Integrating TCA into FX trading workflows

TCA in FX requires highly granular, high frequency data, including tick-by-tick quotes and trades from diverse liquidity sources such as ECNs and bank/non-bank streams. Platforms providing comprehensive, microsecond-level timestamps are essential to measure accurately against dynamic benchmarks like risk transfer price or arrival price.

The decentralised FX market complicates benchmarking due to fragmented liquidity and the absence of a unified pricing reference. Comprehensive data aggregation and normalisation from multiple liquidity sources are essential to create accurate benchmarks reflective of true market conditions. 

“The ability to seamlessly integrate proprietary logic via tools such as AlgoBox further enhances the depth and precision of analytics throughout the trading lifecycle, from pre-trade to post-trade, supporting ultra-low latency execution,” observes John Stead, director of sales enablement and marketing at smartTrade Technologies.

Sophisticated TCA frameworks rely on high frequency, time-stamped data that captures critical microstructural details. But none of this matters unless those metrics are benchmarked against accurate data according to Paul Lambert, CEO New Change FX.

“By aggregating price feeds from across the FX ecosystem and constructing a real-time mid-rate derived from the best bid and offer available, our data enables clients to perform meaningful, apples-to-apples performance analysis,” he says. “We have also developed granular, independent forward and swap benchmarks, which are dynamically adjusted for basis shifts and interest rate differentials.”

Service has evolved significantly

Meanwhile John Crouch, CEO Ideal, observes that TCA has evolved from a simple, ‘check the box’ to a service that improves understanding of technical and operational risk.

“Market participants approach TCA in a different way depending on the extent to which execution costs dominate their P&L,” he says. “If you are a high frequency trader, execution cost represents a large chunk of your P&L so you have to be hyper vigilant, whereas a buy-side person doing one trade a day has different priorities.”

BestX cleans market data and then adjusts it to the type of transaction. It also has a proprietary expected cost model that takes into account the size of the transaction and the liquidity and volatility at that point in time, enabling clients to not only evaluate their trade performance against market data but also the expected cost at the point of execution.

“Specific timestamps around when the trade was sent to the trader, when they began the order and when it was completed give clients a better understanding of their transaction costs,” explains Yusuf Nurbhai, head of BestX.

He observes that there isn’t a single benchmark to evaluate their transaction costs against, which is where technology has evolved to allow clients to define the asset class-specific benchmarks that are most relevant.

“They can achieve a standardised view around execution quality from a cost perspective, but they will have different ways of evaluating trading costs,” adds Nurbhai. “Some may want to evaluate their costs against the WM/Refinitiv 4pm fixing plus or minus a spread because that is what is written into their investment mandates.”

In this scenario, TCA helps identify when and why a trade didn’t meet that criteria.

Analysis contributes to strategy

TCA has evolved significantly from post-trade reviews to becoming an integral part of pre-trade and real-time analytics, empowering traders with proactive execution optimisation strategies explains Stead.

“Predictive analytics enable traders to forecast transaction costs and choose optimal execution methods before trading, while real-time TCA alerts allow immediate strategic adjustments to changing market conditions,” he says.

“By integrating execution logic directly within analytics, our platform enhances traders’ ability to respond dynamically during trades, substantially improving outcomes such as reduced slippage and market impact.”

While early TCA efforts were often regulatory box-ticking exercises, today’s advanced frameworks feed directly into pre-trade decision making. But this evolution comes with a caveat: accuracy is now non-negotiable.

“Furthermore, pre-trade tools only add value if they understand the underlying volatility in the market at the time,” says Lambert. “We have developed an expected spread model that calculates the volatility in the market over the preceding five minutes and adjusts the spreads accordingly. This is useful in the post-trade TCA process as well, as it helps clients understand when costs have risen due to market volatility.”

In terms of implementation, BestX has focused on the areas where it can get the cleanest and most sanitised dataset, spending the majority of its time working with platforms, order management systems, custodians and banks to ensure data consistency.

“If a client trades on Bloomberg, for example, we will get their files directly and because it’s all cloud-based and the install is just a Chrome-based browser, they are ready to go,” says Nurbhai. “The training is more around how to utilise the tool to get the most out of it.”

Well worth the cost

As for how to measure the cost effectiveness of TCA, he says he has spoken to firms who say they have made savings that outweigh the cost of a licence by being able to identify trades that were erroneously booked.

“They are also able to evaluate their counterparties in an empirical, quantitative manner,” adds Nurbhai. “As these counterparties are trying to improve their performance to gain a greater share of trading activity, this also leads to cost savings. Outlier detection drives value by improving counterparty competitiveness.”

Effective TCA implementation requires meticulous data management, justified benchmark selection, comprehensive documentation and independent verification to meet stringent regulatory requirements such as MiFID II and FX Global Code adherence. Leveraging platforms with robust governance, audit and reporting functionalities is critical.

“Trading firms can evaluate TCA cost-effectiveness by establishing clear execution cost baselines, accurately measuring investments in TCA technology and consistently tracking improvements post-implementation,” says Stead. “Translating reduced slippage and lower market impact into monetary terms relative to trading volume demonstrates tangible benefits.”

Scalability and adaptability in TCA require adopting flexible, modular infrastructures capable of handling increased data volumes and market evolution.

“Leveraging AI-driven analytics enhances dynamic responsiveness to changing market conditions,” observes Stead. “Integrating these systems seamlessly requires careful planning, including phased rollouts and interoperable design using standard protocols.”

