Modern capital markets’ structure has been shaped by the adoption of new technologies. Major regulatory shake-ups, such as MiFID and RegNMS, drove huge investments in technology, pushing executions from screens and phones onto electronic trading platforms to take advantage of the fragmenting liquidity landscape. This sparked the rise of algo and HFT trading, which in turn precipitated large-scale investment in high-performance infrastructure.
Post-Credit Crisis regulation impacting proprietary trading momentarily slowed the technology arms race, however, best execution rules have ensured latency remains important. The proliferation of hedge funds and quant trading boutiques means there is still significant order flow being executed via high-frequency trading strategies.
While this has created opportunities, it has also resulted in increased costs and shrinking margins. Compounded by a highly competitive execution landscape, firms are left fighting for share in over-brokered markets. Scale and innovation are often the only means of differentiation, and the ability to move faster in response to new opportunities has become paramount. Cloud is increasingly seen as a key tool to help achieve this.
To that end, several major exchanges have made substantial moves to embrace cloud over the last few years. In 2019 the CME Group teamed up with Google to host many of its data services. More recently, the London Stock Exchange has partnered with Microsoft to migrate key functionality to Azure in a deal involving the software giant taking a 4% stake in the exchange group.
While these initiatives are primarily data-focused, there is also momentum in trading services. In 2021, Nasdaq announced a multi-year partnership with AWS to build its next generation of cloud-enabled infrastructure (although this will use the AWS edge computing solution rather than their public cloud offering). Similarly, Deutsche Börse announced plans to accelerate the development of its proposed digital securities platform in Google Cloud.
Capital Markets has been slow to embrace public cloud. Firms have been wary of issues regarding performance, privacy, and security, all of which can dampen cloud’s appeal even before the complexity of migrating legacy applications is considered. However, broad acceptance by other industries is highlighting some compelling benefits, forcing financial institutions to take cloud more seriously. And while the initial focus was data, we are increasingly seeing trading workflow targeted.
But what do firms’ real-world experiences look like? Are they realizing the benefits promised by cloud vendors?
It may be difficult for cloud to live up to the hype, and the advantages experienced by trading firms aren’t necessarily those they’d anticipated, but many are seeing genuine benefits:
Agility. Perhaps the key advantage, cloud allows firms to rapidly deploy new lines of business, exploit emerging opportunities, or experiment with new quant models while also retaining the ability to exit.
Elasticity. Removes the need to invest in expensive capacity to mitigate unforeseen volatility and data processing needs.
Spin up/down for testing. Enables rapid response to new trading ideas, opportunities, or customer requests without having to rely on slow supply chains or convoluted procurement processes.
Catalyst for innovation. Scalability gives organizations room to innovate, unshackled by on-prem infrastructure constraints, ultimately creating a competitive advantage in terms of IP and time to market.
Ease of Big Data integration. Helps unlock the full value of an enterprise’s data. Cloud can host huge datasets, support the computational power needed to process that information, and apply advanced tools to drive workflow automation and extract value.
Infrastructure cost savings. Cloud offers potential savings given its elasticity, while removing the complexity and overhead of managing on-prem environments and benefitting from the economies of scale that large cloud providers can achieve.
Despite the benefits, cloud adoption for electronic trading represents multiple challenges.
Firms need to be careful if cost is their key driver. High-volume, predictable workloads can work out more expensive, as experienced by a US bank’s municipal bond trading operation. Leveraging cloud when launching its innovative algo analytics capability, the firm achieved rapid time to market. But, once established, predictable workloads for pricing a million securities daily proved more cost-effective using on-prem infrastructure. Such are the nuances of the applicability of cloud for certain scenarios.
Latency performance and predictability are the most likely deal-breakers when it comes to cloud adoption. A common view is that cloud is not yet capable of supporting the latency-sensitive demands of electronic trading. While it is true that achieving ultra-low, deterministic latency is challenging, these concerns are perhaps viewed through the lens of highly liquid markets. Latency is less of an issue depending on the type of asset class and trading model or specific link within the value chain.
Notwithstanding cloud operators’ state-of-the-art stance on data security, concerns remain regarding the deployment of sensitive data into cloud environments. This has led some organisations to designate data as ‘hot’ or ‘cold’, the former retained in-house where security can be controlled by the owner. Financial institutions will have to get comfortable with cloud security and adapt current procedures and policies if they are to fully exploit its potential.
Similarly, data concentration increases business vulnerability. The large cloud providers invest heavily in infrastructure and service continuity, but are not insusceptible to loss of service, as was demonstrated on June 13th this year during a short but wide-reaching AWS outage. EU regulators are trying to address this risk by encouraging the development of multi-cloud services. Interoperability, however, is not straightforward. One exchange we spoke to highlight their implementation of a hybrid cloud/on-prem business continuity model as a possible way forward.
Perhaps the biggest challenge is the initial move to cloud. The complexity of legacy applications and rigid organisation structure can be a major barrier to migration. Along with a lack of requisite skills, the cultural change needed to execute such a transition can be significant. This often leads to a focus on new initiatives as a first step, putting off more complex migrations and operational restructuring until firms have more experience.
Despite these challenges, the inherent benefits of cloud architecture offer significant incentives for firms to move beyond the realm of data and look at opportunities across their trading workflow. When asset class and market structure are also factored into the discussion, it suggests plenty of scope to leverage cloud in the trading arena. Increasingly CTOs understand that there is no one-size-fits-all design for infrastructure – their role is to apply the most suitable technologies to address business requirements within budget and regulatory constraints. Cloud is a key part of their toolset in helping organisations move faster in response to changing client and market needs.