In its November 2020 impact paper, “The State of Capital Markets Post-Pandemic: Major Paradigm Shifts,” Aite Group reported on its findings from interviewing a range of market participants from banks, asset managers, hedge funds, exchanges, and consultants. It concludes that across the financial services industry electronification of regulated markets was already happening prior to lockdown but “…a lot of work needs to be done to expedite the adoption of electronic trading across all asset classes. The COVID-19 crisis amplified the resilience of high-touch trading techniques, particularly to move larger trades in over-the-counter markets, even as 55% of participants feel the crisis will eventually foster more e-trading.”
While the move to FX e-trading was well advanced, Aite’s Senior Analyst Yue Malan points out that the major multibank platforms such as FXall, FX Connect, Bloomberg and 360T have become increasingly active in promoting their automation functionalities. The moves to cloud trading or least cloud data storage and software as a service (SaaS) usage have become increasingly widespread. Linked with this she adds that cyber-security has become “a major focus for large organizations” as remote working remains, and technology vendors continue the migration to the cloud.
Transaction cost analysis (TCA) and the adoption of RegTech technologies have seen greater take up. Data storage, retrieval and analysis have played a key role in pre-trade automation. Looking forward Malan sees more innovation in FX trading. “Data for non-spot FX, NDFs, options and swaps are where there will be growth. There’s a lot of low-hanging fruit for those non-spot products. Less than half of them are electronified, they are still reliant on voice. People are getting into those non-spot products to innovate, to find an opportunity.”
If this gives a taste of what’s to come there is every reason why these things will happen quickly, according to Medan Gabbay, Chief Revenue Officer at Quod Financial a London-based vendor of technology for data-driven automated trading. “You have three things coming together. You have increased trust in technology, you have budget, because many firms are making money and you have lots of pent-up ideas of how they want to make changes and launch projects. I think there’s going to be a big boom in technology change. Many firms have a lot catching up to do after quite long period of Covid-enforced lockdowns and disruption. There’s an element of a huge amount of dead time.” His firm, he adds, has “never been so busy.”
Buy-side - FX hedging
Several people we spoke to in researching this article pointed to two specific areas that were ripe for change. On the one hand is equipping the buy-side with easier to use tools the other is overcoming their legacy systems with new capabilities to encourage them to use improved FX hedging tools. These are areas that HSBC has been addressing.
“Clients are looking to streamline their legacy FX hedging workflows,” explains Vincent Bonamy, Head of Global Intermediary Services, Markets & Securities Services at HSBC. “They are looking to reach utmost transparency and efficiency of their hedging outcomes to ensure consistent and stable performance, as well as to reduce transaction costs and inefficient work flow procedures. This has led to products being in demand that enable clients to achieve economies of scale by using outsourced FX risk management services. This relates to all kinds of market participants but, first and foremost, asset managers, asset owners and corporates.”
He explains that the firm’s FX Hedge Efficiency Analytics Tool (FX HEAT) helps clients to understand how their chosen FX parameters result in the realised returns of their FX hedging strategies. It compares the risk and performance metrics of multiple alternative strategies and breaks down the achieved results of FX activities into various constituents.
“In its first phase, it is to help asset managers to understand the difference of returns in currency hedged vs non-currency hedged investment products. It allows clients to understand how market factors and operational workflow procedures impact the results and efficiency of their FX hedging activities.”
FX HEAT also plays to buy-side compliance and oversight needs. “Market participants, especially those with fiduciary responsibilities, need to ensure consistent and good outcomes for their clients. This requires them to conduct market-related activities in an organised, documented and justifiable framework.
Our products, reporting and continuous service offering support clients in fulfilling their internal or regulatory driven oversight functions.”
Looking forward, Bonamy believes that automated hedging for non-alpha generating activities will be used by most market participants. He adds that there has been a shift by asset managers and owners towards a stronger focus on passive investment strategies (i.e. ETFs and ETCs) and that this is raising questions for many of them. “We believe solutions for these type of challenges will be highly in demand in the coming years.”
Sell-side – could do better
Who you deal with is an important consideration when it comes to e-FX. Not all sell-side offerings are the same but there is generally plenty of room for improvement, especially as multidealer platforms add more functionality of their own.
Describing how the sell-side could up their game, Keith Hill, a London advisory board member at Caplin Systems, a provider of desktop and mobile e-trading technology, describes a push and pull dynamic. On the one-hand banks are under pressure from new bank entrants and FinTechs and want to cut their costs as their margins are narrowing. The buy-side meanwhile is becoming accustomed to well-designed desktop solutions and easy-to-use mobile apps to make payments and do other FX business.
