Remarks of Commissioner Mark P. Wetjen before the Global Derivatives Trading & Risk Management Conference, Amsterdam, The Netherlands
Applying the Insights of Joseph de la Vega’s Confusion de Confusiones to Today’s Derivatives Markets -- a Case for More Transparency
May 21, 2015
Good morning and thank you for having me here to speak at the Global Derivatives Trading & Risk Management Conference. I am honored to be with you in Amsterdam.
Let me start my remarks with a few words about this beautiful city. The city of Amsterdam has a rich history in financial markets. Widely credited with having the first “modern” stock exchange for shares in the East and West India Companies during the 17th century, as well as markets for related derivatives, today’s Amsterdam is an ideal place to host a conference focusing on the trading of derivatives.
The setting of the conference also should inspire us all to bring the proper perspective to a discussion about modern-day trading in derivatives. Indeed, many of the same key dynamics exist today, four centuries later, in financial markets – including those for derivatives – that existed in the 1600s.
Some of you here today surely know of the book, “Confusion of Confusions,” by the author Joseph de la Vega, a 17th Century poet, author, and investor. This book offered important observations and some timeless quotes about trading on the Amsterdam stock exchange, some of which prove to be apt when referencing today’s markets as well.
To begin with, de la Vega categorized the “three classes which participate in the Exchange. The first is constituted of the large capitalists or the princes of the Exchange, the second of the merchants, and the third of the professional speculators.” It would take little creativity to find an appropriate place for today’s derivatives market participants in one of these same three categories.
In describing the stock exchange at that time, de la Vega wrote that the exchange business “is comparable to a game. Some of the players behave like princes and combine strength with tenderness and amiability with intelligence, but there are some participants who lost their reputation and others who lack devotion to their business even before the play begins.”
This observation surely resonates with many here when thinking about today’s markets, and unfortunately so regarding those who lose their reputations. To be sure, there has been great interest in recent arrests related to charges brought by the CFTC against traders accused of violating CFTC rules. I will address this more in a moment.
In the fictitious dialogue of his book, de la Vega’s shareholder character advised that, “What really matters is an awareness of how greed and fear can drive rational people to behave in strange ways when they gather in the marketplace.” In the context of the book, these were words of caution to those participating in the “game” of trading on an exchange.
These words are also useful for policymakers today to remember when crafting rules to govern marketplaces. We must take care to understand human behavior and the incentives that drive it in the context of a financial market place in order to devise the best market structure.
These behaviors and incentives are older than de la Vega’s book and likely will not be eradicated from a financial market, nor should that be the goal – that would be a fool’s errand.
Likewise, interest in using markets to manage and transfer risk is also older than de la Vega’s book, and continues to offer a valuable function to commercial enterprises.
What has changed, obviously, is the technology used to operate and trade in financial markets. In de la Vega’s time, he wrote that “bulls spread a thousand rumors about the stocks, of which one would be enough to force up the prices.” These false rumors were spread through Amsterdam coffeehouses frequented by traders – a rather slow and rudimentary semblance of manipulation. Today, there is software a trader could use to more readily deploy a manipulative device.
With de la Vega’s poignant and timeless observations as a backdrop, I would like to focus my remarks today on the automation of derivatives markets, and how this development has given rise to new policy considerations.
But this development otherwise should not be thwarted by policy makers, and certainly should not be misunderstood. Indeed, technology has brought numerous benefits, including facilitating the entry of new participants that can and do serve as liquidity providers in today’s derivatives markets, increasing trading volume, and narrowing bid-ask spreads. This is a development we should encourage given other developments affecting traditional liquidity providers, so long as we take care to address the attendant risks of an increasingly automated market.
To accomplish the latter, the Commission and other market regulators must better understand the risks posed by automation through better access to market data, including order-message data. Additionally, better access to data must include enhanced coordination among regulators to share data about distinct, but inter-connected, markets. This is especially true for regulators that oversee derivatives markets and those who supervise related cash markets.
Moreover, the CFTC should continue to pursue a parallel track to incentivize new-entrant liquidity providers to the swaps marketplaces, or swap execution facilities (SEFs), that it oversees.
I will cover each of these recommendations in more detail, and explain how recent events commend action.
October 15 Event in U.S. Treasury Markets
The $12.5 trillion Treasury market is one of the most important and liquid financial markets in the world, and is used by U.S. and global retail investors, hedgers, and financial institutions, as well as foreign central banks. By some accounts, over the past 10-15 years, high frequency or algorithmic trading in the cash treasury market has grown rapidly, increasing to over 50% of all trading activity.
