Algorithmic Trading in the Future: 5 Major Trends


1. AI/ML-based dynamic parameter selection

How to effectively utilise the available trading techniques, execution procedures, and algo capabilities that a trader may not be aware with is the challenge facing desks today. Algo trading necessitates continual tweaking and frequent modifications to the trading environment, making it exceedingly challenging to stay current.

To now, it has been the job of the algo providers themselves, whose sales teams try to convince traders of the virtues of particular algorithms in the hopes that the trading desk would select the appropriate algorithm at the appropriate moment. This approach has been shown to be unscalable and unworkable as one unfavourable outcome might cause a trader to choose for predictable and simple-to-understand algo behaviour for months at a time.

More trust between traders and customers will develop as algos become increasingly specialised to the instrument, market circumstances, and available credit present during the deal, opening up a world of options for enhancing performance and motivating algo providers to invest in their systems.

2. Real-time integration of Transaction Cost Analysis with algorithms (TCA)

TCA is now ingrained into both the sell-side and buy-side procedures and workflows, so traders are having a comparable amount of trouble finding out how to use it on their desks.

Orders are routinely assessed against certain performance indicators and standards. TCA, on the other hand, simply measures a result. It offers the trader no guidance on how to improve any unfavourable trading outcomes. It has become nearly hard for a trader to connect poor TCA outcomes to what variables, in a variety of complicated systems, may affect the trading outcome as trading technology and algos have grown more sophisticated.

A greater focus will be placed on real-time tools that translate TCA results into workable system configurations. The key component that every trading desk will need to turn TCA from a pat-on-the-back when things goes well into a fundamental tool affecting trade results is that feedback loop.

3.Algorithms will become increasingly prominent across several asset types

The expanding usage of algos across asset classes, including cross-asset automation, is a noteworthy trend that emerged in 2021/2021 and that we predict will continue to gain steam. This is a continuation of the trend away from basis point fees for Tier-1 brokers and makes use of the fact that multi-asset trading technology can make these trading methods accessible to all desks at no extra cost. Simple automation techniques like Smart Order Routing (SOR) on FX Forwards and comprehensive algo trading methods on derivatives are two examples.

Hayley McDowell and Peter Ward, Global Head of Futures and Options Electronic Execution at JP Morgan, endorse this development and talk about the expansion of the fixed income (FI) futures algorithm in an article published by The Trade. The Market indicates that "Since 2016, JP Morgan's futures trading volumes have climbed 40% year over year. Algos now make up approximately 20% of the bank's whole futures trading volume, a considerable increase from the previous 4-5% I "2016 and 2017."

In another instance, JP Morgan reports that in 2016, customers used algo orders to trade almost 35% of their eFX Spot volume. This grew to approximately 60% in 2020, signalling a turning point in the implementation of FX algorithms. The benefits of FX algos will be difficult to overlook.

4. Pre-trade suggestions for enhanced performance understanding

Traders are experiencing information overload and are limiting their use to their top 4–5 preferred algos as the menu of algos and the range of programmable options continues to grow.

However, it appears that asset managers are less inclined to transmit greater volumes and are choosing to use Implementation Shortfall (IS) or Auctions to send lower notional quantities. In fact, to enhance the quality of their executions and assist their clients in lowering trading costs, market players have chosen algos to divide up huge orders into more manageable chunks that may be disseminated over several venues and executed against multiple liquidity pools.

In this case, any other algo type's child slice may be a peg that employs a dynamic offset from a specified benchmark driven by AI/ML forecasts of micro-volatility and market occurrences. By integrating it with predictive algorithms to foresee minute market movements, the peg itself automatically boosts the efficacy of passive trading within any algo, enabling traders to earn a few more basis points (BP) on all use cases. This enables brokers to add new aspects of algorithmic trading without having to engage in the difficult task of persuading end users to believe in more complex algorithmic trading behaviour.

5. Using algorithms to increase automation and efficiency

Finally, the expansion we have witnessed includes automation. Automation is the only answer as trading desks are under pressure to process customer orders more quickly, with fewer traders, and in higher quantities. Simple routing instructions have developed into context-based rulebooks that take market conditions into account. Volatility, depth, hit ratios, latencies, and real-time customer data like P&L over time, recent orders, and current order books are a few examples.

Building these intricate rules enables complete flow automation and allows for the prescription of certain strategies and paths. In this manner, dealers can hasten the decision-making process at the time of trade. Cross-asset behaviour such as the automatic production of FX legs for orders in non-listed currencies might fall under this category.