What’s driving innovation in FX liquidity management?

Technology and the move from voice to electronic trading have left their indelible marks on the FX market and on liquidity management specifically. Technology has undoubtedly made the market more efficient and transparent but it has also made it more fragmented and more complex. As a consequence, liquidity management has become a much more complicated process, requiring much greater use of technology and third-party services. 

Participants are looking for more diverse liquidity sources to ensure they are both competitive and also compliant with best execution rules and the FX Global Code of Conduct. This has led to the growth of the liquidity management service providers who have added advanced analytics to their offerings to help clients asses execution quality and manage fragmented liquidity more efficiently. 

But what are the next steps in liquidity management? Is there a role for AI and other advanced technologies? And what will greater use of these tools do to the relationship between liquidity takers and makers and the various intermediaries that stand between the two? 

Fragmented market

Liquidity management is becoming more and more complex due to a mix of factors, says Andrea Sanna, head of liquidity management at London-based FX broker Alp Financial. “The FX market is getting increasingly fragmented, with liquidity spread across multiple venues, making it harder to aggregate efficiently. This fragmentation leads to inefficiencies, price discrepancies, and challenges in ensuring the best execution,” says Sanna. 

“At the same time, regulatory pressures are pushing firms to diversify their liquidity sources to comply with best execution rules and the FX Global Code of Conduct. This adds another layer of complexity, as firms must navigate reporting requirements and demonstrate that they are sourcing liquidity in a competitive and transparent way,” he says. 

“On top of this, the growing reliance on electronic trading and algorithmic execution has increased technological complexity. Firms are investing heavily in advanced analytics to measure execution quality and optimise liquidity management strategies. Managing fragmented liquidity effectively now requires improved real-time data processing, AI-powered decision-making, and smart order routing systems, all of which demand significant technological investment,” says Sanna. 

Liquidity providers are no longer just market makers, they are evolving into technology and data providers as well, says Sanna.  “The growing demand for transparency means they need to offer deeper insights into how liquidity is sourced and priced. Clients expect detailed reporting and execution analytics to ensure they are receiving fair and competitive pricing.There is also increasing pressure to improve execution quality. Firms on the buy-side are becoming more demanding, and they ask for lower slippage and reduced market impact. As a result, liquidity providers are investing in providing real-time analytics and algorithmic execution tools to stay competitive,” says Sanna. 

“Additionally, as the market shifts toward multi-asset trading, liquidity providers are expanding beyond FX to offer liquidity across multiple asset classes. This requires them to develop more flexible and adaptive pricing models. The industry is also seeing a push for more customised liquidity solutions, which has led liquidity providers to offer tailored pricing streams and execution models that align with individual trading strategies,” says Sanna. 

At the heart of this transformation is technology, says Sanna. “Liquidity providers that fail to embrace automation, AI, and machine learning risk losing relevance as their competitors use these tools to refine pricing, improve risk management, and provide clients with better execution outcomes. AI and machine learning are revolutionising FX liquidity management by making execution more precise, automated, and predictive. One of the biggest impacts has been in market forecasting. AI-driven models analyse vast amounts of historical and real-time data to anticipate liquidity conditions and price movements. This allows liquidity providers to adjust pricing dynamically, ensuring better execution,” says Sanna.

Smart order routing is another area where AI is making a difference, says Sanna. “With liquidity now highly fragmented across multiple trading venues, AI-powered algorithms help route orders to the best available liquidity pools, reducing costs and improving fill ratios. Risk management is also being enhanced by machine learning. These models can continuously scan market conditions and adjust risk exposure in real time, helping liquidity providers react quickly to market shifts and mitigate potential disruptions.”

Additionally, automated market-making strategies are becoming more sophisticated, says Sanna. “AI is being used to refine bid-ask spreads dynamically based on order flow, volatility, and liquidity conditions. This results in more stable liquidity provision and better pricing for clients. As AI continues to advance, liquidity management is becoming more efficient, data-driven, and responsive, enabling firms to navigate market complexities with greater agility.”

