The World Federation of Exchanges makes recommendations for US Treasury AI regulatory framework
The World Federation of Exchanges (WFE), the global industry association for exchanges and central clearing counterparties, has called for enhancements to legislative, regulatory, and supervisory frameworks applicable to AI in financial services.
The WFE, which today published its response to the US Treasury’s consultation on the uses of AI in the financial services sector, said there are valid concerns about the uncertainties surrounding the evolving landscape of AI technologies, which require a close look at regulation in order to protect investors and other market participants. However, the WFE recommended the Treasury seek an appropriate balance between innovation and protection to ensure that the regulatory framework isn’t too broad, too complex and that there is cohesion and alignment amongst regulators and international standard setters.
If the US framework introduced fails to meet this, the benefits that AI brings to economic growth, productivity, automation and innovation will be at risk.
On behalf of the exchange and clearing industry, representing the providers of over 250 market infrastructures, that see more than $124tr in trading pass through them annually (at end-2023), the WFE advises the Treasury that:
- The definition of AI should be precisely tailored to avoid including more than what is necessary. A broad definition would create onerous restrictions and not be proportionate to the risks that different tools have.
- A definition of AI should focus on computer systems with the ability to make decisions or predictions based on automated, statistical learning.
- AI deployment by malicious actors is an emerging type of risk associated with this technology which financial services firms are well aware of and are tackling.
- Whilst traditional risk management techniques can be used to manage risk of AI systems, more work needs to be done to develop AI specific risk management tools.
- Third parties will be valuable to help develop AI tools and risk management tools, but Treasury is right to be cognisant of the risks around big tech firms utilising their market dominance.
- Regulatory uncertainty is a key concern amongst our members. Regulators should focus on outcomes and use sound judgment, fostering collaboration to support innovation and competitiveness in financial markets.
- Our members favour a principles and risk-based approach to developing a regulatory framework, where requirements are proportional to the level of risk associated with AI applications. This needs alignment among the various financial regulators and must be compliant with international standards.
- Ultimately, government policy should encourage modernisation by promoting the use of cutting-edge technologies like AI, cloud computing, and machine learning in capital markets. This enhances market dynamics, and provides better services to consumers.
Nandini Sukumar, Chief Executive Officer, at the WFE commented, “AI regulation must enhance protection whilst avoiding the curtailment of progress and modernisation. The definition of AI in the President’s Executive Order is overly broad and could create unnecessary complexity by imposing extensive compliance obligations if implemented for financial services. Policy should establish the appropriate safeguards and supervision, but it must also encourage innovation and promote the use of cutting-edge technologies, like AI. It’s through this that we can drive efficiency, enhance market dynamics and provide better services to consumers.”
Richard Metcalfe, Head of Regulatory Affairs at the WFE commented, “While these technological innovations and the associated concerns about managing generative AI are significant, it is important to remember that, as trusted third parties providing secure and regulated platforms for trading securities, our members are already carefully scrutinising tools and establishing controls to govern AI use. The US Treasury should therefore take care to design an AI regulatory framework which is principles based, to maintain flexibility and encourage innovation. We also need to have an incremental approach to AI regulation, allowing for gradual adjustments and learning, ensuring that regulations do not hinder technological progress.”
The full response and policy recommendations can be found here.
For more information, please contact:
Edelman Smithfield +44 7813 407 665
wfe@edelmansmithfield.com
Cally Billimore +44 7391 204 007
Communications Manager communications@world-exchanges.org
About the World Federation of Exchanges (WFE):
Established in 1961, the WFE is the global industry association for exchanges and clearing houses. Headquartered in London, it represents the providers of over 250 pieces of market infrastructure, including standalone CCPs that are not part of exchange groups. Of our members, 36% are in Asia Pacific, 43% in EMEA and 21% in the Americas. The WFE’s 87 member CCPs and clearing services collectively ensure that risk takers post some $1.3 trillion (equivalent) of resources to back their positions, in the form of initial margin and default fund requirements. The exchanges covered by WFE data are home to over 55,000 listed companies, and the market capitalization of these entities is over $111tr; around $124tr in trading annually passes through WFE members (at end-2023).
