OTC Markets Group has named Meighan McGowan Executive Vice President, Corporate Services. According to a press release, McGowan joins from Computershare, where she was Head of Business Development, Investor Engagement North America, a division that includes Georgeson Advisory, Investor Intelligence and IR Suite. Prior to that, she held a senior business development role at Morrow Sodali, where she advised public companies, their in-house counsel and investor relations teams on corporate governance, proxy solicitation, capital markets intelligence, and M&A and activism defense.
Meaghan Dugan
Meaghan Duganhas joined IMC as Managing Director – U.S. Market Structure Strategy Lead. Based in IMC’s Chicago office, Dugan will lead the firm’s U.S. market structure strategy and its engagement with exchanges, regulators, clearing organizations and industry bodies. Dugan joins IMC from Cboe Global Markets, where she most recently served as Senior Vice President, Head of U.S. Derivatives. At Cboe, Dugan’s responsibilities included business strategy, competitive positioning, market structure and market development for Cboe’s U.S. Options and Futures businesses. She brings more than 20 years of experience across listed options trading, market making, electronic trading and exchange leadership. Cboe has not named a successor, although the company told Traders it is undergoing a search and Rob Hocking is serving as interim.
Northern Trust has appointed Rima Qureshi, former executive vice president and chief strategy officer of Verizon Communications Inc., to its board of directors, according to a press release. Qureshi brings nearly four decades of experience shaping strategy, leading transformation and building businesses across global telecommunications and payments companies. At Verizon, which Qureshi joined in 2017, she helped set the company’s strategic direction amid significant industry change. Prior to Verizon, Qureshi spent 23 years at global telecommunications company Ericsson
Meyer “Micky” Malka, founder and managing partner of Ribbit Capital, has been appointed to the Board of Directors of Apollo Global Management, effective January 1, 2027. According to a press release, Malka has more than 30 years of experience building and investing in technology and financial services businesses across the US, Europe, Asia, and Latin America. Malka currently serves on the board of directors of Robinhood Markets, Inc., and several private companies.
The Bank of America Corporation Board of Directors has appointed Jim Fitterling and Bill Thomas as directors, according to a press release. Fitterling is Executive Chairman of the Board of Directors of Dow Inc., a global materials science company. From 2018 to 2026, Mr. Fitterling served as Chief Executive Officer of Dow, and he was elected Board Chair in 2020. Thomas served as Global Chairman and Chief Executive Officer of KPMG International, a global organization of independent professional services firms providing audit, tax and advisory services, from 2017 to 2026.
Fnality has hired Roger Nolan as Group COO based in London, reporting into Michelle Neal, Group CEO. According to a press release, Nolan was previously the COO of LCH, the UK clearing house subsidiary of the London Stock Exchange Group. He has more than 30 years’ experience in investment banking and financial market infrastructure, with considerable expertise managing rapid business expansion at heavily regulatory financial market infrastructure providers.
If you have a new job or promotion to report, let me know at alyudvig@marketsmedia.com
Adena Friedman, chair and chief executive of Nasdaq, said on the third quarter results call on 22 October 2026 that the group achieved new milestones across all three divisions, including record index net inflows, record fintech ARR (annualized recurring revenue) growth, and the establishment of Nasdaq Digital Liquidity Networks within Market Services.
Friedman added: “The shift towards more connected and always-on markets is gaining momentum, and Nasdaq is shaping the next generation of market infrastructure.”
Nasdaq executed several strategic actions to advance this evolution. Digital Liquidity Networks (DLN) was formed to unify the capabilities, investments, and solutions the firm has been developing to advance tokenization and always-on markets under one strategy.
Roland Chai, Nasdaq
Within DLN, Nasdaq completed the acquisition of LeveL Markets, the third largest Alternative Trading System (ATS) in the U.S. by trading volume, on 15 October 2026. LeveL is expected to contribute revenue of between $7m and $8m in the fourth quarter.
Roland Chai, head of Digital Liquidity Networks, said in a statement: “LeveL Markets advances Nasdaq’s vision for more connected, always-on markets by adding a scaled off-exchange liquidity platform and execution expertise to Nasdaq’s technology, market capabilities, and client relationships.”
Friedman continued that the LeveL Markets acquisition strengthens the ability to deliver institutional-grade liquidity and execution solutions to clients with more flexibility through an off-exchange venue. She added: “We also have the ability to expand Level ATS to new markets and to connect its 2,500 institutional and retail brokerage clients to multiple global markets and asset classes.”
In another new capability, HSBC is extending its 15-year partnership with Nasdaq, adding Calypso’s exchange-traded derivatives (ETD) clearing capabilities. This brings clearing across over-the-counter (OTC), exchange-traded derivatives, and repos into a unified platform, helping HSBC manage margin, collateral, and risk more efficiently according to Friedman. She said Nasdaq had very narrow capabilities in ETD clearing and described HSBC as a “great partner” to work with to build out that capability.
“This allows us to go to market with a new module in Calypso and be much more comprehensive in allowing people to manage risk,“ Friedman added.
Her view is that always-on markets and AI will lead to the convergence of markets, asset classes and geographies, so risk management needs to be multi-asset class, global and real time.
“You have to use agentic capabilities to manage risk in real time and take action in real time and all the capabilities we are building in Calypso are related to that,” she said. “In addition to tokenization, it’s been a big investment year for us and we see it unlocking durable demand.”
Adena Friedman, Nasdaq
Friedman described Calypso as the “premier” provider of collateral management capabilities. In March this year Calypso announced an integration with Baton Systems to provide clients with advanced collateral workflow automation and fast, efficient asset mobilization capabilities. Baton Systems operates a real-time network connecting major global banks and futures clearing merchants (FCMs) directly to clearing houses.
“All forms of collateral are moving into a tokenized state, and that allows banks to think about moving from one collateral pool to another in a much more seamless way,” she added.
Nasdaq has completed a proof of concept with two major investment management firms, Vanguard and Wellington, to use Canton, the privacy-enabled blockchain, to move collateral. Friedman said the pilot worked “instantly and perfectly.”
Equities trading
Nasdaq remains on track to go live with 23/5 trading of U.S. equities on 6 December 2026, which will expand global access to the stocks.
“This launch of expanded, nearly continuous weekday trading hours is an important step in the journey to always-on markets, enabling broader investor access and greater transparency for all market participants,” said Friedman.
Nasdaq has also signed a deal with Eqlipse Trading technology to power the 24/7 U.S. equities trading ecosystem. In market technology, Nasdaq said there growing demand for always-on markets. In the third quarter Bruce ATS signed an agreement to use Nasdaq’s trading technology to power the first 24/7 U.S. equities trading ecosystem.
“Expanding continuous U.S. equities trading over the weekends through an institutional-grade ATS is a significant development for the markets,” added Friedman.
Other markets are also using Nasdaq for managed services, to help them modernize their platforms and move to cloud. She said: “It’s a really good demand environment right now.”
In addition, Nasdaq is working with DTCC as the post-trade market infrastructure prepares to launch its tokenization service in October 2026. Nasdaq is on track to launch the Nasdaq equity token in the first half of 2027, which Friedman said will put issuers at the center of tokenization.
In September this year the U.S Securities and Exchange Commission announced an innovation exemption allowing new markets called Tokenized Securities Venues (TSVs) to trade NMS stocks with certain limitations, including market share thresholds. Friedman said: “They are not going to have a huge impact on our business.”
