With John Radle, Global Head of Trading and David Biser, Senior Trader, Campbell & Company
The foreign exchange (FX) market, notorious for being dynamic, fragmented, and competitive, has become increasingly electronic in recent years and now more naturally lends itself to Transaction Cost Analysis (TCA). In the past, trading was driven by voice-based market-making and any sort of implementation shortfall was difficult to quantify. Measuring the quality of FX execution is becoming more attainable and is also becoming a requirement as the market continues to have stricter guidelines around FX trading activities
When trading electronically, the relevant data and trade details are going to be more granular. As technology and analytics improve, so too will the usefulness of TCA. As an example, we gather all the individual fills and can measure to what degree we are passive as opposed to aggressive. But trading is a process more than it is a single outcome; point-in-time analysis is important but it is often more meaningful when comparing traders/liquidity providers/algorithms over longer time periods. Interpretations will vary when evaluating implicit costs, but the point is that we can now take a wider range of data points into account and make modifications as market dynamics change.
Finally, all these technological developments are coupled with increased scrutiny of trading by both the regulators and clients. Transaction Cost Analysis studies have been adopted by many in the market to provide a certain level of confidence in trade execution. This emphasis is the reason it is important to have tools to show best execution.
Cleaning, managing and analysing data presents a significant challenge. However, Campbell & Company has a long history as a strong quantitative manager. We have leveraged our capabilities on the research side to effectively carry out those deep dives on the underlying currency pairs and liquidity providers. FX TCA can be difficult for electronic trading because of the lack of consistency – comparing measurements will be less meaningful if each provider has a different liquidity pool. Furthermore, each algo has its own logic so it may be difficult to determine whether it was poor algo performance that produced a sub-optimal result or whether it was because of limited liquidity during a trade.
As more pressure is exerted on providers, there are certain elements that are becoming more standardised when producing TCA. Particularly when they offer algorithmic execution solutions, banks are increasingly taking prudent measures to evaluate their liquidity sources and to remove those venues that could adversely affect fills. There is an impetus for people to provide that information so that algo users can get the best price possible given the strategy and their respective liquidity pool.
Recently we have been working with the banks and firms like Pragma to get feedback on venue analysis and fill rates and have subsequently used this information to complement our TCA. This gives us a different perspective than solely using sell-side TCA. We rely upon the use of algorithmic trading across a diverse set of venues and methods of execution and believe it is important to have a method to independently evaluate trades.
Using TCA in an increasingly granular fashion has been a great way for us to give feedback to our liquidity providers, particularly our banks. For example, a bank may be streaming into a liquidity pool and believe they are streaming competitive prices. Even if they are away from the market by the smallest of margins, we may not be transacting with them. We have excellent relationships with our banks and in these circumstances know they welcome any feedback. We have seen banks move up the ranks dramatically just by making minor adjustments to their price stream. It has been a good opportunity for us to have productive conversations and to provide valuable feedback based on our unique experience.
The business continues to be driven, in large part, by relationships though now we have substantially more data and metrics to enhance the dialogue. The banks appreciate and understand that the buy-side wants to execute as efficiently as possible and are viewing execution quality more objectively. They are moving towards increased transparency and using the information provided by us to improve executions and reduce market impact.
The spreads provided by the banks are, in part, due to how the market participant interacts with the liquidity provided. As we do more business and as they become more comfortable with our style of trading, the banks are more willing to engage in a mutually beneficial dialog. Listening to feedback makes them more competitive in the long run and we get better execution and more options for our investors. TCA plays such a pivotal role as we want to understand where/why we are dealing and how we can reduce costs for our strategies and for our clients.
The proliferation of non-bank liquidity sources means the market dynamics are changing and adds additional complexity. Relationships with banks will always be important and we value that they are a stable source of liquidity, but these additional liquidity providers are stepping up and taking a significant role in market making.
Increasingly, we are seeing emerging markets that were not traditionally traded electronically made available in execution algorithms. As liquidity improves and the average daily trading volumes increase, we would expect to see more markets begin to trade electronically and already pricing has begun to improve within certain pairs.
It is unlikely that we will ever be totally satisfied with our TCA. We are always looking for ways to improve – to look deeper into the microstructure of the market and to calculate where we can possibly gain an edge and thereby achieve better executions for our investors. We have achieved major successes in TCA already but there is more valuable work to be done. Our aim is to continue to evolve and, as the market changes, ensure that we always get the best execution for our clients by using TCA in the most productive way possible.
We’d love to hear your feedback on this article. Please click here