLE/October%2015%20report/treasury-market-volatility-10-14-2014-joint-report.pdf.
The report discusses the preliminary findings regarding the conditions that may have contributed to the October 15 volatility, particularly in the “event window” that began at 9:33 a.m. ET. Among other potential causes of this volatility, the October 15 Joint Staff Report states that several large transactions occurred between the release of certain U.S. retail sales data and the start of the event window; that there was a significant reduction in market depth following the retail sales data release, which appears to have resulted from a high volume of transactions and bank-dealers and principal trading firms changing their participation in the cash and futures order books; that latency associated with a significant increase in message traffic due to order cancellations increased just before the event window; and there was a higher incidence of “self-trading” during the event window.
Id.
at 4-6.

Notwithstanding the risks described above, several commentators have argued that algorithmic trading results in a more efficient marketplace. A recent study of the equities market concluded that algorithmic trading narrows spreads, reduces adverse selection, and reduces trade-related price discovery.
11

The study also suggested that algorithmic trading improves liquidity and enhances the information provided in quotes. Another recent study of low latency activity in the equities market (typically associated with high frequency trading) concluded that “an increase in low-latency activity reduces quoted spreads and the total price impact of trades, increases depth in the limit order book, and lowers short-term volatility.”
12

11

See, e.g.,
Hendershott, Jones and Menkveld, “Does Algorithmic Trading Improve Liquidity?,” The Journal of Finance, Vol. LXVI, No. 1 (Feb. 2011), available at
http://faculty.haas.berkeley.edu/hender/algo.pdf.

12

See
Hasbrouck and Saar, “Low-latency trading,” Journal of Financial Markets 16 (2013) at 646-679, available at
http://people.stern.nyu.edu/jhasbrou/Research/LowLatencyTradingJFM.pdf.

C. The Proposed Regulations

1. Overview of NPRM

The Commission is pursuing a number of goals in proposed Regulation AT. As an overarching goal, the Commission seeks to update Commission rules in response to the evolution from pit trading to electronic trading. The risk controls and other rules proposed in this NPRM are focused on algorithmic order origination or routing by market participants, and electronic order execution by DCMs. In addition to mitigating risks arising from algorithmic trading activity, the proposed rules are intended to increase transparency around DCM electronic trade matching platforms and the use of self-trade prevention tools on DCMs.
13

Furthermore, the proposed rules are intended to foster transparency with respect to DCM programs and activities, including market maker and trading incentive programs, that have become more prominent as automated trading becomes the dominant market model.

13

See
section IV(Q) below for a discussion of the term “self-trade” and proposed regulations with respect to self-trade prevention.

The Commission notes that Regulation AT generally does not address trading activity on swap execution facilities (“SEFs”). The Commission believes that neither execution nor order entry on SEF markets are sufficiently automated at this time to require the degree of automated safeguards proposed herein.
14

In addition, Regulation AT is not proposing a number of measures discussed in the Concept Release, such as the following: Proposals to implement various post-trade reports (post-order drop copies, post-trade drop copies, and post-clearing drop copies), “reasonability checks” on incoming market data used by firms operating automated systems, policies and procedures for identifying “related” contracts, and proposals to standardize and simplify order types, each of which was discussed in the Concept Release.
15

14
The requirements on DCMs arising out of Regulation AT may ultimately be imposed on SEFs. However, an important consideration for the Commission is that SEFs and SEF markets are much newer and less liquid than the more established and liquid DCMs and DCM markets. While SEFs and SEF markets are still in this nascent stage, the Commission does not want to impose additional requirements that may have the effect of decreasing the number of SEFs or decreasing liquidity. For these reasons, and in light of the lesser degree of automation in SEF markets, the policy considerations underlying Regulation AT are not as critical, at least at this time, in the SEF context.

15

See
Concept Release, 78 FR at 56569-73 for a summary of measures discussed in the Concept Release.

Market participants using automated trading include an important population of proprietary traders that, while responsible for significant trading volumes and liquidity in key futures products, are not registered with the Commission. These unregistered proprietary traders include a number of traders engaged in high-frequency trading (“HFT”). The Commission notes, however, that the risk control requirements under proposed Regulation AT do not vary in response to a market participant's algorithmic trading strategies; the same risk controls would be required in connection with high-frequency and low-frequency algorithmic trading. In particular, HFT is not specifically identified under the proposed regulations, and is not regulated in a different fashion from other types of algorithmic trading under proposed Regulation AT. Instead, the proposed regulations focus on automation of order origination, transmission and execution, and the risks that may arise from such activity. As discussed above, nearly universal electronic order matching at DCMs is

increasingly complemented by algorithmic order origination among market participants. Against this backdrop, the Commission believes that appropriate pre-trade and other risk controls are necessary at the level of market participants, clearing FCMs, and DCMs, in order to ensure the integrity of Commission-regulated markets and provide market participants with greater confidence that intentional, bona fide transactions are being executed.

Principal elements of Regulation AT for market participants and clearing FCMs include: (i) Codification of defined terms used throughout Regulation AT; (ii) registration of certain entities not otherwise registered with the Commission; (iii) new algorithmic trading procedures for trading firms and clearing firms, including pre-trade and other risk controls; (iv) testing, monitoring, and supervision requirements for ATSs; and (v) requirements that certain persons submit compliance reports to DCMs regarding their ATSs. Principal elements for DCMs include: (i) New risk controls for Direct Electronic Access (“DEA”) provided by DCMs; (ii) transparency in DCM electronic trade matching platforms; and (iii) new risk control procedures, including pre-trade risk controls, compliance report review standards, self-trade prevention tool requirements, and market-maker and trading incentive program disclosure and related requirements.

As mentioned above, Regulation AT is not intended to discriminate across registration categories, connectivity methods, or even “high-frequency” or slower trading strategies. Rather, Regulation AT is focused on reducing risk, increasing transparency and disclosure, and related DCM procedures.
16

In developing Regulation AT, the Commission built on the Concept Release and relevant comments received, which are discussed further in section II(B) below. However, interested parties will observe that the Commission has chosen not to pursue certain measures discussed in the Concept Release (as discussed above), while also proposing a small number of new measures not addressed in the Concept Release. In addition, Regulation AT in certain cases seeks only to clarify the scope of existing Commission regulations that may be impacted by the growth of automated trading environments.

16

See, e.g.,
the compliance reports required to be submitted by AT Persons and clearing member firms of AT Persons under § 1.83, the statistics required to be reported by DCMs regarding self-trading that they have both authorized and prevented on their platforms under § 40.23, and the disclosure required of DCMs with respect to market maker and trading incentive programs under § 40.25.

In preparing this NPRM, the Commission has reviewed relevant industry practices, measures taken by other U.S. and foreign regulators, and best practices or guidance set forth by other informed parties. In these sources and comments received in response to the Concept Release, the Commission has identified an emerging consensus around pre-trade risk controls for automated trading and supervision standards for ATSs. The Commission also notes comments received in response to the Concept Release that are supportive of risk controls placed in multiple stages across the life-cycle of order generation, transmission, management and execution (
i.e.,
similar risk controls placed at the levels of market participants, clearing member FCMs, and DCMs). Proposed Regulation AT attempts to balance flexibility in a rapidly changing technological landscape with the need for a regulatory baseline that provides a robust and sufficiently clear standard for pre-trade risk controls, supervision standards, and other safeguards for automated trading environments. The specific regulations and amendments proposed by Regulation AT are discussed in greater detail below.

2. The Proposed Regulations Under Parts 1, 38, 40, and 170

Regulation AT proposes new regulations or amendments to existing regulations in parts 1, 38, 40, and 170 of the Commission's regulations. It proposes to amend part 1 by inserting the following defined terms: § 1.3(tttt)—Algorithmic Trading Compliance Issue; § 1.3(uuuu)—Algorithmic Trading Disruption; § 1.3(vvvv)—Algorithmic Trading Event; § 1.3(wwww)—AT Order Message; § 1.3(xxxx)—AT Person; § 1.3(yyyy)—Direct Electronic Access; and § 1.3(zzzz)—Algorithmic Trading. Regulation AT also proposes to amend existing § 1.3(x), which defines Floor Trader.

In addition, Regulation AT would create a new subpart A in part 1 that includes the following new regulations applicable to AT Persons and their clearing FCMs: § 1.80—requiring AT Persons to implement pre-trade risk controls and other related measures; § 1.81—requiring AT Persons to implement standards for the development, testing, monitoring, and compliance of their ATSs; § 1.82—requiring clearing member FCMs to implement pre-trade risk controls and other related measures for orders from their AT Person customers; and § 1.83—requiring AT Persons and their clearing member FCMs to provide to DCMs annual compliance reports, and to keep and provide upon request to DCMs certain related books and records.

Regulation AT also proposes to amend part 38 of the Commission's regulations. Specifically, it would amend existing § 38.255—Risk controls for trading, to require DCMs to have in place systems reasonably designed to facilitate the FCM's management of the risks that may arise from their customers' Algorithmic Trading using Direct Electronic Access. Regulation AT would also make corresponding changes to the discussion of risk controls in Appendix B—Guidance on, and Acceptable Practices in, Compliance with Core Principles (Subsection (b)(5)—Acceptable Practices for Risk controls for trading). Finally in part 38, Regulation AT would amend existing § 38.401(a) to require DCMs to provide additional public disclosure regarding their electronic matching platforms.

Regulation AT would also amend part 40 of the Commission's regulations. It would create the following new regulations: § 40.20—requiring DCMs to implement pre-trade risk controls and other related measures; § 40.21—requiring DCMs to provide a test environment to AT Persons; § 40.22—requiring DCMs to implement a review program for compliance reports regarding Algorithmic Trading submitted by AT Persons and clearing member FCMs, require that certain books and records be maintained by such persons, and review such books and records as necessary; § 40.23—requiring DCMs to implement self-trade prevention tools, mandate their use, and publish statistics concerning self-trading; and §§ 40.25-40.28—requiring DCMs to provide disclosure and implement other controls regarding their market maker and trading incentive programs. Finally, Regulation AT would make changes to the definition of Rule in § 40.1(i) in response to certain of the changes proposed above.