“In the past,” Hill explains, “banks have focused very heavily on risk systems, and trading, essential for managing their risks and getting prices out to clients. But they’ve under-invested in the tools that sales use when dealing over the phone with clients, trading on behalf of those clients when they call in, for example, for prices on more sophisticated products, or when sales involvement is necessary. So, now, I think they’re really focusing on the sales journey, and helping sales become more efficient, more productive, giving them more information about the client that is calling in. For new entrants, it’s easier to do if you’re starting from scratch. It’s quite difficult to do if you’re dealing with legacy infrastructure however.”
He goes on to say that there has been a ground shift in banks’ attitudes towards mobile applications for professional use. “I think the pandemic has had much to
do with that, in terms of allowing people to work remotely, and if you look back to the start of 2020, there were many banks who said no to sales people selling and traders trading remotely from home, in the professional foreign exchange market. Then, within weeks, it happened. The other point is, it didn’t really have a dramatic impact on the functioning of the foreign exchange market. It still functioned completely smoothly.”
So the sell-side is treading something of a tightrope between the need to cut costs but at the same time improve the convenience and the quality of the customer experience. Customers are used to speed and ease of use in their daily lives, the genie is out of the bottle. The sell-side has to play catch up, probably by outsourcing their development rather than building next-gen offerings in house, which may be quicker and less hampered by legacy considerations.
Automating currency derivative workflows
Automating derivative workflows is a work in progress but there have been significant recent advances according to Bloomberg’s Global Head of Foreign Exchange Electronic Trading, Tod Van Name, “Automation in derivatives has already impacted everything from auto-priced liquidity to account allocations. In addition, unique solutions exist to centralize and simplify the daily FX option expiry process by enhancing communication with counterparties through convenient dashboards, like Bloomberg’s Expiry Manager. The Expiry Manager provides clients with a robust automated workflow solution, replacing the manual process of exercising and expiring FX options and removing the need for re-keying expiry details between systems,” says Van Name.
He points out that there are different degrees of automation. “Many of our clients have already adopted solutions to optimize efficiency, reduce operational risk and reduce transaction cost. And for those who are just starting, they are eager to learn best practices. Even the most sophisticated firms are always looking for innovations to give them a competitive edge.” He explains that automation typically takes two forms:
- Workflow automation includes integration (often 2-way) between the order management system (OMS) and the execution platform, solutions for trade netting, account allocations, order staging, batch and block trading, multi-routing (sending many at one time), and, of course, straight-through-processing.
- Trading automation in its simplest form might be competing quotes via an RFQ process, but can now include API trading, Trading At Best, and rule-based trading that allows users to determine parameters and conditions under which orders are to be executed. Often referred to as low-touch trading, these solutions allow traders to focus on higher value trades while providing reliable and predictable process for routine trading.
“NDFs may be more challenging, given liquidity considerations and the addition of fixing sources, but we are already seeing many liquidity providers starting to stream pricing, offer algorithmic trading, and provide transaction cost analysis. Technology has not only accelerated price creation and distribution, but risk management for market makers,” says Van Name, adding that he sees no technical barriers to leveraging the same solutions available for deliverable currencies.
Thinking ahead, he concludes that in the coming years we are likely to see the increased use of natural language processing and artificial intelligence to create orders, optimize execution efficiency and speed processing. At some point, blockchain solutions may speed settlement.
Disruption – how will the trader’s job change?
In FX trading as in many other fields of work, increasing automation, machine learning and artificial intelligence can appear to threaten the nature of work as we know it. Established job functions and responsibilities morph into something different.
Jill Sigelbaum, Head of FXall Capital Markets, at London Stock Exchange Group’s Refinitiv sees a change in FX trading roles but by no means their disappearance. “The long-term trend towards further electronification is clear, yet we believe there will always be a place for voice execution. And it will be combined with pre-trade and post-trade workflow to streamline operational processes. The importance is to give traders flexibility in how they decide to conduct their price discovery and trade execution, giving them the choice to opt for a fully electronic workflow when appropriate, or to pick up the phone and get some market colour from their liquidity provider before executing a higher-market impact trade - potentially a very large trade, or a trade in an illiquid currency pair for example.”
FXall already offer traders the ability to electronically affirm trades agreed over the phone. This allows them to benefit from electronic trade capture, audit trails, and post-trade capabilities associated with electronic trading.
I asked Sigelbaum, if she could envisage a time when FX trading would be carried out by machines and the entire process from end-to-end would be automated. “While some of the most advanced participants may look to build fully automated end to end trading workflow,” she replied, “for most buy-side market participants, full automation will be for a subset of their trades. We see significant growth in execution methods for rules-based auto-execution workflow: a set of rules that automatically take orders from their resting state in the order management system, implements pre-set workflow, and executes them according to pre-set conditions.