If NMS stock is tokenized by a third party unaffiliated with the issuer, the TSV must provide a written notice to the issuer allowing them to object. A number of companies have already objected to tokens representing their stock being traded on a TSV set up by OKXICE. Nasdaq has said it will not allow its own stock to be traded on TSVs.
In contrast, Nasdaq is working with crypto firm Kraken to deliver a tokenized equity that is fungible with traditional shares and offers real-time settlement. In September 2026 Nasdaq announced an expansion of its relationship with Payward, the parent company of Kraken, including an agreement by Nasdaq Ventures to invest $100m, the continued advancement of the companies’ work on the Nasdaq equity token framework, and a new market surveillance agreement. Nasdaq equity tokens will be designed to connect with Payward’s xStocks ecosystem.
Arjun Sethi, Kraken
Arjun Sethi, co-chief executive of Payward, said in a statement that more than $2 trillion of stock trades run through the U.S. clearing system every day.
“Buys and sells net down by about 98%, and the clearing house holds $10bn to $20bn of collateral against what is left while it waits a day to settle,” Sethi added. “Cutting that wait from two days to one in 2024 released $3 bn. Onchain settlement removes the wait.”
Friedman said Nasdaq will be able to use Level ATS to distribute tokenized equities through Kraken. The Nasdaq equity tokens will settle through a “very innovative” custodian according to Friedman.
“It’s a really nice, elegant solution,” she added. “We will have instantaneous settlements and it also allows the issuer to have line of sight into investors.”
Some clients have said they will be an early adopter of the Nasdaq equity token, which is slated to launch in the first half of next year.
Nasdaq has commented to the SEC that TSVs sit outside the general rules that govern U.S. equities trading and are not transparent today as trades do not get reported to the tape.
“Our view is that tokenization should be focused on all trading and not try to create this sidecar kind of experimentation,” Friedman added. “The good news is the SEC did limit the extent to which these innovation exemptions can have an impact on equities.”
AI
Sarah Youngwood, chief financial officer of Nasdaq, said on the call: “The move towards always-on markets and expanding AI use cases are supporting client demand for the data franchise, which will be durable trends for years ahead.”
Sarah Youngwood, Nasdaq
The data business reported a 34% year-over-year increase in the number of enterprise licenses. Friedman said this is due to Nasdaq focussing on delivering data that is AI-ready, and delivering capabilities that drive agentic capabilities for clients. The group has created MCP connectors to allow clients to absorb and ingest the data in a flexible way and integrate it into agentic workflows.
Friedman said: “Artificial intelligence is becoming increasingly foundational to how institutions make decisions, manage risk, and allocate capital and Nasdaq is extending its leadership through solutions embedded across our platform.”
For example, so far this year 25% of bookings for eVestment, Nasdaq’s cloud-based institutional investment data, analytics, and market intelligence platform, were related to AI use cases.
In Data Link, the cloud-based financial, economic, and alternative data platform, Nasdaq continues to expand distribution as clients develop AI-native workflows. Friedman gave the example of partnering with retail broker Robinhood to embed a new proprietary dataset into its agentic trading platform. She added: “This gives investors access to sophisticated market sentiment and institutional-grade investor insights to power agentic decision making.”
She continued that Nasdaq expanded client adoption of its AI capabilities. For example, the firm signed new clients for calibration copilot, an AI-powered tool built into Nasdaq’s market surveillance and compliance platform.
Nasdaq Verafin, a cloud-based financial crime management software platform, has an agentic AI workforce that is used by more than 850 clients. In the third quarter Verafin announced the general availability of two new AI workers, the agentic AML (anti-money laundering) analyst and the agentic fraud analyst.
The group also launched a new operating environment which enables clients to connect their own AI infrastructure and serve as an access point for the agentic capabilities of Calypso, Nasdaq’s capital markets and treasury management platform.
As clients are thinking about deploying AI, they are moving more of their own infrastructure into the cloud. As a result, Nasdaq is seeing an uptick in cloud signings within fintech solutions. Friedman said: “We had 85% of bookings in cloud, and that allows us to be more of a managed service provider.”
Source: Nasdaq
Index business
The index business reached new records for exchange-traded product (ETP) assets under management and net inflow records.
Index had $14bn in net inflows over the trailing twelve months, including $49bn in the third quarter of this year. At the end of the third quarter, ETP assets under management was a record $1.16 trillion, with the average of $1.1 trillion, both new records.
Nasdaq launched 32 new index products in the third quarter, including 10 in the institutional annuity space. Half of the new products launched in the third quarter were outside the U.S. Approximately one third, 35%, of the trailing 12-month net inflows have been driven by products launched over the last five years.
“We continue to experience early momentum from our partnerships with BlackRock and State Street, with combined assets under management exceeding $1bn,” added Friedman. “We also continue to grow our long standing partnership with Invesco, expanding the Invesco QQQ innovation suite through the launch of the Invesco Nasdaq International Innovators 100 ETF.”
Financial results
Third quarter 2026 net revenue was $1.5bn, an increase of 13% over the third quarter of 2025. Youngwood said Nasdaq delivered “excellent” third quarter results, headlined by outstanding solutions revenue growth, continued operating leverage, and strong EPS (earnings per share) growth.
The Market Services division set a record in third quarter net revenue of $309m, up 2% versus the prior year period. Nasdaq said the business benefited from higher industry volumes across U.S. options and European equities with robust performance in U.S. cash equities, which achieved higher market share and near record capture.
Source: Nasdaq
Solutions revenue was $1.2bn in the third quarter, an increase of 17% versus the prior year period, due to strong growth across capital access platforms and financial technology.
The listings franchise continued its strong momentum from the first half of the year, according to Friedman. She said there has been a modest softening in the IPO environment in recent weeks, with some companies in certain sectors re-evaluating the timing of their IPOs. However, the listings pipeline remains robust, and she said Nasdaq is well positioned to support companies preparing to enter the public markets across a wide range of sectors.
Youngwood said: “Our durable business model and disciplined execution underpin our capital allocation strategy that prioritizes funding investments in long-term growth while returning more than $1.7 bn to shareholders year to date through share repurchases and dividends.”
Artificial intelligence could be one of the most significant developments in hedge fund investing since the emergence of quantitative trading.
What makes the current AI revolution different is that it is no longer limited to quantitative hedge funds employing teams of PhDs and programmers. Generative AI and increasingly sophisticated machine-learning tools are making advanced investment research capabilities accessible to discretionary fundamental managers, credit specialists, macro traders, and even relatively small hedge funds. The biggest change is that AI can dramatically increase the amount of information an investment professional can analyze, while reducing the time and cost required to do it. There are six major areas where AI is transforming the investment process.
1. Fundamental investment research
Traditionally, an equity analyst might spend days reviewing annual reports, quarterly earnings releases, conference call transcripts, industry publications, and competitor filings before forming even a preliminary view on a company. Much of that time is spent locating, extracting, and organizing information rather than analyzing it. AI can now accelerate many of these tasks dramatically.
For example, an analyst researching a publicly traded industrial company could use AI to:
A. Analyze 10 years of SEC filings and earnings transcripts, extracting key financial data, segment results, and management guidance into a structured format in minutes rather than days.
B. Identify changes in management’s language regarding demand, pricing, and margins. Subtle shifts in tone, such as moving from “strong demand” to “stable demand,” or a sudden increase in hedging language, can be early signals that fundamentals are changing before it shows up in reported numbers.