Finally, Regulation AT proposes to amend part 170 of the Commission's regulations. It would require in new § 170.18 that all AT Persons become members of at least one registered futures association (“RFA”). Regulation AT would create a new subpart D in part 170, and require in proposed § 170.19 that RFAs adopt membership rules, as deemed appropriate by the RFA, requiring pre-trade risk controls and other measures for ATSs; standards for the development, testing, monitoring, and compliance of ATSs; designation and training of algorithmic

trading staff; and clearing FCM risk management standards.

II. Background on Regulatory Responses to Automated Trading

A. The Commission's Regulatory Response to Date

The Commission has responded to the development of automated trading environments through a number of regulatory measures that address risk controls within both new and existing categories of registrants, including DCMs, SEFs, FCMs, SDs, MSPs and others.
17

While focused to a degree on financial and related risks, these provisions reflect the Commission's ongoing commitment to maintaining the safety and soundness of automated trading in modern derivatives markets. The Commission has adopted regulations with respect to DCMs and SEFs that require exchanges to establish risk control mechanisms to prevent market disruptions, including mechanisms that pause or halt trading.
18

The guidance and acceptable practices to the SEF and DCM rules in part 37 and 38, respectively, provide examples of acceptable risk controls.
19

In addition, in the DCM final rules, the Commission adopted new risk control requirements for exchanges that provide DEA to clients. Regulation 38.607 requires DCMs that permit DEA to have effective systems and controls reasonably designed to facilitate an FCM's management of financial risk.
20

17
These measures are discussed in more detail in the Concept Release.
See
Concept Release, 78 FR at 56548.

18

See
Core Principles and Other Requirements for Designated Contract Markets, 77 FR 36612, 36703 (June 19, 2012) [hereinafter “DCM Final Rules”]; Core Principles and Other Requirements for Swap Execution Facilities, 78 FR 33476, 33590 (June 4, 2013) [hereinafter “SEF Final Rules”].

19

See
DCM Final Rules, 77 FR at 36718; SEF Final Rules, 78 FR at 33601.

20

See
17 CFR 38.607.

The Commission also adopted relevant regulations for FCMs, SDs, and MSPs. Such firms that are clearing members must establish risk-based limits based on position size, order size, margin requirements, or similar factors for all proprietary accounts and customer accounts.
21

The regulations, codified in §§ 1.73 and 23.609, also require these entities to “use automated means to screen orders for compliance with the [risk] limits” when such orders are subject to automated execution.
22

In addition, § 1.11 requires FCMs to have “automated financial risk management controls reasonably designed to prevent the placing of erroneous orders” and “policies and procedures governing the use, supervision, maintenance, testing, and inspection” of automated trading programs.
23

The Commission also adopted regulations requiring SDs and MSPs that are clearing members to ensure that their “use of trading programs is subject to policies and procedures governing the use, supervision, maintenance, testing, and inspection of the program.”
24

21
17 CFR 1.73(a)(1) and 23.609(a)(1).

22
17 CFR 1.73(a)(2)(i) and 23.609(a)(2)(i).

23
17 CFR 1.11(e)(3)(ii). The Commission notes that the requirements of § 1.11(e)(3)(ii) fall within an FCM's broader obligation in § 1.11 to establish and maintain a formal “Risk Management Program.” Such program must include a risk management unit independent of the business unit; quarterly risk exposure reports to senior management and the governing body of the FCM, with copies to the Commission; and other substantive requirements. Proposed Regulation AT would
not
require FCMs to subsume applicable requirements into their § 1.11 Risk Management Programs. However, the Commission is seeking public comment in the questions below regarding whether, in any final rules arising from this NPRM, FCMs should in fact be required to incorporate elements of Regulation AT proposed in §§ 1.80, 1.81, 1.83(a), and 1.83(c) into their § 1.11 Risk Management Programs. Such incorporation could help improve the interaction between an FCM's operational risk efforts pursuant to § 1.11(e)(3)(ii) and its pre-trade risk controls and development, monitoring, and compliance efforts pursuant to §§ 1.80, 1.81, 1.83(a), and 1.83(c). It could also help ensure that an FCM's §§ 1.80, 1.81, 1.83(a), and 1.83(c) processes benefit from the same internal rigor and independence required by § 1.11.

24
17 CFR 23.600(d)(9).

Finally, the Commission adopted final rules implementing new authority under the CEA to, among other things, broadly prohibit manipulative and deceptive devices and price manipulation.
25

The Commission also provided guidance on the scope and application of CEA Section 4c(a)(5), which makes it unlawful for any person to engage in any trading, practice, or conduct on or subject to the rules of a registered entity that violates bids or offers, demonstrates intentional or reckless disregard for the orderly execution of transactions during the closing period, or is, is of the character of, or is commonly known to the trade as, “spoofing.”
26

25

See
17 CFR 180.1 and 180.2.

26

See
Antidisruptive Practices Authority, 78 FR 31890 (May 28, 2013).

B. The Commission's 2013 Concept Release

Overview of Concept Release.
As noted above, in 2013 the Commission issued a “Concept Release on Risk Controls and System Safeguards for Automated Trading Environments,” which provided an overview of the automated trading environment and discussed a series of pre-trade risk controls, post-trade reports and other measures, system safeguards, and additional protections that could be implemented by Commission registrants or other market participants. The Concept Release reflects the Commission's ongoing commitment to the safety and soundness of U.S. derivatives markets in times of technological change, including the growth of automated trading.

The Concept Release was published in the
Federal Register
on September 12, 2013.
27

The initial 90-day comment period closed on December 11, 2013, but was reopened from January 21 through February 14, 2014, in conjunction with a meeting of the CFTC's Technology Advisory Committee (“TAC”). The Concept Release requested public comment on 124 separate questions regarding the necessity and operation of potential pre-trade risk controls, post-trade reports and other measures, system safeguards and additional protections (such as proposals to identify “related” contracts on trading platforms, and proposals to standardize and simplify order types). The Concept Release served as a vehicle to catalogue existing industry practices, determine their efficacy and implementation to date, and evaluate the need for additional measures. The Concept Release was not a proposed rule, but rather a prior step designed to facilitate a public dialogue and educate the Commission so that it may make an informed determination as to whether rulemaking is necessary and, if so, the substantive requirements of such a rulemaking.

27
Concept Release, 78 FR 56542.

Topics Discussed in Concept Release.
The Concept Release highlighted data on the increased importance of electronic and algorithmic trading across a number of U.S. markets (including equities, futures and fixed income markets). The Concept Release also noted that the infrastructure of automated trading environments has progressively decreased the time necessary to process orders and execute trades, reducing the communication times between market participants and trading venues.
28

One exchange group now indicates that its “median inbound latency for order entry” on its trading platform is fifty-two (52) microseconds within its “four walls.”
29

As discussed in the Concept Release, advances in trading speeds are partly due to the development of dedicated fiber-optic and microwave communications networks that have dramatically reduced transmission times across large

distances.
30

On a smaller scale, co-location and proximity hosting are two common methods for reducing the distance, and thus latency, between market participants and the exchanges. Co-location services are now provided by most large electronic trading platforms within the United States.

28

See id.
at 56546-47.

29

See
CME Group, “The World's Leading Electronic Platform. CME Globex,” (2014) at 3, available at
http://www.cmegroup.com/globex/files/globexbrochure.pdf.

30

See
Concept Release, 78 FR at 56546.

Another important latency-reducing advance in connectivity discussed in the Concept Release is Direct Market Access (“DMA”). For purposes of the Concept Release, the Commission defined DMA as a connection method that enables a market participant to transmit orders to a trading platform without reentry or prior review by systems belonging to the market participant's clearing firm.
31

DMA can be provided directly by an exchange or through the infrastructure of a third-party provider, but in all cases, DMA implies that an order is not routed through a clearing firm prior to reaching the trading platform.
32

For purposes of Regulation AT, as discussed in section IV(D)(7) below, the Commission proposes to define a slightly modified term: “Direct Electronic Access” (“DEA”), as opposed to Direct Market Access. Despite the slightly modified name, the Commission intends that the term “Direct Electronic Access” has a meaning similar to “Direct Market Access,” as such term was used in the Concept Release.
33

31

See id.

32

See id.

33
The Commission notes that the term “direct electronic access” is also used in existing Commission regulation 38.607. Regulation AT does not modify § 38.607, and the term “direct electronic access” in § 38.607 will continue to have the meaning specified in that section.

The Concept Release discussed a set of risk controls that would be intended to operate at the same rapid speed at which trading occurs in the automated trading environment. As the industry reduces latency through improvements in technologies for the generation, transmission and execution of orders or management of other data, there is concern that the drive for ever lower latencies may lead to a competitive race toward progressively less stringent risk controls.
34

A separate, but related, concern is that market participants may simply engage in trading at speeds beyond the abilities of their risk management systems, or those tasked with monitoring their activity. Risk management systems operating at these misaligned speeds could allow an active algorithm to breach its prescribed risk controls and disrupt one or more markets.

34
As noted by the Futures Industry Association's Market Access Working Group, for example: “[p]re-trade risk controls have become a point of negotiation between trading firms and clearing members because they can add latency to a trade.”
See
FIA Market Access Risk Management Recommendations,
infra
note 97 at 8. Similarly, the TAC's Pre-Trade Functionality Subcommittee noted that latency is a key area where trading firms and brokers are competing to gain an advantage.
See
TAC Pre-Trade Functionality Subcommittee, “Recommendations on Pre-Trade Practices for Trading Firms, Clearing Firms, and Exchanges Involved in Direct Market Access” (Mar. 1, 2011) at 2 [hereinafter “CFTC TAC Recommendations”], available at
http://www.cftc.gov/ucm/groups/public/@swaps/documents/dfsubmission/tacpresentation030111_ptfs2.pdf.

