The pandemic has accelerated the trend toward adopting cloud-based market data services and computationally intensive risk analytics, according to analysts on recent capital markets webinars.
Despite past concerns over latency and security with the public cloud, many banks and asset managers are using cloud-based market data services and data analytics for historical back-testing and risk-based calculations.
But this is happening against the backdrop of the COVID-19 outbreak last March when buy-and-sell side firms that had cloud-based applications, access to data, and network connectivity via the cloud, had the flexibility to shift to remote work and disaster recovery sites with little or no disruption.
“We’re seeing cloud as a key element for reducing cost, increasing agility, and increasing the level of responsiveness across organizations,” said David Easthope, Senior Advisor, Market Structure & Technology at Greenwich Associates on a Jan. 13 webinar about top market structure trends in 2021.
The cloud is also associated with access to modern technologies, such as open application programming interfaces (APIs), alternative data, and artificial intelligence. “The potential for AI and machine learning and combining those two trends in the cloud is a powerful combination,” said Easthope.
Market Data in the Cloud
Financial firms tend to test the cloud waters with historical back testing. For example, Refinitiv provides a history of tick data in real time on the Google Cloud, noted Easthope, who also spoke on a Security Traders Association (STA) call about market data and the cloud. It covers both OTC and exchange-traded instruments across 500 venues and goes back to the 1990s.
Among the reasons why cloud has gained momentum, users can get access to larger data sets potentially faster and potentially at lower cost, said Easthope. Just as Netflix grants access to streaming services, like movies and TV shows, what this means is that the data stays in one place and the vendor grants the customer access to data using identity access management.
Even exchanges and data vendors that generate huge revenues from market data are shifting deployment to the cloud, citing lower costs, on-demand access, and no need to procure hardware or install networks. Once a firm is cloud configured, adding a new data set to trade a different asset class such as futures is like adding an additional TV channel.
CME Group is distributing a real-time market data managed service known as CME Smart Street on Google Cloud, by curating certain topics and this is good for streaming analytics, noted Easthope. Last April, Nasdaq launched Nasdaq Cloud Data Service, providing real-time equity, index, and fund data, through cloud delivery, including cloud-based APIs.
According to Brad Bailey, Research Director in Celent’s Capital Markets Practice on the research and advisory firm’s webinar, the cloud has changed the way capital markets operates. “Simply changing the delivery model to the cloud for the massive datasets like the tick-by-tick data has changed the way that quantitative modeling is taking place at the buy side, hedge funds, and sell side, and that is profound,” said Bailey.
The rise of data marketplaces is another way to access large data sets and build on top of them. On March 11, Wall Street Horizon, a boutique firm supplying high-quality corporate events data, said it would deliver its data feed over the IEX Cloud to reach institutional traders and investors. Launched in 2019, IEX Cloud recently upgraded its platform technology and incorporated new data sets to serve a growing list of enterprise customers, professionals and retail investors, according to the news release.
In Wall Street Horizon’s case, institutions who use IEX Cloud will gain access to its clean, normalized events data alongside other bespoke fundamental datasets to help create trading and risk strategies. “The appeal of the IEX Cloud is that it offers a pull-down menu approach,” said Barry Star, CEO of Wall Street Horizon in an interview.
“An institutional trader can go to one place to get this plethora of data across the spectrum. It might be tick data, it might be corporate events, it might be insider trading data, it could be credit card receipts. You get it all in one place,” said Star. Another benefit is that cloud providers building data marketplaces can reach a larger population than a single data provider can reach on its own. For example, IEX surpassed global 150,000 registered users across 120 countries.
However, cloud adoption is moving beyond market data delivery. As the buy side increases its comfort level with cloud security, the technology has expanded to trading applications and other types of financial market infrastructure.
Hedge funds have been among the first movers in adopting software-as-a-service (SaaS)and other types of managed or hosted services, involving the private and/or public cloud. “We talk to hedge funds whose OEMS is more SaaS-based,” said Easthope, referring to order and execution management systems (OMS/EMS).
In February, The Trade reported that Lyxor Asset Management, one of France’s largest asset managers, had implemented the cloud-based EMS FlexNOW from FlexTrade, which combines cloud, desktop, mobile, and API technology.
