Remarks of CFTC Commissioner Rostin Behnam at the ASIFMA 2018 Annual Conference: Developing Asia’s Capital Markets, Singapore

Fintech, Friction, and Formula 1:  A Learning Journey

October 31, 2018
[Delivered in Singapore November 1, 2018]

Introduction

Thank you for the kind introduction.  It is a pleasure and honor to join you today.  I want to thank Mark Austen and the Asia Securities Industry and Financial Markets Association (ASIFMA) for inviting me to both deliver remarks and provide some perspectives on the global regulatory agenda.  Before I begin, please allow me to remind you that the views I express today are my own and do not represent the views of the Commodity Futures Trading Commission (CFTC or Commission) or my fellow Commissioners.

Today marks my last day of a 10-day trip through Asia during which I participated in ISDA’s Annual Japan Conference and visited with fellow regulators and market authorities and participants in Tokyo, Shanghai, Hong Kong, and Singapore.  Since I will be sharing my views on more defined global issues facing our markets in Panel 1, and I will be leaving Singapore this evening before tomorrow’s sessions covering the fintech issues pulsing at the forefront of our markets, I would like to share some thoughts on bringing fintech more conclusively into the regulatory fold.  This would help to ensure that structures are in place to limit—and not amplify—any shocks to the overall financial system in the event of a technology failure.

Level-Setting the Fintech Fit and the Need for Friction

Around the world, there are various approaches to incorporating the latest fintech advances in cloud computing, machine learning, artificial intelligence (AI), distributed ledger technology (DLT), and all manner of virtual currencies and digital assets into existing rules and regulations.  But I think we can all agree that fintech fits within the principles of financial regulation aimed at ensuring the safety and soundness of individual firms and addressing risk to the larger system while incentivizing innovation, participation, and inclusion.

Financial innovations sometimes begin as a means to avoid regulation.  However, institutions, products, and processes that prove valuable to producers, consumers and the public; that are able to integrate with established banking and regulatory systems; and that are subject to clear, enforceable standards can become mainstream in spite of circuitous beginnings.[1]  It is a misconception to believe that the notion of regulation is so offensive that the true revolutionaries, those big dreamers[2] who are innovating at the edge, will flee at the first imposition of some boundaries.  Indeed, regulation creates legal certainty, freeing innovators from concerns that their activities and products may be deemed illegal or banned from the market.[3]  Regulation also creates a little necessary friction during the development stages to ensure that processes and products meet established standards and demonstrate proficiency before being introduced into our financial networks and to the public.  This, in turn, may decrease development costs by minimizing not only the costs of engaging outside counsel to clarify legal status, but also the costs associated with litigation born from compliance failures or consumer harm from innovation gone awry.[4]

Last week in Japan, I included fintech among the list of challenges we are facing, each of which presents an “uncomfortable level of uncertainty, as well as a sense of disdain for the anticipated costs of regulatory friction.”[5]  However—and this was largely directed toward fintech — “Obscurity should not disqualify or deter us from moving forward: the decision must be to act.”[6]

In the U.S., our regulatory fintech agenda has largely been enforcement-driven.[7]  While the steady progression of proceedings in various venues is building some legal certainty into our landscape, enforcement actions don’t always bring about enough thoughtful collaboration between legitimate firms and regulators.[8]  An open, forthright dialogue without direct repercussion allows us to learn from the small errors and make them a part of the regulatory equation so as to prevent and deter bigger ones.[9]  This is especially critical with fintech because mistakes will happen.  If our goal as regulators is to certify that there are structures in place to absorb and buffer the shocks from a trader or technology failure, then our aim must be to ensure those structures—whether rules, standards, or even barriers—are informed by an understanding of what can go wrong.

With our current lack of a full understanding of the promise and perils of fintech, to echo the words of Monetary Authority of Singapore (MAS) Managing Director, Ravi Menon, we are on a “learning journey” with industry.[10]  As we move forward, market regulators will serve an essential role in fintech development by ensuring that innovators are socialized into a culture of regulation and compliance that provides legal certainty.  Indeed, a successful integration into our existing market regulatory structures will promote stability and instill greater integrity, confidence in, and adoption of fintech.

