Vivek Shankar

Pathways to success: Taking steps to optimise Post Trade FX

September 2023 in Special Reports

Market participants typically view the FX market as technologically advanced compared to other asset classes. While this view fits the front end, back-office processes and workflows present a very different picture. The degree of fragmentation in FX poses operational challenges to back offices, particularly post-FX trade processing. Vivek Shankar investigates.

Institutions have long been aware of the financial and operational advantages of improving post-trade processes. While technological innovations are bringing more efficiency, challenges remain. Many of these challenges rise from the structure of the FX market.

“FX has a long-standing history and broad use by a wide-ranging set of market participants,” says Kah Yang (KY) Chong, Acting Head of GlobalLink Post Trade, TradeNexus. “This led to the development of multiple post-trade processes and workflows catered to meet specific requirements, or based on technology available at that time. These processes are embedded in complex ways with many firms, which has made change more difficult to implement.”

Meanwhile Melissa Stevenson, Product Director FX at ION, points to existing infrastructure challenges within firms. “A lot of  banks still have multiple disparate and legacy applications to manage post trade workflows whereby manual intervention is still required,” she says. “Manual intervention introduces human error and potential for points of failure; this, in turn, increases operational risk. Investments to replace these applications are costly and less of a priority than the investments made in the front office applications.” 

So how are firms using technology to handle this complexity, and what are solution providers offering? But first, it is worth examining where pain points lie and the nature of complexity within post-trade processes.

“The FX Global Code is extremely clear on the relevance of post-trade risks..”

Kah Yang (KY) Chong

Post-trade pain points

Andy Coyne, Founder of CobaltFX, is well-positioned to speak about the origins of complexity post-trade. A lack of common standards is to blame, he says. “The FX industry is paying a heavy price for the unnecessary complexity in the post-trade space. There are currently limited standards and everyone uses their own fragmented siloed technology requiring end-less reconciliations.”

“Just like a modern Tower of Babel, we speak our language and employ armies of translators to solve these problems. To solve this we need to set common standards and allow primary regulated market operators to take control of those standards.”

Vincenzo Dimase, Global Director, Sales Strategy & Execution, FX & Post Trade, London Stock Exchange Group (LSEG), points out that technology hasn’t penetrated every area of post-trade processing due to a few critical issues. “While financial markets have accelerated the automation and adoption of machine learning and artificial intelligence, the speed of this change varies across regions and markets,” he says. “We continue to see pockets of manual post-trade processing activities, a lack of genuine straight-through processing, and several workflows solved by physical printing or wet signature. Further, the fragmentation of liquidity across several pools and venues has led to increased complexities across the front, middle, and back office.”

Chong notes that existing industry initiatives go some way towards reducing this complexity and fragmentation. “More recently, significant efforts through regulatory, industry initiatives and availability of new technologies have made progress in streamlining these processes,” he says. 

When asked, he cites examples such as through FMIs like CLS, increased automation through APIs, and voluntary agreements such as the FX Global Code.

“However,” he continues, “challenges to improve will remain. Given that FX is predominantly traded OTC between two parties, and executed for multiple objectives in complex and bespoke ways, changing post-trade across participants of different sizes, types, and sophistication will continue to take time and remain evolutionary, rather than revolutionary.”

“The Bank of England’s The Future of Post Trade report in 2020 highlighted quite a few pain points we still see across our client base,” Stevenson says. “Data collection as part of the client on-boarding process is not standardised, margin calculations, communications and dispute resolution processes differ across firms, information beyond material economic trade data not exchanged early enough in the trade life cycle and may be inaccurate, incomplete, or not standardised, automated interfaces for data exchange are underdeveloped, and the lack of standardised mechanisms for the collective sharing of data on suspicious activity in real time.”

Coyne points out that some pain points are credit-related, a major focus for CobaltFX. “Some of the pain points in the market lie in the over-allocation of credit,” he says. “Among other things, we handle this by addressing the inaccuracy and the systemic risk of over-allocation of credit. We start by putting credit providers back in control of this precious and undervalued asset. We are making the administration of credit far easier and less expensive. We can also address some of the serious flaws with designation notices.”

“Manual intervention introduces human error and potential for points of failure; this, in turn, increases operational risk.”

Melissa Stevenson

Exorbitant prices for silo-like infrastructure is another pain point that Coyne zeroes in on. “These systems are not designed to be multi-tenanted with logical separation and there is a charging advantage to this,” he notes. “Any operational friction leads to more fees which are in the incumbent provider’s interest to maintain. That same friction also leads to costly errors.”

