Articles

June 17, 2026

NeuGroup Quick Takes: 2026 First-Half Peer Group Meetings: DB & DC

# Cash and Working Capital

# Technology

# EMEA

How members are putting pension surpluses to work and getting data ready for AI-driven investment analysis.

Editor's note: NeuGroup Quick Takes will be updated with new takeaways following the completion of each of NeuGroup's spring peer group meetings. NeuGroup for DB & DC Plan Management, June 9-10, New York City Sponsored by MetLife Many plan sponsors are sitting on surpluses; they may have more flexibility than they realize. A number of members are navigating pension surpluses—a good problem to have, but one that requires a defined strategy. One member’s company carries a $6 billion pension surplus in the U.S.; the plan reached 100% funded status during the 2021-22 rate surge alongside most peers and has been in surplus ever since and longer than most. That experience has shaped a clear-eyed view of how the surplus can be leveraged. Rather than treating excess assets as trapped, the member suggests peers consider several strategies, starting with annual Section 420 transfers that allow the company to move a portion of its pension surplus each December to fund retiree health benefits, reducing the need to use corporate cash. It has also implemented a cash account program, similar to the approach IBM adopted in 2024. Under the program, employees continue to contribute to 401(k)s and the DC plan sponsor—instead of making matching contributions—provides an amount equal to 6% of salary to a pension-based cash account (funded by the DB plan surplus) that guarantees a 4% annual return. The arrangement has reduced annual cash costs by roughly $100 million while the pension surplus continues to grow. Advisor Mercer helped design the program. The company also views its sizeable pension surplus as a potential strategic asset in M&A transactions, where it could provide value to a future acquirer or help offset pension-related liabilities in an acquisition. DB sponsors with plan surpluses that are considering full termination face financial pain: Among other costs, 50% of any surplus is subject to excise tax. The mitigation tool for such a situation is a qualified replacement plan, which can reduce the tax hit closer to 20%. AI is ready for institutional knowledge work, but data is a bottleneck. Scott Dunphy, managing director and AI lead at MetLife Investment Management (MIM), made clear that the most pressing question is no longer whether AI can handle complex financial work, but whether firms are organized to take advantage of it. Mr. Dunphy discussed a commercial real estate underwriting model built in roughly 20 minutes using AI—eliminating a task that would otherwise have taken him close to two weeks. MIM also recently completed a proof of concept using 21 AI agents working in sequence to produce a full investment memo, with each agent handling a discrete task: market research, financial modeling, tenant analysis, deal narrative, etc. The result, Mr. Dunphy said, was "pretty amazing." But the harder challenge is upstream, he added. MIM's real estate group, which has operated for over a century, still stores much of its institutional knowledge on shared drives that AI cannot effectively search. Teams are migrating files to SharePoint and using Hebbia, an AI-powered knowledge platform that sits on top of those repositories, to help AI find and work with the information within them. "AI is the easy part," Mr. Dunphy said. "The hard part is the organization around it." Members across the room recognized the same obstacle at their own companies, and the session produced a clear consensus: Teams that treat data governance as a first step, not an afterthought, will have a decisive advantage as agentic AI becomes standard practice.

NeuGroup for Growth Tech Treasurers, May 20-21, San Martin, CA Accessing AI, using it, feeding it data. Insights on AI came from members at companies both deploying and developing AI tools. Some have access to a variety of those solutions while others are limited to either Copilot in Microsoft environments or Gemini at companies using Google Workspace. One participant at “a Google shop” said senior leaders have decreed “only Gemini and nothing else for now.” At a company that uses Microsoft, “We’ve just been told you can’t use any third party, you cannot download, you cannot use a personal cloud or ChatGPT,” one treasurer said. He expressed disappointment with PowerPoint presentations made by Copilot. A member with access to Gemini who is about to get a license for Claude said, “The interesting thing will be the cost factor—if the cost ends up being a blocker in the future.” In the same vein, at earlier meetings this year, many members praised Claude but some expressed concern about the expense of employing it at scale for the full spectrum of use cases. AI options built in-house are also on the horizon for some companies. But limited IT resources create a unique set of challenges for the majority that aren’t in the technology sector. Members at companies developing AI shared their unique perspectives with peers. One said the usefulness of the tools is closely tied to the range of data they can access. Their tool can “access everything” and “all that contextual data is the biggest ‘unlock’ for me,” they said. Rather than using it, say, to simply analyze bank data, “I find where I use it more is for all that contextual data that lives out there: Why was a certain decision made six months ago? Well, the tool knows and can find the doc and tell me why.” Members described treasury environments that still rely heavily on spreadsheets and manual processes, particularly for cash forecasting. Many are evaluating how AI agents could automate forecasts, summarize data and generate insights, but acknowledged that poor data quality remains a significant obstacle. Several members said their AI efforts depend on first improving data quality and accessibility. Projects discussed included building finance data lakes, consolidating bank and ERP data, enriching treasury data in Snowflake and upgrading API connections to improve forecasting and reporting. Cash discipline in action. During one session, several members noted that investors increasingly view cash as the most reliable indicator of financial health because earnings can be managed, but cash cannot. Cash, not growth, is now the higher priority for many businesses. As a result, treasury teams are being asked to provide deeper forecasting, greater explainability and stronger working capital discipline. “The shift coming from the overall market is basically going to award companies who are more disciplined about the cash rather than just focusing on growth,” one treasurer said. The session revealed that the nature of forecasting varies by company stage and business model. Some members are focused on short-term liquidity and covenant management, while others are planning around multi-year capital commitments and uncertain future growth. Regardless of circumstance, treasury’s role is expanding beyond reporting into influencing operational decisions, coordinating cross-functional stakeholders and improving working capital outcomes. Members also emphasized that cash culture must be driven from the top. Treasury teams that have successfully elevated cash awareness often received explicit support from senior leadership, enabling them to challenge assumptions, coordinate business functions and drive accountability around cash generation. The discussion concluded with an example of a treasurer who convinced the CFO to give treasury greater access to committees, data and decision-making processes after arguing that treasury could not explain cash flow variations without being involved. Over time, that access led to more accurate, reliable and predictable cash flow forecasts.