On the morning of October 15, 2014, the Treasury market experienced an unusually high level of volatility, with the benchmark 10-year Treasury yield plunging over 30 basis points between the market open and close. According to a recent Federal Reserve staff speech, a sharp decline in yields started around 8:30 a.m. after an announcement that U.S. retail sales were below expectation. This was followed by an unusual round-trip in yields just after 9:33 a.m. that lasted approximately 10 minutes.
Treasury futures contracts also experienced high volume and price movements that morning. According to publicly available data, prices spiked between 9:33 and 9:45 a.m., followed quickly by prices returning to 9:33 a.m. levels. High frequency and algorithmic trading activity was active that morning in the futures markets, although that type of activity is frequent in the futures markets on a normal day.
Historically, volatility in intraday changes in the Treasury market was associated with important economic events or major policy announcements. However, what caused the volatility on this day is still a subject of discussion. Factors that have been mentioned, in addition to the U.S. retail sales release, include general worries about the global economy, dealers choosing to not provide liquidity during the period of volatility, and investors liquidating a large amount of short positions in shorter-term interest rate futures in the weeks surrounding October 15.
The CFTC does not directly regulate trading venues for Treasury securities, but it does oversee exchange trading of Treasury futures. As this audience is aware, futures exchanges can be important and useful price-discovery markets for the related cash market. Consequently, to fully understand one market, a regulator must understand the other.
Among other topics, this event has informed a larger discussion about the presence and role of high frequency or algorithmic trading in the Treasury futures markets. With respect to the cash markets, and as highlighted by the Treasury Market Practices Group white paper on Automated Trading in Treasury Markets, trading in the most liquid, on-the-run Treasury securities in the inter-dealer market has witnessed an increasing presence of automated trading, and high-frequency trading in particular. Given the relationship between the two markets, I believe the event also has revealed the need for readily accessible and available cash Treasury data.
The CFTC’s Complaint Against Sarao
On April 21, 2015, the CFTC unsealed its civil complaint against Navinder Sarao (Sarao) and his company, Nav Sarao Futures Limited PLC. The CFTC alleges that Sarao and his company manipulated and attempted to manipulate the intra-day prices of the E-mini S&P 500 futures contract (E-minis), and engaged in disruptive trading known as “spoofing” on the exchange (i.e., bidding or offering with the intent to cancel the bid or offer before execution). Also that day, the U.S. Department of Justice unsealed a federal criminal complaint against Sarao alleging that he committed wire fraud, manipulation, attempted manipulation, and spoofing.
The complaints allege that Sarao used automated-trading software on numerous occasions between April 2010 and April 2014 to spoof the market and manipulate the intra-day price for the E-minis. By placing multiple, large-volume orders on the futures market, Sarao is alleged to have created the false appearance of substantial supply in order to induce other market participants to react to this market information, creating downward pressure on E-mini prices. Sarao then modified and cancelled those orders before execution.
Sarao’s activity in the E-mini contract is alleged to also have taken place during the May 6, 2010 “Flash Crash.” The CFTC’s complaint alleges that Sarao contributed to order imbalances in the futures market, and that those order imbalances affected the U.S. stock market.
Although the trading activity began in 2010, the CFTC then, and now, does not have the ability to monitor real-time message and trade data. As a result, it took years for a whistleblower to uncover the activity, and then years thereafter for the CFTC and DOJ to put together the case.
Interconnection and Interplay Across Markets Complicates Surveillance
The October 15 incident shows the interconnection between the derivatives and cash markets. For instance, spread trading, where one seeks out arbitrage opportunities to capitalize on pricing discrepancies between Treasury securities and Treasury futures, is a frequent practice by firms relying on automated systems and trading in both the futures and Treasury markets. The Sarao incident illustrates the interconnection between the derivatives and equities markets.
These incidents also show the prevalence of high frequency and algorithmic trading. From a regulatory perspective, when there is an unusual event, the immediate concern should not be whether high frequency or algorithmic trading was involved, but whether manipulation or disruptive trade practices were involved.
Automated-trading firms do not necessarily react differently from how traders have reacted historically; they just react faster.
For instance, floor brokers behaved similarly when futures trading was conducted in the trading pits. In response to historical market events, and in the face of rapid price movements for unclear reasons, floor brokers hesitated and slowed their responses or pulled back rather than reacting aggressively.
This behavior is not much different from automated trading firms unplugging their algorithms. The difference, of course, is that the impacts of accelerated price movements in one market can be transferred more quickly to another market.
