Your browser is unsupported

Please visit this URL to review a list of supported browsers.

Adding cloud to high performance trading infrastructure

The benefits and challenges

Adding cloud to your high performance trading infrastructure can help offer scalability, faster development and lower capex costs, according to participants at a recent webinar hosted by Trading Tech Insight.

The trend is clear: financial services as a sector is migrating more of its critical systems to the cloud while vendors are offering cloud-based solutions. That’s the consensus from a panel which discussed cloud adoption and challenges for those looking to make the shift.

Since ICE launched cloud-based historical data products and its Cloud Connect Program three years ago, the growth in user numbers and use cases has been strong.

But as high-performance trading businesses start to shift to cloud, can technology keep up with the required speeds of data delivery and ultra-low latency performance? And what’s driving decisions to add cloud to high performance trading?

Audience polls during the webinar found nearly two thirds of respondents had added cloud to their high-performance trading architecture to a ‘good’, ‘great’ or ‘greatest’ extent.

The ability to scale was singled out as the leading benefit of adding cloud, named by 65 per cent of respondents, followed by lower capital spending (18 per cent) and improved competitive advantage (12 per cent). Just 6 per cent of respondents claimed that adding cloud to their trading architecture provided no benefits.

Scalability and time to market are important advantages. Lifting capacity by adding physical servers and networking connections takes time and planning, while spinning up new cloud services can add capacity quickly.

What are the challenges of adding cloud to high-performance trading infrastructure at your organisation? (multiple answer)

But the panel suggested that businesses can find additional, unsung benefits from moving to the cloud, including the use of cloud services for risk management, data analytics, monitoring and artificial intelligence and machine learning.

“Traditionally when people migrate to the cloud, they’re using computing and storage technology,” said Leon Liang, ICE’s director of feeds business development. “But for trading and financial firms, there are a lot more offerings.”

Liang said that the monitoring systems available in the cloud could be very powerful for trading firms, monitoring latency throughout the trading process - a benefit that was sometimes overlooked.

Respondents to the audience poll also identified challenges in adding cloud to their high-performance trading infrastructure. These included the complexity of their trading infrastructure posing challenge for adding cloud, while some said they struggled to access the skills required.

Liang commented that the rapid adoption of cloud was being driven in part by system vendors moving to the cloud and encouraging clients to adopt the trend. He added that the difficulty of finding talent was exacerbated by the attractiveness of multinational cloud providers as employers.

Which of these technologies are you focusing on to manage unstructured data?

In addition to complexity and staffing, there are further challenges - with latency raised as a concern by many users. Panelists noted this could be split into latency to aggregate data, and latency of distribution - the latter arguably a more critical issue for high-performance trading clients.

Location was highlighted as another key decision - whether to store data close to a business, trading venue, or data vendor access point. This was noted as especially pertinent for financial clients considering low latency market data in the cloud. It was recognized that the biggest banks were adopting multi-cloud solutions to handle this, with different systems across the different public clouds. “Is your data vendor able to deliver real time market data into different public clouds?” Liang said. “These are questions you need to answer.”

Liang suggested steps to consider for moving to the cloud:

  1. Prepare data requirements. This involves understanding whether your current data requirements – both real time and static – can be met in the cloud environment you are considering.
  2. Carefully consider location. The physical location of cloud servers can be critical to success.
  3. Map infrastructure. This includes understanding every system in the current deployed environment and whether it will translate to cloud. But it also means knowing about any additional services you may wish to use like AI/ML, data monitoring, data analytics and other SaaS services available in the cloud.
  4. Finally, consider the total cost. The immediate cost offered by cloud providers can be confusing as it sometimes included credits for successful migrations that hide the true ongoing price.