In light of the potential for disruptive trading events related to such high-speed algorithmic trading, the Concept Release addressed 23 potential risk controls and other measures broadly grouped into four categories. The first includes “pre-trade risk controls,” such as controls designed to prevent potential errors or disruptions from reaching trading platforms, or to minimize their impact once they have. A second category of safeguards includes “post-trade reports” and “other post-trade measures.” Examples in this category include reports that promote the flow of order, trade and position information; uniform trade adjustment or cancellation policies; and standardized error trade reporting obligations. The third category of risk controls discussed in the Concept Release is termed “system safeguards,” including safeguards for the design, testing and supervision of ATSs, as well as measures such as “kill switches” that facilitate emergency intervention in the case of malfunctioning ATSs.
35

Finally, the Concept Release presented a fourth category of measures focusing on various options for improving market functioning or structure.

35
As explained in section IV(A) below, the Concept Release used the term “ATS” or “automated trading system” to refer to the algorithms used to automate the generation and execution of a trading strategy. For purposes of this NPRM, the Commission has determined to use the term “Algorithmic Trading” or “algorithmic trading system” (abbreviated as ATS), as opposed to the term “automated trading system.” For purposes of discussing comments to the Concept Release, the Commission may use the terms ATS and automated trading system as such terms were used in the Concept Release.

Comments Received on Concept Release and Commission Response.
The Commission received a total of 43 public comments on the Concept Release, including comments from DCMs, an array of trading firms, trade associations, public interest groups, members of academia, and consulting, technology and information service providers in the financial industry. All comments are available on
www.cftc.gov.
Many of the comments received are detailed and thorough, and the Futures Industry Association (“FIA”) conducted surveys to gauge existing risk-management practices. Other commenters provided academic papers in support of their points of view.

Staff reviewed all comments received and made recommendations to the Commission. This NPRM reflects the Commission's decision to propose regulations in certain areas addressed by the Concept Release, including: Registration of certain entities not otherwise registered with the Commission; enhanced identification of orders placed on exchanges; pre-trade risk controls at exchanges, trading firms and clearing firms; standards for development, testing and supervision of algorithmic systems; trading firm and clearing member FCM compliance reports regarding algorithmic trading; and self-trade prevention tools. Regulation AT also addresses several areas not covered in the Concept Release, including transparency in exchange trade matching systems and market-maker protections, and in certain cases seeks to clarify the scope of existing Commission regulations that may be impacted by the growth of automated trading environments.

C. Other Recent Regulatory Responses

1. SEC Regulatory Initiatives

The SEC has recently taken regulatory steps related to automated trading, aimed at preventing instability in the equities markets. Most significantly, the SEC adopted the Market Access Rule and Regulation SCI.

The Securities Exchange Act Rule 15c3-5—Risk Management Controls for Brokers or Dealers with Market Access (the “Market Access Rule”), adopted in November 2010, requires brokers and dealers to have risk controls in place before providing their customers with access to the market.
36

Specifically, the Market Access Rule requires risk controls that prevent entry of (i) orders exceeding appropriate pre-set credit or capital thresholds in the aggregate for each customer and the broker-dealer; and (ii) erroneous orders, by rejecting orders that exceed appropriate price or size parameters, on an order-by-order basis or over a short period of time, or those that indicate duplicative orders.
37

These risk controls must be under the direct and exclusive control of the broker-dealer (subject to certain exceptions) and regularly reviewed for effectiveness.
38

In October 2013, the SEC brought its first enforcement action under the Market Access Rule, securing a $12 million settlement with Knight Capital in connection with the firm's August 2012 trading incident that disrupted the markets.
39

36

See
Market Access Rule, 75 FR 69792 (Nov. 15, 2010);
see also
SEC Press Release No. 2010-210, “SEC Adopts New Rule Preventing Unfiltered Market Access” (Nov. 3, 2010), available at
http://www.sec.gov/news/press/2010/2010-210.htm.

37

See
Market Access Rule,
supra
note 36 at 69825-26;
see also
SEC, Responses to Frequently Asked Questions Concerning Risk Management Controls for Brokers or Dealers with Market Access (Apr. 15, 2014), available at

https://www.sec.gov/

divisions/marketreg/faq-15c-5-risk-management-controls-bd.htm.

38

See
Market Access Rule,
supra
note 36 at 69826.

39

See
SEC Press Release No. 2013-222, “SEC Charges Knight Capital With Violations of Market Access Rule” (Oct. 16, 2013), available at
http://www.sec.gov/News/PressRelease/Detail/PressRelease/1370539879795
[hereinafter “SEC Knight Capital Release”].

On November 19, 2014, the SEC adopted Regulation Systems Compliance and Integrity (“Reg SCI”).
40

Reg SCI applies to alternative trading systems, certain self-regulatory organizations (including registered clearing agencies), plan processors, and exempt clearing agencies (collectively, “SCI entities”). Under Reg SCI, SCI entities are required to have comprehensive policies and procedures in place for their technological systems. The SCI entities must, among other things, take appropriate corrective action when systems issues occur; provide notifications and reports to the SEC regarding systems problems and systems changes; inform members and participants about systems issues; conduct business continuity testing; implement standards that result in SCI systems being designed, developed, tested, maintained, operated, and surveilled in a manner that facilitates the successful collection, processing, and dissemination of market data; and conduct annual reviews of their automated systems, which must be summarized in a report that is provided to the SEC.
41

40

See
Reg SCI, 79 FR 72252 (Dec. 5, 2014);
see also
SEC Press Release No. 2014-260, “SEC Adopts Rules to Improve Systems Compliance and Integrity” (Nov. 19, 2014), available at
http://www.sec.gov/News/PressRelease/Detail/PressRelease/1370543496356#.VKQS2qxOlaQ.

41

See
Reg SCI,
supra
note 40 at 72437-39.

The SEC has also taken action in the area of enhancing oversight of proprietary trading firms. In March 2015, the SEC proposed a rule that would narrow an exemption that currently exempts certain broker-dealers from membership in a national securities association.
42

The exemption was originally designed to accommodate exchange specialists and other floor members that might need to conduct limited hedging or other off-exchange activities ancillary to their business.
43

Over time, proprietary trading firms were able to take advantage of this exemption.
44

The SEC's proposed rules would amend the exemption to target those broker-dealers for which it was originally designed, and require broker-dealers trading in off-exchange venues to become members of a national securities association. In the securities markets, this association is the Financial Industry Regulatory Authority (“FINRA”).
45

42

See
SEC, Press Release No. 2015-48, “SEC Proposes Rule to Require Broker-Dealers Active in Off-Exchange Market to Become Members of National Securities Association” (Mar. 25, 2015) [hereinafter “SEC Press Release on Broker-Dealer Registration”], available at
http://www.sec.gov/news/pressrelease/2015-48.html#.VSbd9KwpBaQ
; Exemption for Certain Exchange Members, 80 FR 18036, 18042-43 (Apr. 2, 2015) [hereinafter “SEC Proposed Rule on Exemption for Certain Exchange Members”].

43

See
SEC Press Release on Broker-Dealer Registration,
supra
note 42.

44

See id.
The SEC estimates that there are approximately 125 firms exempt from association membership, which includes some of the most active cross-market proprietary trading firms.
See
SEC Proposed Rule on Exemption for Certain Exchange Members, 80 FR at 18042.

45

See
SEC Press Release on Broker-Dealer Registration,
supra
note 42.

The SEC's Chair explained that the proposed rule “embodies a simple but powerful principle of the federal securities laws—the protection of investors and the stability of our markets require that trading is overseen by both the Commission and a strong self-regulatory organization.”
46

In its preamble to the proposed rule, the SEC explained that, in the event that a broker-dealer trades electronically across a range of exchange and off-exchange venues, an individual exchange of which the broker-dealer is a member may be unable to effectively regulate the off-exchange activity of the broker-dealer, because the exchange may lack the resources or expertise to oversee such off-exchange activity.
47

The SEC viewed FINRA, the self-regulatory organization (“SRO”) to which off-exchange trades are reported, as being in the best position to regulate cross-market activity by broker-dealers.
48

46

See id.

47
SEC Proposed Rule on Exemption for Certain Exchange Members, 80 FR at 18042-43.

48

See id.
at 18041-45.

The SEC has taken additional regulatory initiatives in this area. On July 11, 2012, the SEC adopted Rule 613 under Regulation NMS, requiring SROs to submit a plan to the SEC to create, implement, and maintain a consolidated audit trail (“CAT”). This audit trail is intended to increase the data available to regulators investigating illegal activities such as insider trading and market manipulation, and improve the ability to reconstruct broad-based market events in an accurate and timely manner.
49

The SROs submitted the plan on September 30, 2014.
50

In addition, in response to policy recommendations resulting from the Flash Crash events of May 6, 2010, the SEC and the securities industry implemented market-wide circuit breakers as well as a “limit up-limit down” mechanism in order to moderate price volatility in individual securities.
51

The SEC is also working to update its regulatory regime to improve firms' risk management of trading algorithms and to enhance regulatory oversight over their use.
52

The SEC is also developing an anti-disruptive trading rule to address the use of aggressive, potentially destabilizing trading strategies during vulnerable market periods.
53

49

See
SEC Press Release No. 2012-134, “SEC Approves New Rule Requiring Consolidated Audit Trail to Monitor and Analyze Trading Activity” (July 11, 2012), available at
http://www.sec.gov/News/PressRelease/Detail/PressRelease/1365171483188#.VKQkAqxOlaQ.

50

See
SEC, “Rule 613 (Consolidated Audit Trail),” available at
http://www.sec.gov/divisions/marketreg/rule613-info.htm.

51

See
Mary Jo White, Chairman, Securities and Exchange Commission, Enhancing Our Equity Market Structure (June 5, 2014), available at
http://www.sec.gov/News/Speech/Detail/Speech/1370542004312#.VKP_o6xOlaS.

52

See id.

53

See id.