Likewise broker algos: the use of broker algos to automate the execution of orders based on a set of parameters entered manually by the trader. For example, a trader will choose to execute a US$500 million order with a TWAP algo, which will split the order into smaller slices, and spread them evenly for execution over a specified time period, in order to achieve the time-weighted average price and minimize market impact.”
Gathering views from market experts, it soon becomes clear that automation in FX is evolving, and evolving all the faster for the market’s Covid experience. The market however is a broad church. Participants vary enormously in their nature, the results that they aim to achieve in their engagement with the market and the resources at their disposal to do it with.
The possibilities seem infinite however. For every aspect of market activity across pre-trade, execution, post-trade, over all reporting and compliance there are plenty of people with ideas, the knowledge and the skills to build and implement new forms of automation. Only one thing is for sure and that is that ever-increased levels of automation will apply and the chances are if we re-run this story 12 months from now, the picture will already have changed significantly. You only need to look at where we were 12 months ago.
At the end of June 2020, in the immediate aftermath of the first lockdown, the BIS sponsored Global Foreign Exchange Committee (GFXC) took stock of how the market had performed. Likening the market impact to that of the 2008 financial crisis they explained, that spreads had widened, volatility had increased sharply, liquidity had deteriorated for both spot and swap markets and funding costs had surged in many currencies. FX flows relating to asset rebalancing, margin calls, liquidity withdrawals and hedging sparked increased demand for US dollars and the share of internalisation used by sell-side market participants decreased.
“However, what was notable about this latest crisis,” the Committee continued, “was both how sharply these measures deteriorated and how quickly they recovered following timely and decisive actions by central banks and governments around the globe.” In general this assessment is echoed by market participants, now a year on from the volatility in the early months of 2020. Despite the rapid transition to WFH, they say that there was remarkably little market disruption nonetheless there is still work to do, according to AITE’s research.
“The new normal of the decentralized work environment has created the need for more structured communication rather than the split-second communication typical of a trading desk. Additionally, real-time surveillance is becoming more important. For smaller shops, a shift to more reliance on outsourced trading models is expected to gain traction,’ they found.
Smart order routers
In conjunction with the increasing application of algorithms in trading strategies, smart order routing is enabling more efficient sourcing of liquidity and placing of orders in FX.
“SOR is based on all the information that I have available now, and any historical information that I’ve gathered,” explains Medan Gabbay, at Quod Financial. “This includes hit ratio, volatility, cost of rejects, what I’m seeing now in the market. What is the available liquidity and the criteria I’ve been given, what venues are available or not, how do I make the best execution decision. That might be choosing to place the whole amount as a full amount; it might be choosing to split it across multiple streams, and those microsecond decisions are totally impossible for a trader to make, or even understand. They can review them in a TCA post-trade process, but to actually make those decisions as an executing trader is impossible.”
He goes on to add SOR gives confidence, both to the buy-side and the investors, but it’s currently being used unevenly in the market. “No sell-side would consider entering the market without smart order routing or algo-trading technology,” says Gabbay. “But most buy-sides really don’t think that they need that at all to set up a shop, and perhaps they don’t because of the nature of the market at the moment. However that’s definitely where it’s going, and they need to start considering how they can introduce this into their workflows safely and cost-effectively.”
SOR makes sense for all parties in the market, in their quest to get best terms at the optimal moment for their trade. There are still some cultural issues to overcome but Gabbay and his team at Quod believe that it will be only a matter of time before all participants use SOR tools.
From Siri and Alexa to on-screen bots, we are familiar with digitals assistants. They are still relatively rare in FX however and BNP Paribas is working to change this for its customers, making their FX trading lives easier. In March the bank announced that ALiX, its digital trading assistant that it launched in 2019, was expanding to cover its Cortex FX trading platform. Though not yet voice-activated, it can respond to queries on spot, forwards, swaps, options, and orders, in addition to algos.
In essence, a small ALiX widget sits in the corner of your screen and will respond to queries in a variety of language formats using Ai and natural language processing (NLP) technology to access data and pricing. Joe Nash, Chief Operating Officer for digital FX, at BNP Paribas explains that financial services have been a bit slow to adopt digital assistant technology that has become fairly common in the tech sector.
“We want people to be able to ask, “what’s EURO / Dollar market colour looking like today” and be able to get right up-to-date with what’s happening. Before ALiX, a client would have to pick up the phone and speak to someone. Now it’s at their fingertips,” he states.
ALiX’ NLP capabilities can handle colloquialisms and respond to pretty much any day-to-day market query thrown at it using its access to data in the bank’s FX research and market data systems. ALiX is evolving and learning, the more queries it answers.
Asked if he could envisage a time when digital assistants could execute transactions, perhaps with other digital assistants on the other side of a trade, Nash didn’t rule it out. ALiX has long way to go before it can do that but for the moment it’s a pretty clever digital tool that other market participants will be likely to emulate.