The relationship between liquidity providers and their clients is shifting from a simple transactional model to a more data-driven, collaborative partnership. One of the key factors driving this change is the increasing demand for transparency. Clients want more insight into execution quality, pricing, and the true cost of liquidity. As a result, liquidity providers are now offering more detailed execution analytics and post-trade reporting to meet these expectations.

Another major change is the way firms evaluate their liquidity providers. Instead of simply looking at spread costs, they are analysing execution performance, market impact, and the consistency of liquidity provision. This has led to increased competition among providers to deliver not just liquidity, but also smarter, more adaptive execution tools.

Multi-venue and multi-asset liquidity are also playing a role. As firms look to diversify their liquidity sources and trade across different asset classes, providers must adapt by offering cross-asset and multi-market liquidity solutions. This has led to a more customised approach, with providers tailoring their offerings to meet the specific needs of different trading strategies.

These shifts are creating a more dynamic and competitive landscape, where liquidity providers must continuously innovate to stay relevant.

While size has historically been an important factor in liquidity provision, it is no longer the sole defining characteristic of a top-tier provider, says Sanna. Technology, execution quality, and transparency are becoming even more crucial.

“Firms are prioritising providers that can deliver high-quality execution with minimal slippage and market impact. Clients want to understand how pricing is determined and whether spreads are fair. Providers that can offer robust reporting and post-trade analytics are gaining an advantage. Customisation and flexibility are also becoming essential. Instead of rigid liquidity solutions, firms are looking for providers that can tailor pricing streams and execution models to match their specific needs. In this evolving landscape, investing in emerging markets and innovation, while keeping the business costs lean, are proving to be more valuable than sheer size alone,” says Sanna.

The future of FX liquidity management will be shaped by AI-driven execution, real-time liquidity optimisation, and seamless multi-venue integration. Advanced execution algorithms and predictive analytics will play a key role in optimising liquidity sourcing and reducing market impact, says Sanna. 

“Machine learning will enhance risk management by dynamically adjusting liquidity provision based on real-time market conditions. Firms will also demand deeper data insights, leading to greater use of analytics dashboards that provide real-time execution performance metrics. In this next-generation environment, liquidity providers will need to offer a combination of cutting-edge technology and customisation to stay ahead in an increasingly complex market.”

Regulatory and competitive pressures

One complicating factor influencing liquidity management is the potential for pricing feeds to be contaminated while passing from price maker to price taker, says John Stead, Director of Sales Enablement, Strategy, and Marketing at smartTrade.

“For example if the feed passed through a venue that charges the maker then inevitably the maker will have to make a call if they will forgo some potential profit and leave pricing unchanged; add some additional markup to compensate; stop pricing via that venue or take other non-price action such as reducing depth or increasing sensitivity of last look controls. Only a transit route such as direct bank connectivity provided by smartTrade or a venue like FXSpotStream will give clean pricing feed upon which one can build an accurate book and the basis for best execution,” says Stead. 

A combination of regulatory and competitive pricing pressures are redefining participants’ roles in terms of liquidity management, especially well-known market makers that must prove their pricing is legitimate and auditable, says Stead. “A number of years ago we added a significant amount of granular control, logic and audit logs to our last look controls to ensure that if banks chose to apply such checks on flow that it was done in a manner entirely consistent with their FX Global Code of Conduct obligations and any service declarations made to clients. We saw that clients needed to extract more and more data via APIs to record how client pricing streams had been build and orders filled both to allow them to improve the service they offered but also to enable them to prove best execution and best pricing was followed at all times,” says Stead. 

On the competitive pricing side, smartTrade’s market-making clients are looking in ever more granular detail to differentiate their pricing, says Stead. “Once upon a time it was enough to adjust pricing by client tier, pair, tenor and size. Those days are long gone for the market making clients looking to win new business.”

Now, says Stead, the demand is for more tailored liquidity and pricing. “The competition might result in tighter spreads but that is a well-trodden path.  Other aspects of pricing that can be adjusted to give extra value include our maker bank clients offering pricing with longer hold times for selected end clients where due to some underlying business reason they need a fixed price to check the economics of a deal, of course makers offering larger sizes is on the rise again as an alternative to clearing large amounts with algos. On the payments side of the smartTrade business we see banks being more competitive when pricing cross border payments in an attempt to provide a better all-round solution for clients not just hedging risk but also paying bills.”