C. Compare financial metrics with 20 competitors, including growth rates, margins, return on capital, working capital trends, and capital allocation decisions, to see whether a company’s performance reflects company-specific strengths or simply industry-wide conditions.
D. Identify inconsistencies between management commentary and reported financial results, such as optimistic statements about pricing power while gross margins are declining, or claims of strong demand alongside rising inventories and receivables.
E. Build preliminary financial models and sensitivity analyses, showing how earnings and valuation respond to changes in key assumptions such as volume, pricing, input costs, and interest rates.
F. Highlight potential investment catalysts and risks, including upcoming product launches, contract renewals, regulatory changes, debt maturities, litigation, and shifts in customer or supplier concentration.
AI can also extend research beyond a single company, mapping supply chain relationships, summarizing expert network and industry interview notes, and flagging developments at customers or suppliers that may affect the company being analyzed. This allows an analyst to cover more companies with greater depth, or to go much deeper on a concentrated set of positions. It also reduces language barriers. An analyst can have AI read annual reports, filings, and earnings releases in Japanese, German, or Spanish and report back in English, allowing investment teams to look across a wider range of markets.
A research project that historically required several days of preliminary work may now be substantially accelerated. However, data validation and investment judgment remain essential. AI tools can misread tables, misattribute figures, draw on outdated information, or present a confident answer that is simply wrong. Analysts must verify key data against source documents, and firms need clear policies on data security, the handling of material non-public information, and documentation of how AI-generated analysis is used in the investment process. These governance requirements apply to every use of AI discussed in this paper.
The investment implication is that analysts can spend less time collecting and organizing information and more time developing differentiated investment insights. Because these tools are becoming widely available, simply having access to AI will not be a lasting competitive advantage. The edge will increasingly come from how a firm integrates AI into its research process, the quality of its proprietary data and judgment, and its ability to identify insights the rest of the market has not yet recognized.
2. Alternative data analysis
Hedge funds have long used alternative data, but AI greatly expands the types of information that can be analyzed, and allows it to be processed far more quickly and economically. Examples include satellite imagery, credit card transaction data, shipping activity, job postings, consumer reviews, web traffic, app downloads, and social media.
Much of this data is unstructured, meaning it comes in the form of images, text, or raw activity logs rather than neat tables of numbers. Historically, turning it into something usable required teams of data scientists and engineers. AI can now handle much of that work: counting cars in parking lots from satellite images, reading millions of product reviews to measure shifts in customer satisfaction, mapping job postings to specific companies and roles, and cleaning and aligning datasets from different vendors so they can be analyzed together.
Consider a hedge fund analyzing a retailer. AI could combine credit card transaction trends, online pricing and promotional activity, hiring patterns, consumer sentiment, web and app traffic, and inventory indicators to develop a view of revenue growth before quarterly earnings are released. It could also reveal why sales are changing, for example whether growth is coming from new customers or higher spending by existing ones, whether discounting is driving volume at the expense of margins, or whether a competitor is gaining share in a particular region or product category.
The competitive advantage increasingly comes from combining multiple datasets and extracting useful signals rather than simply purchasing access to the data. A single dataset viewed in isolation can be misleading. Credit card panels may over-represent certain demographics, web traffic may not translate into purchases, and satellite data can be distorted by weather or seasonality. Cross-checking several independent sources helps confirm whether a signal is real.
Alternative data also carries specific risks. Many datasets have short histories, making it difficult to test how reliable they are across different market environments. Data quality and vendor methodology can change without notice. Legal considerations include consumer privacy rules and the terms under which data was collected or scraped. Strong vendor due diligence and documented compliance procedures are becoming essential.
The investment implication is that AI may make alternative data more accessible to smaller managers while increasing competition for previously proprietary signals. As more managers buy the same data and apply similar tools, the insights it produces are likely to be priced into markets faster and decay sooner.
3. Quantitative investment strategies
AI is expanding the tools available to systematic hedge funds. Traditional quantitative models often depend on predefined statistical relationships, such as linear factor models built on value, momentum, quality, or carry, where the researcher specifies in advance which variables matter and how they relate to returns. More advanced machine-learning models can detect nonlinear relationships and interactions across many variables. A signal that works only when volatility is low and credit spreads are tightening, for example, may be invisible to a linear model but identifiable by a machine-learning approach. Natural language processing also allows systematic funds to convert text from earnings calls, news, regulatory filings, and social media into quantifiable signals, alongside the alternative datasets described above.
Potential applications include:
A. Identifying changing market regimes, such as shifts between risk-on and risk-off environments or between inflationary and disinflationary periods, so models can adjust when historical relationships stop working.
B. Forecasting volatility more accurately, improving position sizing and risk budgeting.
C. Refining trading signals by combining large numbers of weak signals into a stronger composite and identifying when individual signals are decaying.
D. Dynamically adjusting risk exposures as correlations, volatility, and factor crowding change.
However, this is also one of the areas where AI can be most dangerous. Financial markets have a low signal-to-noise ratio, and the relationships within them change over time as market structure, regulation, and investor behavior evolve. Complex models may fit historical noise rather than discover durable relationships. The more variables and parameters a model has, the easier it is to produce an impressive backtest that fails in live trading. Other risks include look-ahead bias, data errors, models that are difficult to interpret, and crowding, as multiple managers using similar data and techniques arrive at the same trades and face the same losses when they unwind.
The winners will not necessarily be managers with the most sophisticated AI models, but those with the strongest data, research discipline, and out-of-sample validation. That means requiring an economic rationale for why a signal should work, testing on data the model has never seen, accounting realistically for transaction costs and capacity, and monitoring live performance against expectations so decaying signals are identified early.
4. Portfolio construction and risk management
AI can help managers understand risks that traditional portfolio models may overlook. Conventional risk systems typically measure exposure through sectors, countries, market capitalization, and standard factors such as value, momentum, and size. These are useful, but they can miss the economic relationships that actually drive returns.
For example, a portfolio containing 50 different securities may appear diversified, but many positions could be exposed to the same underlying economic driver. A portfolio holding a semiconductor maker, a utility, an electrical equipment manufacturer, a commercial real estate developer, and a copper miner spans five different sectors, yet all five may be heavily dependent on continued spending on AI data centers. If that spending slows, positions that looked unrelated could decline together.
AI can help identify these common exposures by analyzing company fundamentals, news, supply chains, earnings sensitivity, and trading behavior. It can map customer and supplier relationships, measure how each company’s revenues and earnings respond to changes in interest rates, commodity prices, currencies, or consumer spending, and group securities based on how they actually trade together rather than how they are classified.
Other applications include:
A. Stress testing, including hypothetical scenarios that have no direct historical precedent, such as an escalation in tariffs, a sharp move in a specific currency, or the failure of a major counterparty, with AI helping to trace how the shock would ripple through each position.
B. Liquidity forecasting, estimating how long positions would take to exit under normal and stressed market conditions, and how that compares with the fund’s redemption terms.
C. Correlation analysis, recognizing that correlations are not stable and often rise sharply during market stress, when diversification is needed most.
D. Scenario modeling, quickly evaluating how the portfolio would perform across a range of economic, policy, and market outcomes.
E. Crowding analysis, identifying positions widely held by other hedge funds that could face forced selling if those managers reduce risk at the same time.
AI may improve portfolio efficiency and risk monitoring, particularly for complex multi-strategy funds. These firms may run dozens of independent investment teams, each managing its own book. AI can aggregate exposures across all teams in close to real time, revealing when multiple portfolio managers have unknowingly built similar positions or exposures to the same theme, and helping the central risk function reallocate capital more effectively.