Effective integration depends on flexibility and interoperability. TCA systems must be able to plug into existing trade management systems, OMS platforms and data lakes without disruption.

Don’t underestimate regulatory requirements

Regulators are increasingly wary of conflicted benchmarks – using timestamps that are off by even hundreds of milliseconds is akin to flying blind and firms need to apply TCA methodology consistently across trades and be able to explain their logic months later in an audit or client review.

“Many desks focus on spot FX but regulators expect the same best execution standards for forwards and swaps,” says Lambert. “They will want proof that execution policies are not only documented but also demonstrated, which means showing how TCA results feed back into execution decisions and counterparty reviews.”

Legacy TCA systems were often built for static, low frequency analysis and don’t scale well. New Change FX rebuilt its TCA platform from the ground up using a cloud-native architecture, which allows the system to scale horizontally and vertically without compromising performance or user experience.

“The platform automatically consumes client execution data, matches it against high frequency benchmark data and delivers both post- and pre-trade insights via API or dashboard,” explains Lambert. “This scalability makes the system future-proof, capable of adapting to the evolving dynamics of global FX markets.”

Crouch explains that when it comes to achieving seamless integration with existing systems, Ideal utilises proprietary, automated tools to minimise client effort, making the integration process as effortless as possible. “Our tools normalise the data from internal systems and vendors – we typically get clients live in days or weeks,” he says.

Buy-side firms under pressure to accomplish more with fewer resources have limited bandwidth for trying new trading tools. When they start doing TCA in-house, they often need to source new data, requiring technical work and budget.

“If we look at where best execution started, the only way you can show you are doing a good job is by getting data from loads of counterparties – and sourcing that data is a complicated process,” says Crouch. “Contractually, some of these providers don’t want to give data to other firms. In this scenario, a third party can play a valuable role by showing firms where they sit in the broader market ecosystem without disclosing sensitive information.”

The power of data

The impartiality of the data used to produce TCA reports has been a contentious topic in the FX market for some time. However, Crouch says reporting has become much more transparent over the last decade.

“At this point, showing inaccurate analysis would require malicious intent and I just don’t think any platform would do that,” he says. “That said, the benefits of full independence are focus and flexibility. Independence allows us to utilise data from many sources to produce a more complete assessment.”

Crouch also emphasises the importance of data privacy. “We keep our data fully segregated, whereas other firms co-mingle data for peer analysis,” he adds. “The extent to which an analytics firm is keeping a user’s data completely private drives many client decisions. Firms seek Ideal solely because we safeguard their privacy.”

Lambert suggests the most important question in TCA is deceptively simple (what are your trades being measured against?) and adds that venue-provided TCA reports often benchmark execution against the provider’s own prices, offering only a narrow view of performance.

“As an independent provider, we benchmark trades against a market-wide, regulated data set. This objectivity provides a clear, unbiased view of execution quality across the full FX landscape, meaning firms can be confident they are performing well relative to the best available market conditions.”

Venue-provided TCA reports often carry inherent conflicts of interest, risking biased benchmarks and selective data representation, says Stead. Independent TCA providers counteract these issues by delivering impartial, comprehensive data analyses using standardised methodologies.

smartTrade complements independent analysis by offering transparent, impartial benchmarking tools, supported by its flat-fee model, ensuring that data used for TCA remains clean and unbiased.

As clients increasingly trade across multiple venues, aggregating TCA reports across venues to get a holistic view of trading performance becomes very difficult – and if they are only evaluating performance based on the available market data of an individual venue, they may be missing opportunities to perform better by trading on other venues.

“The buy-side can also opt to have their data aggregated, anonymised and put into a community data pool to allow them to evaluate their trading performance against their peers,” says Nurbhai. “So you can see the value of data based on verified trades rather than trades that have been handpicked by a venue or a liquidity provider.”

Integration the way forward

The future of FX TCA points towards greater integration with predictive analytics and AI, providing real-time insights and hyper-personalised execution strategies across asset classes. Future-oriented TCA systems will dynamically adapt trading strategies mid-execution and provide deeper analytical insights into liquidity provider behaviours and algorithm performance.

“We are leading this evolution by integrating advanced AI tools such as Copilot into trading workflows, leveraging modular microservices and customisable APIs to ensure continuous adaptation to market trends, ultimately delivering enhanced execution intelligence and superior operational effectiveness,” adds Stead.

According to Nurbhai, the ‘sand in the gears’ of TCA is when a busy buy-side trader has to log into multiple systems to get trading information before going back into their OMS or EMS, which is why his firm created a tool that will interrogate its database to provide a ranking of counterparties.

“TCA can also be used to show the best performing algos historically based on fill speed and spread, helping clients decide whether to trade RFQ or algo,” he says.

Crouch emphasises the importance of being able to assist firms as they trade a wider range of asset classes, utilising asset agnostic analysis and addressing nuances of each asset class.

“Ideal focuses on automation, enabling users to understand the variables that drive profitability and risk – that simplified process saves time across management, risk, sales, and trading teams,” he says.

Lambert reckons the future is live TCA with immediate feedback loops, informing pre-trade analytics such as the cheapest date to roll a swap to, because the future of trading is already here with machines trading with machines.

“The future will be about real-time, machine-readable live TCA, not reports at T+1,” he says. “We are talking live analytics that plug directly into automated execution engines, enabling adaptive behaviour on the fly.”