NeuGroup for Large-Cap Treasurers and Mid-Cap Treasurers Symposium, May 19-20, New York City Bank relationships need to be earned. Some companies rotate debt capital markets assignments among banks, but one member argued that approach can undermine performance by removing incentives for banks to compete for business. The member, a former banker, recently led a $1.5 billion bond deal, and said companies often end up weakening bank performance by predetermining lead roles. One lesson was the value of creating competition among banking partners rather than relying on fixed rotations. “You lend to me, that's your ticket of entry. Everything after that you need to earn,” the member said. “Every single bond deal is a jump ball.” The member said banks should be rewarded for bringing ideas, providing useful market intelligence and helping treasury teams prepare for transactions long before issuance. In contrast, rigid rotations can encourage bankers to "check out" because they know their role and economics are already determined. The presenter described setting clear expectations that lead roles, economics and future opportunities depend on the quality of coverage. Convertible financing: A 19th-century instrument booming once again. Convertible bonds have been around since the railroad era, but the market is experiencing a modern resurgence. Today, a combination of strong equity markets, elevated volatility and abundant investor demand is driving another surge in issuance, according to Raj Imteaz, partner and head of convertible and equity derivatives advisory at ICR Capital, the meeting’s sponsor. “This year we’re on pace for $140 billion,” Mr. Imteaz said, referring to U.S. convertible issuance, after noting that annual volume had fallen as low as $30 billion to $40 billion in the 2010s. “Just so you understand how massive it is.” The session explored why converts have become increasingly attractive in a higher-rate environment. Mr. Imteaz discussed how issuers can achieve substantially lower coupons than in traditional debt markets while preserving flexibility around dilution through tools such as capped calls. Discussion also covered sizing considerations, settlement methods and common execution pitfalls. One key takeaway was that successful convertible transactions begin long before a company is ready to issue. He encouraged members to understand the major structural decisions in advance so they are not learning the product under the pressure of a live deal. While market windows matter, preparation and education are what ultimately drive successful outcomes. Inventory finance moves into the mainstream. Inventory finance is rapidly gaining attention among large corporates looking to improve working capital and strengthen supply chain resilience, said Chris Howarth, CFO of Falcon Group. Mr. Howarth explored how supply chain disruptions, geopolitical uncertainty and tariff pressures have pushed companies to carry significantly more inventory than they did before the pandemic. This has the potential to strain commercial relationships as counterparties may get into a sort of “wrestling match” over who holds the inventory on their balance sheet. That shift has increased interest in solutions that allow companies to preserve liquidity while maintaining larger safety stocks. Inventory finance, he said, is emerging as the “third pillar” of working capital optimization alongside receivables and payables programs. The discussion quickly moved beyond working capital metrics. Members discussed how inventory finance can help companies maintain larger inventory buffers without committing additional capital, a tradeoff that has become more relevant as supply chains remain vulnerable to disruption. Members also examined practical implementation considerations, including accounting treatment, supplier engagement and the differences between bank-led structures and models involving specialized trading companies. Inventory, Mr. Howarth said, is “a little bit like a hot potato. Everybody wants to eat it, but nobody wants to actually hold it.”

NeuGroup for ERM (May 19, Foster City CA) and NeuGroup for Internal Audit (May 20-21, Palo Alto)

ERM pushes for earlier input in strategic discussions. Many risk management teams are working to move into strategic planning conversations rather than operating primarily as reporting or compliance functions. The problem at some companies is timing. As one member put, “The strategic choices happen, and then ERM gets involved." But at several organizations, ERM leaders are participating earlier in strategic discussions, helping leadership teams challenge assumptions, identify vulnerabilities and evaluate potential disruptions before plans are finalized. These members described efforts to embed risk discussions into annual planning cycles, M&A evaluations, capital allocation decisions and executive operating reviews.