AI should support rather than replace risk judgment. Models built on historical data may underestimate the impact of events without precedent, and risk outputs that cannot be explained are difficult for a risk committee to act on.
5. Trade execution and transaction costs
AI is being used to improve the timing and execution of trades. Transaction costs extend well beyond commissions. They include bid-ask spreads, market impact (the extent to which a fund’s own buying or selling moves prices against it), and opportunity cost when an order is delayed or only partially filled. These costs are often invisible in a fund’s reported expenses, but they reduce returns directly. For larger hedge funds trading substantial volumes, even modest improvements in execution can have a meaningful impact on performance.
Applications include:
A. Predicting liquidity, forecasting intraday volume and depth of the order book for each security so orders can be sized and timed appropriately.
B. Selecting execution algorithms and brokers, learning from the fund’s own trading history which algorithms, brokers, and trading venues have delivered the best results for different types of orders and market conditions.
C. Estimating market impact before a trade is placed, helping portfolio managers judge whether the expected return on an idea justifies the cost of implementing it.
D. Determining when to execute orders, deciding whether to trade aggressively to capture a short-lived opportunity or patiently to minimize impact, and adjusting in real time as conditions change.
E. Reducing information leakage, detecting when other market participants appear to be trading ahead of a fund’s orders, and changing execution patterns in response.
These tools are not limited to equities. In fixed income, credit, and other less liquid markets, where many securities trade infrequently and prices are less transparent, AI can help estimate fair value, identify likely counterparties, and improve the pricing of requests for quotes.
For example, a manager purchasing $500 million of securities might use AI-driven execution models to reduce market impact by several basis points. A hypothetical five-basis-point saving would represent $250,000 in reduced transaction costs. The benefit compounds with turnover. For a hypothetical $5 billion fund that trades $15 billion of securities a year, the same five-basis-point improvement would save approximately $7.5 million annually, or roughly 15 basis points of additional return for investors. AI can improve net returns even when it does not generate a single new investment idea.
The benefits are not evenly distributed. Managers with large assets, high turnover, or less liquid strategies stand to gain the most, while smaller firms will largely rely on tools provided by brokers and technology vendors. As adoption spreads, execution is also becoming an arms race in which yesterday’s advantage quickly becomes today’s standard.
6. AI investment agents and automated research workflows
This may be the most transformative development over the next three to five years. Instead of simply asking AI questions, investment firms are beginning to develop AI agents capable of executing multistep research workflows.
The difference is significant. A conventional AI tool answers one question at a time, and the analyst decides what to ask next. An AI agent is given an objective and works toward it on its own: breaking the task into steps, retrieving filings and data, running analyses, checking its results, and deciding what to investigate next based on what it finds. Rather than functioning as a faster search engine, it begins to resemble a junior research analyst that can work through a full assignment.
Illustrative AI investment research workflow:
Stage 1: Investment hypothesis. The analyst or portfolio manager defines the idea, such as rising electricity demand from data centers benefiting certain utilities and equipment suppliers.
Stage 2: Agent-driven research. The AI agent identifies the companies most exposed to the opportunity, analyzes their filings and earnings transcripts, evaluates competitors, processes relevant alternative data, and builds financial scenarios showing how earnings and valuations might respond under different assumptions.
Stage 3: AI-generated investment research package. The agent delivers a structured report with sources, financial models, key risks, and potential catalysts, with each conclusion linked back to the underlying documents and data so it can be verified.
Stage 4: Portfolio manager review and decision. Humans validate the work, apply investment judgment, and approve any position within the firm’s risk limits.
Agents can also be used for ongoing monitoring. Once a position is in the portfolio, an agent can track every new filing, earnings release, news story, management change, and relevant data point, and alert the investment team when something challenges the original thesis.
A human analyst could supervise several such workflows simultaneously. The important distinction is that automated research does not mean autonomous, unsupervised investing. Because each step builds on the one before, an early mistake, such as a misread figure or an incorrect assumption, can carry through to the final conclusion. Firms adopting these tools need clear limits on what agents are permitted to do, a complete audit trail of the sources and steps behind each conclusion, and human sign-off before any research influences a portfolio decision.
The productivity of a small investment team could increase substantially, and it may change how hedge funds build their teams, with fewer analysts devoted to gathering and organizing information and more focused on judgment, insight, and supervising AI-driven research. There is also a risk that if many firms rely on similar agents analyzing the same information, their conclusions will begin to converge, making genuinely independent thinking more valuable than ever.
What this means for fund managers: AI could change the competitive balance between large and small hedge funds
Historically, large hedge funds enjoyed substantial advantages in research staffing, technology budgets, alternative data, and analytical infrastructure. A multi-billion-dollar firm could employ dozens of analysts, data scientists, and engineers, purchase expensive datasets, and build proprietary systems that smaller managers simply could not afford.
AI could reduce some of these advantages by making advanced analytical capabilities available at much lower costs. Powerful AI models, cloud computing, and specialized vendor tools can now be accessed on a subscription basis, without the need to build infrastructure from scratch. A five-person investment team may eventually be able to perform research tasks that previously required a much larger team.
The impact extends beyond investment research. AI can also streamline many of the non-investment functions that weigh heavily on smaller firms, such as compliance monitoring, investor reporting, responding to due diligence questionnaires, and preparing marketing materials. Lowering these costs could reduce the level of assets a manager needs to run a sustainable business, making it more viable for emerging managers to launch and grow.
Smaller managers may also have structural advantages in adopting AI. They can often move faster, with fewer layers of approval and less legacy technology to replace. And as AI makes widely followed signals and large, liquid markets more efficient, the remaining opportunities may increasingly lie in niche or capacity-constrained strategies where smaller managers are better positioned to invest.
However, the largest firms may still retain significant advantages in proprietary data, computing infrastructure, specialized AI talent, and the ability to build customized systems. If widely available AI tools allow everyone to analyze public information equally well, the value of data that others cannot access will rise, and large firms are best positioned to secure exclusive datasets and data partnerships. They can also afford to compete for scarce AI and machine-learning talent, invest in experiments that may not pay off, and benefit from the vast amounts of data generated by their own trading activity. Large multi-strategy platforms, in particular, may use AI to further strengthen their ability to allocate capital, manage risk, and attract talent across many investment teams.
The competitive outcome is not predetermined. AI is likely to narrow the gap in some areas while widening it in others. Smaller managers that use AI effectively may be able to compete on research quality and operating efficiency in ways that were not previously possible, while the largest firms may extend their lead in the most data- and technology-intensive strategies.
What this means for allocators: evaluating managers’ AI implementation in due diligence
For investors, AI adds a new dimension to manager due diligence. Because these tools are becoming widely available, the key question is not whether a manager uses AI, but whether it is genuinely enhancing the investment process and whether the risks are properly controlled. Firm size alone may also become a less reliable indicator of a manager’s research capabilities, making it important to evaluate how effectively a manager uses AI relative to its resources and strategy.
Useful questions include how AI is integrated into research and decision-making, and what controls govern its use, including data validation, the handling of material non-public information, limits on AI agents, and human accountability for final decisions. Allocators should ask which alternative datasets a manager relies on, how they are validated, and what the manager does with that data that competitors are not. For systematic managers, the focus should be on how the firm guards against overfitting, how live results compare with backtests, and whether the team can explain why its models make the decisions they do. On risk management and execution, allocators should understand how a manager identifies concentrated economic exposures, how its stress tests are designed and influence portfolio decisions, and how it measures and improves transaction costs.