Finally, while not directly relevant to Commission-regulated markets, the SEC is working with equities exchanges and FINRA to minimize latency between different market feeds. Specifically, exchanges must not transmit data directly to customers any sooner than they transmit data to a securities information processor (“SIP”), the system that consolidates market feeds from all platforms and publishes the public price ticker. In addition, the technology used for transmitting data to the SIP must be on a par with what is used for transmitting data to direct feeds.
54

Finally, the SEC is working to address concerns associated with the fragmentation of trading venues, dark trading venues, and broker conflicts.
55

54

See id.

55

See id.

2. FINRA Initiatives

In addition to the SEC, FINRA is developing rules focused on automated trading and transparency in the equities markets. In March 2015, FINRA published a Request for Comment proposing to require registration (as a “Limited Representative—Equity Trader”) persons that are (1) primarily responsible for the design, development

or significant modification of an algorithmic strategy; or (2) responsible for supervising such functions.
56

FINRA explained that given today's highly automated environment (according to FINRA, where firms trade using automated systems that initiate pre-programmed trading instructions based on specified variables, referred to as algorithmic trading strategies), it is concerned that persons involved in preparing or supervising algorithmic trading may lack adequate knowledge of securities rules and regulations, which could result in algorithms that do not comply with applicable rules.
57

Accordingly, FINRA believes such persons should meet the same minimum competency standards for knowledge of securities regulations that apply to individual traders.
58

56

See
FINRA, Regulatory Notice 15-06, “Registration of Associated Persons Who Develop Algorithmic Trading Strategies” (Mar. 2015), available at
http://www.finra.org/sites/default/files/notice_doc_file_ref/Notice_Regulatory_15-06.pdf.

57

See id.
at 3.

58

See id.

In March 2015, FINRA published a regulatory notice (15-09) providing guidance on supervision and control practices for algorithmic trading strategies in the equities markets.
59

The notice offered guidance on practices in five general areas: General risk assessment and response; software/code development and implementation; software testing and system validation; trading systems; and compliance. Among other practices, the notice recommended that firms should consider: Implementing a development and change management process that tracks the development of new trading code or material changes to existing code; implementing a basic summary description of algorithmic trading strategies that enables supervisory and compliance staff to understand the intended function of an algorithm; conducting testing to confirm that core code components operate as intended and do not produce unintended consequences; implementing controls, monitors, alerts and reconciliation processes that enable the firm to quickly identify whether an algorithmic is experiencing unexpected results; and providing for adequate communication between supervisory and compliance staff related to the function and control of algorithms such that the firm meets its regulatory obligations.
60

59

See
FINRA, Regulatory Notice 15-09, “Equity Trading Initiatives: Supervision and Control Practices for Algorithmic Trading Strategies” (Mar. 2015) [hereinafter “FINRA Notice 15-09”], available at
https://www.finra.org/industry/notices/15-09.

60

See id.

3. European and Other Regulatory Initiatives

a. ESMA

The European Securities and Markets Authority (“ESMA”) is an independent EU Authority established in January 2011. ESMA published guidelines on automated trading in February 2012, which became effective across the European Union on May 1, 2012.
61

The ESMA guidelines addressed the operation of an electronic trading system by a regulated market or a multilateral trading facility; the use of an electronic trading system, including a trading algorithm, by an investment firm for dealing on its own account or for the execution of orders on behalf of clients; and the provision of direct market access or sponsored access by an investment firm as part of the service of the execution of orders on behalf of clients.
62

61

See
ESMA, “Systems and controls in an automated trading environment for trading platforms, investment firms and competent authorities” (Feb. 24, 2012) [hereinafter “ESMA Guidelines”], available at
http://www.esma.europa.eu/system/files/esma_2012_122_en.pdf
and accompanying public statement, available at
http://www.esma.europa.eu/system/files/2012-128.pdf.

62

See id.
at 3.

Among other elements, the ESMA guidelines recommended that trading platforms should have: Arrangements to prevent the excessive flooding of the order book; arrangements (such as throttling) to prevent capacity limits on messaging from being breached; and arrangements (for example, volatility interruptions or automatic rejection of orders which are outside of certain set volume and price thresholds) to constrain trading or to halt trading in individual or multiple financial instruments when necessary.
63

The ESMA guidelines also recommended that trading platforms should have procedures in place to identify potential market abuse in an automated trading environment, such as ping orders, quote stuffing, momentum ignition, and layering and spoofing.
64

63

See id.
at 13.

64

See id.
at 16-17.

In addition, the ESMA guidelines recommended that investment firms should make use of clearly delineated development and testing methodologies prior to deploying an electronic trading system or a trading algorithm, and should monitor their electronic trading systems, including trading algorithms, in real-time.
65

ESMA also recommended that investment firms implement price and size parameters, systems that control messaging traffic to individual trading platforms, financial risk controls, and controls that block a trader's orders if they are for a financial instrument that the trader does not have permission to trade.
66

As to orders submitted via direct market access and sponsored access, ESMA recommended, among other things, that such orders be submitted to the same pre-trade risk controls that it recommends for investment firms (including, for example, price and size parameters).
67

65

See id.
at 10.

66

See id.
at 14-15.

67

See id.
at 21-23.

On March 18, 2015, ESMA released a report finding that all 30 participating European Economic Area members have incorporated the Guidelines into their legal framework, and all except three have incorporated it into their supervisory framework.
68

The report went on to identify challenges to further enhancing compliance including: Market complexity, IT-knowledge, additional on-site inspections of markets, testing of trading halts, and setting up ring-defense against cyber-attacks.
69

68
ESMA, Automated Trading Guidelines: ESMA Peer Review Among National Competent Authorities (Mar. 18, 2015), available at
http://www.esma.europa.eu/system/files/esma-2015-592-automated_trading_peer_review_report_publication.final_.pdf.

69

See id.
at 9-10.

As discussed below, ESMA has performed additional work in the area of automated trading, such as developing technical standards for the requirements of MiFID II.

b. MiFID II

The European Commission published a new Directive on markets in financial instruments (“MiFID II”) on June 12, 2014.
70

The Directive contains a definition of both `algorithmic trading' and `high-frequency algorithmic trading technique,' which is defined as a specific type of algorithmic trading. Among other requirements, the Directive requires that an investment firm engaged in algorithmic trading must have effective systems and risk controls to ensure that its trading systems are resilient and have sufficient capacity, are subject to appropriate trading thresholds and limits, and prevent the sending of erroneous orders or other system activity that may create or contribute to a disorderly market.
71

Such a firm must also have effective business continuity arrangements to deal with any failure of its trading

systems and must ensure its systems are fully tested and properly monitored.
72

Furthermore, an investment firm that engages in a high-frequency algorithmic trading technique must store in an approved form accurate and time sequenced records of all its placed orders, including cancellations of orders, executed orders and quotations on trading venues and make them available to the competent authority upon request.
73

70

See
European Commission, “Updated rules for markets in financial instruments: MiFID 2” (June 12, 2014) [hereinafter “MiFID II”], available at
http://ec.europa.eu/finance/securities/isd/mifid2/index_en.htm.

71

See id.
at Article 17(1).

72

See id.

73

See id.
at Article 17(2).

The MiFID II Directive also requires a regulated market to be able to temporarily halt or constrain trading if there is a significant price movement in a financial instrument on that market or a related market during a short period. In exceptional cases, a regulated market must be able to cancel, vary or correct any transaction.
74

In addition, the Directive requires a regulated market to have in place effective systems, procedures and arrangements, including requiring members or participants to carry out appropriate testing of algorithms. A regulated market must also provide environments to facilitate such testing, to ensure that algorithmic trading systems cannot create or contribute to disorderly trading conditions on the market. The Directive requires a regulated market to implement systems to limit the ratio of unexecuted orders to transactions that may be entered into the system by a member or participant, to be able to slow down the flow of orders if there is a risk of its system capacity being reached, and to limit and enforce the minimum tick size that may be executed on the market.
75

74

See id.
at Article 48(5).

75

See id.
at Article 48(6).

The European Commission requested that ESMA develop technical and implementing standards for MiFID II. On May 22, 2014, ESMA published a consultation paper seeking comments on certain topics in connection with MiFID II, including “micro-structural issues” such as testing and risk control requirements for investment firms engaged in algorithmic trading and trading venues.
76

ESMA published another consultation paper on December 19, 2014, seeking further comments on technical and implementing standards in connection with the implementation of MiFID II and summarizing comments received in response to ESMA's May 2014 paper.
77

The comment period for the December 19, 2014 consultation paper closed in March 2015. In late 2014, ESMA released a final report covering technical advice in certain areas, including the definition of algorithmic trading, HFT, and direct electronic access.
78

In July 2015, ESMA released final technical advice relating to investor protection topics, including procedures for financial services firms to apply for authorized status, information required of firms applying to passport into other jurisdictions, and co-operation between regulatory authorities.
79

On September 28, 2015, ESMA released a final report on draft regulatory and implementing technical standards for MiFID II (“2015 Final Draft Regulatory Standards”).
80

This report provides regulatory standards for investment firms engaged in algorithmic trading as well as for trading venues that allow algorithmic trading. Details regarding ESMA's standards are discussed below as relevant to the Commission's proposed regulations relating to risk controls and other measures that AT Persons, clearing member FCMs and DCMs must implement.

76
ESMA, “Consultation Paper,” (May 22, 2014), available at
http://www.esma.europa.eu/system/files/2014-549_-_consultation_paper_mifid_ii_-_mifir.pdf.

77
ESMA, “Consultation Paper,” (Dec. 19, 2014) and accompanying Annexes A and B, available at
http://www.esma.europa.eu/system/files/2014-1570_cp_mifid_ii.pdf.

78
ESMA, “ESMA's Technical Advice to the Commission on MiFID II and MiFIR,” (Dec. 19, 2014) [hereinafter “ESMA Technical Advice Final Report”], available at
http://www.esma.europa.eu/system/files/2014-1569_final_report_-_esmas_technical_advice_to_the_commission_on_mifid_ii_and_mifir.pdf.

79
ESMA, Final Report: MiFID II/MiFIR draft Technical Standards on authorization, passporting, registration of third country firms and cooperation between competent authorities, Art. 6(g) (June 29, 2015), available at
http://www.esma.europa.eu/system/files/2015-esma-1006_-_mifid_ii_final_report_on_mifid_ip_technical_standards.pdf.

