With the implementation of MiFID II more than a year ago, European hedge fund trading “underwent a tectonic shift,” according to a Greenwich Associates report. The new regulations, which unbundled research and other “soft” services from the commissions bulge bracket brokerage firms charged institutional clients, were designed to take make brokerage commission costs transparent and lesson the role relationships played in decisions. These changes are likely to come to the US, Richard Johnson surmised in the recently released report “Trends in Global Equity Electronic Execution.”
But could the free market already be doing the job?
The regulatory changes in Europe have had their desired effect to various degrees. Relationships are “less valued on a relative basis,” Johnson wrote, pointing out European traders value low-touch service and market structure expertise to a higher degree than their US counterparts.
MiFID II is responsible for focusing decisions based execution performance and electronic trading methods, lowering costs and thus improving returns performance for pension funds, endowments and other retirees served by institutional fund managers.
But due to technical advances, this switch to lower cost execution methods is already taking place in the US and Canada without the prompting of regulators.
Technological advancements are driving meaningful changes in North American trade execution. Liquidity sourcing algorithms use cold mathematical analysis to route equity trades to the lowest cost venue regardless of relationship or value-added services – and such methods are on the rise.
In 2018 electronic trading comprised 41% of total volume in the US and 35% in Europe, up 5% in each region when compared to 2015. What is meaningfully different is how technology rather than humans are making the decisions.
Leading this charge is the “algo wheel,” a methodology that uses artificial intelligence to monitor trading conditions on a wide variety of electronic venues and determines the lowest cost execution at any given moment. When conditions change – a bid / ask spread on one venue slightly adjusts, for instance – the wheel automatically changes trade venue. The algo wheel is not designed to take away human responsibility, as parameters around the selection process can be customized by the trader – which includes sending trades to high-touch human brokers if that is the best option. Of those using electronic trading methods, 22% are using the method which is far from a random selection process as the term “wheel” may indicate. In fact, a benefit of the technology is that it justifies trade execution decisions based on best, lowest cost trade execution.
While the algo wheel has the potential to level the playing field for any trade venue that offers best execution, currently the bulk of electronic trading flow is centered on three primary brokers, 66% and 71% in the US and Europe respectively.
In addition to algo wheels, artificial intelligence and machine learning techniques, colloquially known as “algo bots,” are being used to “trade on the fly” based on incoming data, a practice showing promise in trade execution.
As brokers become AI ready, so, too are the trading venues. IntelligentCross is an ATS that is “more than a dark pool,” Johnson noted. The trading venue utilizes machine learning to optimize the timing of trades with the goal to minimize market impact while improving liquidity.
The use of advanced AI-infused algorithms is not mainstream, yet. Fully 73% of the 256 institutional investors surveyed by Greenwich Associates note that they have not used advanced algorithms for trade execution. But a large percentage of those who have tried the algorithms say they performed better than standard algorithms.
If traders deploy increasingly sophisticated algorithms to make trade routing decisions, will a North American version of MiFID II be needed?
This article first appeared on ValueWalk Premium