AI also offers allocators an opportunity to apply the same analytical tools to their own portfolios, identifying hidden overlap in economic exposures and crowded positions across the hedge funds they invest in.
Conclusion
Across the hedge fund industry, artificial intelligence is making information cheaper to gather and analyze. Signals built on public data will be priced in faster, and backtests and scenario analysis will be easier to produce. Research that once took an analyst a week can now be compiled in a day.
Implementing AI is a balancing act. Firms that do nothing could find themselves left behind on productivity and insights, while others that move too quickly could end up trading on a hallucination or implementing a flawed strategy from an overfit backtest. What will separate managers is what AI cannot supply: the discipline to validate the research that these tools produce, data that others cannot access, and the judgment to act on only the best insights.
Firm size will tell allocators less than it once did, and the quality of a manager’s process will tell them more as smaller firms become capable of research that used to require a large team. Allocators must understand how their current and prospective fund managers are using AI to create a more efficient and differentiated investment process, and what controls exist to mitigate the additional risks that come with it.
Artificial intelligence is changing how investors access, interpret and use market data, creating new opportunities for retail traders to benefit from information and analytical capabilities traditionally available to large financial institutions, according to Brandon Tepper, Senior Vice President and Global Head of Data at Nasdaq.
“AI is really helping investors gather data in a more easy to use fashion to help tell stories from that data so that they can make informed investing decisions,” Tepper told Traders Magazine.
Brandon Tepper
According to Tepper, Nasdaq is seeing new clients, applications and use cases emerge as financial institutions, technology providers and individual investors explore how AI can extract greater value from existing datasets.
The shift extends across retail investors, institutional traders and quantitative funds, with AI becoming increasingly important to how market information is delivered and consumed.
While AI is changing how investors interact with information, Tepper emphasized that the quality of the underlying data remains fundamental. “In order to have good AI, as you know, you really have to have clean, trusted, and accurate data,” he said.
Nasdaq operates 18 exchanges globally, according to Tepper, providing market data that can be combined with complementary and alternative datasets to support investment analysis. He described this as a layered approach, beginning with foundational exchange data before incorporating additional information.
Tepper acknowledged that AI systems are not always accurate and that their predictions depend heavily on the quality of the information they receive. “The value of the underlying data is the basis for good AI, and there are risks of using incorrect data,” he said.
He emphasized the importance of appropriate controls, security measures and understanding what AI applications can and cannot do.
The growing accessibility of institutional-grade market intelligence is another important development, with Tepper pointing to the narrowing gap between retail and professional investors. “I think the gap is getting very narrow,” he said.
Tepper noted that investors previously needed expensive financial terminals or proprietary datasets to obtain certain information, while technology platforms are now making that information easier to access.
One example is Nasdaq’s expanded partnership with Robinhood, which is bringing additional institutional-grade market intelligence into the brokerage’s AI-powered investment tools.
According to Tepper, the relationship began more than a decade ago and has expanded from equity market data to more sophisticated datasets, including Nasdaq TotalView.
More recently, Nasdaq began supplying alternative datasets through Robinhood’s new Agents ecosystem, allowing users to interact with AI agents that draw on market intelligence to answer investment-related questions. “We’re not just streaming live data, we’re helping investors utilize it in a more agentic way,” Tepper explained.
He said the development could eventually allow individual investors to establish algorithmic trading strategies based on predefined market conditions.
Tepper also pointed to the growing sophistication of retail traders, including investors described as “pro-tail,” who increasingly seek more advanced market information.
Despite the opportunities, he emphasized that AI-powered investing requires appropriate safeguards, particularly as applications become capable of supporting automated trading activity.
He cited Robinhood’s approach, which requires users to establish and separately fund an account for its agent-based functionality, with restrictions on access to information outside the platform. According to Tepper, the data is also not used to train the underlying AI models.
“If you have the right rails and the right security and risk profile, and understand what it can and can’t do, it’s going to just get better and easier to use,” he said.
Tepper said Nasdaq has been preparing its datasets for AI-driven applications for several years, including developing AI-ready data delivery capabilities and support for Model Context Protocol (MCP).
“AI is embedded in every single conversation we have with existing clients, new clients, startups, individual investors,” he said.
Market data moves toward an always-on environment
Beyond AI, Tepper identified expanding trading hours as another development shaping the market data industry. He said Nasdaq is diversifying geographically and across product categories, with more than 250 products available through Nasdaq Data Link.
The company is also preparing for growing demand for market information outside traditional trading hours as the industry moves toward 23-hour, five-day trading and potentially continuous markets.
According to Tepper, Nasdaq has been testing new data feeds in preparation for a December 6 launch of 23/5 trading.
Tepper linked the move toward extended trading hours to changing consumer expectations, noting that people have become accustomed to accessing services at almost any hour. “Why not have the ability to invest in your favorite company at all hours of the day?” he asked.
He sees opportunities for retail and institutional investors as trading hours expand, while emphasizing appropriate risk controls. “I see AI as sort of the next wave of investing,” Tepper said, predicting that improved access to information will support increasingly sophisticated agent-based investment applications.
CME Group, the derivatives exchange operator, reported record average daily volume for the third quarter and said volumes in the year to date are higher than the record volumes at the same time in 2025.
On 21 October 2026 CME reported that average daily volume rose to a third-quarter record of 29.4 million contracts, up 16% from the same period last year.
Non-U.S. average daily volume also reached a third-quarter high this year of nine million contracts. This was driven by growth in energy products, up 40%, agricultural products increasing 28% and equity index products rising 24%.
The derivatives exchange also reported all-time high average daily volumes for July of 27 million contracts, up 23% from a year ago, and 31.8 million for September, an increase of 22% from a year ago. Average daily volume in October 2026 is currently tracking at highest October ADV level on record.
Lynne Fitzpatrick, CME
Lynne Fitzpatrick, chief financial officer of CME Group, said on the results call on 21 October 2026 that fourth quarter volume is surpassing last year’s record fourth quarter by 17%, and the group is tracking 12% ahead of last year’s record volume in the year to date.
She added: “We remain firmly on track to deliver another record year and our focus remains on delivering deep liquid markets and maximizing capital efficiencies for our clients and strong earnings for our shareholders.”
Terry Duffy, chairman and chief executive of CME Group, said on a results call that beyond the strong volume growth, the underlying health of the business is best reflected in open interest and the significant capital efficiencies the group provides to customers.
Open interest reached 128 million contracts at the end of the third quarter, the highest open interest at the end of September in CME’s history according to Duffy, with open interest continuing to grow in October.
Duffy said product innovation remains central to CME’s strategy and will continue to be a key driver of performance and the ability to serve clients effectively.
“However, innovation must always be anchored in market integrity, financial safeguards, and client protection,” he added. “Innovation without real risk management is not progress and has the potential to create systemic risk.”
CME has sued the Commodity Futures Trading Commission over the regulator’s approval of crypto perpetual futures in the U.S., arguing that the contracts are swaps because they use a funding rate. Duffy said: “We feel very strongly about our case and we will abide by the court’s decision.”
Compute futures
CME is working closely with Silicon Data, which is backed by trading firm DRW and provides GPU market intelligence and benchmarking, to launch a compute futures market, pending regulatory review. The CFTC has extended the comment period on CFTC’S filing for another 45 days, which Duffy said takes it to 9 November 2026.