80
ESMA, Final Report: Draft Regulatory and Implementing Technical Standards MiFID II/MiFIR (Sept. 28, 2015) [hereinafter, the “ESMA September 2015 Final Draft Standards Report”], available at
https://www.esma.europa.eu/system/files/2015-esma-1464_-_final_report_-_draft_rts_and_its_on_mifid_ii_and_mifir.pdf
; ESMA, Regulatory technical and implementing standards—Annex 1 (Sept. 28, 2015) [hereinafter, the “ESMA September 2015 Final Draft Standards Report Annex 1”], available at
https://www.esma.europa.eu/system/files/2015-esma-1464_annex_i_-_draft_rts_and_its_on_mifid_ii_and_mifir.pdf
; ESMA, Cost-Benefit Analysis—Annex II, Draft Regulatory and Implementing Technical Standards MiFID II/MiFIR (Sept. 28, 2015), [hereinafter, the “ESMA September 2015 Cost-Benefit Annex II”], available at
https://www.esma.europa.eu/system/files/2015-esma-1464_annex_ii_-_cba_-_draft_rts_and_its_on_mifid_ii_and_mifir.pdf.

c. Other European Regulatory Initiatives

In May 2013, Germany enacted the Act on the Prevention of Risks and Abuse in High-frequency Trading (the “High-frequency Trading Act”).
81

The High-frequency Trading Act requires that firms engaged in high-frequency trading must be licensed.
82

In summary, high-frequency trading is defined to include each of the following four elements: (i) Trading for one's own account, or by proprietary trading firms; (ii) trading algorithmically without human intervention; (iii) trading using low-latency infrastructures; and (iv) trading that generates a high intraday message rate.
83

In addition, exchanges must impose, on a product-by-product basis, an excessive system usage fee and an order-to-trade ratio limit intended to prevent unnecessary messaging.
84

Finally, the High-frequency Trading Act requires identification of algorithmically generated orders and trading algorithms, which is intended to enhance monitoring of manipulative activity.
85

81

See
online summaries of High-frequency Trading Act (2013), available at
http://www.bafin.de/SharedDocs/Veranstaltungen/EN/WA11_20130430_hft_workshop_en.html.

82

See id.; see also
Morgan, Megan, Tabb Forum, “Decoding the German HFT Act: A Guide to Regulating Electronic Markets” (Oct. 17, 2014), available at
http://tabbforum.com/opinions/decoding-the-german-hft-act-how-to-regulate-electronic-markets.

83

See id.

84

See id.

85

See id.

In May 2015, the Bank of England's Prudential Regulation Authority (“PRA”), the United Kingdom's prudential supervisor of major trading firms, announced that it would assess the adequacy of existing risk measurement and management practices with respect to trading algorithms, including whether controls around algorithmic trading are “fit for purpose.”
86

The PRA discussed the growth of automated trading in financial markets, which has included incidents of extreme volatility. For example, volatility seen in the Swiss Franc exchange rate on January 15, 2015, following the Swiss central bank's decision to remove a floor to the exchange rate, may have been exacerbated by high-frequency trading.
87

86

See
Bailey, Andrew, Bank of England, “Financial Markets: Identifying risks and appropriate responses,” at 9 (May 15, 2015), available at
http://www.bankofengland.co.uk/publications/Documents/speeches/2015/speech814.pdf.

87

See id.
at 5-6; Binham, Caroline, “High-frequency trading faces tougher Bank of England scrutiny,” Financial Times (May 15, 2015), available at
http://www.ft.com/intl/cms/s/0/f7d4e438-fb20-11e4-9aed-00144feab7de.html#axzz3aUzgwb2N.

Finally, in July 2015, the United Kingdom's Financial Conduct Authority issued a consultation paper addressing strengthening accountability in

banking.
88

The proposed rule specifically set out to capture individuals responsible for the deployment of trading algorithms in its Certification Regime.
89

Pursuant to the proposal, individuals responsible for: (1) Approving the deployment of a trading algorithm or a material part of one; (2) approving the deployment of a material amendment to a trading algorithm or a material part of one, or the combination of trading algorithms; and (3) monitoring or deciding whether or not the use or deployment of a trading algorithm is or remains compliant with the firm's obligations would be captured and subject to the Certification Regime.
90

88
Financial Conduct Authority (“FCA”), CP15/22 Strengthening accountability in banking: Final rules (including feedback on CP14/31 and CP15/5) and consultation on extending the Certification Regime to wholesale market activities, at 46 (July 2015), available at
https://www.fca.org.uk/your-fca/documents/consultation-papers/cp15-22.

89

Id.

90

Id.

d. The October 15 Joint Staff Report

As discussed above in section I(B), on July 13, 2015, five regulatory agencies issued the October 15 Joint Staff Report on the unusually high level of volatility and rapid round-trip in prices that occurred on October 15, 2014 in the market for U.S. Treasury securities, futures and other closely related financial markets.
91

In addition to discussing the events of October 15, the report includes an Appendix C that summarizes many of the risks of automated trading. These risks include the following: Operational risks (ranging from malfunctioning and incorrectly deployed algorithms to algorithms reacting to inaccurate or unexpected data); market liquidity risks (arising from abrupt changes in trading strategies even when a firm executes its strategy perfectly); market integrity risks (automated trading can provide new tools to engage in unlawful conduct); transmission risks (shocks based on erroneous orders impacting multiple markets); clearing and settlement risks (as more firms gain access to trading platforms, trades may not be subject to sufficient settlement risk mitigation techniques); and risks to effective risk management (the speed of trade execution may make critical risk mitigation devices less effective).

91

See
October 15 Joint Staff Report,
supra
note 10.

D. Industry and Regulatory Best Practices and Recommendations

Widely recognized organizations and governmental entities or agencies have issued “best practices” for automated trading, including the National Futures Association (“NFA”), the FIA, ESMA, and the International Organization of Securities Commissions (“IOSCO”), among others.

1. NFA Compliance Rule 2-9: Supervision

NFA, a registered futures association under Section 17 of the Act, has provided guidance regarding ATSs to industry participants since 2002. Specifically, NFA Interpretive Notice 9046 addresses the “Supervision of the Use of Automated Order-Routing Systems” in the context of NFA's overarching supervision requirements in Compliance Rule 2-9 (Supervision).
92

The Commission believes that Compliance Rule 2-9 and Interpretive Notice 9046 are especially relevant because of their wide applicability as NFA membership rules, binding on FCMs, IBs, CPOs, CTAs, and other NFA members. In addition, these provisions and interpretations have been in place since at least 2006, such that NFA members—and by extension many AT Persons—will have been subject to regulatory requirements concerning algorithmic trading for many years.

92
NFA, “9046—Compliance Rule 2-9: Supervision of the Use of Automated Order-Routing Systems,” (Dec. 12, 2006), available at
https://www.nfa.futures.org/nfamanual/NFAManual.aspx?RuleID=9046&Section=9.

Compliance Rule 2-9 requires each NFA member to “diligently supervise its employees and agents in the conduct of their commodity futures activities for or on behalf of the Member.” Interpretive Notice 9046, first issued in 2002 and revised in 2006, states that NFA's board of directors “firmly believes that supervisory standards do not change with the medium used. How those standards are applied, however, may be affected by technology.” To fulfill their supervisory responsibilities, NFA members “must adopt and enforce written procedures to examine the security, capacity, and credit and risk-management controls provided by the firm's automated order-routing systems (AORSs).” Interpretive Notice 9046 applies to systems “that are within a Member's control, including AORSs that are provided to the Member by an application service provider or an independent software vendor.” NFA acknowledges that NFA members will not control an AORS chosen by an NFA customer, such as direct access systems provided by exchanges. In such circumstances, the NFA member must nevertheless adopt procedures “reasonably expected to address the trading, clearing, and other risks attendant to [their] customer relationship[s].”

Among other requirements, Interpretive Notice 9046 addresses the following standards for automated systems:

• Pre-Execution Controls (including both credit and “fat-finger” protections): “An AORS should allow the Member to set limits for each customer based on commodity, quantity, and type of order or based on margin requirements. It should allow the Member to impose limits pre-execution and to automatically block any orders that exceed those limits.”
93

93
Interpretive Notice 9046 does not require NFA members to “impose pre-execution controls on all customers, however. The Member should review the customer's sophistication, credit-worthiness, objectives, and trading practices and strategies when determining whether to impose controls pre-execution or post-execution and deciding what levels to use when setting limits.”

• Post-Execution Controls: “For customers subject to post-execution controls, the Member should have the ability to monitor trading promptly. The AORS should generate alerts when limits are exceeded through that system. The system should also allow the Member to block subsequent orders, either in their entirety or by kind (
e.g.,
to block orders that create a new position or increase an existing position but not orders that liquidate some or all of an existing position).”
94

94
The Interpretive Notice adds that “[t]his ability can be provided by the AORS or through other risk-management systems.”

• Direct Access Systems: “When authorizing [customer] use of a direct access system that does not allow the Member to monitor trading promptly, the Member should utilize pre-execution controls, if available, to set pre-execution limits for each customer, regardless of the nature of the customer.”

• Review: “Members should use AORSs in conjunction with their credit-review/risk-management systems and should evaluate the controls imposed on each customer as part of their regular credit and risk-control procedures.”

A number of the controls summarized above are in keeping with the Commission's proposed requirements for AT Persons, including proposed § 1.80, which requires pre-trade risk controls and other measures reasonably designed to prevent an Algorithmic Trading Event, including but not limited to maximum order message and execution frequencies per unit time; order price parameters and maximum order sizes; and certain order cancellation capabilities. The Commission notes once again its intent in much of Regulation AT to build on

existing regulatory requirements and industry practices so that its proposed regulations facilitate an ongoing transition to effective risk controls in algorithmic trading. The Commission believes that the existence of related regulatory standards enforced by NFA since 2002 and updated in 2006 would help minimize any potential disruptions or burdens that would otherwise be associated with a number of the Commission's proposed rules for AT Persons. The Commission also believes that NFA's prior experience in this area will assist in complying with the requirements of proposed § 170.19, discussed in detail in section IV(F) below.

