BestEx Research is set to launch a new framework which allows users to create bespoke implementation shortfall (IS) algorithms for each client’s unique alpha profile and execution risk preferences, The TRADE can reveal.
The new tool, Adaptive Optimal IS, has the front end integrated with a research-driven framework, allowing users to build and test multiple versions of IS algorithms against each other to iteratively improve performance.
Adaptive Optimal IS seeks to provide a solution to two primary challenges that buy-side traders face with respect to IS algorithms. Namely, the opacity of execution algorithms and their impact of their internal design on overall transactions costs; as well as the approach of sell-side execution providers, which tend to focus on marketing features as opposed to identifying the best solution for individual clients.
“Adaptive Optimal IS addresses both of these issues by completely redefining how IS algorithms are designed and customised – from the ground up,” Hitesh Mittal, chief executive of BestEx Research, told The TRADE.
“It provides all the foundational elements needed to define an IS algorithm. For instance, the designer can specify whether the algorithm should be opportunistic, driven by percentage of volume (POV), schedule-driven, or use a hybrid approach. This can be configured based on factors such as order size, volatility, asset class, and more. In the case of equities, the designer has multiple options on how to integrate dark liquidity into the algorithm. Users can also specify how the algorithm adjusts its speed in response to varying market conditions.”
Mittal highlighted that one of the reasons IS algorithms typically fall short is due to the “noisy” nature of their measurement. Primary factors that contribute to higher IS – such as market impact costs, spread costs, rapid alpha decay during order execution and adverse selection costs – may be difficult to discern for a small order pool due to “noise” surrounding the measurement of performance. As a result, it can often be unclear whether IS resulted from natural price fluctuations in the underlying instruments or from these cost factors.
“When measurement is lacking, an algorithm’s implementation is often inadequate,” added Mittal.
“A core component of the Adaptive Optimal IS framework is its capacity for A/B testing, which supports our measurement-driven approach to algorithm design. Randomised A/B testing allows for more fair, direct comparison of strategies and helps determine which design decisions work best for each firm and trader. We can confidently design tailored strategies when we have evidence to support our design decisions.”
The launch of Adaptive Optimal IS follows the appointment of former head of platform sales at Citi, Matt Cousens, who joined BestEx Research as its head of EMEA equities.
As part of his role, Cousens holds responsibility for driving the rollout of BestEx Research’s equities execution algorithms for US and Canadian trading to the European customer base as well as their general globalisation.