A MATTER OF BALANCE.
Jason Hughes, Global Head of Sales at ADSS argues that while the impact of new analytical tools to the FX trading desk is growing, human knowledge and experience are still irreplaceable.
How would you characterise the maturity of transaction cost/quality analysis today?
The adoption of execution quality and cost analysis in the market is increasing rapidly, but the product itself is still in an early growth phase. We are seeing more independent providers coming to the market and a gradual acceptance from existing technology providers that they must broaden their product suite to include transaction cost analysis (TCA) reporting. The fragmented state of liquidity in the FX market has highlighted the importance of monitoring execution quality and transactional costs. The market has welcomed the introduction of readily available analytics which can help identify providers guilty of recycling liquidity.
Is the use of different data affecting this analysis?
Analysing your liquidity providers (LPs) performance for trade execution has been done for some time, but often in reports provided by each LP, and potentially with different price data referenced in each of them. There has been a need to understand the data sources that vendors are using in their analysis so that one can then easily compare them, and also to ensure that they are using an independent and unassociated price source to benchmark. This has been simplified as there are now solutions that allow for this to take place by consolidating LP data and benchmarking each LP against the same pricing data source. This allows for better-defined benchmarking and the ability to slice and dice performance metrics in a more meaningful way across all of your LPs, using the same reference data.
Who within buyside firms (client firms) is making use of execution analytics today?
This type of analysis is becoming more broadly demanded from a variety of clients as it can be used for analysis across a number of scenarios within electronic trading environments, such as liquidity pool optimisation, market impact assessments for order execution and algo optimisation when looking at things such as automated risk management. Where there has been a need to reduce latency, analysing the performance of your algo at different venues to ensure you are achieving optimal returns has become a key part of optimisation process. This has therefore given firms the ability to compare, on what is essentially a level playing field with known data, and has become key in terms of giving a fair assessment of true performance at each venue.
What are the greatest challenges they face in applying quantitative outputs to what can be qualitative decisions?
This type of in-depth analytics is often used to help refine the automated process churning along in the background. This would help developers or quants improve their coding by understanding how the same thing works at different venues or in different environments.
This means in most cases that the qualitative information does not need to be considered as part of the automated process because there is a human layer in between to help interpret the qualitative aspect into something more quantitative. In the future, as machine learning and the use of AI becomes more prevalent, this is likely to change.
Can best execution policies be effectively turned into more quantitative processes?
Best execution policies are often viewed as more of a compliance requirement than something that is considered on a trade-by-trade basis. There are a number of factors here that can change the interpretation of best execution from a trader’s perspective, such as whether you wish to execute your order in as large a size as possible, in the quickest time possible or are you more concerned by the price achieved and would accept a partial fill so the size executed is secondary. All these factors materially change what you are looking for in terms of execution. Other factors can also come into play.
For some participants, the market impact for very large orders again changes what can be perceived as best execution, if for example the idea is to enter or exit a position with minimal price impact to the underlying market. For many, this means there will always be both quantitative and qualitative factors when considering whether they are receiving best execution.
What complications are created by execution analysis across multi-asset trading? How can analytics help to overcome these?
With all asset classes you need to understand the dynamics of the market you are trading and FX trading across different venues may differ as much as executing orders in equities versus trading in commodity futures. It is always key to understand the dynamics of the market you are trading in and to be aware of the behaviour of other traders or trading systems because they may impact the order book you are looking to trade. While analytics can help refine your approach, sometimes there is still nothing that replaces the knowledge and experience built up by experts and professionals who have been trading in those markets for many years.