The WFE is the definitive source for exchange-traded statistics and publishes over 350 market data indicators. Its free statistics database stretches back more than 40 years and provides information and insight into developments on global exchanges. The WFE works with standard-setters, policy makers, regulators and government organisations around the world to support and promote the development of fair, transparent, stable and efficient markets. The WFE shares regulatory authorities’ goals of ensuring the safety and soundness of the global financial system.
With extensive experience of developing and enforcing high standards of conduct, the WFE and its members support an orderly, secure, fair and transparent environment for investors; for companies that raise capital; and for all who deal with financial risk. We seek outcomes that maximise the common good, consumer confidence and economic growth. And we engage with policy makers and regulators in an open, collaborative way, reflecting the central, public role that exchanges and CCPs play in a globally integrated financial system.
Website: www.world-exchanges.org
Twitter: @TheWFE
From Latency to AI Algo Driven Capital Markets
By Kelvin To, Founder and President, Data Boiler Technologies
Over half a century has passed since Electronic Communication Networks (ECNs) disrupted the traditional floor-based model of stock exchanges and ushered in the era of electronic trading. Since Regulation National Market System (Reg. NMS) was adopted almost 20 years ago, the honorable goal of promoting venue-by-venue competition and fair price formation across securities markets turned into a latency arm race. This has pushed market data and connectivity costs to rise exponentially.
In the quest to gain an edge over competitors, trading venues offer different rebates (e.g. enhanced market-making discount), introduce a speed bump (e.g. liquidity enhancing access delayed), proliferate order-types (e.g. midpoint-extend-life order), come up with new business models (e.g. market-on-close) and create other privileges (e.g. exclusive access to certain pegging orders). People jockeying around to make money by altering the queuing and wait times at the “checkout counters”. This has widened the gap between the haves and have-nots in the past.
It is a fitting moment to reflect on the evolution of capital markets. Several trends are emerging that could push today’s market structure from a focus on latency to one driven by algorithmic and artificial intelligence (A.I.) technologies:
If you can’t beat them, join them (outsourced execution):
A decade ago, former US SEC Chair Mary Jo White famously said, “deemphasize speed as a key to trading success.” Whether it is regulatory inaction, or the current administration’s proposing a wrong prescription, market participants today still find it hard to beat the high-frequency trading (HFT) firms. So long as the Consolidated Tape is NOT a reasonable compromise if not a close substitute of Exchanges’ proprietary feeds, everyone is and will continue be subservient to telecom infrastructure vendors. There is not enough alpha for the HFTs to justify the costs and benefits in doing just proprietary trading. Thus, HFTs get into the business of outsourced trading execution services. With stock exchanges optimally restricting access to price information by exploiting the inelasticity in demand of proprietary products, in turn, many choose to collaborate with the Haves for outsourced execution rather than compete.
This “collaboration” aims to maximize/ segment order flow for negotiation of tier rebates and other incentives that pose potential conflicts of interest and BestEx issues. The latency arm race would only subside if time-lock encryption is adopted to protect time sensitive data from being decrypted prematurely. The voluntary collaboration among the outsourcers and HFTs at echo chambers further complicates the markets. Aside from using transaction cost analyzers, liquidity sourcing and other tools or “bandages” to fabricate the fragmented markets, market participants are increasingly leveraging advance technologies, such as A.I., algo wheels, quant models, etc. to navigate and stay afloat in the markets.
Despise echo-chambers’ polarization and trends toward a more diverse group of participants
The 20th century marks the beginning of a cyber punk era. Elites (Corpo) advertise their selected truths popular among those who rely on corporate thinking rather than developing an ability to think independently. In the opposite spectrum, the rebellious mobilize the online communities (street kids) to orchestrate significant market movements. This is exemplified by the MEME stock phenomenon, where the naïve were feeling enlightened or motivated by their “leaders” to do a gag that would otherwise be prohibited if it occurred at a broker-dealer, then top market-makers were being lambasted to advance the rebellious’ controversial agenda.