2. FIA Reports on Automated Trading

On March 23, 2015, FIA released the “FIA Guide to the Development and Operation of Automated Trading Systems” (the “FIA Guide”), which provides recommendations concerning appropriate risk controls at the trader, broker and exchange levels.
95

Risk controls recommended by FIA include maximum order size limits, maximum intraday position limits, market data reasonability checks, price tolerance limits, repeated automated execution limits, exchange dynamic price collars, exchange market pauses, exchange message programs, message throttles, self-trade prevention tools, kill switches, cancel-on-disconnect service and exchange-provided order management tools. FIA also recommended audit trail procedures that identify automated trading system operators; certain post-trade measures to monitor for potential credit events or unintended trading; measures related to co-location services; and disaster recovery and business continuity procedures. Finally, FIA recommended measures related to automated trading system development and support, including general principles related to testing; policies and procedures related to security; systems monitoring procedures; and documentation procedures. Consistent with the approach the Commission intends to pursue in Regulation AT, the FIA Guide states that, “[c]are should be taken to avoid implementing overly prescriptive standards or rules that impose a one-size-fits-all approach to all entities.”
96

95
FIA, “FIA Guide to the Development and Operation of Automated Trade Systems” (Mar. 23, 2015) [hereinafter “FIA Guide”], available at
https://fia.org/sites/default/files/FIA%20Guide%20to%20the%20Development%20and%20Operation%20of%20Automated%20Trading%20Systems.pdf.

96

Id.
at 6.

The Commission encourages industry participants to consider FIA's recommendations. In the event that the FIA Guide recommends best practices that are not proposed in Regulation AT, the Commission encourages industry participants to consider implementing the FIA best practices if they are appropriate to their business and are reasonably designed to prevent an Algorithmic Trading Event. FIA's recommendations may also serve as a useful starting point for an RFA considering potential measures in response to proposed § 170.19, discussed in section IV(F) below.

FIA has issued several additional reports related to the appropriate best practices that should be implemented with respect to automated trading. In April 2010, FIA issued a report addressing the risks of direct market access and providing recommendations for risk controls to be implemented by exchanges and applied across all trading firms.
97

In November 2010, FIA's Principal Traders Group (“FIA PTG”) released a report setting out recommended risk controls for trading firms that have direct access to exchange matching engines,
98

as well as a global survey of futures exchanges to determine what controls were in place to manage the risks in providing trading firms with direct market access.
99

In March 2012, FIA PTG and FIA European Principal Traders Association issued recommendations to assist trading firms in establishing internal procedures, processes and controls for the development, testing and deployment of trading software.
100

Finally, in September 2013, FIA released recommendations for increasing the usefulness of drop copy systems in exchange-traded markets.
101

97
FIA, “Market Access Risk Management Recommendations,” (Apr. 2010), available at
http://www.futuresindustry.org/downloads/Market_Access-6.pdf.

98
FIA PTG, “Recommendations for Risk Controls for Trading Firms,” (Nov. 2010), available at
http://www.futuresindustry.org/downloads/Trading_Best_Pratices.pdf.

99
Sutphen, Leslie, “Exchange Survey Finds Wide Range of Risk Controls in Place,” (Jan. 2011), available at
http://www.futuresindustry.org/downloads/RC-survey.pdf.

100
FIA PTG & EPTA, “Software Development and Change Management Recommendations,” (Mar. 14, 2012), available at
http://www.futuresindustry.org/downloads/Software_Change_Management.pdf.

101
FIA, “Drop Copy Recommendations,” (Sept. 2013), available at
http://www.futuresindustry.org/downloads/FIA-Drop_Copy(FINAL).pdf.

3. IOSCO Reports on Electronic Trading

IOSCO is an international body of securities regulators. IOSCO develops, implements and promotes adherence to internationally recognized standards for securities regulation. Its membership regulates more than 95% of the world's securities markets in more than 115 jurisdictions.
102

In October 2011, IOSCO released recommendations to promote the integrity and efficiency of markets in order to mitigate risks posed by the latest technological developments.
103

Among other things, IOSCO recommended that regulators ensure that trading venues have in place suitable trading control mechanisms such as trading halts, volatility interruptions, and limit-up/limit-down controls to deal with volatile market conditions, as well as trading systems that have the ability to adjust to changes in message traffic (including sudden increases).
104

In addition, IOSCO recommended that all order flow of trading participants, regardless of whether they access the market directly, be subject to appropriate controls, including automated pre-trade controls. IOSCO also recommended that regulators should identify any risks arising from currently unregulated direct participants of trading venues and take steps to address them.
105

102

See
IOSCO's public Web site, available at
https://www.iosco.org/about/?subsection=about_iosco.

103
Technical Committee of the IOSCO, “Regulatory Issues Raised by the Impact of Technological Changes on Market Integrity and Efficiency: Final Report,” (Oct. 2011), available at
http://www.iosco.org/library/pubdocs/pdf/IOSCOPD361.pdf.

104

See id.

105

See id.

More recently, in April 2015, IOSCO released a consultation report entitled “Mechanisms for Trading Venues to Effectively Manage Electronic Trading Risks and Plans for Business Continuity.”
106

The report compiles the results of a survey that IOSCO sent to trading venues across more than 30 different jurisdictions. Based on the information compiled, the report proposes best practices that should be considered by trading venues when developing and implementing risk mitigation mechanisms. These practices are intended to promote the integrity, resiliency and reliability of trading systems and business continuity plans. With respect to managing risks originating from market participant technology, the report explains that most trading venues have policies, procedures and tools to detect and address the operational risks associated with electronic trading. These tools

include, among others, pre-trade risk controls (such as price and volume controls or filters and order entry controls), the ability to block, suspend or disconnect a user (
e.g.,
a kill switch), measures to halt trading in the event of sudden price movements, and throttles that constrain the number or frequency of messages from any given participant.
107

IOSCO also explained that many trading venue participants use pre-trade risk controls such as order volume, price per security, credit, notional value of order, order value, capital, position checks, price deviation thresholds, and regulatory integrity checks.
108

Finally, IOSCO addressed direct market access by referring to a previous report it issued in 2010, called “Principles for Direct Electronic Access to Markets.” In that report, IOSCO recommended that intermediaries (including clearing firms) have adequate operational and technical capability to appropriately manage the risks posed by DEA.
109

106
IOSCO, “Mechanisms for Trading Venues to Effectively Manage Electronic Trading Risks and Plans for Business Continuity: Consultation Report,” (Apr. 2015) [hereinafter “IOSCO 2015 Consultation Report”], available at
http://www.iosco.org/library/pubdocs/pdf/IOSCOPD483.pdf.

107

See id.
at 20-21.

108

See id.

109

See id.
at 22-23. IOSCO uses the term DEA or “direct electronic access” to mean an arrangement where a client of an intermediary obtains access to the market through the intermediary's infrastructure or access without using the intermediary's systems.
See id.
at 20 n.56.

4. CFTC TAC Subcommittee

In 2011, the Pre-Trade Functionality Subcommittee (“TAC Subcommittee”) of the CFTC's TAC issued recommendations for pre-trade controls for trading firms, clearing firms and exchanges which use, or provide, direct market access.
110

The TAC Subcommittee recommended the following risk controls for trading firms: Quantity limits on individual orders; price collars; execution throttles; message throttles; and a kill switch that would cancel all existing orders and prevent the firm from placing new orders. The TAC Subcommittee further recommended that clearing firms trading on their own behalf should comply with those risk controls. In addition, clearing firms should confirm that their client firms are implementing such controls, approve the parameters used by the trading firm, and have access to the kill switch. Exchanges should implement, and require trading firms to use, pre-trade quantity limits on individual orders; intra-day position limits; price collars; and message throttles. The TAC Subcommittee also recommended that exchanges implement clear and consistent error trade policies, order cancellation policies that allow for automatic cancellation of orders on disconnect, and the ability for clearing firms to view their firm's orders and to cancel working orders.

110

See
CFTC TAC Recommendations,
supra
note 34.

5. FIX Risk Management Working Group

Additional organizations have released best practices documents, including FIX Protocol Ltd.'s (“FIX”) Americas Risk Management Working Group. FIX is a non-profit, industry standards association that owns, maintains and continuously develops the Financial Information eXchange (FIX) Protocol in response to market requirements. In 2012, FIX released risk control guidelines for algorithmic trading orders and direct market access orders.
111

FIX identified typical order scenarios that brokers attempt to detect, which include the following: An order for an exceedingly large quantity; an order that will adversely impact the market for a given security; an order with incomplete or conflicting instructions; an order that is potentially duplicative or unintentionally repeating; an order where adverse or favorable price moves impact the order while it is working; and an order that may be stale or may have been replaced by the client or a system.
112

FIX explained that the absence of appropriate risk controls can result in market dislocation, failure to settle/deliver, conflict between the client's intent and order execution, and trading the wrong product.
113

FIX provides a recommended matrix of risk controls, which includes maximum order quantity, average daily volume checks, price limit checks, favorable/adverse price move checks, position limits, credit checks, and stale, runaway, and duplicate order checks.
114

111

See
FPL Americas Risk Management Working Group, “Recommended Risk Control Guidelines,” (2012), available at
http://www.fixtradingcommunity.org/mod/file/view.php?file_guid=32127.

112

See id.
at 5. Other scenarios include an order where the symbology cannot be resolved to a single security and large accrued long or short positions that may result in settlement and/or delivery risk if the client cannot settle the trade.

113

See id.

114

See id.
at 22.