Echoing Coyne’s earlier view about a lack of uniform standards,  Chong notes that some of these complexities are deep-rooted, making fast-paced changes impractical. “For example,” he says, “asset managers managing a large number of funds have built complex workflows to meet requirements across multiple parties such as their technology providers, counterparties, custodians, asset owners, and regulators in all the jurisdictions where they operate.”

“The flow from order management to execution and post-trade allocations, and settlements work in a closely coordinated fashion which means any improvements to post-trade require a firm’s time and effort to analyse and implement.”

Chong also points out that data quality issues arise from siloed infrastructure, a significant problem given the volume of internal and external interactions systems have. Typically, a firm’s front office platform directs data to a different database from its post-trade system.

“Within post-trade, this can lead to delays in resolving issues such as trade discrepancies, settlement delays, or assist in decision-making processes where data is involved,” he says. “To make sense of data accurately, this may require manual intervention by the user which leads to inefficiencies in the process, and increase financial costs since issues are resolved late.” “Clients rely on our solutions to provide front-to-back workflows from execution to post-trade, with minimal manual touch on the client’s end,” he adds. “This includes innovations within the data space, where an increasing number of clients are looking to automate their data feeds that will allow them to gain insights across the trade lifecycle.”

Stevenson, Chong, and Coyne agree that the costs of common post-trade solutions are out of line with the issues they cause. Coyne lists issues like credit risk charges through balance sheet usage for derivatives, operational reconciliation, the unnecessary duplication of allocations, and the continual movement of cash as reasons for high costs. “We see excessive costs in the maintenance of these fragmented post-trade solutions,” says Stevenson. “Banks must spend on implementing and maintaining custom or bespoke integrations for client on-boarding and connections to various confirmation, settlement, reporting or ERP channels. Management of all these custom/bespoke integrations requires a lot of support overhead.”

“The FX industry is paying a heavy price for the unnecessary complexity in the post-trade space”

Andy Coyne

The next-gen post-trade solution picture

“We are starting to see aggregation solutions aimed at solving liquidity fragmentation,” LSEG’s Dimase says. “A reduction in point-to-point venues connection and a move into one single aggregator for post-trade solves for operational risk as it reduces single point of failures and the number of single connections to risk management and position keeping solutions.”

Coyne believes eliminating silos is a critical step to removing post-trade processing inefficiencies. “The real game changers here are twofold,” he says. “It is the on-ledger mutualisation of all trade life-cycle events and the move away from siloed infrastructure toward a shared middleware that creates credit, middle and back office efficiency.”

“This gives banks the option of using middleware to make existing silos more efficient whilst simultaneously creating the option to increase efficiency and reduce carbon footprint. The ultimate decision is whether to reduce or eliminate the post-trade silo.”

Stevenson believes banks must identify areas where automation can be introduced and leverage APIs. “New services deliver more automated workflows across all aspects of the post-trade process and provide a more consolidated view of complex Post Trade activities,” she says. “At ION, our next generation back office solution provides truly automated “no-touch” post-trade processing from client on-boarding, trade registration through to settlement.”

And is AI present in these workflows, given recent technological advances? “AI and machine learning are being leveraged to automate workflows and provide greater transparency, client analytics, and alternative methods of settlement,” Stevenson says.


A big part of the post trade FX budget has been allocated to comply with industry, regulatory and trade 
repository guidelines

Chong says partnerships and application interoperability are critical and cites a few examples of TradeNexus putting these into action. “We have partnered with CLS to deliver a solution that enables clients to view their trade instructions in CLSSettlement within our consolidated dashboard,” he says. “Likewise, we have also partnered with LCH to deliver a clearing workflow whilst maintaining flexibility for our clients to choose between clearing or remaining bilateral.”

And what about interoperability? “GlobalLink recently launched its GlobalLink Digital (GLD) platform to provide a single front-end for our clients who interact with multiple applications,” Chong says. “GLD unifies the desktop environment so that market participants can create and operate easy-to-deploy and customizable trading and workflow solutions. This allows clients’ trading systems to dynamically communicate with internal and external systems to optimise exception management, trade operations and risk management all based on tailored metrics and data.”

“For example, a client can perform best execution/TCA analytics, execute trades and monitor post-trade metrics all within a single dashboard, where the data flows continuously and consistently through the lifecycle of the trade.”