To bring risk perspectives into decision-making when assumptions can still be challenged and plans adjusted, ERM leaders depend on credibility, trust and executive sponsorship. Influence often matters more than formal authority, especially where ERM remains a lean function with only one or two dedicated resources. In those environments, members said that when they frame ERM as a way to challenge assumptions, test strategic plans and identify execution risks earlier in the process, they gain greater access to leadership discussions.

Audit reconsiders what skills are needed. In the IA meeting, members grappled with the implications of AI on the makeup of their teams, repeatedly returning to a fundamental question: What will the audit team of the future look like? The answer is inextricably tied to identifying tasks where AI will play a major role. Examples ranged from AI-assisted workpaper reviews and report drafting to automated SOX testing and emerging agentic tools capable of performing increasingly complex audit activities. While members remain divided on the maturity of specific solutions, many agreed that the long-term direction is clear: audit teams leveraging AI will need different skills than they have today.

That shift is already influencing hiring and talent strategies. One member described broadening recruiting criteria beyond traditional accounting and finance backgrounds to include analytics, engineering and other technical disciplines. Others discussed creating dedicated AI resources, partnering with universities or investing in upskilling programs focused on data, automation and AI tools. Several participants emphasized that future teams will likely blend technical specialists, subject matter experts and auditors rather than relying exclusively on traditional audit career paths.

AI adoption may outpace governance and readiness. AI, of course, was a key topic throughout both of the risk-focused meetings as organizations experiment with new tools and workflows. Members discussed using AI for horizon scanning, survey analysis, audit testing, report drafting, risk assessments and workflow automation. Several teams are already piloting agentic solutions and building internal applications to automate routine activities. As one participant observed, "Everyone is trying and learning."

At the same time, members acknowledged that governance structures are struggling to keep pace. There are lingering questions around ownership and accountability, particularly as AI-related decisions increasingly span legal, compliance, audit, risk, technology and data teams. One participant noted that technology is moving faster than the ability to deploy products and govern them, while another described AI as "a blessing and a curse at the same time" because of the opportunities it creates alongside the risks.

The conversation also highlighted concerns around data governance, model drift, third-party AI tools and agent-to-agent interactions. Several members suggested that existing governance models were built for traditional software environments and may not be sufficient for rapidly evolving AI ecosystems. While organizations are moving at different speeds, the broader discussion suggested that most remain in an experimentation phase—balancing enthusiasm for AI's potential with caution around the risks and controls needed to support it.

NeuGroup for Capital Markets 2026 H1 Peer Group Meeting, May 12-13, Menlo Park, CA Dealing with bond deal dynamics. Members at the meeting sponsored by Wells Fargo delved into the timing, sizing and currencies of investment-grade bond deals amid a flood of hyperscaler debt issuance to fund the AI buildout. More issuers are swapping portions of fixed-rate debt to floating rates and Wells reported increased interest in longer-dated swaps as the yield curve has steepened. One capital markets manager gave peers a detailed description of an offering to fund M&A, notably the decision to “go first” with a deal after a rare, short-lived (three-day) absence of offerings in mid-March sparked by the U.S. war with Iran. One lesson and key takeaway she emphasized: Flexibility is critical amid volatile conditions: Her company “decoupled” USD and euro offerings as European markets took longer to recover. That’s also an example of another lesson learned: “Execution sequencing drives outcomes—not just pricing.” Treasury also dropped plans for a floating-rate tranche in response to investor sentiment. The overarching message: prepare, prepare, prepare. Uncertainty over geopolitical risk and market volatility combined with unpredictable M&A closing timelines means it pays to be ready to act on very short notice. This treasury team increased its revolving credit facility by exercising the revolver’s accordion; secured a large delayed-draw term loan; “and then we started to layer into strategically pre-issuance hedging,” the member said. In the end, she reported a “really strong result,” with a large oversubscription and significantly tighter pricing than initial price talk.

Another issuer shared a similar experience financing a large acquisition. Internal debates centered on whether to delay issuance to reduce interest expense on financing if the deal closed later, but treasury pushed to move earlier because of geopolitical uncertainty. The decision ultimately proved beneficial, the director of capital markets said, as market conditions deteriorated shortly afterward.

Revolvers: sizing, drawing and the CP backstop debate. Revolving credit facilities generated ample food for talk, specifically how to size them, whether to draw on them, and how rating agency requirements around commercial paper backstopping constrain their utility. One member described failing to convince a rating agency to decouple the CP program size from the revolver commitment—not to eliminate the revolver, but to allow CP issuance to scale beyond the revolver ceiling without adding more bank commitments. Other members confirmed they'd hit the same wall. One company resolved it by adding a 364-day revolver to match expanded CP program ambitions. That’s a potential solution, but an expensive one in terms of bank wallet share. On sizing methodology, one member preparing to refinance a revolver is benchmarking against peer liquidity ratios and the sources-and-uses framework that rating agencies use. Another described stress-testing against worst-case quarterly cash flow periods, particularly seasonal windows when payables draw down liquidity. A third took a more philosophical stance: treat the revolver as cheap insurance against volatility and get it as large as the bank group can reasonably support.

