The definition of insanity is doing the same thing over and over again and expecting a different result.
If anything, the Tick Size Pilot reinforced that passively waiting for the rebate will have a negative impact on performance. This should not come as a surprise and would be true of any algorithm with static behavior built on prioritizing rebates and not optimized to adapt to a changing market microstructure.
Several studies on the impact of the TSP have been released, including a recent one which projected a large shortfall to ordinary investors. We believe brokers who were unsuccessful in optimizing their own trading logic based on microstructure changes resulting from the pilot have failed to paint the entire picture of the overall industry’s performance in trading those TSP securities. We also believe there were some valuable lessons learned from the TSP which can be applied to the Access Fee Pilot and other market structure studies in the future.
Some logic in other providers’ TSP studies that we find to be flawed include assessing only child order performance and aggregating results across all three TSP test groups.
Assessing only child order shortfall performance versus arrival price is a less-stable metric than looking at performance of the parent order. Child order performance is heavily dependent on how an algorithm’s trading logic works. If there are no changes made to the algorithm based on the observed shifts in microstructure or if the algorithm is not capable of dynamically adapting, then it should be expected that performance of that algorithm will be negatively impacted because its trading logic is not optimized. Looking at child order performance reflects that algorithm’s performance and does not necessarily reflect overall execution quality.
In our own analysis, Tick Size Pilot Year in Review, we examine each of the three test groups versus the control group to identify scenarios where widening the tick could improve market quality and to learn from those scenarios. For example, in our analysis, we show data that suggests Test Group 3 (the most restricted test group) had little to no improvement on impact costs and overall performance was inferior compared to Test Group 1 and Test Group 2. Aggregating the results from all test groups, as other providers have done in their analyses, provides no clear assessment which can be used to support the viability of any of the test groups.
Many brokers also cited significant costs associated with preparing for the anticipated challenges the wider tick would present, but simply increasing usage of midpoint orders without truly optimizing an algorithm’s behavior was not money well spent. Brokers also needed to ensure their algorithms would capture liquidity at the most opportune times and not only on the near-side of the spread. During the pilot, liquidity had clearly shifted and there were opportunities to leverage inverted venues, remove liquidity as it pooled at the midpoint and opportunistically use aggressive order types to mitigate impact.
An estimated 90% of institutional executions engage with algorithms and those that are hard-wired to capture the rebate saw a decrease in performance with a widened tick. The TSP exposed how algorithms designed to “chase the rebate” instead of efficiently sourcing liquidity regardless of rebates and opportunistically crossing the spread miss out on the best-quality executions. While the Tick Size Pilot may not have completely achieved its goal, it demonstrated that brokers who were able to adapt their algorithms for shifts in market microstructure were able to realize positive impacts on performance.
To discern meaningful results and optimize trading performance during microstructure shifts, brokers must be fully transparent and provide their clients with real-time analysis and the ability to control and customize their routing protocols. Not adapting and using the same trading logic under different microstructure conditions has proven to cost end-investors, so let’s stop the insanity and do something different with the Access Fee Pilot.