As we study these incidents, it is clear that there are some regulatory gaps that need to be addressed. Historically, trading practices and regulations have evolved by product, by trading platform, by regulator, and/or by country. But given the interconnection of our now global markets, we need to be thinking holistically about the interconnections between markets and the speed with which markets move, as well as the impacts of cross-market trading strategies and firms.
Markets are becoming increasingly electronic, and high frequency and algorithmic trading is increasing, if not dominant, in some markets. But the markets are subject to different or inconsistent levels of regulatory oversight, even where there are participants running substantially similar algorithms across multiple markets.
Partnering with Trading Venues Would Improve the CFTC’s Surveillance Capabilities
To properly surveil the derivatives marketplace, the CFTC needs to have an accurate picture of market participant activity. We do not have such a picture today because we do not have regular access to order-book and message data at exchanges, as illustrated by the Sarao case.
The CFTC cannot retain the public’s trust in its ability to perform the agency’s congressionally mandated mission without the proper view of what’s occurring in the marketplace.
Surveillance with executed transaction data alone is not enough, especially considering the changes in market structure and technological innovation. In order to detect other types of manipulation like spoofing, layering, and flipping, plus new types of gaming strategies, receiving order book and message data is necessary.
Additionally, only with regular order book and message data can the commission develop a comprehensive and thorough understanding of how markets evolve and the types of market participants in our markets, including what strategies they employ, what other markets they participate in, and where they hedge versus take profits. It will help the commission understand, for instance, whether high frequency and algorithmic traders are engaging in market making strategies and providing liquidity, or engaging in alpha-seeking strategies that consume liquidity.
It is only with regular order book and message data that the CFTC can build baseline analytics abilities with respect to human capital and technology, and improve our understanding and detection of manipulative and disruptive trading strategies, such as spoofing.
However, getting the data is only part of surveillance; we also need to have the right resources to make it useful. The CFTC has human-capital capability, but staff looks at order book and message data ad hoc, through specials calls or during an enforcement investigation. There are staff across divisions that have been and are capable of analyzing such data, but a truly effective surveillance program that successfully deters illegal behavior, as well as instills trust in the marketplace, will only materialize with regular consumption of order-book data by a sufficiently resourced and staffed surveillance program.
The commission faces one key obstacle standing in the way of such a surveillance program: insufficient resources. The commission’s needs relate to both technology as well as human-capital constraints.
We cannot wait for additional funding from Congress to establish a more effective surveillance program. I propose that, as a start, the CFTC consider a partnership with the exchanges it oversees to create a joint surveillance function that analyzes order-book and message data.
Such a joint effort could be designed to rely on the exchanges providing their technological tools, analytics software, as well as access to the data itself. The joint effort should also be carefully designed to address any cybersecurity or other concerns. As self-regulatory organizations (SROs), the exchanges and SEFs are the first line of monitoring and enforcement of their trading rules and the CFTC’s prohibitions against manipulative and disruptive trading practices, and have available or proprietary software tools to pursue that mission. Accordingly, such a joint effort could be designed to rely on the exchanges providing their technological tools, analytics software, as well as access to the data itself. Indeed, there is precedent for information sharing between the SROs and the government.
However, access to exchange and SEF data in this context should be viewed as distinct from turning over the data to the agency – at this juncture, the CFTC does not have the requisite technology to effectively collect, aggregate, and analyze the information. The benefits and lower costs of cloud data storage must be considered and potentially harnessed here as well.
But, the CFTC could devote its human capital to this joint effort, with the goal of expediting the training of CFTC staff and building up the agency’s surveillance capabilities over time. Ideally, additional staff would be hired to help with the effort, but in the near term, existing staff could be re-assigned or devoted to the mission.
Perhaps as importantly, the exchanges and SEFs also would benefit from providing access to order book and message data in order to more likely achieve a common understanding between the CFTC and the SROs of what should be viewed as illegal or manipulative activity. Today, given the relative asymmetry in access to the data between the CFTC and SROs, all too often, different conclusions can be drawn about whether conduct should be considered illegal.
This serves no one. A more common view would streamline both the surveillance and enforcement functions of the CFTC as well as the SRO, which in turn will help maintain the public’s trust in how our markets are run and supervised.
I should note that the Securities and Exchange Commission’s implementation of the MIDAS reporting framework could be very instructive and serve as a model for the effort I describe today, insofar as it leverages cloud storage and partnership with private entities.