To draw or not: One member raised whether periodic draws for operational readiness rather than necessity could help keep the banking syndicate engaged and responsive. The room was split. Most participants said they never draw on the revolver. One company drew during the pandemic. Several noted the decision ultimately comes down to CFO tolerance for the optics, since any draw invites questions from investors and analysts.

One company is building its first revolver from scratch, converting bilateral bank loans into a revolving structure. The member connected with peers who have navigated bank syndicate negotiations from the ground up.

Process automation in the age of AI. The fast and furious advances of AI are inspiring and to some extent prodding treasury teams to probe their processes and decide whether to redesign them, and whether and how to use agents. “You can think about AI as kind of like an automating tool for every individual process that you do,” one member said. “What we're trying to do is figure out if we can redesign the processes themselves in a systematic way that is being enabled by the presence of the new tools. In other words, like a complete system redesign as opposed to automating personal productivity of our teams and treasury.”

Other members echoed that approach. “In treasury, we're looking at how can we introduce process automation and AI—because they're different,” one member said. “In big projects, cash forecasting, FX, KYC, bank account maintenance, things like that. And smaller projects, very day-to-day operational tasks the team are doing. I think there's 50 plus projects going within the treasury team alone on automating processes and taking out the manual actions.”

Another member examining processes offered a nuanced view of using agents, emphasizing that token usage—and therefore computing cost—should be taken into account. “We don't necessarily want to go the route where we automatically apply agents to every single process we are doing, just thinking about the tokens we might use; so we're thinking first, how can we automate the process and then apply agents on top of it?” he said.

Beyond tokens, though, lies a view of what requires computer coding (made easier by AI) and what requires AI. “I like to look at my processes by [whether] I need a deterministic solution or something more intelligent. If it’s really deterministic in nature, a Python code is the best way to go, in my opinion, because every time I run it, I know I have exactly the same results.”

“For instance, the FX settlement process, it's running the same way every single time. So I don't need an AI agent to come in and do my settlement process if I have a Python code that can do it exactly the same in a very automated way.”

By contrast, he said, “If I have something more analytical, let's say some kind of analysis where I want to have deeper insights, maybe also brainstorm like with AI, I think that's where AI agents really and all kind of LLMs come in handy.”

NeuGroup for Regional Bank Treasurers 2026 H1 Peer Group Meeting, May 6-7, New York City Sponsored by Morgan Stanley Liquidity management: a real-time discipline. Regional bank treasury teams are sharpening their focus on liquidity visibility and deposit governance, as lessons from the 2023 Silicon Valley Bank crisis remain fresh. Across the meeting, members described investments in intraday monitoring capabilities, depositor analytics and escalation frameworks designed to identify changes in customer behavior faster and respond more decisively. Several banks discussed enhancing dashboards to track deposit concentrations, ACH activity and uninsured balances in near real time, while others highlighted efforts to improve communication protocols between treasury, business lines and executive leadership during periods of stress. One member said treasury teams are now expected to “know where the money is moving during the day, not after the fact,” while another described liquidity management as “much more operational than it was even two years ago.” Members also noted that regulators and boards continue to expect stronger governance around contingency funding planning, stress testing and liquidity assumptions, further reinforcing the need for more dynamic monitoring tools and processes. Flexibility drives balance sheet strategy. Portfolio and balance sheet discussions reflected a continued move away from strategies centered primarily on yield optimization toward approaches emphasizing flexibility and optionality. Members described reassessing investment portfolios, funding structures and hedge strategies through the lens of liquidity resilience, capital impacts and adaptability under changing market conditions. One member said treasury teams are increasingly evaluating investments based on “how quickly you can reposition” under stress scenarios, not simply projected yield. Several treasury teams discussed maintaining greater balance sheet flexibility by prioritizing securities that are easier to pledge or hedge while carefully evaluating AFS and HTM positioning. Others highlighted ongoing restructuring efforts aimed at improving future earnings capacity without compromising liquidity or capital objectives. The conversations underscored how regional banks continue balancing pressure to improve profitability with the need to remain agile in uncertain rate and funding environments. Treasury teams increasingly see optionality itself as a strategic asset, particularly as deposit competition, regulatory expectations and interest rate volatility continue to influence balance sheet decisions. Treasury AI efforts shift toward practical applications. Artificial intelligence discussions across the meeting focused less on experimentation and more on practical workflow improvement. Members shared examples of using AI tools to support forecasting processes, automate reporting tasks, streamline reconciliations and improve preparation for presentations to executives and the asset-liability committee. Several banks described early pilots involving Copilot, Power BI integrations and secure internal AI environments designed to help treasury teams reduce manual work and improve efficiency. One participant said the most successful projects are “the boring ones that save people time every day,” while another noted that treasury teams are now under pressure to “show actual productivity gains, not just interest in AI.” At the same time, participants consistently emphasized that governance remains the primary constraint on broader adoption. Data security, model oversight and approval processes continue to shape how quickly banks can scale AI initiatives.

