Good Morning, Bom dia, Tudu bem.
It is a great pleasure to join you for the fourth edition of Regulation Week, jointly organized by the Fundação Getulio Vargas Center for Law and Regulation in Rio de Janeiro and the George Washington University Regulatory Studies Center here in Washington, D.C.
For those I have not met, I am Mel Gunewardena, Director of International Affairs to Chairman Mike Selig at the U.S. Commodity Futures Trading Commission.
The views I express today are my own and do not necessarily represent those of Chairman Selig or the Commission.
When I began my career at Goldman Sachs in 1994, markets were largely manual. Trading pits, hand signals, brokers, phones and handwritten tickets remained central to execution and price discovery.
Then markets moved to screens. Electronic trading broke down geographic barriers and computer networks helped move institutional trades from phones to electronic platforms. Options moved from pits and voice markets to screens. Markets became faster, more transparent and increasingly global.
The relationship between cash and derivatives also began to reverse. When I started, derivatives were largely priced from underlying cash markets. Today, across many of the world’s most important markets, derivatives increasingly drive price formation, price discovery and risk transfer. A simple lead-lag analysis demonstrates that Treasury futures lead price discovery in interest rates; WTI and Brent crude oil derivatives establish reference prices used throughout physical oil markets; and equity-index futures determine where cash equities reopen after overnight developments.
Global derivatives markets now exceed $1.2 quadrillion, and the CFTC oversees roughly $600 trillion across interest rates, equities, credit, currencies and commodities.
The next stage was the computer age. Mathematical models, expanding data and greater computing power moved markets from predominantly human execution toward algorithmic strategies operating across asset classes and geographies.
We are now entering another stage: markets that are not merely electronic or algorithmic, but increasingly continuous, programmable, composable and autonomous.
In 2017, during President Trump’s first term, the CFTC became the first regulator in the world to establish regulated crypto asset derivatives markets, with the launch of bitcoin futures.
Today, under President Trump, the United States is leading the next transformation. The GENIUS Act has established a federal framework for payment stablecoins, while the CLARITY Act is advancing a broader framework for crypto asset markets.
At the CFTC, Chairman Selig is translating that leadership into global derivatives markets—bringing innovation onshore and adapting the agency’s framework to a fundamentally different market architecture. The Commission approved the first “true” onchain perpetual futures contract, set expectations for 24/7 trading, clearing and settlement, and began putting rules in place for prediction markets.
The global financial markets are moving toward programmable and composable financial assets, continuous settlement and machine-executable infrastructure.
Tokenized securities, commodities, funds and collateral can operate alongside stablecoins and tokenized deposits. Smart contracts, atomic delivery-versus-payment, oracle networks, cryptographic attestations and cross-chain interoperability could transform how liquidity, collateral and risk move through the global financial system.
AI is also transforming who—or what—trades these markets. When I began at Goldman Sachs in 1994, Fischer Black was still at the firm, and the Black-Scholes model he developed with Myron Scholes remained the gold standard for modern quantitative finance. It applied probability and stochastic mathematics to the pricing of uncertainty.
The next generation used statistical models and greater computing power to identify relationships across markets and execute algorithmically. I saw that progression firsthand while covering Renaissance Technologies, then at the frontier of model-driven trading.
Today, artificial intelligence and large language models are taking it further—from models governed by predefined rules toward systems capable of interpreting unstructured information, reasoning across multiple inputs, making decisions and increasingly acting autonomously.
These are not separate developments. A tokenized asset can trade continuously, settle through a stablecoin, rely on smart contracts and oracle networks, reference a traditional derivatives benchmark, move across interoperable protocols and ultimately be traded by an autonomous AI system.
Collateral that was once valued and moved periodically can be revalued, pledged, released and rehypothecated in real time. Smart contracts can automatically adjust margin, liquidate positions and redirect collateral across interconnected markets and protocols. This may reduce settlement risk and improve capital efficiency, but it can also accelerate procyclicality, trigger cascading liquidations and transmit risk across markets before either regulators or market participants can intervene.
This convergence could make markets more efficient, expand access to capital and create new ways to transfer risk. But it could also fundamentally reshape the speed, scale and channels through which prices and risk moves across products, platforms and jurisdictions.
But this structural transformation represents a major and growing gap in the international regulatory agenda. International regulation remains organized around individual institutions, products and technologies while the market itself is becoming integrated, continuous and increasingly autonomous.
Much of the international regulatory architecture was built for markets with identifiable products, institutions and jurisdictions; defined trading hours; distinct trading, clearing and settlement functions; and ultimately, human decision-making.
Those assumptions are breaking down.
Markets may remain legally separated by jurisdiction, venue and product, but economically they are increasingly interconnected.
A Brent-linked contract trading on an online platform in Singapore can affect Brent in London and transmit into WTI in the United States. Volatility in a Korean equity can be amplified through an onchain derivatives platform operating from the British Virgin Islands and transmitted back into Korea.
The jurisdictions may be different. The venues may be different. The products may be different. Economically, they are part of the same market.
For example, a technology interface in one jurisdiction may be a gateway to a global derivatives market, with liquidity, leverage and risk distributed across an onchain ecosystem. Regulators should understand the economic market behind the interface, not simply the entity inside their legal perimeter.
