The CFTC announced its new automated trading rules, marking a major shift on how automated trading will be developed and implemented in the coming years.
The Commodity Futures Trading Commission (CFTC) took another step toward regulating automated trading systems used in the futures markets with sweeping rules changes that will affect a large number of traders in the futures markets.
On Tuesday, the agency approved a proposed rule called Regulation Automated Trading, or Regulation AT, which involves market participants using automated trading such as: clearing member futures commission merchants (FCMs) as well as anyone registered as an FCM, floor broker, swap dealer, major swap participant, commodity pool operator, commodity trading advisor or introducing broker engaged in algorithmic trading. (For definitions of participants and summary of the rule, go to MarketsReformWiki).
The rule also addresses risk controls that need to be in place by market users and exchanges, as well as other safeguards. Among them are: maximum order message and order size parameters, standards for development, testing and monitoring of ATSs and new reporting requirements regarding risk controls and other ATS procedures that will be reviewed by exchanges.
One of the major components of the CFTC rule proposal is a registration requirement for proprietary traders that use algorithmic trading through direct market access on exchanges. The agency also called for all automated trading participants to become members of a so-called “registered futures association,” or RFA. The CFTC added that the RFAs would be required to “consider membership rules addressing algorithmic trading for each category of member in the RFA. Taken together, these provisions would allow RFAs to supplement elements of Regulation AT in response to future industry developments.”
The registration component of the new rules caught some industry professionals off guard. In the meetings leading up to the CFTC meeting, FIA officials and members repeatedly emphasized that registration, which is now part of similar European rules, was unnecessary. To industry participants, current identification procedures are already in place and can identify any trader via firm registrations and ID policies. The registrations are expected to affect around 100 firms.
The Futures Industry Association’s Principal Traders Group released a statement after the meeting stating its hope that the CFTC will follow industry best practices and a data driven approach going forward. (The FIA has no immediate comment, as it was still looking over the proposal’s details). The PTG’s chairman, Rob Creamer, who is also president and CEO of Geneva Trading, issued the concern over the CFTC’s access to proprietary algorithms developed by trading firms.
“We’ll pay particularly close attention to the issue of access to proprietary algorithms and the intellectual property concerns raised by (CFTC) Commissioner Giancarlo,” the statement said.
Giancarlo said in his statement he was “particularly concerned that Regulation AT requires that each market participant keep a source code repository for algorithms and make it available for inspection to any representative of the Commission or the U.S. Department of Justice for any reason.”
“I am unaware of any other industry where the federal government has such easy access to a firm’s intellectual property and future business strategies,” Giancarlo said in his statement. “Other than possibly in the area of national defense and security, I question whether the federal government has similarly unfettered access to the future business strategy of any American industrial sector…I am unclear why either the Commission or the U.S. Department of Justice needs access to source code information without a subpoena, especially the Justice Department, whose only use for such information would be in criminal proceedings. Does today’s rule proposal presume that the use of automated trading technology makes a trading firm more likely to engage in criminal behavior than a manual trading operation?”
The CFTC has said that there may not be any need for firms to actually give up their source code for algorithms, but rather inspect them in certain cases.
___
**Editor’s Note: This section below and supporting interviews were conducted prior to the CFTC meeting on November 24. However, much of the information from the rule proposal is addressed in these interviews. JLN will update this information as needed.
It is apparent the CFTC drew some of its guidance from a key Futures Industry Association industry report called The Guide to the Development and Operation of Automated Trading Systems. The report was developed over the course of a year by a committee of the FIA Market Technology Division, along with the FIA Principal Traders Group and the FIA European Traders Association (FIA EPTA).
Leslie Sutphen, president at Financial Markets Consulting LLC, and Greg Wood, a director at Deutsche Bank, who co-chair the FIA Market Technology Division, Automated Trading Committee, led the team that published the 48-page report in March. It addresses key components of automated trading systems such as pre-trade risk controls, post-trade analysis, development and design, security, operations and documentation policies.
Leslie Sutphen, president at Financial Markets Consulting LLC, speaking prior to the CFTC rule proposal on November 24, talks about the guidelines for automated trading systems, found in The Guide to the Development and Operation of Automated Trading Systems.
While some of this may seem dry to the average industry participant, this is one of the most anticipated rulings from the CFTC in months, if not years. Ultimately, the CFTC and Securities and Exchange Commission are each trying to prevent major market disruptions from automated trading systems that have shaken investor confidence in recent years.
“I think what everybody worries about is market disruption,” said Leslie Sutphen, one of the authors of the FIA paper. “What if the automated trading algorithm doesn’t do what it is supposed to do? What if we’re not even aware that it’s doing something that we don’t know. Most of this paper is around taking steps to prevent the algorithm from doing something you didn’t intend.”
The FIA Market Access Working Group focused on two major elements associated with “disruption.” One is the issue of the algorithm running wild. The other is intentional manipulation of the market. About one-third of the paper deals with pre-trade risks and methods.