NeuGroup for Tech Treasurers 2026 H1 Peer Group Meeting, May 6-7, Half Moon Bay, CA

Sponsored by Standard Chartered

ROI: AI agent investment recommendations. The treasurer of a large technology company and a tech specialist working with her team demonstrated how AI agents can move beyond productivity gains and directly improve financial outcomes. The treasurer described an investment agent designed with Concourse trained on treasury policy, duration targets and portfolio parameters. By using some of the recommendations generated by the agent for “bond rotation” trades, the team has generated an incremental gain in net interest income over the next two years of $2.5 million. “So positive NPV, positive ROI trades, that shows the value in terms of what the agent's doing in 10 minutes every morning versus an analyst looking at a Bloomberg screen” for up to 90 minutes,” she said. In addition, treasury also deployed AI to monitor global cash pools and alert analysts when excess balances could be invested. That initiative has generated “another $500,000 of incremental interest income annually,” the treasurer said.

Building an AI-ready workforce. This company’s determination to “move the needle” on AI initiatives with direct P&L impact or risk mitigation is clear from the CFO’s mandate on finance function hiring of interns and early-in-career AI analysts: only recruit candidates with quantitative, data analytics and AI backgrounds. More broadly, the treasurer said, “We are recruiting across the entire company where it is expected and mandated to include AI questions as part of your interview.” Employees are now required in self-assessments to describe how they are using AI. The treasurer described a mandatory AI training day last December that will be repeated next month. She said treasury had demonstrated enough “momentum” and “traction” with AI projects to win approval for a dedicated AI analyst headcount. She also described a “build and buy” approach in which the team develops some tools internally while also experimenting with outside AI-agent vendors and platforms.

Another emerging best practice: encouraging AI champions who set a high standard for experimentation and achievements. This company has a quarterly CFO organization all-hands meeting where someone receives a monetary recognition for being an AI champion. “It's really the recognition,” the treasurer said. “We're really trying to do as much as we can to push forward the fluency, the leadership and that mindset.”

NeuGroup for Life Sciences Treasurers 2026 H1 Peer Group Meeting, May 6-7, New Brunswick, NJ

Sponsored by Societe Generale

Foreign financing structures draw interest from life sciences treasury teams. Several life sciences treasury teams discussed using foreign financing entities to fund M&A, diversify investor access and create additional capacity to efficiently issue lower-cost non-US-dollar debt. One member described using an overseas financing entity to issue euro debt that helped refinance higher-coupon debt, free up roughly $5.5 billion for business development and generate about $85 million in interest expense savings. The structure also diversified a debt stack that had been almost entirely dollar-denominated and accelerated rating agency deleveraging commitments by roughly two quarters.

Much of the discussion focused on the mechanics behind these structures, including the need for financing entities with real local substance: people, operations and activity that make them more than shell companies. One member said the issuing entity also needed to sit under a parent with enough operational scale and capacity to repay the debt, even with a corporate guarantee in place. Several also noted that once the debt is issued through the structure, companies may need to keep refinancing it over time rather than quickly unwind the arrangement.

The accounting complications proved just as challenging for some as the financing itself. One member cited an example in which they did not want to change existing FX accounting policy to accommodate a transaction, forcing treasury to convert euro proceeds back into dollars through a spot trade that created roughly $7 million of P&L volatility as currencies moved during the trading day.

Brainstorming next-wave AI projects. Treasury teams used the meeting to compare notes on larger AI projects that reach well beyond drafting emails or summarizing documents. Much of the discussion centered on whether AI can be connected directly into treasury infrastructure—including SAP, Bloomberg, Kyriba, Clearwater and internal data lakes—to automate processes that are still heavily manual.

One member described building an AI-driven process for FX exposure gathering and cash flow hedge analysis, replacing spreadsheets and manual data collection with a system tied into treasury data feeds. Another said the company wants AI continuously monitoring counterparty news against positions held in Clearwater that automatically identifies potential risks, exposures and recommendations. Others discussed projects to use AI to automate end-of-day cash positioning, trigger investment proposals based on bank limits and accelerate DSO monitoring by identifying lagging countries as soon as consolidated financials close rather than weeks later.

Several examples focused on connecting AI to existing systems. One member said Copilot reduced a recurring SAP-based investor relations support process from roughly a day and a half to about 10 minutes by extracting, formatting and organizing data automatically. Another treasury team used ChatGPT to cut a Kyriba transaction-tagging process from about 40 hours to two.

The discussion repeatedly returned to the same challenge: how to move from isolated experiments to scalable treasury infrastructure that teams can maintain long term.

NeuGroup for Global Cash and Banking 2026 H1 Peer Group Meeting, April 29-30, Las Vegas

A deeper look at bank fees. Members said bank fee analysis, a perennial treasury issue, starts with a basic problem: many teams do not have clean, reliable visibility into what banks are charging, what the charges mean or whether they match contracted pricing. Several described going back to basics: manually reviewing statements, reconciling charges and cleaning up data before attempting to automate.