As markets are increasingly interconnected, I notice four important gaps across the various international standard setting work and the various bilateral regulatory work undertaken by international regulators
The first is supervisory cooperation.
Many regulatory Memoranda of Understanding and supervisory cooperation agreements were designed for technical cooperation and information sharing. They were not built for markets capable of transmitting risk across jurisdictions in seconds, and many lack clear requirements for market disruption, cyber incidents, infrastructure failures, liquidity stress, defaults and timely supervisory notification.
Information sharing after something has gone wrong is not crisis management. Modern supervisory cooperation requires early notification, real-time coordination and clear responsibility before and during market stress.
The second gap is financial-stability oversight.
The international financial stability architecture remains heavily weighted toward central banks and cash-securities regulators, including jurisdictions with limited derivatives markets. Derivatives regulators and experts overseeing the markets where global price discovery and risk transfer increasingly occur remain underrepresented.
That imbalance shapes the analysis.
Non-bank credit illustrates the problem. Years have been spent pursuing hypothetical risks that have not materialized, while structural changes in AI, tokenization, new products and derivatives have been examined too often as isolated workstreams rather than as interacting components of a changing market structure.
The result is product analysis rather than systemic analysis, frequently driven by what has already entered the news cycle. Examined separately, these developments can appear novel but contained. Viewed together, they reveal how leverage, positioning, concentration, collateral and liquidity connect markets and transmit and amplify risk across the financial system.
Financial stability cannot be understood solely from economic statistics, balance sheets and cash markets—or from yesterday’s headlines in The Wall Street Journal. By the time a vulnerability becomes front-page news, it is no longer a warning. It is an event.
Understanding that risk requires derivatives expertise, market-level data, a practical understanding of how markets operate and a holistic view of the financial system. Without them, the system is structured to understand the next crisis only after it arrives.
That is not forward-looking financial stability oversight. It is institutionalized crisis management.
The third gap is global standard-setting governance.
Authorities come to the table with vastly different markets, expertise and access to information. Yet leadership and representation can reflect geography, institutional convention and established relationships rather than market scale, expertise, data and responsibility for the risks being supervised.
Consensus and cooperation then become a substitute for judgment. The objective becomes finding what everyone can agree to rather than identifying what global markets require. The result can be lowest-common-denominator regulation, comfortable, broadly acceptable and behind the risk before a standard is published.
A global market standard setter cannot operate like a diplomatic institution. Market relevance, expertise and risk—not convention—should determine who leads, what gets prioritized and how standards are shaped.
It should not be controversial to expect an institution that sets governance standards to practice good governance itself—or one that sets global standards to focus on the most consequential global risks.
If that sounds like a high standard, it should. Standard setters, of all institutions, should not be afraid of one.
The fourth gap is in our market safeguards.
Much of today’s framework was built for the computer age—fat-finger trades, erroneous orders, computer glitches and algorithmic disruptions. Price limits, and circuit breakers were designed for those risks.
But the next disruption could come from autonomous systems functioning exactly as designed. It could begin in an unsupervised market trading through a weekend or overnight, when liquidity is thin and traditional markets are closed, establishing prices that are transmitted into regulated markets when they reopen.
The market may be closed. Price discovery is not.
Our current markets safeguards were not designed for markets that are autonomous, decentralized, and continuous.
These gaps cannot be addressed retrospectively. Regulators must understand the technology, anticipate the new risk topology created by composability, interoperability and automated execution, and establish coherent standards that permit responsible innovation while protecting markets.
The greatest regulatory risk may no longer be simply that markets move outside the regulatory perimeter.
It may be that the regulatory perimeter itself no longer describes the market.
At the CFTC, under Chairman Selig, the agency is reviewing the architecture through which foreign markets and institutions access the United States and the cross-border arrangements supporting that access.
Participation in the world’s deepest financial markets should be supported by modern supervisory arrangements that protect customers and market integrity while addressing financial stability, economic and national security risks.
International cooperation therefore becomes more important, not less. But it must evolve with the markets it is intended to govern.
The choice is not between innovation and regulation. It is whether innovation develops within deep, transparent and well-regulated markets—or beyond the reach of frameworks that were too slow to understand it.
The next generation of markets is already being built: assets that are programmable, infrastructure that is composable, trading that is continuous and participants that are increasingly autonomous.
Capital is already moving through this architecture. Price discovery is already occurring within it. And risk will move through it faster than our institutions unless regulation evolves.
Markets will not wait for regulation to catch up.
The defining regulatory question of the next decade is therefore not whether these markets will emerge. It is whether those responsible for protecting the global financial system will understand them early enough to shape them.
The United States has repeatedly led the world through transformations in market structure—from manual markets to electronic markets, to the computer age.
Today, as global markets stand at the threshold of another great transformation, we again have both an opportunity and a responsibility.
Under the leadership of Chairman Selig, and with responsibility for roughly half of the $1.2 quadrillion global derivatives market, the CFTC will help lead this transformation.
Because the future of markets will be shaped by those who recognize the opportunity, understand the risks and have the vision to realize its extraordinary benefits.
