Beyond the Checkbox: Reshaping Sell-Side Execution Analytics
By Nazed Mannan, Head of Quant Products, BMLL
In today’s markets, execution analytics has evolved from a best execution reporting chore into a front-line differentiator. What was once a reactive, box-ticking TCA process is now a strategic capability shaping trading decisions, broker selection, and investment workflows. At the inaugural BMLL Client Summit EMEA in March, leading practitioners discussed how exploding data availability, buy-side sophistication, and increased scrutiny on implicit costs are driving this shift.
Firms are moving from generic benchmarks to highly bespoke, regime-aware analytics that reflect each client’s unique flow, enabled by unified, high-quality market data. The conversation also highlighted a future where AI and natural language interfaces will make advanced analytics more universally accessible.
Here’s what we heard and learned on the day.
From Compliance Chore to Strategic Consultancy
Historically, Transaction Cost Analysis (TCA) was treated as a reactive compliance exercise. However, as the buy-side has become increasingly quantitative and demanding, clients now expect proactive execution consultancy rather than standard reports on VWAP or arrival price. They look for strategic advice on whether to wait for the US market open or how to navigate spreads during periods of volatility.
Furthermore, the appetite for deep market microstructure transparency has expanded beyond systematic hedge funds. Sovereign wealth funds, long-only funds, and family offices now ask complex questions about execution mechanics. As buy-side desks help portfolio managers discover alpha, sell-side brokers are adapting by providing transparent toolsets that directly align with client intentions.
Taking Control: The Power of a Single Source of Truth
A major historical pain point for the sell-side was reliance on third-party TCA providers, which lacked the flexibility to support bespoke benchmarks. When clients questioned specific numbers, brokers struggled to explain the provider's black-box calculations.
By adopting BMLL’s comprehensive, normalised market data, together with its applications, firms gain a holistic data science environment that provides accurate insight into any research requirement. This ‘single source of truth’ gives brokers complete control and consistency, ensuring the exact same data used for live trading algorithms and pre-trade models is also used for intra-trade monitoring and post-trade client reporting.
This internal trust and control has allowed at least one firm to completely replace its external TCA and best execution providers within three months, removing a significant, direct cost from its balance sheet.
Turning One Quant Into Five: The Power of Accessible Data
The panel emphasised the operational efficiency gained from modern data infrastructure. A decade ago, analysing historical tick data required massive internal servers, redundancy, and dedicated data-cleaning teams, meaning a new business idea could take up to six months to validate.
Today, easy access to [BMLL’s] cloud-based, clean data allows firms to rapidly spin up compute resources, test ideas, and fail fast without heavy upfront investments. This acts as a significant force multiplier; one panellist noted their new data solution provided a 5x performance improvement, effectively turning one quantitative analyst into five.
Driving Innovation: Regime Analysis and Fragmented Liquidity
Firms are using BMLL’s granular, normalised data to conduct sophisticated market structure analyses that directly drive algorithmic product innovation:
Regime-based analysis: Brokers recognise that algorithms cannot be judged in a vacuum. Instead, they cluster market regimes by factoring in volatility spikes, spread widths, momentum, and imbalances to determine which algorithms and liquidity sources perform best under highly specific conditions.
New product launches: Deep analysis of order book build-ups and price formation into the close has enabled firms to successfully design and launch specialised algorithmic strategies tailored for closing auctions.
Discovering fragmented liquidity: In fragmented spaces like ETFs, comprehensive data allows brokers to stitch together the true liquidity landscape, proving to investors that high-quality liquidity exists on traditional lit screens, not just via bilateral risk platforms.
Bridging the Gap: Elevating the Broker-Client Partnership
Dynamic, clean data removes the traditional barriers between the buy-side trading desk and the sell-side. Rather than engaging in subjective debates over why a specific flow underperformed, brokers can proactively dive into the data to identify the exact root cause. For example, one firm quickly showed a client that their flow underperformed peers simply because they started sending orders earlier in the day when spreads were naturally wider. Additionally, this data is used to dynamically evaluate bilateral risk provider quotes, analysing fill rates to determine algorithm interactions based on broader order book health.
Ultimately, this transforms execution analytics into a collaborative partnership where the sell-side actively provides tools to optimise performance.
Looking Ahead: LLMs, Shared Languages, and AI
When asked about the future of the industry, panellists highlighted major frontiers defining the next generation of analytics:
A shared language: The industry struggles with an overly complex landscape. There is a strong desire for an agreed-upon standard for trade classification, allowing the sell-side and buy-side to finally speak the exact same language when discussing metrics like addressable volume.
AI and Large Language Models (LLMs): Integrating LLMs and Natural Language Processing (NLP) with structured market data will drastically lower barriers to entry. This allows non-technical users, such as heads of desks, to intuitively query complex execution data using plain English.
Deeper Machine Learning: Firms expect to lean heavily into advanced machine learning to provide deeper, more predictive execution analytics to clients.
