Our Innovation Journey: The Story of Liquidnet Smart Blocks
Innovation has been the name of the game in trading for many years but improving the correlation between innovation and alpha generation is one of the fastest growing frontiers in the world of trading. As a leading unconflicted agency broker, Liquidnet has consistently been recognized for being a leader and trailblazer in electronic trading technology and was most recently voted Best Algorithmic Trading Provider at the 2021 Waters Rankings1.
The real question is why.
Being successful in this arena comes down to two components — delivering tailored solutions that truly address the needs of the buy side, which can only be achieved through trust and close collaboration, and solving targeted industry challenges through the lens of innovation and the application of financial technology.
It’s by following this approach that we created Smart Blocks.
Liquidnet Smart Blocks allows traders to intelligently search for block liquidity using transient performance opportunities while still tracking to traditional benchmarks. The logic, available in select Liquidnet algorithmic strategies, dynamically seeks block liquidity based on a securities trading performance relative to a specific benchmark.
This is the story of how we got there.
The problem
From listening to our institutional clients, one of the things that stood out most was that, while trading in a schedule-based algorithm, they were looking for an automated solution to capture block liquidity at opportune times based on their trading objective and their risk appetite for deviation from the benchmark. Subconsciously or consciously, most traders will manually look for such opportunities, but in a non-systematic manner and when time permits. Smart Blocks seeks such opportune times in a quantitative and systematic manner.
While volume weighted average price (VWAP) and percentage of volume (POV) are amongst the most commonly used algo strategies in the market today2, savvy traders look to take advantage of short-term performance opportunities when individual stocks deviate from a benchmark.
With heightened market volatility and increased automation, the solution was evident. We needed to build a model designed to automate these decisions — strategically seek block liquidity to maximize the overall performance of an algorithmic strategy and minimize market impact.
Leveraging insights and analytics
By analyzing our trading data, we observed a number of things.
First, benchmark strategies seemed to be used more as default during times of heightened levels of volatility in the market, when there is a lack of clarity around price and/or direction.
Secondly, while some are still looking for opportune times to deviate from a pre-defined trading schedule in order to capture alpha, the heightened market volatility seen since the start of the pandemic has led them to ‘trade the average’ instead, as they are concerned with being caught on the wrong side of price moves. However, having a quantitative, systematic approach to identifying opportunities to capture alpha, where one could adjust their willingness to deviate from a target benchmark, would be welcomed.
Below is an illustration of how a schedule-based algorithm can take advantage of a stock’s short term price dislocation from a highly correlated basket of peers (Figure 1). Trading a block at such prices can lead to significant performance relative to the VWAP benchmark.
The process
Developing a tool that is truly innovative requires collaboration from many departments, as well as with our clients. So, we put together a team of experts, tapping talents from across our global Technology, Products and Execution and Quantitative Services teams, and formed working groups with select Liquidnet Members. Our goal was to not only identify and fully map out the problem, but to also define a market-driven solution.
Although the feedback we received was positive — the solution we had built made sense to our clients — we still had to improve.
To enhance performance, we developed a robust simulator and back-testing environment to more optimally set configurations of Smart Blocks to achieve the end user’s desired behavior. These environments provide the end user with clear expectations as to the frequency of opportunities along with expected performance relative to several key trading benchmarks, as demonstrated below (Figure 2).
By seeking continuous feedback and collaborating closely, we reached what we believe is an optimal solution — one that leverages conditional orders to help minimize disruption to overall trading performance and intelligently seeks block liquidity only at opportune times, with customizable parameters that a trader can control.
For example, a higher urgency to maintaining a pre-defined trading schedule will seek out opportune times for blocks less frequently and potentially with higher thresholds to a benchmark, while a lower urgency to maintaining that same trade schedule would more frequently be seeking opportunities for block liquidity and potentially with lower thresholds to a specified benchmark.
Smart Blocks is one of the proprietary signal driven frameworks included in Liquidnet’s premium suite of algos, which are specifically designed for institutions and their unique workflows. As our offering continues to evolve to address the shift in the market towards automation and data-driven investing, we constantly review and enhance our services and solutions to help our institutional clients compete and be ahead of the curve.
Available in select Liquidnet algo strategies, we’re in the process of introducing Smart Blocks globally. Please speak to your Liquidnet representative to explore the functionality in more detail, or to enable Smart Blocks by default.
1 Waters Rankings 2021: https://www.waterstechnology.com/awards-rankings/7854076/waters-rankings-2021-all-the-winners
2 The Trade 2021 Algorithmic Trading Survey: https://www.thetradenews.com/the-trade-magazine-spring-2021/
Written by Mike Capelli, Head of Americas Execution and Quantitative Services, and Scott Kartinen, Global Head of Algorithms