“We think this is a product that needs to get out when you look at all the conversations around data centers, AI, hyperscalers and chips,” Duffy added. “People are contacting us about how they can manage risk in this ever-changing landscape, so I would hope that the government would allow this product to move forward sooner rather than later.”
Delaying compute futures has a real cost for US innovation. Some of the largest data center lenders have hit their risk limits. Give them a liquid market to hedge compute price risk and they can free up lending capacity. It's the lack of a hedging instrument that has caused…
Don Wilson, chief executive and founder of DRW Trading, said on X: “It’s the lack of a hedging instrument that has caused credit spreads to blow out. We support the CFTC’s commitment to getting this right. The good news is we already have standards for building these markets properly: independent benchmarks, transparent methodologies and regulatory access to the underlying data. Apply them to compute and let well-designed futures move forward.”
Derek Sammann, global head of commodities markets at CME Group, agreed on the call that compute futures have been a “tremendous source” of client engagement and demand as the group has been working on the product for almost two years, including looking for a robust index provider.
Sammann argued that Silicon Data can provide data at huge scale across dozens of platforms and thousands of data points in all types of transactions, including the forward curve, which is crucial for developing an index that will be the reference for a futures contract. He added: “Another important point is that their index is already IOSCO compliant, meaning it is globally ready to be adopted by the community for the benchmark product.”
Source: CME
24/7 trading
In the third quarter CME also started 24/7 trading for 1-oz gold futures on 24 July 2026 and 100-oz silver futures on 11 September 2026. Cryptocurrency 24/7 trading launched on 29 May 2026 and CME said more than 361,000 contracts ($3.7bn) have traded since launch, with an average of just over 20,000 ($206m) per weekend.
“CME Group’s 1-oz gold and 100-oz silver contracts are the largest pool of regulated weekend trading volumes compared with offshore perpetual markets and other futures contracts,” said the firm.
Average daily volume for 24/7 trading of 1-oz gold futures was more than 15,000 over the weekend timeframe making up nearly 20% of total 1-oz gold futures ADV. For 24/7 trading of 100-oz silver futures, average daily volume was over 5,000 over the weekend timeframe, making up more than 30% of total ADV.
However, CME withdrew its filing to launch 24/7 trading for a 10-barrel oil future which it said would provide a transparent, regulated alternative to the 24/7 oil contracts that other venues have introduced in the marketplace. For example Hyperliquid, a decentralized blockchain derivatives protocol, provides 24/7 trading of perpetual oil futures outside the U.S. Hyperliquid gained attention when tokenized commodities and perpetual futures continued to trade on the decentralized finance (DeFi) platform after the U.S. and Israel launched joint strikes on Iran in February when traditional exchanges were closed over the weekend.
Terrence Duffy, CME Group
“U.S. participants are trading overseas on markets where they are not supposed to be participating,” said Duffy. “But we have not seen any evidence of policing of that activity so I’m surprised our 24/7 oil contract got held up.”
Duffy continued that the filing was withdrawn because commercial firms want to focus on bigger contracts and did not want to staff up another desk.
In the third quarter CME also launched several new products including single stock futures on 27 July 2026 and CME Group FutureSports Indexes (CME FSPI) on the National Hockey League.
More than 12 retail platforms are actively offering CME’s single stock futures. CME said volume is strengthening with nearly 350,000 contracts ($3.4bn notional) traded since launch, and average daily open interest reaching 16,000 contracts in September. Tim McCourt, global head of equities, FX and alternative products at CME Group, said on the call that single stock futures are off to a “strong” start and he is encouraged by the global participation in this new product.
Julie Winkler, chief commercial officer at CME Group, said on the call that it has been a busy year for product launches as CME has launched 368 new products so far this year. She said: “It proves that we can move quickly and we are the only traditional exchange offering 24/7 products.”
Capital savings
Winkler said CME is also advancing its execution technology and introducing industry-first functionality such as Treasury Link.
The latter connects CME’s interest rate futures with the cash Treasury liquidity pool on BrokerTec, enabling transparent, centralized spread trading on CME’s technology platform Globex. Clients can use Treasury Link to transact the differential between both markets with one atomic transaction, eliminating legging risk, reducing friction and maximizing capital efficiency. Clients have been testing Treasury Link which is due to go live in the first quarter of 2027, pending regulatory review.
Source: CME
The group generated a new high of $100bn in average daily margin savings in the third quarter. The increase in margin savings in the third quarter was primarily driven by higher equity index volumes and increased fixed income, currencies and commodities cross-margining, which grew from $1bn earlier in the year to $2.3bn on average in September.
Duffy added that the firm is on track to launch CME Securities Clearing on 7 December 2026 in advance of the SEC’s Treasury clearing mandate coming into effect and with U.S. debt at record levels. The group said this offers another capital-efficient clearing option “precisely when the U.S. Treasury market needs it most.”
Winkler said the new single stock futures also contribute to capital savings for clients.
Julie Winkler, CME
“On the institutional side this is a capital and margin efficiency play,” she added. “The ability to cross-margin single stock futures against equity index products is helping them unlock additional equity financing capacity.”
She continued that buy-side firms are also embedding single stock futures into their systematic QIS strategies and long/short equity models.
Cloud migration
Duffy added: “We also made strong progress in our cloud migration, making CME Group the only exchange to offer clients the choice of both co-location and cloud-managed services to expand access to our markets for years to come.”
Initial performance testing for CME Globex has been completed in the cloud and the group said it is reaching new milestones in the ongoing migration of CME markets, including faster round-trip times for trading.
CME said it is working to ensure clients can access both Google Cloud-managed services and co-location capabilities in Aurora in late 2027. Livestock market migration to the cloud will take place out of the new Aurora Region in late 2027.
Financials
CME reported record third-quarter revenue of $1.7bn, up 13% versus the same quarter in 2025, including a new all-time quarterly record for market data revenue of $243m. Fitzpatrick said market data revenues were up 20% and delivered the 34th consecutive quarter of year-over-year revenue growth.
Adjusted net income of $1.1bn, 13% ahead of the third quarter of 2025.
“Delivering double-digit volume growth across every single asset class drove double-digit gains in both revenue and earnings, resulting in the second highest revenue, adjusted net income, and earnings per share of any quarter in our history,” added Fitzpatrick. “Through the first three quarters of this year, our revenues have grown by 9% and our adjusted earnings per share by 11%.”
The Office of the Comptroller of the Currency (OCC) today announced a final rule to implement the Guiding and Establishing National Innovation for U.S. Stablecoins (GENIUS) Act.
Scott Bessent
“The GENIUS Act marks a watershed moment in American finance, further cementing the United States as the crypto capital of the world. Just as President Abraham Lincoln’s National Currency Act of 1863 established a national framework for bank-issued currency, the GENIUS Act establishes a federal framework for payment stablecoins in the digital age,” said U.S. Secretary of the Treasury Scott Bessent. “The OCC’s final rule carries that landmark legislation forward, providing a clear, practical foundation for private-sector innovation, competition, and American leadership in digital assets.”
“Today’s final rule advances President Trump and Secretary Bessent’s vision for American leadership in digital assets,” said Comptroller of the Currency Jonathan V. Gould. “We are providing a clear, practical framework for payment stablecoin innovation, with strong standards for reserve backing, redemption, and risk management. The OCC is drawing on its longstanding supervisory expertise to support responsible growth and innovation for this new industry.”