“Integrating a new EMS in our complex workflow is never easy but with FlexNOW, we have replaced our old solutions and greatly improved the efficiency and operational safety of our workflows, especially when working with larger baskets,” stated Samy Debbah, Head of Dealing Desk at Lyxor Intermediation. The system gives firms the ability to trade thousands of stocks and futures across brokers and algos in seconds.
Even though banks are conservative and highly regulated, several big banks have embraced the cloud. At SIBOS, a large industry conference, JP Morgan Chase CEO Jamie Dimon told the global audience that 20% of the bank’s applications are cloud-based. In December, Deutsche Bank signed a multi-year, strategic agreement with Google Cloud to accelerate the bank’s transition, citing faster application development, use of AI and data analytics, and flexibility to respond to pressing challenges.
Deciphering the Cloud
While banks are migrating applications to the cloud, they are not necessarily rewriting everything. In fact, there are multiple ways to implement cloud. For example, there is a difference between plugging into a SaaS-based application through an Internet connection and building on top of cloud-based infrastructure, explained Celent analysts on the webinar about cloud in the capital markets and adoption speeds.
“You can migrate to it, you can consume it, and you can build upon it,” said Monica Summerville, Head of Capital Markets at Celent in London, describing the various methods on the firm’s cloud webinar. For example, infrastructure-as-a-service or IaaS is a technique that firms use to “lift and shift applications.” This means firms can provision applications by abstracting the underlying infrastructure. Another variation of cloud computing known as Platform-as-a-Service (PaaS) includes toolkits that developers can use for application development, web streaming, and analytic tools.
Rather than leaping into the public cloud, many banks and brokers are running operations on private or hybrid clouds. Private cloud consists of a set of technologies which a firm runs on their own premises, and they are still managing it. “You are getting the benefits from the agility and elasticity point of view, and teaching organizations the culture of cloud,” said Summerville on the webinar. Since hybrid cloud provides a seamless interface between public and private cloud, firms have the option of “bursting into public cloud if you run out of resources,” said Summerville.
To avoid lock-in or risk exposure to any single vendor, financial institutions are taking a multi-cloud approach. “They have made an enterprise decision to use different vendors for different things,” said Summerville.
Nasdaq’s cloud systems and apps are hosted in data centers operated by third-party tech giants, including Amazon.com, Microsoft Corp., and Alphabet’s Google, “which enables users to scale computing needs based on demand, more easily than in their own data centers,” reported the Wall Street Journal in ‘Nasdaq Tech Chief Credits Cloud with Helping Manage Market Frenzies.”
Analysts agree the message for capital markets is multi-cloud and cloud agnostic. “It’s not just Google and Amazon, but IBM is a big cloud provider too,” said Easthope. In the real-time market data realm, Xignite is offering market data APIs through the IBM cloud, while ACTIV Financial is offering data through a BT Radianz managed service, and MayStreet, which runs the SEC’s Market Information Data Analytics System (MIDAS), provides cloud-based data delivery.
The Journey Forward
However, analysts maintain that banks and asset managers are in the early innings of migrating to cloud computing. “Right now, less than 1% of cloud loads are in front office. The bulk [of applications] in the sell side are with risk and the buy side with quantitative analysis,” said Bailey. In a high regulated industry with multiple applications and complexity, migrating to the cloud is going to involve shifting a web of connectivity and infrastructure, he noted.
Still, analysts predict the journey to cloud computing will continue.
While banks still need to be conservative, said Easthope, in today’s remote world, they also need to get on-demand computational resources to solve specific issues within banks across the trading and risk domain, he said.
As banks get more cloud friendly, Easthope said he expects the rise of trading and risk infrastructure, and particularly software applications and systems to support this function moving to the cloud. This includes connectivity from SaaS-based models and FPGAs (field programmable gate arrays) moving into the cloud to speed up risk calculations.
However, cloud migration does not always immediately reduce costs, cautioned the Greenwich analyst, who said this usually occurs over a period of years. “It gives agility relatively quickly, but the cost savings happen over time,” said Easthope.