Trading in the Cloud – Has Its Time Finally Come?

The structure of capital markets and how they are traded are shaped through the technology adopted by market participants over the past two decades. And while some firms remain cutting edge, many still operate legacy infrastructure characterised by high cost and low flexibility. Facing falling margins and intense competition, can cloud technologies help provide trading firms with the agility they need to survive and prosper going forward? Let’s look at the state of trading in the cloud.

Major regulatory shake-ups (MiFID and RegNMS) have required significant investment in technology. This has pushed executions from screens and phones to electronic trading platforms as firms took advantage of the fragmenting liquidity landscape. Thus we witnessed a rise of algorithmic and high-frequency trading, which in turn precipitated large-scale investment in high-performance infrastructure, co-location, and connectivity technologies.

Subsequent post-Credit Crisis regulation restricted proprietary trading at sell-side firms, momentarily slowing the technology arms race. However, best execution rules – enshrined within MiFID II and Dodd-Frank – 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, ultra-low latency trading strategies.

While this changing landscape has created opportunities, it has also seen a major increase in costs and shrinking sell-side margins adding to trading firms’ competitive pressures. The situation is not helped by current inflationary, energy and geopolitical concerns.

Regulatory overhead, increased infrastructure investment driven by liquidity fragmentation, and a highly competitive execution landscape have left sell-side firms fighting it out for market share in over-brokered markets. Scale and/or innovation are often the only means for firms to differentiate themselves, and the ability to move faster in response to new opportunities has become paramount. Agility is no longer a nice to have; it’s a necessity.

To that end, several of the world’s largest exchange operators have made substantial moves to embrace the public cloud. CME Group teamed up with Google Cloud Services (GCS) to host many of its data services in a cloud environment. The London Stock Exchange Group (LSEG) has partnered with Microsoft to migrate key functionality to its Azure cloud infrastructure. The deal involved the software giant taking a 4% stake in the exchange operator.

While these initiatives are primarily focus on data, we are also seeing momentum in trading services. In 2021, Nasdaq announced a multi-year partnership with Amazon Web Services (AWS). The goal was to build its next generation of cloud-enabled capital markets infrastructure (although this will use the AWS edge computing solution rather than their public cloud offering). Similarly, Deutsche Börse announced this February plans to accelerate the development of its proposed digital securities platform with Google Cloud’s secure infrastructure. Aquis offers a public cloud version in AWS of its exchange platform through its technology sales division.

But despite these partnerships appearing to endorse the model, at least for exchanges, others remain unconvinced. ICE recently said it had no plans to migrate any of its critical infrastructure to the cloud.

So given the challenges and drivers, is cloud – and in particular the public cloud – going to play an important role in the evolution of trading infrastructure?

According to a survey of 20 leading capital markets organisations commissioned by Rapid Addition, the answer is broadly yes. The caveat is that public cloud isn’t for everyone, everywhere, all the time. But it does have an increasingly important role to play.

Parts of the trading workflow can move to the public cloud and other parts cannot. Cloud may be appropriate depending on function, with more latency-sensitive applications generally perceived as less suitable.

Or it may be appropriate for certain, often fast-evolving asset classes or subgroups. One respondent, for example, described a massive cloud infrastructure in production for the firm’s US municipal bond trading operation. The firm found cloud to be an appropriate platform for rapidly deploying algo analytics for munis when it first entered the space. Then, it needed the flexibility and scalability to quickly scale up operations.

However, with the business now established, however, this respondent questioned the ongoing need for cloud. They suggested that its now predictable workloads for pricing around a million securities overnight may better be suited to on-premises infrastructure since the elasticity element is less relevant. Such are the nuances of the applicability of cloud technologies to certain situations.

Concerns regarding latency and security have restricted some firms’ cloud-based activities to non-production environments or supporting functions in the wider trading ecosystem, with survey respondents describing use cases such as tick data storage, certain software applications, user acceptance testing (UAT), and middle-office functions like P&L calculation and risk management. However, the overfocus on latency (or deterministic latency) is perhaps overshadowing the growing adoption of cloud in certain asset classes and other critical aspects of electronic trading workflow.

Answering the question ‘Trading in the cloud – has its time finally come?’ is not, therefore, straightforward. Two things are clear: for certain use-cases, cloud can work; but it’s not the silver bullet for everything, particularly when it comes to cost.

This article originally appears on Trading Tech Insight as part of our partnership with The A Team Group.