The final rule establishes requirements for permitted payment stablecoin issuers and foreign payment stablecoin issuers under the OCC’s jurisdiction, as well as certain custody activities conducted by OCC-supervised entities. It addresses all of the regulations the OCC is required to promulgate under the GENIUS Act other than those related to the Bank Secrecy Act, anti-money laundering, and Office of Foreign Asset Control sanctions, which will be addressed in a separate rulemaking in coordination with the Department of Treasury. In the course of rulemaking, the OCC received hundreds of comments in response to the proposal from a wide variety of stakeholders, including crypto-focused companies, community bankers, trade groups, academics, and more.
The OCC will continue to coordinate with other agencies responsible for implementing aspects of the GENIUS Act. The agency will announce in a separate news release when it will begin accepting and processing applications and waiver requests from prospective permitted payment stablecoin issuers. Given the nascency of the industry, the OCC will closely monitor its evolution and structure and adjust the supervisory and regulatory framework for payment stablecoin issuers to the extent necessary.
The final rule and the GENIUS Act take effect on January 18, 2027.
Intercontinental Exchange, Inc. (NYSE:ICE), one of the world’s leading providers of financial market technology and data powering global capital markets, and TradePMR by Robinhood (“TradePMR”), a leading technology and custodial services provider for registered investment advisors (RIAs), announced today that access to the web-based version of the ICE Connect Wealth offering has been integrated into Fusion®, TradePMR’s purpose-built advisor platform. 1
The ICE Connect Wealth offering provides comprehensive tools to analyze portfolio and market data, provide trade support for private labels, and access to ICE-provided tools and data. TradePMR’s Fusion® platform provides RIAs with tools to help manage their business and support their clients’ evolving needs.
“ICE and TradePMR have a long-standing relationship working together to provide solutions for the wealth management community, including the ICE Bonds Private Label solution which allows thousands of investment managers to search and trade fixed income securities,” said Chris Edmonds, President of Fixed Income & Data Services at ICE. “Client access to ICE Connect Wealth through TradePMR’s Fusion® platform unites the power of market research and data with a full suite of trading, execution and client management solutions on one convenient platform.”
“Financial advisors need technology that makes it easier to manage their clients’ investments and spend more time delivering value,” said R. Scott Victoria, President and COO, TradePMR by Robinhood. “By integrating ICE Connect Wealth with our Fusion® platform, we’re giving advisors access to additional research and trading tools within their existing workflows, helping them make more informed investment decisions, operate more efficiently and ultimately deliver a better experience for their clients.”
ICE Connect brings together a range of cross-asset real-time data, news and analytics for global markets. The platform helps customers manage price and currency risks and streamline workflows. ICE Connect is part of ICE’s Fixed Income and Data Services business, which provides comprehensive fixed income execution, clearing and data solutions that can help enhance market insights, manage risks and uncover investment opportunities.
22 October 2026 (London): The FIX Trading Community has announced the publication of its 24-Hour Markets Industry FAQ, a practical guide developed by industry practitioners to help firms navigate the growing shift toward extended-hours and 24-hour trading.
Enrico Cacciatore
As exchanges, trading venues and market participants continue to explore longer trading sessions, the FAQ addresses many of the questions firms are facing across trading, transparency, connectivity and post-trade workflows. Key topics covered include guidance on trading sessions and sub-session flagging, trade date considerations, order management, post-trade processing, clearing implications and operational resilience. The document also highlights how existing FIX data fields can be leveraged to support 24-hour trading models while promoting consistency across the industry.
“Once the session never closes, basic questions like which trade date a fill belongs to, or which session it printed in, become where reconciliation breaks,” said Enrico Cacciatore, Co-Chair of the FIX 24-Hour Markets Working Group and Co-Founder and Chief Product & Strategy Officer of CalcGuard Technologies. “Most of the answers already exist in FIX. This FAQ shows firms how to use the fields they already have consistently, before every venue invents its own convention.”
John Greenan, Co-Chair of the FIX 24-Hour Markets Working Group and Head, Technologies at Aquis Technologies, added: “Extended-hours trading impacts far more than matching engines and order books. It touches everything from trade processing and risk controls to client workflows and market operations. This FAQ provides a common framework that firms can use as they assess and implement longer trading hours.”
“The industry has been looking for a clear resource on this topic,” said Jim Kaye, Executive Director of FIX Trading Community. “This FAQ document reflects the expertise of market participants from across the ecosystem and demonstrates how established FIX standards can help the industry adapt to new trading models without reinventing the wheel.”
The Securities and Exchange Commission (SEC) is advancing a regulatory agenda spanning equity market structure, tokenization, extended trading hours and capital formation, as the agency considers how existing rules should adapt to changes in U.S. capital markets.
Speaking at SIFMA’s 2026 Annual Meeting on Tuesday, October 20, Jamie Selway, Director of the SEC’s Division of Trading and Markets, and Sebastian Gomez, Deputy Director of Legal and Regulatory Policy in the Division of Corporation Finance, discussed the Commission’s priorities with SIFMA President and CEO Kenneth E. Bentsen, Jr.
One of the Commission’s priorities is the proposed rescission of Rule 611 of Regulation NMS, commonly known as the Order Protection Rule, alongside the prohibition on locked and crossed markets under Rule 610(e). Selway said the SEC had received approximately 1,200 comments as part of its review, which included two industry roundtables. Market participants raised concerns about the potential impact on the National Best Bid and Offer (NBBO).
The agency is also examining how changes to Rule 611 should align with the modernization of best execution requirements. “There’s lots of reasons to update best execution, not just 611, but also after-hours trading, overnight trading, or tokenization,” Selway said.
He emphasized that best execution remains a FINRA responsibility and that the SEC is working closely with the organization. Selway indicated that FINRA could initially issue guidance alongside SEC action on Rule 611, with additional rulemaking potentially following. “I think that 611 for us is a Q1 event in all likelihood,” he said, referring to the first quarter of 2027.
The expansion of U.S. equity trading hours is approaching a milestone, with exchanges preparing for 23-hour, five-day trading beginning December 6. Selway said the transition would introduce exchange trading, public market data and clearing during overnight hours. “What’s new here is exchanges offering 23 by 5 trading, and there being public market data and clearing overnight,” he said.
He identified corporate actions and investor protection as two areas the SEC will monitor closely. “Corporate actions will now be compressed into the hour pause between eight and nine Eastern,” Selway said, highlighting the operational challenges involved.
The SEC will also monitor the effectiveness of overnight limit-up/limit-down protections. Despite these developments, overnight trading volumes remain relatively small. Selway estimated that overnight activity currently represents approximately 1% of consolidated equity trading volume, compared with roughly 20 basis points one or two years earlier.
He said estimates from market participants vary, with some expecting overnight trading volumes to reach 5%, while one projection of 30% struck him as “outlandish.” He pointed to demand from international retail investors, particularly in Asia-Pacific, as a potential source of growth. Selway also suggested that fully continuous trading could arrive sooner than anticipated, noting plans by alternative trading systems to introduce 24/7 operations.
Tokenization and Investor Protection
The SEC’s September 17 innovation exemption for tokenized securities was another focus of the discussion. The exemption permits trading of tokenized NMS stocks through automated market makers on eligible tokenized security venues, subject to conditions including volume limitations and issuer notification requirements.
Selway said the Commission intends to use the framework to assess market activity before moving toward permanent rules. “We want to do something exemptive, collect data, see how the market reacts, and then head to formal rulemaking,” he said.