“We spent a lot of time on that because if you can keep things from causing problems before it hits the marketplace, obviously that’s better for everybody,” said Keith Fishe, managing partner at TradeForecaster Global Markets, who also worked on the FIA study and spoke with JLN prior to the CFTC rule proposal on November 24. “We spent a lot of time exploring important practices we think groups should try to incorporate in their operational procedures and their technology.”
Some of the FIA paper focuses on well-known controls such as order-size limits, or position limits.
“There’s no reason why an order should not have some sort of limit quantity, or at least have a double check if it exceeds a quantity before it ever leaves the trading participant,” Fishe said. “But there are also quantity limits that could be done at the exchange.”
Fishe added that the group looked at messaging as well. In a world where thousands of messages are often sent per second to an exchange by one participant, just how much is too much, or how can a firm adjust so as not to inadvertently manipulate the market?
“You don’t want too many cancellations versus orders, however there are always exceptions,” Fishe said. “Sometimes the market is moving because there is a related exchange or a related marketplace that is moving dramatically. Nonetheless, you kinda want guidelines.”
There are also descriptions in the FIA report of dynamic price collars, market price validation checks to make sure the market data is correct, self match prevention and kill switches, all of which would stop an erroneous or disruptive order from hitting the exchange.
“We think having dynamic price collars, particularly at the exchange level, are really important because that is what helped the futures market during the Flash Crash in May of 2010,” Sutphen said. “The CME actually had a reality check in place about that pricing so it actually paused the market when things looked out of whack.”
Other market disruptions such as Knight Capital Group’s meltdown in 2012, which cost the firm $440 million in a single day, and the so-called flash crash on October 14, 2014 in the bond markets, have raised concerns in Congress and among regulators. Sutphen agrees that some improvements there are required.
“We think the Treasury market could mature a bit and adopt some of the controls that are available on the futures and equities side as well,” Sutphen said. “We’re hopeful this standard applies not just to futures but to other asset classes that trade electronically.”
The CFTC’s new Regulation Automated Trading rule proposal calls for more rules to prevent self- trading. Mike Forst, director of product management at Trading Technologies talks about the issue and how his firm is trying to address it.
The CFTC’s new Regulation Automated Trading rule proposal aims to prevent self- trading. John Lothian News spoke with Lisa Dunsky, attorney Sidley Austin, prior to the CFTC’s rule approval to discuss some of the issues surrounding self- trading, as well as how spoofing may factor in.
Lisa Dunsky, an attorney with Sidley Austin, told JLN that participants will now be forced to use exchange or third-party technology to prevent self- matching.
“With the vast majority of trading occurring on platforms such as Globex, with many more proprietary trading firms where you have multiple traders or multiple trading systems trading in the same products, you have many situations where people have no intent at all to be on the same sides of the same trade, but it happens inadvertently and it would still be considered a wash trade,” Dunsky said in an interview that preceded the CFTC rule announcement.
One thing that the group is not recommending or advocating is a complex pre- check of every order and order type before it is sent to an exchange. Sutphen said that is not practical because of the diversity of asset classes and markets.
“This is more about preventing something really disruptive – not credit events necessarily, not runaway algos that would hit every exchange in the world, but things that are easy to implement and work 99 percent of the time,” Sutphen said.
Robo Data: Automated Trading On CME
The chart illustrates the percentage of trades that are done on CME via automated trading systems, or those without any human interaction versus the percentage of electronic trades entered manually, also done as point-and-click or other manual methodologies. The CFTC study, called Automated Trading in Futures Markets, published in March 2015, examined trade data on CME from 1.5 billion trades on 805 futures products and close to 362,000 individual accounts, from November 12, 2012 to October 31, 2014. Source: CFTC
Another area of focus is on the post-trade side of the equation. While that seems a little out of step with preventing disruptive trades in the first place, Sutphen and Fishe said that the group focused on it because it allows participants to take an even deeper look into what is happening. The paper calls for things such as “drop copy feeds” from the exchange which go back to the firms, so they can reconcile trades with cleared and settled trades.
“If you get what the exchange knows you did, compared to what you think you did, and you can reconcile that in near real time, then you can find problems much much quicker,” Fishe said.
The trick to tackling ATS issues is to strike a balance that gives clarity on guidelines without being too prescriptive and harming innovation. Fishe said that most firms use so-called “agile” software development methods, and this could be one area where regulators and participants more closely align themselves.
“In our industry surveys we found that most firms have adopted what is called agile methods, which are more modern ways of developing software,” Fishe said. “There is no one way or prescribed way for people to develop software, but there are some good, important practices,” Fishe said. “You want to be able to know what went into that software. So there are some important suggestions we made in our guidelines for regulators to think about.”
Ultimately, the goal for industry participants is to get on the same page as regulators so the guidelines work. What FIA and others in the industry hope for are rules that help provide guidance but do not hamper or harm trading approaches that are safe.
Original article can be found here.