There are the digital nomads, infusing “trust” (analogous to MBS credit enhancement) into digital assets while they are building related financial infrastructures to rent seek on transactions flowing across. Foreign adversaries would like to see the US engaged in “unhealthy” competition to erode the US’s prominent market position. Many do not seem to realize the emerging threats against capitalism or dare to admit it. DeFi and De-dollarization movements are on the rise and reap benefits out of chaos. Witnessing this dynamic, market participants would need their algo and A.I. to guide them in discerning who’s who and doing what, that impacts the markets.
One size does not fit all and mass customization
Different trading venues (lit exchanges, dark pools, systemic internalizers, single dealer platforms) are like different streaming platforms providing various contents (e.g. block trading, exotic, passive, conditional) that fit the appetite of respective subscribers. While we are generally supportive of standardization and harmonization of regimes, policy makers should also be warned that open access and interoperability could inadvertently be a path to a monopoly or reinforcing the elites’ oligopoly that hurts fair competition.
SEC Commissioner Peirce reminds “Hardwiring a technology into a rule runs the risk of preserving that requirement far after that technology’s expiration date” [amid Bloomberg is offering the open source FIGI reference data for FREE] with respect to the US joint data standards proposal in implementing the Financial Data Transparency Act. Also, no matter how Data Expert Group(s) are working with regulatory authorities in calibrating the consolidated tapes, reference price arbitraged is inevitable when there is more than one de facto NBBOs under the US competing consolidators model (or EBBOs in the UK and EU). If the public markets opt to serve only the most liquid and static form of investments, the private markets will prevail for its flexibility to dynamically customize deals for the investing communities.
From automated intelligence for economy of scale to decentralized federated learning
Some said “A.I. stands for Automated Intelligence”. We agree that automation benefited the elites to achieve economy of scale in the past. However, federated learning and real-time analytic platforms (RTAP) are more efficient and effective to provide timely decisions and rapid responds in modern dynamic markets.
Today, many resources are devoted or wasted in sourcing alternate data. Too much data and inherent problems (timestamp synchronization) of data imprecision causing centralized intelligence to take forever to achieve ‘golden source’. Then, people are juggling to insert or remove nodes for tunning/ data corrections (e.g., Support Vector Machine requires labeled data), while not knowing valuable insights may inadvertently be removed from the dataset. In turn, inexactitude in trade sequencing caused analytic results based on vector measurement or visualized heat-maps to be erroneous (false +/-).
The trend is empowering decision points to the field. Non-developers are now able to use no/low-code generation with A.I./ large language models to do some ad-hoc tasks for WYSIWYG rather than relying on the development team. Whereas the development team uses these tools to take on more complex projects, concentrate on quant modeling, reverse engineering of others’ algorithms, better Boolean logic, taxonomy, and whatnot.
Market Volatility, A.I. Hallucinations and the surprising similarities between Music and Trading:
Trading teams are poking holes in the initial results of A.I., which they are right to do so because there are different machine learning models. Firms should not go blindly with neural network, deep learning black boxes, then go about craving for evermore data to feed the models. While extraordinary market volatility and suspicions of A.I. hallucinations persist, there is also a growing recognition of the opportunities presented by newfound signals and liquidity amid chaos.
The performance of machine learning will improve over time. It may discover unknown and hidden opportunities which were previously nonsensical to humans. It is a paradigm shift to transition from a latency-driven capital market to one powered by algorithmic and A.I. machine learning technologies. Be open minded to explore the surprising similarities between Music and Trading. The sound that data makes is well suited to time series.
***
Data Boiler is a Type C organization member of the European Commission’s Data Expert Group. Between my patented inventions in signal processing, analytics, machine learning, etc. and the wealth of experience of my partner, Peter Martyn, we are about Market Reform, Governance, Risk, Compliance, and FinTech Innovations to create viable paths toward sustainable economic growth.