He stressed that existing securities protections would continue to apply, including trading halts, antifraud and anti-manipulation requirements, and applicable broker-dealer best execution obligations.
Gomez addressed concerns from public companies about the impact of tokenization on their securities and corporate actions. He explained that issuers can object to participation on particular tokenized venues following notification, with a 30-day period to respond. “The issuer can say no,” he said.
Gomez also said tokenization was unlikely to significantly change existing processes for shareholder communications and dividends. “Nothing in the order changes Schedule 14A. Nothing in the order changes the communications requirements and the delivery requirements to shareholders,” he said. He explained that much of the activity is expected to involve tokenization of securities entitlements rather than changes to the underlying shares.
The SEC is also working with the Commodity Futures Trading Commission (CFTC) on securities-based derivatives, including single-stock futures and perpetual futures. Selway identified several platforms preparing to introduce securities-based perpetual futures in early November, including Kalshi, Coinbase, Bitnomial and affiliates of Kraken and Crypto.com. “We’re communicating frequently and we’re in a bit of a hurry,” he said.
The agencies are working with FINRA and the National Futures Association to develop monitoring arrangements for the products. Selway noted that the earlier market for single-stock futures had failed to gain traction, adding that the objective is to facilitate innovation while maintaining investor protections.
Gomez outlined the Commission’s efforts to address the declining number of publicly listed companies and the tendency of businesses to remain private for longer. He said one proposal would simplify issuer reporting classifications by establishing a $2 billion public-float threshold for large accelerated filers, while extending disclosure accommodations to smaller and mid-sized companies. Another proposal would expand Form S-3 eligibility, allowing newly public companies to raise additional capital without repeating the more extensive Form S-1 registration process, he added.
Gomez also discussed proposed changes to the accredited investor definition, which currently relies heavily on income and net worth. The proposals contemplate additional pathways based on professional qualifications and a potential accredited investor examination. “Can we actually allow someone to qualify to participate in one of these offerings, not just because they’re rich, but because they actually have the sophistication to do so?” Gomez said.
He also highlighted the potential impact of expanded Form S-3 eligibility on investment research, allowing broker-dealers to initiate coverage of newly public companies earlier under the Rule 139 safe harbor. “My hope is that we’ll see more research coverage among those newly IPO companies,” he said.
The discussion also covered mandatory central clearing of U.S. Treasury transactions. Bentsen referenced cash Treasury clearing at the end of 2026 and repo clearing in the middle of 2027. Selway expressed confidence in preparations for cash clearing but acknowledged the additional complexity of repo transactions. “We feel very good about cash. Repo is a complicated asset,” he said.
Selway said the SEC has been working with SIFMA on implementation issues, including an affiliate exemption, and expects further questions to emerge as repo clearing preparations continue. “We’re absolutely intent on implementing next year, but I think we’re still in learning mode, and we’ve got more to do,” he said.
The Commission’s agenda also includes electronic delivery, transfer agent modernization and updates to research-related rules. “Very, very busy, good progress, but plenty more to do,” Selway said.
Both officials emphasized that they were speaking in their personal capacities rather than on behalf of the Commission.
By Joanna Fields is the founder and CEO, Aplomb Strategies
In honor of Halloween, I was recently asked for a good horror movie recommendation, which got me thinking: does anyone remember The Ring? The 2002 horror classic started with an innocent act: someone pressed play on an unmarked videotape. Does anyone remember videotapes? The moment it was watched, the damage was done, the phone rang, and the clock started ticking. The only way to escape the curse was to copy the tape and pass it on, spreading it further with every attempt to contain it.
Joanna Fields
More than twenty years later, the cursed tape may have a modern sequel: unsecured AI tools that could start a chain of events with ominous outcomes – kicked-off by an unaware agent whose seemingly benign actions spin beyond their control. What if:
➢ A trader under investigation types privileged details of his case into a public chatbot to assess his own exposure.
➢ An engineer uses AI to rewrite model parameters in a database outside the secured and controlled system.
➢ An investment banker joins a confidential call about a client’s unannounced acquisition while the firm’s AI assistant records, transcribes, and summarizes it.
In each case, a single keystroke could send confidential information beyond the organization’s control and trigger a series of seemingly inevitable consequences.
The depth of the nightmare lies partially in the AI tool’s terms, including: what data can be stored, reviewed by third parties, or used to train future models. As soon as it has been “seen” by AI – it can’t be “unwatched.” A series of events could be initiated where terms like, “privilege, confidentiality or privacy” may be argued to be waived, trade secrets may lose protection, or novelty for a future patent may be compromised. Like the cursed tape in The Ring, the understated consequences have heavy implications and arrive on their own schedule.
Heroic Attempt to Break the Curse from Detroit
Just when every AI chat log seemed like a cursed tape bound for discovery, a Detroit court recently offered a plot twist. In Warner v. Gilbarco, Inc. (E.D. Mich. Feb. 10, 2026), Magistrate Judge Anthony P. Patti held that a self-represented plaintiff’s use of ChatGPT did not waive work product protection. AI, he reasoned, is a tool, not a person, and waiver requires disclosure to an adversary.
But like any good horror sequence, the relief was short-lived. One week later, Judge Jed Rakoff in New York rejected privilege protection for a criminal defendant’s AI communications. Because Warner limited the ruling by addressing only work products, it offers little shelter for privileged communications, trademark strategy, or unfiled patents.
With courts splitting outcome decisions, individual states writing their own AI and privacy rules, and regulators racing to catch up – there is no single map for protecting confidential, privileged or proprietary information in the U.S. The rules change depending on where the tape is played.
The Phone Rings in a Securities Fraud Case
Like the victims in The Ring, former GWG Holdings executive Bradley Heppner never suspected that pressing play could seal his fate. Charged with securities and wire fraud for allegedly misappropriating about $150 million before GWG’s bankruptcy, (which caused roughly $1 billion in investor losses), he used the consumer version of Claude to map out potential defenses and legal arguments. Federal agents later seized 31 of those AI-generated documents.
When his lawyers tried to bury the tape, Judge Rakoff refused in United States v. Heppner (S.D.N.Y. Feb. 17, 2026)**. He ruled an AI chatbot is not an attorney, sharing information with a commercial AI platform is not confidential, and forwarding the outputs to counsel later could not make them privileged.
The phone had rung. His private strategy landed in prosecutors’ hands, and he was found guilty of federal fraud counts. For anyone in the financial markets, a private brainstorm with a chatbot may be a recording waiting to be played back in court.
The Firm Chatbot Will Not Save You
Now picture a trader on a Chicago options desk facing an SEC market manipulation investigation. This time, typing late-night questions into the firm’s own AI chatbot instead of a public app. The trader’s position improves, but the curse isn’t broken.
While an enterprise tool with contractual confidentiality and no training on inputs can fix the “leaky tape” problem that sank Heppner – confidentiality, unfortunately, is only one element. Privilege also must be considered, and privilege still requires communication with a lawyer for legal advice. Courts have stressed that this relies on a trusting human relationship that cannot exist with software. So, when the trader asks AI whether order patterns look like spoofing, technically, legal counsel has not been sought.
Adding to the horror, those chats now sit on the firm’s servers as its records, which are required to be retained. Because privilege from an internal investigation typically belongs to the firm, the firm can waive it and hand the logs to SEC enforcement staff for cooperation credit. Unless counsel directs and supervises the AI use, the trader has simply traded a stranger’s tape for one the employer holds, and the SEC may be the next to press “Play”.