One member said a global bank fee review uncovered nearly $500,000 in savings, including a Mexico payroll process that was generating a $1 fee for each confirmation email sent by the bank. Another discovered a bank account transaction fee error that led to excess charges of about $100,000 a month that had gone unnoticed. Others said even a simple pushback to banks produced results, including one company that achieved a 40% reduction in fees after initiating discussions.

Some teams are using AI to interpret the data. One member has a Gemini agent that reads bank statements, maps charges to AFP codes and compares them against master fee agreements, flagging exceptions for review and follow-up. The approach is still evolving, but it reflects a broader shift: once the baseline is established, teams are looking to automate the recurring review rather than repeat the same manual analysis each month.

Embedding AI into workflows. This same pattern is showing up beyond bank fees: members are aiming AI at the parts of treasury that are repetitive, manual and time-consuming. The focus is inserting AI-enabled tools into specific steps that involve checking, reconciling, documenting or following up. Early results are measured in time saved and cleaner execution.

One member said an agent checks whether expected interest payments arrived from dozens of banks, flags gaps and drafts outreach to banks. Another described using AI to validate FX rates in the company’s ERP and flag discrepancies before they create downstream issues. Others pointed to account-opening support, using AI to pull requirements, draft emails, review bank responses and reduce the back-and-forth involved in gathering documentation.

Treasury teams are starting to rewire around AI. As practical use cases take hold, members said the AI shift is beginning to show up in how teams are structured and what skills are needed. Many described formal coordination across finance or the enterprise, with treasury expected to participate, even if resources and data readiness lag. Others pointed to emerging roles—AI champions, data specialists or embedded technical resources—helping bridge the gap between treasury and IT.

The conversation around the future of treasury is shifting toward what work remains distinctly human. Members raised concerns about junior staff losing opportunities to build judgment if foundational tasks are automated. At the same time, several said treasury roles are requiring skills outside of traditional treasury knowledge, moving toward interpretation, cross-functional coordination and influencing decisions across the business. One participant framed this as treasury becoming more consultative, connecting data to action and helping other teams understand implications. That emphasis on communication echoed a theme from last week’s Working Capital Optimization meeting (see below): storytelling is a critical treasury skill, especially as automation increases the volume of data but not necessarily the clarity of decisions.

Key takeaways from NeuGroup for Working Capital Optimization and NeuGroup for Cross-Border Credit & Collections.

Cash culture catalyst: storytelling. A session at the spring meeting of NeuGroup for Working Capital Optimization sponsored by SAP Taulia that included treasury, procurement, AR and AP leaders underscored that creating a cash culture requires effective storytelling that makes clear to non-finance teams how their day-to-day decisions affect working capital and cash flow as well as DSO, DPO and DIO. “The art of that is just a really big deal,” said one member who has designed courses that connect the dots between the company’s P&L, balance sheet and operating cash flow.

The ultimate goal is to demonstrate how cash decisions flow through to return on invested capital, total shareholder return and the company’s stock price. To do that, finance teams need the ability to tell stories in a way that people “can digest and create action,” the member said. Otherwise, terms like leverage, cash conversion cycle or working capital can sound abstract to sales, procurement and inventory teams.

It’s important not to overwhelm people with data. “You don’t want to have 85 million words on a slide,” she said. Instead, the messaging should answer basic questions and define terms. When teams can see how one day of DSO, DPO or DIO translates into real dollars, they are more likely to change behavior. Cash culture sticks when finance turns numbers into a story people remember and act on.

“I am fortunate to have a few storytelling experts on my team,” the member said. “We talk about how you actually approach a story in order to tell it to people, down to how you storyboard, how you actually pull things together to deliver it. Otherwise, people won’t remember.”

Credit and collections challenge: selling to startups. Among the many topics discussed during the pilot meeting of NeuGroup for Cross-Border Credit & Collections in Dublin, members shared how they manage the risks of selling to startups that do not have a track record of traditional credit metrics. Some participants assess startup risk as a portfolio, not just a single-customer credit decision. They said traditional financial-statement analysis is often insufficient for young AI/cloud startups seeking large, multiyear compute contracts. Instead, they look at funding quality, investor support, internet/news sentiment, customer concentration, geopolitical exposure, banking relationships and available collateral.

They also emphasized risk appetite. Some companies favor macro-level portfolio management: track how risky the startup book is becoming, set acceptable loss pools and decide how much exposure the company is willing to tolerate. Others noted that credit teams do not get venture-style upside, so they cannot simply accept many failures in hopes of one winner. The upside is sales revenue, not equity returns.

Mitigation options included shorter terms, collateral, guarantees, insurance, channel financing, selling through partners or banks taking part of the risk. Several participants said emerging markets often require a different approach, with more use of partners, securitization or bank-supported structures. Software suppliers generally face less downside if a startup fails because they can cancel or suspend subscriptions, while hardware makers may have to repossess equipment that has already been installed.

