Watch our webinar with Dynamic Beta Investments to see how improving execution can increase capacity
Request Recording

Designing Optimal Implementation Shortfall Algorithms with the BestEx Research Adaptive Optimal (IS) Framework

June 1, 2023

Minimizing Implementation Shortfall (IS) remains the ultimate goal for both buy-side firms and sell-side execution providers. Every bank or broker offers an implementation shortfall algorithm, each claiming to provide the “optimal” way to trade while minimizing cost relative to the arrival price benchmark. But most traders understand that there is no one-size-fits-all solution to minimizing IS, since each trader’s order flow and measurement of success are unique.

BestEx Research’s new Adaptive Optimal (IS) Framework is designed to address the shortcomings of conventional IS algorithms and to help traders and providers create optimal trading tools for their order flow. Adaptive Optimal is built on the following principles:

Customizable Algorithm Design

Adaptive Optimal accommodates various algorithm designs, including POV-based, Schedule-based, and Opportunistic (schedule-less) designs, to cater to the preferences and execution risk profiles of different traders. 

Benchmark Price Sensitivity

Adaptive Optimal can be configured to exhibit different levels of price sensitivity, based on the trader's beliefs about expected price momentum, execution risk profile, and custom benchmark preferences. 

Dynamic Urgency Levels

Instead of relying on predefined “Low”, “Medium”, and “High” urgency settings, Adaptive Optimal can be configured to adjust its urgency level based on order size, liquidity, and the specific needs of the trader. 

Optimized Must-Complete Behavior

Often, Opportunistic IS algorithms suffer from poor design that creates two orders in two algorithms working simultaneously toward completion, increasing the overall market impact of an order. Adaptive Optimal tackles traders’ need for order completion by dynamically switching between algorithms depending on order size and remaining liquidity expected when “must complete” is selected.

Customizable Dark Liquidity-Seeking Behavior

Exposing 100% of an order to off-exchange liquidity may not always be optimal. For liquid stocks, this behavior can create high adverse selection costs, but for larger orders or less liquid stocks, the benefits may outweigh the costs. Adaptive Optimal allows customization of dark liquidity-seeking behavior based on traders’ needs and preferences.

Minimized Adverse Selection

Algorithms built in the Adaptive Optimal (IS) Framework are designed to mitigate the adverse selection risk inherent in Opportunistic IS algorithms and avoid offering liquidity when prices are expected to improve. 

How do experimentation and consulting contribute to optimized IS performance?

Understanding the true impact of various design approaches on trading costs is difficult when trading decisions are based on historical data; a historical view shows only what happened and not what might have happened if an algorithm’s trading had taken a different shape. The Adaptive Optimal (IS) Framework incorporates experimentation through A/B testing into its design, allowing traders to compare the performance of multiple implementations to determine the best execution strategies for various types of order flow. For example, differentiating between short-term alpha and market impact can be extremely challenging when looking back on a trader’s execution performance. But by randomly assigning orders to various algorithmic approaches simultaneously in our AMS’s Strategy Studio 2.0 for fairer comparison of outcomes, we can partner with our clients to uncover the patterns that drive costs and deliver optimized algorithm design for each client’s specific needs. Each of the design considerations above (e.g., schedule selection, real-time reaction to market conditions, etc.) demands evidence-based decision-making, and experimentation through A/B testing can reveal significant performance improvement. 

Underlying the flexibility of Adaptive Optimal is BestEx Research's award-winning execution technology, including our proprietary limit order model for optimized child order placement, our FastPass Smart Order Router (SOR) that improves queue position on exchanges, and our measurement-driven approach to segmentation of dark liquidity for reduced toxicity. The combination of customized algorithm design and research-driven order placement technology drives improved execution for buy-side institutions. 

Sell-side firms can recreate virtually any IS algorithm they currently offer–such as liquidity-seeking algorithms or arrival price algorithms–while they continue optimizing performance with new design choices and A/B testing when they build their algo platforms with the BestEx Research AMS. The Adaptive Optimal Framework in combination with our AMS enables sell-side firms to partner with their own clients to build and optimize algorithms for the unique use cases of each–without the development time of traditional customizations.

The Adaptive Optimal (IS) Framework by BestEx Research is a more sophisticated, more customizable approach to minimizing implementation shortfall. By incorporating advanced capabilities for meaningful optimization such as A/B testing, dynamic urgency levels, and customizable algorithm design, Adaptive provides traders with greater control over their execution, the evidence they need to make data-driven decisions, and the ability to adapt to changing market conditions in real-time, yielding improved outcomes.