Treasury Today Country Profiles in association with Citi

FX Global Code driving treasury adoption of algo trading

Double exposure image of person and virtual financial graph

As the foreign exchange market embraces the FX Global Code, how can algorithmic trading and transaction cost analysis tools be used by corporates to demonstrate greater transparency, best execution and ethical conduct?

Over the past 12 months, participants in the wholesale foreign exchange (FX) market have welcomed the launch of the FX Global Code and its set of six principles that promote the integrity and effective functioning of the market.

March marked an important milestone when Shell became the first corporate treasury department to publicly commit to the FX Global Code. Shell’s commitment is significant for a number of reasons.

Firstly, non-financial institutions, such as corporates, are active and sizeable market participants, trading approximately US$127trn a year, according to the Bank for International Settlements.

Secondly, when the Bank for International Settlements (BIS) established the Foreign Exchange Working Group to begin writing the FX Global Code in July of 2015, it was primarily directed at those firms providing trading services to asset owners, like corporates.

“For a corporate treasurer to now commit to the Code sends a public message to its service providers – banks, brokers and vendors – of the behaviour it expects from them,” says Curtis Pfeiffer, Chief Business Officer at Pragma Securities. “Doing so opens and encourages new ways of doing business, and advocates for greater use of algorithmic trading and TCA [transaction cost analysis].”

Setting the benchmark

While early versions of execution algorithms were limited to automating relatively straightforward trading instructions, and traded only on one venue, the technology and logic underpinning algorithms has advanced significantly over the past decade as the FX market has evolved.

“Today the speed of trading, coupled with the number of venues and increasing variety of order types, make it impossible for a human trader to replicate algo behaviour on one order, let alone if the trader has multiple orders to trade simultaneously,” explains Pfeiffer.

This means a high performing execution algorithm will assess the current market situation, the available sources of liquidity at a given point in time, and execution strategies available. The system will then select the optimal routing decision and execute a trade in the most efficient manner while minimising market risk – all without the involvement of a human.

In the recent Greenwich Associates’ study, 58% of traders (including corporates) found that algorithms materially reduced overall trading costs. Greenwich also found that over a quarter of FX traders believed algorithms enabled them to have more time to spend on complex orders.

“Thus, in addition to improving execution quality on an order-by-order basis, algorithms have the indirect potential to improve performance on more complex orders by allowing corporate treasurers to spend more time on them,” says Pfeiffer.

Execution quality

The Code is also leading to greater demand from corporates to measure and evaluate execution quality. One way of achieving this is through TCA tools.

“Algorithmic trading lends itself well to TCA because the entire order chain, from the first order sliced to the market, to the last order slice, is logged in databases and can be easily extracted in order to analyse the orders’ performance versus the quotes in the market,” says Pfeiffer.

He adds that historical trading data can also be reviewed against several metrics and factors. These include price benchmark, trade duration, venue traded, currency pair and trade size. “This enables corporates to measure and understand if they are achieving high-quality execution on trades and where they can improve their trading processes.”

A solution for all?

As the market becomes increasingly electronic, algorithmic trading continues to increase in popularity.

Currently, it represents only 10% of dealer-to-client FX volume, but one in four of the largest institutions use algos where it can represent 25 to 30% of their volume.

It is also noteworthy that the strongest increase in algorithmic trading comes from smaller corporates trading less than US$1bn a year, according to another Greenwich Associates study into the topic. This refutes the argument that algos are only relevant for the largest corporates.

“The evolution of FX trading that has led to the use of execution algorithms will continue to gain momentum,” concludes Pfeiffer. “The execution quality and the degree of transparency available to traders through algorithmic trading are in line with the spirit and letter of the FX Global Code and provide a useful framework for corporate treasurers to demonstrate best practices.”

Algorithmic execution – the core benefits for corporates

  • Breaking up a large order into multiple smaller pieces can lead to paying less than for one large trade (or principal trade) where there is a premium paid for immediate execution and risk transfer.

  • Algorithms that access multiple liquidity pools effectively narrow the spreads being traded on.

  • Systems aid transparency by offering time stamps on orders and an easily accessible audit trail, making it easy to measure execution quality against a corporate treasurer’s benchmark.

  • Algo trading can handle more orders reliably, freeing up treasurers’ time to focus on areas where human intelligence and judgement add the most value.

  • Algo trading reduces information leakage as the wider market is not aware of the full size of the order. An algorithm only sends small slices of a trade to the market. This is especially important for larger orders which could impact the price of the currency if the market knew about the full order size. This is especially important for corporates who have more predictable trading patterns than asset managers and hedge funds.