Next generation technologies such as artificial intelligence (AI) and machine learning (ML) are also playing their part in enabling FX liquidity management providers to develop more precise and dynamic tools. While a number of providers are only a few years into the release of their AI-driven services and tools, they are already ahead of the conservative approach taken by banks when it comes to offering automated pricing controls, says Stead.

“Our clients are very happy to see us investing heavily in such technologies but generally they see more immediate opportunities in other areas such as post trade service, increasing general automation, higher levels of internalisation and new services such as options,” says Stead.  “Where we do see most interest is in terms of allowing clients to use our inbuilt algos to configure automated risk book sensitive skews to selected clients or groups.  By combining various algos in the platform many clients have been able to raise their internalisation rates, gain more flow against their peers and monetise their flows more efficiently than before.”

There has also been an evolution in the relationship between FX liquidity providers and consumers, says Stead. “People are more aware of the multifaceted and interconnected nature of the market than ever before.  A bank is generally both a maker (provider) and a taker (consumer). If regulations and balance sheet allows, they have many more opportunities to monetise their flows than ever before thanks to technologies provided by vendors,” he says.

“They are also aware of how their actions may impact the flows they receive and how to educate their own clients to behave when interacting with them.  There will always be those who think that they can get away with double hitting liquidity providers or trying to surreptitiously sweep a full amount stream but they are quickly caught for the most part,” says Stead. 

When it comes to selecting FX liquidity providers the volumes do matter, says Stead. But it is a more multifaceted issue than simply scale. “There is plenty of academic research that where there is an asymmetry of information in the FX market participants who have access to more flows have a better understanding to the true value and direction of any given currency, or to put in in a less academic phrasing clearly size does matter!” says Stead. 

“But this is a multifaceted market and there are many other factors.  An LP will often have a holistic relationship with their clients. FX spot, the most competitive product, will be just one product transacted with clients. Maker LPs that can offer multiple services for their clients will have more chances to retain clients’ business and when they offer benefits that go beyond a simple haircut on an already tight spread,” says Stead. 

While the vision for next-generation FX liquidity management includes AI-driven intelligence, real-time visibility, and seamless multi-venue integration, the reality is that widespread adoption within banks may be more gradual, says Stead. “Recognising the inherent complexities and regulatory hurdles, banks, traditionally slow to embrace new technologies, will likely proceed with cautious implementation. However, vendors are poised to play a crucial role, actively exploring and supporting these advancements. They will provide adaptable solutions, bridging the gap between cutting-edge technology and banks’ operational realities, offering modular deployments and progressive integration of AI, enhanced analytics, and streamlined connectivity,” says Stead.

Furthermore, the landscape will see a blending of traditional FX liquidity with emerging sources, such as FX futures, which are gaining traction as a valuable component of liquidity management strategies, says Stead. “Crucially, we anticipate a significant intensification of competition for FX flows, particularly if peer-to-peer buyside matching gains momentum. This shift could lead to a scenario where certain takers are willing to pay a premium for liquidity streams that are not mined for information or used to train data models, seeking to preserve their trading strategies and reduce potential information leakage. This dynamic will necessitate advanced tools and services that can facilitate access to diverse, and potentially more discreet, liquidity pools, even if full transformation takes longer than initially predicted,” says Stead.   

Meeting clients’ goals

Liquidity management in FX has become a multi-faceted issue, according to John Crouch, founder and CEO of Ideal, a provider of liquidity management and other services for institutional trading firms.

“The goal is to create a bespoke experience because clients have different credit profiles and some are natural liquidity takers vs makers,” says Crouch. “A lot of liquidity management is about understanding your goals and monitoring your liquidity over time to make sure that each client’s goal is met.”

Despite his technical background, the development of the liquidity management service has made Crouch a stronger believer in the value of human relationships. “Ultimately it is people that will state problems with your liquidity or your service.”

