Growth in fixed income futures algorithmic trading at JP Morgan has accelerated rapidly in 2020 as buy-side traders globally turned to the investment bank’s machine learning-equipped algos to grapple with intense market volatility.
Speaking to The TRADE, Peter Ward, global head of futures and options electronic execution at JP Morgan, explains that while the volatility contributed to recent growth, adoption of futures algo trading has picked up pace with clients significantly in the last few years.
Since 2016, futures volumes traded via algos at JP Morgan has increased 40% year-on-year. In fact, algos now comprise of almost 20% of the bank’s total futures trading flow, up significantly from roughly 4-5% in 2016 and 2017, figures seen by The TRADE have revealed.
The period of intense volatility in 2020 due to the global pandemic played a key role in the cumulative buy-side adoption of futures algos as traders became more accustomed to on-screen execution and liquidity.
“When liquidity is harder to source and there is more volatility, execution performance becomes challenged,” Ward explains. “Clients are driven to look at the problem areas in executions and that’s when we consult with them to figure out ways to bring in that performance. Maybe they should consider trading at higher volume at the open or close, or perhaps sitting out the first five minutes on the cash open because of the noise. All of that we can customise for them.
“I think the more challenges clients see in execution, the more opportunity there is for us to come in and help them, and the solution is increasingly the customised algorithm.”
Customised algorithms have become particularly popular with traders in 2020 and in recent years. Volumes on customised algos at JP Morgan have roughly tripled in each of the past three years, alongside a 21% increase in the number of custom algos in 2020 to almost 50 customisations, up from close to zero in 2017.
The bank’s flagship liquidity-seeking algorithm, known as Aqua, is the most common foundation for modified client parameters and customisations. A classic example of customisation is where a client wants to follow a particular trading pattern but then switch urgency or strategy based upon predefined triggers.
JP Morgan rebuilt its algo platform around five years ago to provide the buy-side with more choice about the parameters they can set on their side for algorithms, and there are further customisations that the bank’s electronic traders can configure on behalf of clients. Ward adds this has allowed his team to have a “richer” dialogue with clients and demand is clearly there.
“There has always been demand for customised algos, even 10 years ago there was a lot of demand,” he says. “We just didn’t have a scalable way back then to adapt an algo to what a client really wanted. The reason for that is when a client wants something different, we needed developers to code that and then release it for implementation in the client’s platform, which takes a lot of time.”
Reinforcement learning
The Aqua algorithm has been a particular area of focus for JP Morgan recently. It uses a technology referred to as reinforcement learning to create advanced signals on order routing and placement.
With reinforcement learning, which is a form of machine learning, the algorithm essentially learns from itself over time by looking back at previous signals that it has generated and evaluates performance. The signals will dictate whether the algo crosses the market or stays passive.
Reinforcement learning technology was first applied to a recently launched model of Aqua that is focused on navigating quarterly roll dates when futures contracts expire. It can be a high-volume period and volatile time for traders as everybody is typically rolling in the same week to the next expiration date. In recent years, this activity has evolved from manual, voice-based trading to more electronic, low-touch trading.
“Previously, a lot of this business was executed through voice desks and one reason for that was because trading systems out there couldn’t handle multi-leg products,” he says. “As those systems have been developed in the last few years, we found more of that activity moving to electronic channels.
“A lot of volume goes through on calendar rolls and the challenge is around optimising that experience for the clients rather than imposing a model of trading without looking at the particular client objective.”
In response to the trend and client demand, JP Morgan developed a model of its Aqua strategy, known as the Roll Algo, which went live not long ago for the most recent US treasury roll in February. It has been especially popular with buy-side traders, according to Ward.
“The Roll Algo model focuses on maximising liquidity and pricing opportunities by using signals that help it understand when to cross the spread. It’s the most important area we are working on and has peaked the greatest interest from clients.
“It performed really well in February and there was a lot of client use in that period. With that, the algo learned a lot along the way so we can expect the performance in the next quarter’s roll to be improved.”
The Roll Algo is not the only new addition to JP Morgan’s new strategy line-up. Advanced strategies like Target to trade around the cash or futures close, Multi Leg Strategy for trading multiple instruments at the same time across futures and US treasuries, and options algos have also been developed by the bank.
Volumes in options on futures surged in 2020 as trading floors at major derivatives exchanges like CME that facilitate options trading were forced to shut down. As a result, liquidity shifted to low-touch and electronic channels and JP Morgan’s clients began to ask more questions about trading options through algorithms.
“Options on futures volumes have seen significant growth in the industry over the past few years and 2020 was a breakout year for liquidity on-screen,” Ward adds.
“With that said there are still challenges and nuances to trading them and that’s where we see opportunities to innovate and help our clients with their execution. This can be through simpler Peg and Cross type strategies and ultimately more targeted strategies using a delta or volatility reference.”
JP Morgan expects buy-side adoption of futures algo trading to continue increasing in the near future, having been driven by ongoing market developments and trends over the past few years.
Explicit regulatory requirements on best execution and growing appetite among the buy-side to address challenges in futures and options market structure have been instrumental in the growth of this trend. Best execution essentially forces traders to establish benchmarks to measure performance and trading through algorithms can provide an effective way to do this.
New products have also entered the market where liquidity is shared on multiple markets, which presents challenges in trading those products. Nifty derivatives, for example, are now tradeable in both Singapore and India after the Singapore Exchange (SGX) and India’s National Stock Exchange ended a two-year dispute which put SGX’s futures into question.
Other developments such as extended hours in futures markets also means there are now more hours to trade what is often the same amount of volume. Add periods of decreased liquidity and increased volatility to the mix, traders have progressively sought algorithmic strategies and automated solutions for consistent execution in volatile products, and when targeting cash settlement periods, for example.
It’s not just JP Morgan that is doubling down on efforts in futures algo trading. In January, rival investment bank Citi rolled out a suite of execution algorithms, including its flagship Arrival strategy, for futures markets across all major exchanges in the US, Europe, and Asia Pacific.
In contrast to JP Morgan, the electronic traders at Citi handle all of the algo customisations on behalf of clients. Head of EMEA futures electronic execution at Citi, Gordon Ball, said at the time clients don’t want to enter numerous parameters to execute an order. He added: “the complexity of operating an intelligent algorithm and fine-tuning customisations sits with us, so our clients can focus on their overall investment and trading objectives”.
Elsewhere, a start-up founded by former global head of trading at AQR Capital Management, Hitesh Mittal, launched its own suite of execution algorithms in early 2020 that aims to reduce costs for the buy-side with customised and high-performance strategies. In December, BestEx Research secured $5 million in funding as it prepares to roll out its algos in futures markets.
Amid the arms race in this space, JP Morgan’s Ward predicts the pace of fixed income futures algo trading adoption, particularly customised algos, will continue apace in 2021. It remains a significant focus at JP Morgan as different buy-side clients are also now using algorithms to trade futures.
In the past few years, the type of buy-side client seeking algorithmic execution has shifted from being a relatively small number of large hedge fund clients to the more traditional managers, including pension funds, asset managers and insurance companies.
“Five years ago, there were pockets of interest in executing this way, depending on the specific trader or firm’s appetite. It’s now become far more mainstream, driven by broader electronification in fixed income markets as well more investment firms adopting more explicit execution benchmarks,” Ward concludes.