The strongest theme was balance: Protect assets while enabling strategic sales. One participant framed this as knowing when to prioritize asset protection and when to support revenue. For startups, that means combining automated scoring and external data with human judgment, especially when the exposure is large, multiyear or tied to rapidly changing markets.

NeuGroup for Cash Investments, EMEA Regional Treasurers and Asia Treasurers

KYC AI agent buzz. One member's description of an AI agent that will one day receive and respond to the flood of KYC requests from banks generated visible and vocal interest from peers at the spring meeting of NeuGroup for Cash Investments. KYC, of course, is one of treasury’s most repetitive, inefficient and vexing burdens. For now, the member explained, the agent built by a junior staffer with solid AI skills pulls the KYC data and fills in the forms sent by banks; the staffer emails the forms to the banks. Ultimately, the goal is for the banks to email the corporate’s agent directly for KYC requests. And yes, at some point, bank AI agents will in all likelihood be the ones emailing the corporate’s agent.

Continuous improvements, disruptive events. All of the sessions at the meeting for NeuGroup for EMEA Regional Treasury in Amsterdam addressed “continuous improvements punctuated by disruptive events,” said group leader Paul Dalle Molle. Improvements are usually spurred by organic corporate expansion and the availability of new technologies; the disruptive forces are often M&A and divestitures followed by big changes to corporate culture that must be reflected in treasury.

Two members told peers about complex spin-off journeys leading to the creation of new global treasury organizations suited to evolving needs, including internal growth and ambitious acquisitions. One member’s focus: the creation of a cash culture and the decisions leading to an optimal EMEA cash pool. The other presenter zeroed in on treasury optimization requiring technology decisions and implementation. One key finding from an in-session poll: Members consider their treasury processes highly standardized and reliable but in need of more automation.

Gift City hype. Members at NeuGroup for Asia Treasury cautioned banks and other service providers not to oversell solutions such as Gift City in India, especially with headquarters, before regional treasury has vetted whether they fit the company’s actual structure, funding model and operating needs. Gift City, the special economic zone and financial hub, recently doubled the tax holiday for businesses establishing operations there to 20 years. And while this incentive makes the use case potentially more valuable, several members said the current value proposition remains narrow and highly fact-specific.

NeuGroup Founder and CEO Joseph Neu attended the meeting in Singapore and observed that “this fits into another theme about restructuring legal entities as a strategic role for regional treasurers as part of their business partner responsibilities. To maximize liquidity and facilitate repatriation, business structure and capitalization often needs to be rethought in collaboration with tax, accounting and legal.” He added that one member at the meeting said the structures for rapid investment in new markets are often less ideal for pulling money out of them as the business matures.

Women in NeuGroup (WiNG) Salon Dinner, April 14, New York City

Women in NeuGroup brings together women members for candid peer exchange, professional connection and mutual support across the NeuGroup Network. At an April 14 salon-style dinner in New York sponsored by CIBC, members joined a conversation facilitated by NeuGroup's Julie Zawacki-Lucci on how AI is shaping day-to-day work, leadership and life outside the office. AI: moving from curiosity to workflow. What stood out most for attendees was how quickly AI has become part of everyday work. Some companies are leaning into the shift proactively, like one member’s “AI Tuesday” initiative, which dedicates time for experimentation and learning, with an emphasis on curious exploration and permission to fail. Even inefficiencies, like duplicated work, are being reframed as part of the learning process and an investment in improving AI outcomes over time.

“The AI train has left the station,” one member said. “You’re either on it, or you’re left behind…or worse: under it!”

Claude drew the strongest endorsements among the various LLMs discussed, particularly for Excel-heavy tasks, data analysis and explaining variances. Several members noted it has already become embedded in their day-to-day workflows.

Trovata’s TMS platform was also singled out as easy to implement and intuitive to use, with members noting that both Trovata’s native AI capabilities and Claude have helped teams identify and interpret exceptions more quickly.

One member said the solution had improved payment matching accuracy from 80% to 96%.

Artificial intelligence, human judgment. Much of the discussion turned to what may be lost as AI takes over more of the work that once helped junior employees build judgment, technical fluency and confidence. Members said efficiency gains are real, but so is the risk of developing professionals who can prompt effectively without fully understanding the underlying work. “AI is helping people work faster, but not necessarily smarter,” one member said.

That concern centered on whether younger team members will be able to challenge faulty outputs, detect hallucinations or explain their reasoning if foundational tasks are increasingly automated away. Members returned repeatedly to the idea that human judgment remains the real control layer and that AI should be treated as a starting point, not the answer.

Event Week: March 23-26, 2026, San Diego

A packed week of insights shared at Qualcomm’s headquarters kicked off with members of NeuGroup for Foreign Exchange engaging in proactive, outcome-oriented discussions on AI, forecasting, hedging and market volatility. The FX meeting began the same day as the publication of the NeuGroup Guide: FX Hedging, a timely companion to conversations about how FX teams are refining programs, improving forecasting and rethinking processes in a more uncertain market environment.