6. Senior Supervisors Group (SSG) Briefing Note

In April 2015, the Senior Supervisors Group (“SSG”), composed of the staff of banking and other financial regulatory agencies from ten countries and the European Union, issued an “Algorithmic Trading Briefing Note.”
115

The Note focused on how large financial institutions currently monitor and control for the risks associated with algorithmic trading during the trading day. The Note identified several risks that SSG believes are common to algorithmic trading across jurisdiction and asset class: (i) Systemic risk may be amplified; (ii) algorithmic trading desks may face a significant amount of risk intraday without transparency and robust controls; (iii) internal controls may not have kept pace with speed and market complexity; and (iv) without adequate controls, losses can accumulate and spread rapidly.
116

The Note provided a list of principles for supervisors to consider when evaluating controls over algorithmic trading at banks: (a) Controls must keep pace with technological complexity and trading speeds; (b) governance and management oversight can limit exposure to losses and improve transparency; (c) testing needs to be conducted during all phases of a trading product's lifespan, namely during development, rollout to production, and ongoing maintenance; and (d) when assessing control depth and suitability, management should ensure sufficient involvement of control functions (including compliance, technology, legal, and controllers), as well as business-unit management.
117

115

See
Senior Supervisors Group, “Algorithmic Trading Briefing Note,” (Apr. 2015) [hereinafter “SSG 2015 Note”], available at
http://www.newyorkfed.org/newsevents/news/banking/2015/SSG-algorithmic-trading-2015.pdf.
The SSG includes staff from the following organizations: Canadian Office of the Superintendent of Financial Institutions, the European Central Bank Banking Supervision, the French Prudential Control and Resolution Authority, the German Federal Financial Supervisory Authority, the Bank of Italy, the Japanese Financial Services Agency, the Netherlands Bank, the Bank of Spain, the Swiss Financial Market Supervisory Authority, the United Kingdom's Prudential Regulatory Authority, and, in the United States, the Office of the Comptroller of the Currency, the Securities and Exchange Commission, and the Federal Reserve.

116

See id.
at 2-3.

117

See id.
at 3-4.

7. Treasury Market Practices Group Best Practices

In June 2015, the Treasury Market Practices Group, a group sponsored by the Federal Reserve Bank of New York, and comprised of legal, compliance and business representatives from institutions related to U.S. Treasury market primary and secondary trading, released a white paper on Automated Trading
118

and an updated Best Practices document for trading in U.S. cash Treasury securities markets.
119

The

Best Practice updates, among other things, expanded the scope of recommended risk controls that address the risks of automated trading (automated trading, for purposes of the Best Practices document, means the subset of electronic trading that relies on computer algorithms for decision-making and execution of order submissions), including the documentation of internal policies and procedures, additional transparency in exchange or trading platform market data, error trade rules and exchange provided services, expanded design and testing environments at firms and exchanges, and updated risk controls that align with the speed of trading technology. The white paper notes that these updates were issued in a period when cash Treasury securities markets, like many other asset classes, have experienced a strong increase in automated trading on electronic platforms.

118

See
Treasury Market Practices Group, “Automatic Trading in Treasury Markets,” (June 2015), available at
http://www.newyorkfed.org/tmpg/TPMG_June%202015_automated%20trading_white%20paper.pdf.

119

See
Treasury Market Practices Group, “Best Practices for Treasury, Agency Debt, and Agency

Mortgage-Backed Securities Markets,” (June 2015), available at
http://www.newyorkfed.org/tmpg/TPMG_June%202015_Best%20Practices.pdf.

III. Recent Disruptive Events in Automated Trading Environments

The Concept Release discussed malfunctions in automated trading systems, in both derivatives and securities markets, that illustrate the technological and operational vulnerabilities inherent to automated trading environments.
120

As an example, the Flash Crash of May 2010 involved an automated trading system with a design flaw that impacted both the derivatives and securities markets. According to the CFTC/SEC joint report on the Flash Crash, an automated execution algorithm did not take price or time variables into account. Given the parameters of the program, the algorithm continued to send orders even as prices moved far beyond traditional daily ranges.
121

In another example, in 2012 a securities trading firm, Knight Capital Group, made a coding error in an automated equity router, and then incorrectly deployed new code in the same router.
122

Because of these coding errors, the firm's automated trading system inadvertently built up unintended positions in the equity market, eventually resulting in losses of more than $460 million for the firm.
123

The malfunction impacted the broader market, creating swings in the share prices of almost 150 companies; these price swings were high enough to trigger pauses in the trading of five stocks.
124

120

See
Concept Release, 78 FR at 56548-49.

121

See
U.S. Commodity Futures Trading Commission and U.S. Securities and Exchange Commission, “Findings Regarding the Market Events of May 6, 2010,” (September 30, 2010) [hereinafter, the “Flash Crash Report”], available at
http://www.cftc.gov/ucm/groups/public/@otherif/documents/ifdocs/staff-findings050610.pdf.

122

See
SEC Knight Capital Release,
supra
note 39.

123

See id.

124

See
Strasburg, Jenny and Bunge, Jacob, “Loss Swamps Trading Firm,” Wall St. J. (Aug. 2, 2012), available at
http://online.wsj.com/article/SB10000872396390443866404577564772083961412.html
and Valetkevitch, Caroline and Mikolajczak, Chuck, “Error by Knight Rips Through Stock Market,” Reuters (Aug. 1, 2012), available at
http://www.reuters.com/article/2012/08/01/us-usa-nyse-tradinghalts-idUSBRE8701BN20120801.

Foreign markets have also experienced disruptive events in recent years. For example, in May 2012 in Mexico, a “fat finger” error by a market participant resulted in the execution of 1.13 million shares (representing U.S. $3.78 billion).
125

In February 2015, there was a five minute delay in opening futures and options on the Eurex exchange in Germany because a market participant's system was transmitting duplicate orders.
126

In February 2014, trading in three-year Korean treasury bonds was halted for almost two hours at the Korea Exchange due to a system malfunction resulting from an improper order from a brokerage house.
127

On October 26, 2011, the Bombay Stock Exchange had to cancel all derivatives trading due to unusually high volumes and price volatility as a result of a flawed algorithm used by a member firm.
128

125

See
IOSCO 2015 Consultation Report,
supra
note 106 at 1 n.6.

126

See id.
and “Technical Failure Delays Eurex Trading in Futures, Options,” Bloomberg (Feb. 17, 2015), available at
http://www.bloomberg.com/news/articles/2015-02-17/eurex-futures-options-opening-delayed-after-technical-problem.

127

See
“Treasuries trading system disrupted,” Korea Times (Feb. 14, 2014), available at
https://www.koreatimes.co.kr/www/news/biz/2014/09/488_151619.html.

128

See
“Sebi probes Muhurat trading mishap on BSE,” Business Standard (Nov. 12, 2011), available at
http://www.business-standard.com/article/markets/sebi-probes-muhurat-trading-mishap-on-bse-111111200083_1.html.

Goldman Sachs was recently fined $7 million by the SEC for violating its Market Access Rule and causing a disruptive trading event.
129

On August 20, 2013, a configuration error in one of Goldman's options order routers erroneously sent thousands of limit orders to the options exchanges prior to the start of regular market trading.
130

By the time the creation of additional orders was disabled, and efforts to cancel unintended orders were taken, approximately 1.5 million unintended orders (representing 150 million underlying shares) had been executed on the market.
131

The existing risk management controls and supervisory procedures in place at Goldman failed to stop the erroneous orders, and human error and failure to follow best practices exacerbated the errors.
132

While some erroneous orders were able to be cancelled, Goldman's loss ultimately totaled $38 million.
133

129

See
In re Goldman, Sachs & Co., No. 3-16665 (SEC June 30, 2014) (order instituting administrative and cease-and-desist proceedings).

130

Id.
at 2.

131

Id.

132

Id.
at 3.

133

Id.
at 2.

Disruptive events illustrate the importance of effective risk controls. The risk controls contemplated in Regulation AT are intended to limit the extent of market disruption caused by ATSs or trading platform malfunctions. For example, a pre-trade risk control such as a message throttle will prevent submission of orders that exceed a predetermined frequency per unit time. Such a control could be operated by the market participant generating orders, the clearing firm guaranteeing its trades, or the trading platform on which orders would be executed, and would limit the impact of an algorithmic trading system not operating as intended. As another example, monitoring and supervision standards for algorithmic trading may help ensure that human supervisors intervene quickly when automated systems experience unexpected or degraded performance, and that supervision staff have the both the authority and knowledge to take appropriate steps in this scenario.

IV. Overview of Regulation AT

A. Concept Release/Regulation AT Terminology

The Concept Release used the term “
automated
trading system” (abbreviated “ATS”) to refer to the algorithms used to automate the generation and execution of a trading strategy.
134

In discussing comments to the Concept Release, the Commission will continue to use the term automated trading system. However, for greater precision, the proposed rules and preamble for Regulation AT instead refer to “
algorithmic
trading system” (also abbreviated “ATS”). This change is intended only as a change in in nomenclature. ATSs as described herein should not be confused with alternative trading systems in equities markets.

134
Concept Release, 78 FR 56542, 56544.

B. Commenter Preference for Principles-Based Regulations

As an initial matter, the Commission notes a preference expressed in comments to the Concept Release for principles-based, as opposed to prescriptive, regulations. Fifteen

commenters advocated a limited or “principles-based” approach to any regulation arising from the Concept Release.
135

Commenters indicated that prescriptive requirements will become obsolete, stifle innovation, discourage self-reporting of technological failures, may not account for the unique characteristics of market participants, and would result in participants designing around such measures.
136

135
The Futures Industry Association (“FIA”) Comment Letter (Dec. 11, 2013) at 2, 12; CME Group (“CME”) Comment Letter (Dec. 11, 2013) at 3, 41-42; Gelber Group, LLC (“Gelber”) Comment Letter (Dec. 9, 2013) at 1-2; KCG Holdings, Inc. (“KCG”) Comment Letter (Dec. 11, 2013) at 3; The Alternative Investment Management Association (“AIMA”) Comment Letter (Dec. 11, 2013) at 1; The Minneapolis Grain Exchange (“MGEX”) Comment Letter (Dec. 11, 2013) at 1; CBOE Futures Exchange, LLC (“CFE”) Comment Letter (Dec. 11, 2013) at 2; Managed Funds Association (“MFA”) Comment Letter (Dec. 11, 2013) at 2; Holly Bell (Bell”) Comment Letter (Dec. 11, 2013) at 3; Virtu Financial LLC (“VFL”) Comment Letter (Jan. 10, 2014) at 2-3; Chris Barnard Comment Letter (Jan. 29, 2014) at 2; Susquehanna International Group (“SIG”) Comment Letter at 2; IntercontinentalExchange Group, Inc. (“ICE”) Comment Letter (Feb. 14, 2014) at 1-2; 3Red Trading LLC (“3Red”) Comment Letter (Feb. 14, 2014) at 2; OneChicago, LLC (“OneChicago”) Comment Letter (Feb. 14, 2015) at 5.