Lessons from the Learning Journey

As I mentioned in my recent remarks to ISDA Japan, I spent a surprising amount of time during my first year as a CFTC Commissioner examining fintech related issues, especially crypto-assets and DLT.[11]  As part of a self-directed listening tour, I met with designers, developers, providers, and marketers of both prominent and newer, perhaps less mainstream technologies.

Thousands of miles from Washington, D.C., it was not surprising to hear that Washington is “not necessarily open-minded” and that regulators need to “scrap everything and come up with a new vision, refresh our approach, a new paradigm.”  These individuals want regulators to think of fintech as more of an opportunity than a risk.  They asked for acceptance of their method of using failures as a component of their development process.  In the perceived relative safety of the cloud, these visionaries of the moment boasted that making errors and being misunderstood is part of your persona.  They support a slow approach to regulation to give the innovative process time to develop, and to bring early adopters into compliance first to lead the way for others to come on board.  They recognize that their products can only go so far without access to the financial network, and that access requires some loss of independence.  Many also believe that the right principles already exist in the financial regulatory system and that we can make tremendous progress by taking stock and asking whether our principles and regulations make sense for other applications.

I agree that regulation—or regulatory certainty—should not come so early that it stifles innovation or risks arbitrarily picking winners and losers.  To the extent innovations focus on new technologies as opposed to new activities, relationships, or conduct, there may be less room for uncertainty.  But in determining whether our cumulative regulatory stock is fit for fintech purposes or requires further development, we cannot engage so late that we leave our markets, infrastructures, and the public unprotected.  I agree with those who recognize that, “The urge to ‘keep pace’ with technology is thus not a call for law to grow exponentially, but a call for laws that better reflect our current technological capacity.”[12]

MAS consistently demonstrates markedly proactive thought leadership in the fintech space.  MAS stands out for its transparent fintech agenda, its early collaborative efforts and entry into the first bilateral fintech cooperation agreements,[13] and its thoughtful consideration of the appropriate engagement and deployment of regulatory tools to address risks in a targeted manner without discouraging innovation.[14]  I am pleased that our own Commission recently signed a Cooperation Arrangement on Financial Technology Innovation with MAS.[15]  As a first step toward that collaboration, in just a few weeks, the CFTC’s LabCFTC, our dedicated initiative aimed at promoting responsible fintech innovation, will participate in the Singapore fintech Festival 2018.  We are looking forward to participating in Singapore’s pragmatic and flexible approach.

Keeping up with Progress

In terms of surveying our own regulatory toolbox, in 2015 and 2016,[16] the CFTC issued rule proposals commonly referred to as “Reg AT” (Regulation Automated Trading) comprised of risk controls, transparency measures, and other safeguards aimed at enhancing the safety and soundness of automated trading on CFTC-registered designated contract markets (DCMs).  Reg AT was intended to establish pre-trade risk controls to mitigate the potential dangers of an unchecked automated trading system.[17]

It has been several years since Reg AT was a Commission priority and our immediate regulatory agenda indicates that we are unlikely to move forward on Reg AT in its current iteration.

Reflecting on my learning journey to date, I agree that the CFTC, in terms of knowledge and data, and the markets, in terms of innovation, eagerness, and perhaps even acceptance, have moved beyond the original bounds of Reg AT.  It is time to work collaboratively with industry and our fellow regulators on new regulatory initiatives aimed at establishing appropriate principles and structures in furtherance of well-reasoned, targeted, and timely oversight—which may or may not ultimately take the form of regulation—to address finech beyond algorithmic trading.  We need to consider more comprehensive and inclusive structures that address the role of technology in all of its iterations within the scope of our regulatory and supervisory structures.

I appreciate the Chairman’s openness to working with me and my fellow Commissioners, market participants, and the public to determine whether there are elements in Reg AT that could serve as a basis for a new and more effective rule. [18]  I believe that certain elements of Reg AT remain relevant and could serve as the basis for new initiatives and proposals, and that we should leverage our ever growing data resources and expertise in LabCFTC, our Market Intelligence Branch, Office of Chief Economist, Division of Enforcement, and Commissioner-sponsored advisory committees towards any future proposals.  We need action and we need ideas.  As for me, I have started thinking of an idea as to where to start…

