Worldwide Business Research (WBR) recently hosted its annual TradeTech FX USA symposium dedicated to all things foreign exchange. Rather than feeling the calming Miami breeze blowing through one’s hair, this year’s event brought industry professionals together in a virtual setting – a sign of the times as the global pandemic rages on. During the event, panelists brought a range of foreign exchange (FX) topics to light. Discussions were centered on what attendees might expect from the markets and technology in the future once Covid-19 subsides. During the event, analysts from Aite Group, and independent research and consultancy firm, used a series of polls to gauge market sentiment across a number of subjects including the use of advanced technology in trading and data analytics. This report focuses on key takeaways from TradeTech FX USA through the lens of attendee survey responses. A total of 773 attendees logged into Zoom from February 9th to February 10th. Of that total figure, 211 were from the buyside and the remaining 562 were other market participants from sellside firms, the vendor community, media outlets, and research firms.
Market participants attending this year’s TradeTech FX USA virtual symposium were inundated with rich discussions highlighting numerous developments in FX. The following section focuses on these topics and survey responses to Aite Group’s polling questions. Keeping up with technology proves to be more difficult than keeping up with the Kardashians.
As attendees readied themselves for the first day of the event, industry challenges were clearly on everyone’s mind and were certainly the focus of several panels. When asked which concern was most pressing, industry experts overwhelmingly pointed to the need to keep up with technology (Figure 1). Other items that go bump in the night included the lack of adequate data and regulations. Although still important, market volatility – an area the FX market navigated fairly well in early 2020 – and market fragmentation were lower on the list.
Finding smarter ways to interact with liquidity is the only way to interact with liquidity. There is much debate over what the trading desk of the future might look like. For instance, some traders believe their relationships with sellside banks that can internalize trades are the key to unearthing liquidity, particularly in times of market stress and volatility. On the other hand, proponents of electronic execution via electronic communication networks (ECNs) and automated trading look to central limit order books (CLOBs), algos, and various trading protocols to find the liquidity they need when they need it.
Most often, traders are keen to adopt improved pre-trade analytics to make better execution decisions regardless of the execution method chosen (Figure 2). Industry participants desire more robust algos, particularly beyond the trading of spot FX. Non-deliverable forwards (NDFs) certainly could use more love. The adoption of trade cost analysis (TCA) continues to motivate the development of more robust solutions to better identify optimal trading conditions and liquidity. Lastly, the emergence of peer-to-peer trading and the development of an FX tape are areas attendees believe will also lead to the improved sourcing of liquidity.
What does the buyside really, really want? FX is on the move from a quote-driven market to an order-driven market. As a result, relationships are at the heart of this shift. Knowing which counterparty a buyside firm can trust orders with in unpredictable markets has become a paramount consideration. Investors are tasked to find out if banks are doing the “right thing” or simply showing phantom liquidity. Meanwhile, banks desire high internalization rates and must prove the way they manage client flows is better than whatever their peers are up to. In all of this, the sellside has continued to find ways to add value by providing liquidity, transparency, and analytics.
Customers want more price transparency across FX products from the sellside (Figure 3). However, other demands were nearly as important. For example, investors desired instant liquidity and market microstructure influencing trade execution. Stronger connections with counterparties to control information leakage was also held in high regard as the early days of the pandemic put relationships to the test. Unsurprisingly, with increased demand for data across capital markets, requests for more data-centric solutions tied to liquidity analytics and macro data are on the rise. Lastly, on-demand tools to aid in the search for liquidity, such as pretrade analytics, continue to gain popularity and have shifted from nice-to-have to must-have status.
Algo adoption continues to grow as the need for more automation continues. As financial markets strive to find ways to automate more workflows, algos will continue to play a pivotal role in that process. The evolution towards pure automation is leading to richer data which, in turn, supports more automation. One can envision a world where order management and execution management flow seamlessly. Pre-trade analytics drive the routing of trades and algo selection, feeding post-trade TCA analytics. Some of this is already happening with smaller, low value trades. However, a fully-automated utopia may be some years away.
A majority of survey respondents believe the events of 2020, including the U.S. election, the global pandemic, and resulting volatility spikes, have motivated the adoption of more algo trading (Figure 4). This has led to the desire for changes in trading infrastructure including the adoption of algo and automated trading. Notably, more robust algos designed for the trading of NDFs and more electronification in the FX options market were areas which attract strong interest. The way restricted currencies traded with custodians and third parties are being executed is also coming into focus as investors look to maximize efficiencies and control costs.