Notwithstanding these solutions, Coyne thinks more work needs to be done. “First, a common approach to nomenclature needs to be agreed,” he says. “Although currently not widely used, we (Cobalt FX) have a requirement to use LEIs which all entities must have, so we use this. We have simply appended city and country codes to the LEI to add additional necessary detail.”

He further explains that matching needs to be highly flexible for all matching requirement types. “We do this upfront as we believe that potential market exposure always exists before a full match is achieved. There’s no need for confirmation matching to sit downstream.”

Coyne also points out that despite CLS reducing settlement risks, Herstatt risk still exists for non-CLS currencies and counterparties. “The first question to be asked is ‘Do I need to settle given my funding situation and the maturity profile of my portfolio with my counterparty?’ The on-ledger settlement question is mutualised.” 

Dimase chimes in and says that the availability of “… advanced analytics can enhance governance around post-trade processes and allows for the detection, escalation and remediation of issues, and the ability to better manage reputational risk.”

“We continue to see pockets of manual post-trade processing activities and a lack of genuine straight-through processing….”

Vincenzo Dimase

Regulatory impact

The regulatory picture has constantly been changing, and post-trade workflows have witnessed several knock-on effects. SA-CCR, UMR, and the move to settle trades T+1 are impacting post-trade processing. “A big part of the post-trade FX budget has been allocated to comply with industry, regulatory and trade repository guidelines such as  MIFID II, CFTC/EMIR rules re-write, and the FX Global Code,” ION’s Stevenson says. “Specifically for post-trade FX, firms are placing investment in resources and technology to improve STP workflow, record keeping and reporting, data encryption, which need to meet a common standard across industries and organisations.”

Chong explains that SA-CCR has led asset managers to ensure their portfolios’ post-trade positions are as optimal as possible. “As optimization is currently undertaken within post-trade, this has led to greater collaboration between the front and back-office to undertake more frequent novation, compression, or rebalancing exercises,” he says. “As frequency increases, we are seeing demand from our clients on adopting a more automated approach.”

He goes on to explain that TradeNexus has partnered with Capitolis to deliver a solution to clients to automate trade submission, conduct analysis and monitor statuses in real time within a single dashboard. Also, TradeNexus has set up a T+1 Settlement taskforce to review the post-trade processes clients undertake as the post-trade window will narrow for T+0 and T+1 FX. This includes looking at how to improve matching metrics (such as speed and accuracy), and evolve the post-trade landscape with the objective of greater transparency and automation for clients and custodians. 

“This includes bringing in new functionality to alert clients when a settlement indicator mismatch occurs between themselves and their counterparty, enhancements to our rules engine, automation of actions, broker performance reporting and better exceptions monitoring to improve speed and accuracy of post-trade FX,” he says.

Chong also notes that UMR has prompted more firms to choose between clearing their NDF transactions, optimising their bilateral positions to either remain out of scope, or reduce their bilateral margin exchange.  

These changes have occurred against the backdrop of the FX Global Code identifying post-trade risks. “The FX Global Code is extremely clear on the relevance of post-trade risks and contains two Leading Principles (out of six) dedicated to Confirmation & Settlement and Risk Management & Compliance,” Dimase says. 

“The code also suggests how transactions should be promptly and accurately captured,” he continues, “so that risk positions can be calculated in an accurate and timely manner for monitoring purposes. Finally, the code suggests that market participants keep a timely, consistent, and accurate record of their market activity to facilitate appropriate levels of transparency and auditability and to have processes in place designed to prevent unauthorised transactions.”

Coyne feels the Code’s presence is a reminder of how the industry must self-regulate. “There is no reason why all aspects of the code cannot be adhered to as the requests are not onerous,” he says. “The more the industry shares the burden of self-regulation, the less the need for regulatory enforcement. We need to continue to take collective responsibility for safe market operations.”


AI and Machine Learning are being leveraged to automate workflows

Analytics and implementation costs

More firms seek transparency in their trade lifecycles, and post-trade processes are subject to analytics and data more than ever. What was once a back-office process is now receiving significant attention from stakeholders as firms realise the benefits of moving beyond simply improving operational efficiency.

Stevenson cites a few benefits of adopting analytics. “Real-time analytics generates insights on the business activities of each customer,” she says. “Settlement providers can provide real-time analytics and accurate reporting for customers. Firms can prevent trade settlement failures by identifying potential trades most likely to fail even before the failure occurs.” 

