News

The Technology Behind Successful Algo Trading Implementation

At this time of major disruption for energy trading and technology teams, it can be challenging to allocate quality time to think more broadly about what’s happening to markets, risk and operations – and consider the implications of the macro trends facing the industry, including digitalisation, automation, AI and algo trading.  Luckily, that’s exactly what I was able to do at Energy Trading Week, which took place over two days, entirely online of course, and I have since been reflecting about the content of the event.

 

 

By Rajesh Jethwa

01/07/2020

At this time of major disruption for energy trading and technology teams, it can be challenging to allocate quality time to think more broadly about what’s happening to markets, risk and operations – and consider the implications of the macro trends facing the industry, including digitalisation, automation, AI and algo trading.  Luckily, that’s exactly what I was able to do at Energy Trading Week, which took place over two days, entirely online of course, and I have since been reflecting about the content of the event.

 

 

What struck me the most about the conference were the discussions about algo trading.  The session on “Secrets to a Successful algo Trading Implementation”, featuring industry experts Alex Raguz, Head of Algorithmic Trading at ElbOil, and Erdem Sezer, Manager of Analysis and Trading Strategies at Aksa Enerji, was particularly thought-provoking.

There was an interesting discussion on the benefits of algo trading, including hedging risk within revenue streams, the removal of human and emotional factors in trading, and the increased speed algo trading enables, and a firm consensus emerged that algo trading needs to be underpinned by an effective risk strategy.

Both Erdem and Alex agreed that “you need a robust, well-structured IT infrastructure” before embarking on algo trading while maintaining a healthy appetite for risk.  Critical issues to look out for include governance – who owns and makes decisions about infrastructure – and the assessment of on-premise versus cloud solutions – you can have a bit of both, with computation for analysis on cloud and execution in-house.

It was also telling that the participants surveyed noted the “lack of good quality data” as the top reason for not attempting algo trading.  Indeed, on several occasions, the panellists spoke about the importance of having clean data for reliable signals and the necessity of assessing the “actual” versus the “theoretical” algorithmic performance and course-correcting accordingly.

Much of the discussion centred around data.  Getting clean data is a crucial element of the secret behind performing effective algo trading and this marries up with our experience at Digiterre.  We have been helping energy trading companies establish a high quality “single version of the truth” data stream, from multiple data sources, to derive insights from analytics, algorithmics and the robust governance and human oversight around these platforms.  These themes came up repeatedly during the panel session and the networking afterwards.  It will be fascinating to see how they develop in the future.