A Stress Test Track

Autonomous or self-driving cars are one obvious analogy when discussing bringing fintech into the regulatory fold and our financial networks.  After several high-profile crashes, the most notable of which involved a self-driving Uber car with a backup driver who was looking at her phone at the time of the crash, self-driving leaders are taking a more restrained approach to integrating their technologies into our existing infrastructure so as not to further diminish public enthusiasm and doom the technology before it can truly take off.[19]  They are beginning to discuss the potential costs and benefits of additional friction[20] to slow them down while dealing with the very immediate costs that must be paid when systems are allowed to move too quickly, without that bit of resistance.[21]  Like a lot of the fintech already in use, self-driving cars rely on machine learning, they “learn” to classify and respond to situations based on datasets.[22]

Much like I learned from developers in the machine learning and AI space, it’s hard to find data—images and behaviors in the driverless car space—of every sort of situation that could happen in the real world.[23]  Solving the problem is a matter of capturing as much of the unpredictable cases and errors as possible, and then figuring out how to train systems to deal with them.[24]  As one MIT engineer noted, it’s also a matter of accepting that humans are generally terrible overseers of highly automated systems.[25]  I heard this too during my journey.  There is and there will be no perfection.  At least we know where some of the weaknesses are and, as compared to the openness of our roads and diversity of drivers, it’s a little more feasible to create structures that will limit the impact of a fintech failure.

Taking this all in, I think that we are talking about a step towards standardized proficiency requirements for innovative developments to ensure that whatever products are introduced into our markets, they meet minimum standards of resiliency and stability.  Assuming that our regulatory objectives include confirming that these innovations are able to withstand unpredictability and error and ensuring that fintech does not disrupt functioning markets, either because its design undermines market rules or misfires under extreme conditions, I think we should leverage some of our prior thinking from Reg AT on stress testing,[26] a simulation technique common in the financial industry.

As an initial step, I think we should engage in a collaborative exercise within our industry such that regulators, self-regulatory organizations, and private industry develop a closed environment for simulation, stress testing, and training new technologies.  Within this environment, fintech providers would stress test their models to demonstrate market readiness.  This process should include sharing feedback on the environment to enhance and continue the environment’s evolution and inform any plans for standard setting or best practices.  Of course, the environment would include privacy protections for users’ techniques, practices, products, and processes.  In effect, we could create an open-source stress test on a closed test track for fintech in our markets.

The devil will be in the details, and I am planning on giving this idea some more thought and socializing it within the Commission and some of my new companions along the learning journey, but I think it is worth considering.  Successful supervision does not rest solely with the regulators.  The fintech and financial industries and other stakeholders should engage, exchange their ideas, explain what they are trying to achieve and how they intend to do it.

Conclusion

Getting back to cars, I would like to end with some thoughts on the Marina Bay Street Circuit, otherwise known as the Singapore Street Circuit.  Following its inaugural race, the 2008 Singapore Grand Prix,  the Street Circuit was widely criticized by F1 drivers.  The bumps and harsh curbs, likened to tortoises on the road, raised concerns among drivers that they would suffer suspension or other damage, or worse, be pitched into the wall on the outside of a corner.[27]  The Federation Internationale de L’autombile or FIA responded to drivers’ concerns and made modifications to the curbs.  While the intent was to make the curbs safer, the modification in some instances may have created new danger points by narrowing corner entries.[28]

The Street Circuit has made several modifications over the years, but continues to hold the record for having at least one safety car appearance at every race to date.[29]  As well, there remains what’s described as a “quirky little technical problem” specific to the Street Circuit, the origin of which has never been ascertained.[30]  Near the Anderson Bridge, the cars pass over something underground that creates electrical interference, causing the cars’ sensors to show strange readings and compromising the drivers’ control over the throttle position and clutch.[31]  In spite of being a closed circuit and system where alterations can be made to respond to risks, all risks cannot be eliminated because of unknown, uncontrollable factors —in spite of the presence of human overseers of awesomely sophisticated technology.  However, it sounds like the drivers who take on the challenge of the Singapore Grand Prix are aware of the risks and demonstrate the requisite talents to be on the road.

Regulators must always balance managing risk while facilitating innovation; providing certainty while remaining flexible; promoting growth while protecting the little guy—and we must maintain this balancing act while promoting transparency, engaging with stakeholders and the public, and maintaining full accountability.  How does that saying go, “At least we have each other.”  In all seriousness, as progress continues in all facets of fintech, we will increasingly need to consult, collaborate, and compare notes with one another in order to get the right level of friction.  I’m looking forward to working with you and those big dreamers.

Thank you.