Qualcomm then hosted a joint meeting of NeuGroup for Mega-Cap Assistant Treasurers and NeuGroup for Large-Cap Assistant Treasurers where members dug into many of the same priorities—including AI, liquidity, capital allocation and M&A execution—from a broader perspective. The four days offered a concentrated look at how treasury teams are trying to improve execution while preserving flexibility.

AI is getting practical, fast. AI was the most persistent theme, with conversations ranging from high-level planning on transformational projects to automating small, everyday tasks. The discussions ran through many practical use cases: summarizing materials, building dashboards, reviewing policies, generating analyses, creating lightweight agents and accelerating work that used to sit in Excel or email chains.

FX members narrowed into this, with a dive into one member’s heavily automated hedging process, primarily through so-called vibe coding using LLMs. Another FX hedging manager described using AI assistants to condense reports and pull together analysis more quickly.

Forecasting is the finance nerve center. Forecasting was at the heart of conversations all week, especially among FX managers focused on what better forecasts enable and what weak ones still limit. The discussion ranged from making hedging programs more dynamic to improving exposure gathering and reducing manual work.

It became clear how much structure sits behind mature programs, with one presenter describing a process built around long-dated forecasts, regular review forums and confidence scoring tied to both data and judgment. Forecasting still depends on the quality of inputs as much as the tools used to analyze them.

Managing liquidity amid uncertainty. In discussing the macroeconomic outlook, one session leader from FX meeting sponsor Chatham Financial said, “There’s no longer a base case,” as the war in Iran adds another layer of uncertainty. In the AT meeting, an expert from sponsor TD Securities said the conflict was adding to inflation concerns through oil prices, making markets less confident about rate cuts and more sensitive to how long the disruption might last. As he put it, the “fulcrum for the market is, how long will this last? And how long will oil prices stay high?"

That uncertainty fed into more practical questions around liquidity and capital allocation. One member is spending “lots of time” balancing lower cash, strategic investments and capital return, calling it “capital allocation balanced with appropriate levels of liquidity.” Another said 2026 “looks and feels different,” and her team is working through the “bookend of scenarios, plan of action.”

M&A gets harder when treasury joins late. Three ATs shared how they managed the complexity of recent M&A deals. One compared some separations to a “messy divorce.” The discussion highlighted how deals that look straightforward at a high level can become far more complicated once treasury gets into bank accounts, debt, leases, systems and legal entities.

The biggest takeaway from the session: treasury needs to be involved early. By the time teams are asked to help execute, many of the decisions that shape cash, funding and bank structure have already been made.

Liability management gets a practical review. During a Spotlight Session in the AT meeting, Loop Capital Markets focused on liability management as a set of usable tools rather than a niche capital markets exercise, with discussion centered on upcoming maturities, above-market coupons and ways to smooth out debt towers before they become a problem.

One member said Loop’s analysis was compelling not just for interest-expense savings but also as a way to show balance sheet discipline, even if the final call still depends on cash levels and refinancing plans.

NeuGroup for Mega-Cap Treasurers H1 Peer Group Meeting. March 17, 2026. New York City. Using stablecoins to free trapped cash. One treasurer almost succeeded in using stablecoins to move trapped cash out of Venezuela and volunteered to share the details with peers. (A local stablecoin exchange house license had expired, thwarting the plan for now.) While some members would love to use stablecoins to free cash trapped in countries including Russia, others voiced concerns about regulatory uncertainty, tax implications, accounting treatment and operational readiness. One treasurer is interested in establishing a working group to resolve resolve stablecoin challenges: “My CEO is asking me all the time, ‘Why aren’t we using stablecoin?’” Tech, talent and treasury 2030. Technology, talent and strategy are converging in a broader treasury transformation toward a “2030” vision. AI adoption—particularly in cash forecasting and automation—was a recurring theme, but participants stressed that success depends on foundational investments like clean data and robust data lakes. TMS decisions remain complex, with no single system meeting all needs, reinforcing the importance of defining future-state requirements before selecting tools. “I find it very challenging to go through the treasury management system search right now as we also are changing how we do things with AI,” one treasurer said. At the same time, talent management remains critical: teams must evolve their skillsets while navigating retirements and organizational change.

Working on working capital. Working capital management emerged as a shared priority, with companies focused on both payables and receivables optimization. Several participants discussed extending payment terms (often toward 90 days), though success varies by vendor leverage, industry norms and geography. Members discussed supplier financing and dynamic discounting, with mixed adoption and questions around ROI and operational burden. For receivables, efforts included accelerating collections through ACH, direct debit, and reducing check usage. Virtual cards are seen by some as a practical tool delivering both rebates and extended payables. A recurring challenge is data quality and cross-functional alignment, as procurement and business teams often prioritize P&L over cash flow outcomes.

NeuGroup Quick Takes: 2026 First-Half Peer Group Meetings: DB & DC

How members are putting pension surpluses to work and getting data ready for AI-driven investment analysis.

Popular

Related