Post-trade processes are now subject
to analytics and data more than ever before

Chong cites an example. “In addition to executing on best price,” he says, “a firm may include metrics from post-trade such as counterparty post-trade positions, operational metrics (e.g. average time from execution to confirmation) and settlement type (e.g. CLS vs Gross) to select their counterparty. By bringing these data points together from the front office to post-trade, a firm may obtain a better outcome based on execution price, risk, and operational considerations.”

Coyne feels firms must examine their technological foundations before proceeding with automation, given the risk of automating defective workflows. “Like all data models,” he says, “the old adage of garbage in garbage out applies. It is therefore important that participation is based on accuracy and commitment.”

“Any missing information makes modelling difficult and adds more guesswork. These problems are very resolvable and data APIs play an important role, ensuring that all data is submitted, normalised, and made available on a shared basis.” And what role are APIs playing right now? “APIs combined with standards are essential and increasingly easy,” Coyne says. “Our clients are given UI access, however, most want to be able to send and receive data sets over APIs.”

Chong confirms that API access requests are more common than ever. “We believe the future lies in a hybrid approach where repetitive tasks will move towards more API-based automation,” he says, “whilst complex and exception-based workflows will remain UI-based for better control. We believe for UI, there will be a shift towards more interoperable applications to improve data quality and reduce the need to access multiple applications.”

Implementing all of this sounds complex and expensive. However, Coyne explains that new working models reduce the total cost of ownership significantly, along with its carbon footprint. 
“New models use common standards, golden copy trades, and mutualised life-cycle events. This reduces the cost of ownership and enables much smarter compression and less operational friction.”
“It also shares operational risk with peers and enforces a strong set of standards among the user group.” The benefits don’t end there. “Execution profitability is enhanced with more efficient credit models,” he continues, “not just for the ability to see more prices and deeper liquidity but also risk-weighted assets.”

“Allocated costs are also reduced because the cost of running such infrastructure is shared.” There is a regulatory and transparency impact too, with regulators requesting better market access controls. “Compliance with these requests is very important to the front office business owners,” Coyne says. “We have seen recommendations from bodies such as the GFXD, a subdivision of GFMA, and also other regulators who are keen to resolve the systemic issues such as credit over-allocation. The good news is that this technology is now live and being used.”

The big question when evaluating post-trade
solutions is: Should a firm choose a third party solution or build one in-house?

The way forward

The big question when evaluating post-trade solutions is: Should a firm choose a third party solution or build one in-house? Cloud usage is common enough currently to render the in-house option ineffective from an ROI perspective.

Chong says that firms looking for technology partners must examine whether the partner’s vision is aligned with theirs. “At TradeNeXus, we believe in the value of an ecosystem and collaboration,” he says. “To deliver the best solutions to clients, we assess how we can deliver the best value, which includes partnering with other solution providers to deliver a single dashboard to clients to manage all their post-trade processes.”

LSEG’s Dimase lists a few factors that firms should look at. “….a venue agnostic approach, the ability to adopt market standard protocols (mainly FIX), and the provision of light or zero IT footprint solutions. The availability of advanced analytics to extract value from post-trade data is also a key factor and the ability to deliver workflow solutions that integrate with proprietary solutions.”

“Firms should look for an agile technology provider who is committed to evolving as market factors change,” Stevenson explains. “The technology provider should be in partnership with their client providing a robust service model that can easily accommodate changes in the market but also  can keep costs sustainable.”

“The technology provider should be able to provide a flexible approach to integration and an open architecture that allows firms to easily incorporate common or bespoke workflows. Also, the technology provider should bring a high-level of expertise on all aspects of the post-trade operations.”

Solution provider choice aside, Coyne says the industry is witnessing positive steps that aim to address existing gaps. “For example, there needs to be a better linkage between settlements and credit,” he says. “We will fine-tune optimisation models to not only give enhanced market access but reduce credit exposure in real-time at very low cost. All events must be alerted to those people who have the responsibility to run orderly markets.”

“Industry bodies such as the GFMA,” he continues, “are making meaningful inroads to aspects of the market with long-standing issues that need addressing. The most recent is their publication and call to address over allocation of credit by using dynamic credit from a single point. A product that CobaltFX is live with and expanding the user group.”

Post-trade processes are evolving thanks to technology and firms’ willingness to use it for goals beyond operational efficiency. As the market continues to evolve, post-trade will undoubtedly witness more interesting development.