Such a joint effort certainly raises a number of legal issues that would need to be considered carefully. But again, the principle aim of a partnership with exchanges would be to satisfy the needs of the agency to pursue its surveillance function and build up its baseline analytics capabilities, improve the common understanding of the commission and the SROs of what constitutes illegal activity, and ensure the trust of both market participants as well as the public in the integrity of the derivatives markets.
CFTC Would Benefit from Greater Transparency of Related Cash Markets and Enhanced Collaboration
Financial regulators have a responsibility and should work together to prevent gaps in authority or oversight from allowing the public to lose trust in our systemically important markets. Congress and other regulators should consider whether the October 15 event and the increasing speed and impact of volatility across markets necessitate additional transparency of the Treasury market.
The Treasury market has grown to look and feel more like the equity and futures markets, but there is no routine post-trade reporting framework to allow regulators to verify this, or otherwise view the market on a regular basis. Venues for trading Treasury securities, such as ICAP’s BrokerTec and Nasdaq’s eSpeed, are not subjected to a standard reporting scheme providing the official sector with reliable post-trade data. Regulators are able to access that data similar to the way that the CFTC currently accesses order book and message data for futures: ad hoc through special calls or other enforcement tools.
Meanwhile, Treasury-market participants are able to liaise with trading venues’ data providers to access data for analysis to benefit their own trading purposes. These are distinctive and appropriate commercial arrangements, but illustrate that if data can be made available routinely to market participants, routine post-trade data could be accessed for regulatory purposes as well.
Other important considerations about appropriate timing for such post-trade reporting, and which agency would be best situated and equipped to build and oversee such a framework, must be analyzed. The U.S. policymaking community as a whole would need to answer those questions, but the CFTC and the markets it oversees would be well served by such a framework.
Collaboration among regulators also should be enhanced. Price discovery occurs more quickly and the market functions more efficiently because of high-speed and algorithmic trading and arbitrageurs, but risk and volatility are also transferred across different markets. For instance, an intentionally manipulative strategy, or even an inadvertent error, made by a high frequency or algorithmic trader could quickly cascade and impact the derivatives and related cash markets (and vice versa) across different jurisdictions. This systematic risk requires collaboration among regulatory bodies.
As part of this enhanced collaboration, the CFTC should make it a long-term goal to be able to share market data and conduct systematic surveillance in near real-time in collaboration with other financial regulatory agencies. Many large market participants have a presence in markets that are overseen by different agencies, and enhanced collaboration would give regulators a more comprehensive picture of their activity. Sharing data and analytics across agencies is also a more feasible and less costly goal with the advent of cloud technology. Further, collaboration between the surveillance teams among different agencies would greatly improve the CFTC’s understanding of its own markets, and the financial system as a whole.
CFTC Should Continue Pursuing Policies that Incentivize Non-Traditional Liquidity Providers to SEFs
As the CFTC works toward accessing additional data sources, it should continue its pursuit of policies to attract non-traditional liquidity providers to SEFs. Part of the reason certain futures contracts enjoy so much liquidity is that non-traditional liquidity providers are incentivized to contribute their share through market structure policies affecting exchanges.
The reasons behind the retreat of global banks from fixed-income markets, including related derivatives, are complex, and their capacity as liquidity providers likely will not be fully replaced by non-banks with much smaller balance sheets. Nevertheless, part of the solution to liquidity challenges in derivatives must come from this latter category of participants.
In order to encourage liquidity on SEFs, in particular on SEF electronic central limit order books, I believe the CFTC should revise its floor-trader exemption. This exemption was designed to promote market-making activities by non-traditional liquidity providers on SEFs, and recognized that swap-dealer registration was not necessary or appropriate when all the dealing activity was conducted on regulated exchanges and cleared.
The conditions for the floor-trader exemption need to be revised to make compliance practicable while ensuring that floor traders do not pose an increased risk to the marketplace.
Toward this end, the CFTC also must follow up on its Concept Release on Risk Controls and Systems Safeguards for Automated Trading Environments, which it published in September 2013. Automated trading in the derivatives markets is not just a trend, it is here, and has been for many years, in fact. The CFTC must ensure those systems have adequate pre-trade risk and other risk controls.
To conclude, de la Vega also wrote, “whoever wishes to win at this game must have patience and money.” While his advice applied to trading on the Amsterdam stock exchange, the notion appropriately applies to the regulatory response I have mentioned today, and the CFTC’s surveillance mission as well. With time, agency collaboration, and creative use of resources provided by the government or other stakeholders, the CFTC and other market regulators will succeed in their mission to understand, perform surveillance of, and ensure integrity in, markets that are becoming increasingly automated.
Thank you again for allowing me to be with you today.
Last Updated: May 21, 2015