Data, data, and more data are front and center in the FX ecosystem. Third-party solutions providers, the sellside, and the buyside have their work cut out for them when it comes to handling excessive bursts of activity in the market. The metrics and processes which underly workflows lie in data centers. The basic refinements or flexibility of systems which need to respond to market structure and liquidity functions that happen very quickly –what used to take months now happens in hours – are enabling investors to switch from low volatility environments where everyone can make prices to more tailored strategies to get business done by relying on relationship value.
These dynamics have driven market participants to invest in data science infrastructure and the use of data science platforms such as Databricks, H2O, and DataRobot. However, it is unlikely data science platforms will supplant data and analytics derived from more traditional sources like ECNs, sellside firms, and vendors in the near-term. Investors see this evolution happening in phases where data science platforms coexist with these sources for some time and adoption moves at a slower pace.
The emergence and adoption of the cloud has fostered faster analytical calculations and shorter implementation time for solutions designed to wrangle the vast amounts of data supporting FX trading and analytics. The cloud’s efficiency and flexibility has led to more adoption, particularly in cases where adoption has led to a more cost-effective business model.
While the majority of attendees Aite Group surveyed use public cloud platforms like Amazon AWS, Google Cloud, and Microsoft Azure for their data warehousing needs, there are many market participants which still rely on on-premise data storage and solutions. (Users of kdb+ databases tend to be part of the high-frequency trading crowd that would need the ability to store, analyze, and access very large datasets quickly.) Interestingly, a fairly large chunk of survey respondents indicated they aren’t using any of these data warehouse options.
The need for advanced technology grows; it’s time to give it some sunshine and water. Market participants want to get smarter about the way they execute their trades. As market conditions change, which sometimes happens very rapidly, buyside traders have to simultaneously adjust in-flight execution. This relationship sparks the need for more artificial intelligence (AI) and machine learning (ML). On the other side of the trade, liquidity providers rely on the use of AI when making prices. In the future, it is likely liquidity management becomes 100% automatic which will be supported by AI and ML.
Near-term, surveyed respondents believe the use of AI and ML will be concentrated in execution algos and, to a lesser degree, pre-trade analytics over the next 12- to 18-months (Figure 5). This intuitively makes sense as more of the execution process is becoming automated and will rely on AI and ML as market conditions change. Survey participants see areas like trade idea generation and risk management as being less influenced by AI and ML over the same period of time. Again, this also intuitively makes sense as the regulatory implications have a way to go before being worked out, especially on the risk management side. Regardless of the time line, it is evident that the use of AI and ML is set to increase.
Investors want to bring their TCA “A Game”. The electronification of FX that emerged over a decade ago has fostered a rise in more reliable data and TCA metrics for a range of FX products. Today, FX market participants are able to use TCA from a pre-trade perspective (mostly for algos) so they can rate how each provider has done, understand the market impact of their order, and feedback data into the selection process when they choose which providers they want to execute certain types of trades with.
The majority of respondents indicated they use a TCA solution for FX trading (Figure 6). However, there is a divide when it comes to build versus buy (Figure 7). Some firms use internal TCA tools when they prefer to have complete control of the choice of benchmark and how they measure execution quality. These firms have invested in data science platforms to understand the data from the venues. The majority of firms rely on third party solutions where the calculations are outsourced. This can be a very practical way to leverage a vendor’s quant team, data, business intelligence tools, support, and other resources. Finally, several firms indicated they employ both an in-house and third-party TCA system. This is usually due to data imitations and the need for customization.
TCA has moved far past check-the-box compliance. The majority of survey participants mainly use TCA on the trading desk to decide which brokers they want to send certain trades to. The rise in interest in peer data, which allows buyside firms to make apples to apples comparisons of their execution quality along the lines of firm size, trade size, strategy, and so on is a differentiator for third-party TCA providers.
Market participants are on the hunt for better ways to interact with liquidity. The solution to the liquidity problem is two-fold: buyside firms need to cultivate strong relationships with their counterparties who work to protect their trades and control information leakage while, at the same time, investing in superior analytics and data to make better decisions.
The sellside will continue to be challenged to find ways to add value to clients. While high internalization rates are attractive and often lead to better pricing, clients are still very concerned about price transparency across different FX products. In response, liquidity providers are tasked to offer better data and analytics in addition to proving best execution. Data is growing at an exponential rate. Recent bursts of market activity borne out of the pandemic and other externalities have pushed firms to invest in data science platforms.
Additionally, the reliance on the cloud has grown as firms shift away from on-premise data storage. Data will continue to be one of the most transformative drivers of change in companies and market structure over the next several years.TCA has been elevated by the proliferation of FX data and solutions available to both the buyside and the sellside. The adoption of TCA is expected to continue to grow as firms rely on these metrics for actionable insights, rather than just compliance box checking. Investors should continue to find ways to utilize pre-trade analytics to make execution choices like broker selection and choice of algos.