How has the year begun from an FX market perspective?
Ben Weinberg.
Weinberg: It’s been a very exciting start, honestly. We came into the year thinking volatility might stay low and conditions would be relatively calm, but it’s been the opposite. There are a lot of themes already driving activity, like tariffs related to Greenland, or trading partnerships with China, Japan headlines and JGB dynamics and even the macro uncertainty we’ve seen around political risk and shutdown narratives.
So there’s been much more movement than we expected even just a week or two ago, and that has created a very active backdrop for clients as we begin the year.
Beyond the immediate macro picture, what are the biggest trends shaping client priorities in FX execution right now?
Weinberg: One theme is the growth in FX algo adoption. Advanced market tracking algo strategies along with innovative in-flight analytics are giving clients the confidence to use algos during high volatility periods of time. If we look at the high volatility period around tariffs in 2025, we saw a higher percentage of our overall franchise using algos. NDF algos are another big area of growth. I’ll admit, a few years ago I wasn’t sure the liquidity environment would support it. But the market has evolved rapidly new venues have emerged, internalisation has grown, and clients are now actively seeking solutions in that space.
Another theme we see is client demand for better tools to guide their execution decisions. Whereas in the past, the idea of curating their own liquidity venues was in vogue, clients today are too busy and rely heavily on their liquidity providers to ensure the algo ecosystems have strong controls, and can be tailored to a client’s specific execution needs.
There’s also a real shift from wanting endless features and functions toward wanting smarter outsourcing, clients asking banks to help manage complexity across liquidity venues and aid with execution choices.
Gholizadeh: I’d agree. Across both systematic and real money clients, demand for stronger analytics has accelerated. Clients want better tools pre-trade and post-trade, not just to review performance, but to actively shape execution decisions.
That applies across client types: sophisticated systematic firms, traditional asset managers, and everyone in between. The focus is on measurable improvement, market impact, markouts, speed, and control.
Ben, you mentioned real money clients adopting new approaches. What does that look like in practice?
Weinberg: We’re seeing even typically conservative real money firms explore more innovative execution methods. From clients shifting from time-based algorithms to more sophisticated market tracking algorithms. As well as conditional orders, that weren’t widely used historically, but are increasingly being adopted to manage volatility, and maintain control in fast-moving markets.
There’s also a much stronger focus on quality of fills, internalisation versus fills on liquidity venues. Clients want to understand how to minimise their market footprint.
When we last spoke, the conversation was about technology advancements improving liquidity access. What are the major strides JP Morgan has made on this front over the past year?
Gholizadeh: A key development has been the launch of what we call Adaptive 3.0, the third generation of our flagship Adaptive algorithm.
The central innovation is dynamic execution speed. Adaptive 3.0 adjusts urgency and pacing automatically depending on market conditions. It leverages a simulation framework where we replay client orders across different urgency factors using historical market data.
We look across volatility regimes, liquidity environments, and currency groupings to determine what speed of execution generally delivers the best outcome – balancing market impact, markouts, and the client’s need for timely completion.
Importantly, we do this continuously, typically looking at the most recent six months of history, so the optimisation stays current.
And how does customisation fit into that?
Reza Gholizadeh.
Gholizadeh: That’s one of the biggest shifts. Historically, speed customisation required more manual tuning. Now, Adaptive 3.0 allows this optimisation to happen dynamically, and in some cases specifically for an individual client.
For clients with sufficient flow, we can isolate their trading behaviour and identify optimal execution speeds tailored to them. That may differ from broader market averages, because every client’s execution objectives and footprint are unique.
So clients can increasingly outsource speed decisions to JP Morgan, knowing the algorithm is adapting based on rigorous optimisation.
Weinberg: And that really changes the dialogue with clients. The conversation used to be “let me test this.” Now it’s “how do I optimise execution across my entire workflow?”
Clients are under more pressure than ever to demonstrate best execution, and they want to continuously refine.
We can take a client’s flow from last year, run it through our simulation and optimisation tools, and show: if your priority is beating arrival mid, these are the right algos. If your priority is reducing market impact, here’s the approach. It becomes much more tailored and consultative.
Beyond Adaptive 3.0, what other innovations are you bringing into the algo ecosystem?
Gholizadeh: Another major area has been enhancing how clients interact with our internal liquidity.
We’ve improved our internal-only algos and introduced greater flexibility around spread capture levels. Clients can now have more control over how they access JPMorgan franchise liquidity, which is particularly valuable for those prioritising minimum signal risk, minimal footprint, and reduced market impact.
That’s been a strong solution for clients seeking discretion and efficiency.
How are clients behaving during periods of market stress or high-volume events?
Weinberg: Historically, during high-volume periods, many clients would shift immediately to risk transfer – voice trading or streaming prices – just to get it done quickly.
But what we saw during major volatility events like Liberation Day last year was a substantial increase in algo usage. Clients were comfortable executing electronically even through extreme conditions.
That speaks to two things: the advancement of execution algorithms, and the growing confidence clients have that algos can perform even when markets are moving sharply.
Are regulatory or market structure shifts influencing algo development as well?
Weinberg: Definitely. One area is embedding more market-event awareness into pre-trade tools.
Most participants know the obvious events like non-farm payrolls, but FX is full of less predictable volatility points, especially in less frequently traded currency pairs.
We’re refining tools that help clients anticipate those conditions and incorporate volatility and volume profiles into execution planning.
Gholizadeh: We’ve also continuously improved the volatility and volume profiles used inside our algos around specific events. Those profiles are directly correlated with performance, so it’s a key area of ongoing development.
How else are your evolving your algorithmic capabilities in light of new technology advancements?
Gholizadeh: One practical application is real-time alerting. We’re exploring how newer AI and LLM-based tools can enhance monitoring of algo execution.
For example, if an order is taking longer than expected, or if execution cost deviates from what models predict, AI-driven alerting could allow sales and trading teams (and clients) to respond immediately. The goal is to optimise execution not just after the fact, but in real time.
Weinberg: And clients are demanding more data than ever. They want deeper insight into what they’re trading, liquidity conditions, execution outcomes – all of it.
We’re providing extensive analytics already, but we’re also looking at new AI-enabled ways to deliver that information more efficiently and more intuitively.
Finally, what’s on the horizon for 2026? Where are the next frontiers?
Gholizadeh: Expansion into NDFs is one of the major focus areas this year. Many of the sophisticated tools and features available in spot FX have not fully migrated into the NDF space yet, and that’s a key development path.
We’re also thinking cross-asset. There’s an opportunity to bring successful FX execution frameworks into other asset classes – rates, commodities, and beyond. That’s something we expect to speak more publicly about later this year or into next.
Weinberg: I’d add that a lot of our innovation comes directly from client engagement. Clients come to us with specific workflow needs, and we look at how to incorporate that into our broader offering.
The execution landscape keeps evolving, and we’re committed to staying at the forefront, whether through Adaptive 3.0, internal liquidity enhancements, AI-driven monitoring, or the next generation of data and analytics tools.
