By Mike Powell, CEO, Rapid Addition.
Latency / determinism of performance. Across the board, latency and performance was identified as a potential deal-breaker when it came to deciding whether to adopt a cloud-based strategy. A common view was that cloud is not yet capable of supporting low-latency or latency-sensitive demands. But with latency trading accounting for a small proportion of workflows, a hybrid model is possible, survey respondents suggested.
Data security / privacy concerns. Notwithstanding cloud operators’ state-of-the-art stance on data security, respondents were reluctant to place sensitive data into cloud environments. Many said their organizations were happy to designate their data as ‘hot’ and ‘cold’, the former of which would be kept in house where security could be controlled by the owner.
Need for control. Organizations feel compelled to maintain control over their trading infrastructure, and feel that cloud infrastructure hurts their autonomy. “We want to have more visibility down to the tin with regards to bursts and determinism,” said one respondent.
Complexity of migration. The trading firm may simply be unable to make the jump. Respondents cited the complexity of their own organizations and lack of requisite skills as potential barriers to migration. The culture change needed to execute such a transition was also cited as a handicap. “Adopting DevSecOps and the requisite people, skills, processes, tools and cultural change needed to fully gain the advantages of cloud, together represent a significant barrier,” opined one participant.
Cloud operators’ lack of clarity. Cloud operators have been criticized for providing too little clarity about the own technology and business plans, and that has been a hindrance to respondents’ efforts to assess the applicability of cloud environments.
Potentially higher operating costs. Cost and scalability are among the most common reasons cited for companies adopting a cloud-based strategy. But observers noted that once target levels of scale had been reached, cloud ceased to offer value. Having helped to start up lines of business or to enter new markets, the elasticity offered by cloud has become no longer necessary for some respondents and from then on has represented an unnecessary cost.
Concentration risk. When data is concentrated in a single or a small handful of locations, it becomes potentially vulnerable to power outages and the loss of business that could entail. “Concentration risk is the killer – the more we load in, the bigger the impact of an outage,” said one respondent. “If we have 20 apps in the cloud, it feels like a data center outage if it goes down. Fundamentally, the model is flawed. It’s like having everyone working on the same mainframe.”
The ‘why bother?’ factor. For some respondents, cloud felt like a solution looking for a problem. Yes, it was possible to use cloud for certain trading-related functions, but what was the purpose when there was already an acceptable solution in place? One example of this was order confirmations: “Why bother? You already have a connection for the original order message so why not use it for confirms? Handling them separately seems like an unnecessary complication.”
We’ll be exploring which trading functions are best suited to cloud, and what the future looks like in upcoming blogs, so stay tuned. Or else you can register your interest in the full report here and we will contact you to let you know when it is available later this month.
Subscribe to our newsletter