Global Head of Trade Automation and Analytics at Bloomberg L.P.
What is separating the best in class Fixed Income leaders from the pack in terms of efficiency?
The leading firms show a genuine desire and aptitude for adopting new tools. They have an acute awareness of the technology available to them, internal or external, and are taking action to identify areas they can automate. This means they can free up time for more complex trades. Time is a rare, precious commodity on the trading desk.
The key concern for those looking to automate their trading desk is data quality. How should this be addressed?
Automation is only as good as the data you’ve got, and implementation becomes much more challenging as data quality goes down. All the market wants better data but there’s no silver bullet here. It’s a combination of trying to get as many different data sources as you can and then scrubbing that data. It’s an ongoing challenge.
How should firms approach the implementation of automation?
Firstly you need to have a clear idea of what you want from a technology provider, and then you can draw up an automation strategy based on your priorities in terms of flexibility, control and cost to deploy. Have clear targets and define the variables with which you measure success. The solution needs to be integrated into your existing architecture, as the orders for the automation need to come from the order management system. Finding a company with a good track record and support team will make this collaboration and integration much easier.
There are so many opportunities for automation that it can be difficult to know where to start. To escape the ‘analysis paralysis’, the easiest way to begin is by running a pilot. Start with something which is liquid, like treasuries, and start with small-size orders, then create simple rules and monitor them carefully. It’s human nature to fear the unknown, but automation doesn’t happen overnight. By phasing automation in and closely monitoring progress, firms will mitigate risk greatly.
What do you think is the biggest change we’ll see over the next five years to the trading desk?
A much higher percentage of trades will be done through automation across the globe, that’s a fact. The role of the trader is going to have to shift to accommodate this. As technology becomes more advanced and continues to free up time on the trading desk, the skill sets firms require from traders are going to be more clearly divided into complex, high-touch trades which can’t be taken on by machine automation, while on the low-touch side of things, automation implementation is going to become much more technical, and traders will spend more time deploying, monitoring and optimising roles.
Why do you think Europe is such a powerhouse for financial innovation around AI and ML?
There is more of an appetite here to utilise technology, and to gain the competitive edge through innovation, and a greater percentage of trades are automated in Europe, which puts it in a good position to start experimenting with automation. The regulatory environment has played a big part in this attitude towards efficiency when it comes to automation.
I think the general perception of AI is very healthy in Europe. New technologies are seen as a necessity to stay competitive, and, while traders are aware that this is something that they need to be retrained for in order to make the most of it, they are less fearful of automation posing a threat to jobs.
What is the biggest challenge facing Heads of Trading who are seeking to implement automation, and how should they approach it?
One of the biggest concerns is of unintended consequences through the deployment of automation in what is a highly complex environment. This can be mitigated through the appropriate use of software and planning. The trading desk has a lot of moving parts, so implementing automation requires a long-term plan.
By comparing the first few automated trades to manual trades, you can find out whether they were optimally executed. You can analyse how a human trader would have acted differently and feed that information back into the rule that you created. While managing risk across the desk is a highly iterative process, the investment in establishing a strong framework can give traders greater confidence and control, enabling them to apply more complex logic to fully realise the benefits of their automation strategies.
To help clients implement fixed income automation strategies, Bloomberg has developed automation tools, including Bloomberg Rule Builder (RBLD), which enables fixed income execution traders to create rules that facilitate fully automated trading on Bloomberg venues. With complete control, Rule Builder can break down orders into high- and low-touch, augmenting the trader’s role and freeing you to focus on more strategic activities.