The demand for bespoke liquidity management was always there but the tools to deliver it were not always available, says Crouch. “In addition, market participants’ needs change. For example, when a new head of FX comes in, the strategy might change – more internalisation for example. Some of those demands might be difficult to deliver but they have become easier to articulate,” he says. 

As an example of how liquidity demands change, the early 2000s and the birth of e-trading for FX saw a number of firms look to drive volume above all else – sacrificing profitability in order to become top of the trading charts. However, this eventually changed when banks realised they had to make money. Similarly, ECNs began to limit access to some high frequency traders because it was costing them money. 

The development of tools has also made the relationship more quantitative than qualitative, says Crouch. “In the old days, when all trading was phone-based, the feedback loop was based on picking up the phone and hearing a client say ‘You suck!’. But now the conversations are based around data on spreads or similar actionable insights.”

“Our tools bring more objectivity to this conversation and elevate the relationship between liquidity makers and takers. We have seen this dynamic across all asset classes,” says Crouch. “So in this way, tools have made the relationship more important. Both makers and takers are talking the same language based on common standards for discussing metrics. It is no longer possible to be blind to the metrics.” 

Crouch is also concerned about the role of AI and ML. “We use the technology to some extent and there are a lot of benefits but also a lot of things to consider. My goal is to help the clients and if AI can help me get to the answer faster, that’s great,” he says.

But when it comes to the cons, there are three that stand out, says Crouch – privacy, privacy and privacy. “We are really thoughtful about clients’ data – it belongs to them and it has to be strictly private. We don’t comingle clients’ data and if they want that, they will have to go somewhere else. Respecting our clients’ privacy is one of our guiding principles and we have designed our architecture around it.”

When it comes to the next developments in liquidity management, Crouch talks about the monitoring and inclusion of more data and not just spreads and trading costs. For example, Ideal is evolving analytics for operational risk. 

“So far clients have prioritised profitability,” he says. However, awareness of the importance of so-called ‘operational alpha’ is growing among FX participants. For example, a company may be sending many orders that are off market and/or short-lived that have little chance of matching, but utilise compute and network resources.

“Ultimately it is about asking the right questions and then finding the best way to solve them. By bringing data to the discussion, we’re able to get to the answers quickly” says Crouch.  

Technical innovation

According to Paul Jackson, joint head of sales EU, UK & LATAM, for Finalto, technical innovation has made liquidity management more competitive. “Advances in technology mean that orders can be processed more quickly, over a broader geographic area. LPs need to implement integrated digital solutions and robust operational strategies that keep latency low and optimise execution performance for all clients.” says Jackson. 

“There’s also pressure from rising costs. Technology and operational costs are rising in a context of FX yields remaining relatively low comparative to other asset classes. This dynamic makes the space even more competitive. The core goal has to be maintaining the highest standards of execution, with responsible risk controls, while remaining competitive.”

Leading liquidity providers are innovating and growing by entering new markets, continuing to develop enhanced tools, and providing service that better suits clients’ needs, says Jackson. “Practically speaking, in our experience, that involves expanding into emerging markets (Mena, LatAm, South Africa), building risk applications and enhanced flow analytics, and offering customised liquidity and sophisticated execution logic.”

At the same time, clients are becoming more sophisticated, supported by advances in technology, and with increased market access, says Jackson. This is fuelling the demand for customised liquidity. “Their needs and expectations are highly variable, and liquidity providers need to be agile to take on different types of flow. For example, the needs of a retail broker might be very different to a systematic hedge fund. A ‘one size fits all’ approach is no longer adequate,” says Jackson.

Relationships between FX liquidity providers and liquidity consumers are also changing, says Jackson. “Over the last decade, brokers started mass aggregating to compress spreads and compete against peers. Increased costs have seen brokers pulling back to work closer with select LPs that can offer a broader suite of products and a more cost-effective approach by centralising risk. Cost, product, service and security are key deciding factors for liquidity consumers. Liquidity providers need to continue investing in core technology to remain competitive and provide the highest levels of service. While speed remains crucial, clients also increasingly expect technologies that offer richer data and help optimise their own operations, including advanced analytics and automation,” he concludes.