136
FIA at 2, 12; CME at 3-4, 7; Gelber at 1-2; Tellefsen and Company, L.L.C. (“TCL”) Comment Letter (Oct. 31, 2013) at 5, 18; AIMA at 1, 2; CFE at 2; VFL at 3; Bell at 3.

More specifically, FIA
137

and CME Group, Inc. (“CME”) suggested that the best way to achieve standardization of risk controls is through implementing “best practices” developed through working groups of DCMs, FCMs, and other market participants.
138

Similarly, IntercontinentalExchange, Inc. (“ICE”) indicated that “exchanges are able to better implement and update risk controls on a market-by-market basis than through a Commission rulemaking,” and should be allowed flexibility in designing exchange risk controls.
139

Susquehanna International Group (“SIG”) stated that the Commission should “allow the exchanges to work with firms on tailoring the rules for implementation in ways that best consider the technical intricacies between firms and exchanges.”
140

Virtu Financial LLC (“VFL”) suggested that “mandating risk controls and supervisory systems that are `reasonably designed' or `provide reasonable assurance' of protection would allow participants to tailor these controls to the specific risks associated with their business.”
141

137
The Commission notes that six entities submitted letters in support of FIA's comment letter: RGM Advisors, LLC, Allston Trading LLC, Geneva Trading USA, LLC, Tibra Trading America LLC, DRW Trading Group and IMC Financial Markets.

138
FIA at 63; CME at 41.

139
ICE at 1-2.

140
SIG at 2.

141
VFL at 3.

In addition, five commenters indicated that the Commission already has robust regulations in place to address the risks of automated trading.
142

Such comments cited the DCM and SEF Core Principles;
143

Commission regulations 1.73, 23.609, 38.255, and 38.607;
144

and CEA and Commission market manipulation and disruptive trading practices rules.
145

142
CME at 3; FIA at 5; MFA at 6; Gelber at 2, 5, 20; Bell at 2, 4.

143
Gelber at 21; CFE at 1; MFA at 6.

144
MFA at 4; CFE at 1.

145
Gelber at 2, 5, 20; CFE at 3; CME at 3; MFA at 6; Bell at 2.

In contrast to a limited or principles-based approach to regulation, several commenters supported a more prescriptive approach to a rulemaking addressing the risks of automated trading.
146

These commenters include the Institute for Agriculture and Trade Policy (“IATP”), Better Markets, and Americans for Financial Reform (“AFR”). For example, IATP stated that unless the Commission receives documentation that the risk controls of firms and exchanges are consistent and effective, the Commission should assume that regulatory standardization will be beneficial for each risk control and at each phase of the trade lifecycle.
147

In addition, several academic commenters discussed concerns with automated, high speed trading and advocated specific changes to the trade matching or order submission process to increase market liquidity and efficiency.
148

146
The Institute for Agriculture and Trade Policy (“IATP”) Comment Letter (Dec. 11, 2013) at 4; Better Markets Inc. (“Better Markets”) Comment Letter (Dec. 11, 2013) at 1; Americans for Financial Reform (“AFR”) Comment Letter (Dec. 11, 2013) at 6.

147
IATP at 4.

148
Eric Budish et al. Comment Letter (Feb. 14, 2014) at 1; Brian F. Mannix Comment Letter (Dec.10, 2013) at 2; Elaine Wah et al. Comment Letter (Dec. 11, 2013) at 2.

As discussed below, consistent with comments received, the Commission has taken a balanced approach to the regulations it believes are necessary to manage the risks of algorithmic trading. For example, the Commission proposes a principles-based approach to its risk controls requirements, in that it would require particular controls but allow the relevant entity—a trading firm, clearing member FCM, or DCM—discretion in the design of such control and the parameters that would be used.

C. Multi-Layered Approach to Pre-Trade Risk Controls and Other Measures

In response to the Commission's questions in the Concept Release about the appropriate location for risk controls and other measures, commenters generally supported a multi-layered approach to risk controls, with each level—trading firm, clearing firm, and exchange—implementing risk controls that are adjusted depending on circumstance.
149

149
CME at 8-9; FIA at 61; Federal Reserve Bank of Chicago (“Chicago Fed”) Comment Letter (Dec. 11, 2013) at 2; AIMA at 7; KCG at 2; VFL at 2.

For example, FIA commented that “[i]ntroducing redundant risk controls at more than one focal point in the trading lifecycle may increase the integrity of the marketplace when careful consideration is given to their differences in roles, implementations and configurations.”
150

However, FIA also stated that “we caution against a mandated proliferation of redundant risk controls because the existence of similar but not identically implemented risk controls may do more harm than good. Each new implementation of a control will increase complexity and may cause misunderstanding between traders and risk managers as a consequence of conflicting risk limits.”
151

As an example of a control that may be appropriately implemented at multiple levels, FIA stated that maximum order size limits may be implemented at both market participant and FCM levels without redundancy because they reflect the different responsibilities of each participant.
152

FIA further explained that if an FCM has implemented customer-specific controls within its infrastructure, it would be redundant to use the same controls at the DCM level, though as an additional protection, it is permissible to set higher limits at the DCM that apply across all customers.
153

150
FIA at 61.

151

See id.

152
FIA at 62.

153

See id.

CME cited the TAC Subcommittee's “Recommendations on Pre-Trade Practices for Trading Firms, Clearing Firms and Exchanges involved in Direct Market Access,” and commented that “each level of the `electronic trading `supply chain' (trading firms, clearing firms, and exchanges) must share in the effort to preserve market integrity through the implementation of effective risk controls, no matter if that participant has direct market access or is routing to the exchange via its clearing member firm”.
154

Specifically

with respect to kill switch functionality, CME indicated that kill switch functionality deployed at multiple levels should not be considered redundant.
155

CME further suggested that while multi-layered kill switch functionality is not necessary for effective risk control, it is nevertheless beneficial as it adds additional measures of protection.
156

CME made a general point that registrants should establish controls appropriate to the nature of their business that are reasonably designed to control access, effectively monitor trading, and prevent errors as well as other inappropriate activity.
157

CME indicated that, regardless of whether orders are entered manually through an electronic system or entered through an automated trading system, such principles are equally important, because the method of order entry does not lessen the impact of a particular order on the market.
158

154
CME at 7-8.

155
CME at 22.

156

See id.

157
CME at 43-44.

158

See id.

Other commenters supported a multi-layered approach to risk controls. AIMA indicated that risk controls should be “broadly similar” and applied at the trading firm, clearing firm, and exchange levels.
159

KCG stated that “risk management is most effective when it is multi-layered and overlapping.”
160

VFL stated that a “multilayered system of risk controls is a key ingredient to protect the market from disruptive events.”
161

159
AIMA at 7.

160
KCG at 2.

161
VFL at 2.

The Commission agrees with the comments above that it should adopt a multi-layered approach to regulations intended to mitigate the risks of automated trading. As explained below, the Commission proposes to impose requirements at multiple stages of the lifecycle of an order. The Commission acknowledges FIA's comment that the different role of entities at various stages in the trade lifecycle must be carefully evaluated. While Regulation AT requires the same types of pre-trade and other risk controls to be implemented by different entities, the Commission notes that the proposed regulations allow for discretion in the appropriate design and parameters of each risk control. Accordingly, a trading firm, clearing member FCM, and DCM may each choose to design and calibrate the same control in different ways, depending on how the control is used by each entity to manage risks.

The Commission notes that ESMA's 2015 Final Draft Regulatory Standards require pre-trade risk controls at both investment firms and trading venues.
162

ESMA acknowledged commenter disagreement with such redundancy and stated, “ESMA believes that at least two lines of defence are appropriate in this complex business and thus continues to require the pre-trade risk controls conducted by both investment firms and trading venues.”
163

ESMA's regulatory standards further provide that where a client is granted market access either through an intermediary's systems, or directly without using the intermediary's systems, the direct electronic access provider must apply the required pre-trade risk controls.
164

Regulation AT requires pre-trade and other risk controls at both the AT Person and clearing member FCM level (as well as the DCM level) based on its understanding that the risks—and the resulting calibration levels of the controls—may be different given those entities' distinct priorities and understanding of the risks to themselves and their customers.

162
ESMA September 2015 Final Draft Standards Report,
supra
note 80 at 201.

163

See id.

164
ESMA September 2015 Final Draft Standards Report Annex 1,
supra
note 80 at 218.

Below is a summary of each element of Regulation AT. For each element, the Commission addresses relevant Concept Release comments, summarizes the proposed regulation, and asks questions concerning the proposed regulation.

D. Codification of Defined Terms Used Throughout Regulation AT

1. “Algorithmic Trading”—§ 1.3(zzzz)

a. Concept Release Comments

The Concept Release requested comment concerning whether the Commission should define ATS or algorithm for purposes of any ATS identification system. Commenters disagreed on whether the Commission should adopt a definition of “ATS.” FIA and CME opposed a regulatory definition, arguing that industry already has a definition of automated trading system.
165

FIA and CME indicated that the definition of ATS is self-evident and has been in use for a long time, and that ATS, or automated orders, are orders that are generated and/or routed without human intervention. This includes orders generated by a computer system as well as orders that are routed using functionality that manages order submission by automated means (
i.e.,
execution algorithms).
166

Another commenter, Gelber Gro

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3A2015-30533. Public record. Not legal advice.