For years, the conversation surrounding FX execution algorithms was dominated by workflow improvements. For the buy-side, the primary goal was operational efficiency. Trade Cost Analysis (TCA) in the beginning was just another TLA (three letter acronym). However, we have entered a new era. The market is no longer satisfied with simply flying autopilot on an order; the focus has shifted entirely to the quality of the execution and the data that proves it.
As we look toward the next frontier of trading, pre- and post-trade analytics have moved from being a nice-to-have administrative task, to the absolute centre of the execution workflow. But there is a caveat: for analytics to be valuable, they must be digestible and actionable.
From “put-in-the-drawer” to actionable intelligence
In the past, Trade Cost Analysis (TCA) was often a check-the-box exercise, which usually meant a report generated at the end of the month, reviewed briefly and filed away. Those days are over. The primary driver for the recent surge in algo adoption is that analytics have finally become actionable.
Today’s buy-side traders are not necessarily looking for more data; they are looking for better data to improve execution outcomes. Most traders on a busy desk don’t have the luxury of a dedicated TCA team. They need analytics that give them immediate recommendations to improve execution. If you are monitoring ten different algo banks, which is a common scenario in today’s fragmented market, you need to know instantly who the outliers are.
If one bank, or even a specific algo, consistently underperforms your Key Performance Indicators (KPIs), actionable TCA allows you to dive in by looking into the main drivers of those KPIs (i.e., aggressive vs. passive, order placement, execution speed and most importantly, liquidity venues). By turning data into a communication tool, traders can have more meaningful conversations with their algo providers around data driven topics such as speed of execution, order placement and more, and then request changes to execution parameters.
The quest for quality liquidity
At LMAX Group, we believe that the primary pre-trade analytic isn’t a complex formula. Instead, it is the trader’s experience and common sense applied to their understanding of available liquidity pools. The market often discusses volume, but we don’t hear enough about the quality of liquidity. Banks often internalise algo flow against offsetting interest within their network. In a recent publication the FXPA (with input from real-money clients) issued very good guidelines about transparency within internalisation. If external venues are needed (i.e. to maintain the desired execution speed/end time of an algo) most of the algo banks allow the clients to “curate” their liquidity venues based on execution goals.
One could make an argument that the FX market is an almost “perfect” market, despite the often complained about fragmentation. Fragmentation in FX does not necessarily mean different liquidity sources spread across different venues. It is often the case of the “same” liquidity in multiple pools, and the skill and experience of knowing the fewest number of pools to access to give the most optimal results is the art. This is especially true when focused on the most important characteristics for real money needs, such as market impact and information leakage.
The goals for an execution can shift dramatically. Clients might want to save spread on smaller orders (via RFQ or posting in lit markets) but as soon as a certain parent order size is reached, the priority shifts entirely to avoiding information leakage and the consequential market impact. This is where dark, skew-safe firm, bank only liquidity pools, like the LMAX Buyside Pool, become essential. They allow real money firms to match against multiple different offsetting interest (firm 2-way bank liquidity, passive algo banks orders, peg to mid order type, etc.) without creating a footprint that alerts the rest of the street. Keeping the “taker” – side of this pool limited to buyside flows incentivizes liquidity providers to show tighter spreads and more aggressive skews.
Redefining KPIs: risk transfer and volatility
To assist asset managers in demonstrating best execution, we must look at the metrics that impact the bottom line. In my experience, there are three KPIs that stand above the rest:
- Performance vs. risk transfer price (RTP): this is the ultimate benchmark. It asks if you could have achieved a better result by simply paying the spread at the start of the order. If your algo isn’t outperforming the RTP, it isn’t adding value.
- Execution within one standard deviation: we must consider market volatility. If you execute 95% of your volume within one standard deviation, you have done well, even if there was some slippage. If you are consistently outside that deviation, it is a clear sign of information leakage, suggesting your order is creating volatility that wouldn’t exist without your presence in the market.
- The “simple TWAP” benchmark: a good TCA should compare an “adaptive” algo against a simple Time Weighted Average Price (TWAP). If the smart algo isn’t beating a basic schedule, you are either paying for complexity that isn’t delivering results or you are setting limits that don’t improve your execution outcomes.
The power of negative mark-outs
One of the most misunderstood concepts in TCA is the mark-out. In a typical trading environment, a positive mark-out means the price moved in your favour after the trade, which is a yield for the trader. But for the buy-side, we look for the opposite.
We have created a unique ecosystem at LMAX where we focus on a complete alignment of interest. In most cases the buyside wants to trade “against” the market (i.e. buying in a falling market or selling in a rising one.), resulting in a negative mark-out (yield curve). Liquidity providers are happy with this kind of flow as it closes itself passively with positive results. In this scenario the buyside becomes an appreciated passive “liquidity provider” which leaves all market participants happy (incl. the algo banks as they show risk transfer price outperformance). This non-predatory flow is why we get better pricing from banks because they know that the interest in our pool isn’t looking to “scalp” them. It is real-money interest looking for a fair price without a footprint. In depth TCA provides an overview across venues for pre- and post-execution mark-outs.
A warning on “last look”
It is impossible to discuss best execution without addressing “last look.” Our stance is clear that last look practices have no place in a real-money execution strategy. While the global code of conduct provides a framework, and there are certainly many positive use case scenarios where it is appropriate, it doesn’t go far enough to protect buy-side interests.
If you are a real-money firm, trading against last look liquidity is a common-sense failure. It creates a window for information leakage that can be used against you. By using analytics, firms can identify which providers are rejecting trades or creating slippage during that last look window and shift their flow toward firm, no last look liquidity.
into a learning experience
Independence and the human element
There is a difficult conversation in our industry about independence. Many banks provide their own TCA reports, but these are naturally limited. At LMAX, we made a strategic decision not to build our own TCA product from scratch to avoid the perception of a conflict of interest. Instead, we partnered with Horizon to provide an independent, cross provider, strategies and venues view.
However, data alone is not enough. I believe every TCA report should come with a built-in consulting service, which provides the human element. A trader needs to know not just how they did, but how their peers did. Understanding what other real money clients are doing differently transforms a report into a learning experience. The LMAX Buyside team is proud to have this unique buyside advocate position.
The future: digital assets and stablecoins
Finally, we must look at the convergence of TradFi and digital assets. LMAX is at the forefront of this, moving toward a complete cross-asset marketplace. Whether it is spot crypto or the burgeoning use of stablecoins for cross-border trade, the buy-side will eventually demand the same analytical rigour they expect in FX.
While my current concern is whether we have enough critical mass of data in the crypto market to draw high-conviction conclusions, the trajectory is clear. By 2026, the distinction between an FX desk and a digital asset desk will blur. The firms that succeed will be those that apply these same principles, including digestible data, market-impact reduction and verification, across every asset they trade.
Analytics are no longer just the secondary part of the trade. By shifting from passive reporting to active, digestible and consulting-led intelligence, buy-side firms can finally take control of their execution quality. In a market where every basis point counts, understanding your footprint isn’t just about compliance; it is about performance.
