To The Clouds and Beyond: A New Distribution Channel
Exchanges are developing cloud-based platforms, allowing customers to pick and choose from data products.
Exchanges are ramping up investment in cloud-based computing in order to gain economies of scale, simplify the access and delivery of market data, decrease time-to-market for new products and add new revenue streams from new and under-serviced segments.
“Cloud is the next wave in terms of how firms will handle server virtualization and server capacity planning,” said Don Henderson, chief technology officer at NYSE Technologies, the technology arm of transatlantic exchange operator NYSE Euronext. “That is something almost every exchange and most companies should think about investing in.”
With the cloud, exchanges can offer customized products that allow customers to select and consume only the data required for their specific applications, modeling or back-testing.
“Cloud technology allows you to provide services for customers that can be deployed on demand, in what we call hybrid services,” said Henderson. “A customer may want something that might not be permanent and want to rent it, or they may want something specific for back-testing purposes. Or they may simply need some extra capacity for a variety of reasons either planned or unexpected.”
U.S. exchange operator Direct Edge this year launched EdgeBook Cloud, a suite of on-demand services providing depth-of-book historical trade and quote data from Direct Edge’s two national securities exchanges—EDGA Exchange and EDGX Exchange.
This dataset is delivered via the Xignite cloud technology platform and supports efficient market analysis, back-testing and compliance review. The suite includes EdgeBook Replay, EdgeBook Flex Download and EdgeBook Snapshot and is designed to provide deep data queries that can be easily integrated with end-user systems, without the burden of storing mass quantities of data.
EdgeBook Cloud provides complete proprietary book data where users can quickly obtain the specific information they need, when they need it. Analysis can be performed that isolates a particular stock, or looks at its performance in comparison with others in a real-time simulation.
India’s Bombay Stock Exchange (BSE) has FasTrade on Web (FOW), a cloud-based system developed by Marketplace Technologies, a wholly-owned subsidiary of the BSE, that doesn’t require any servers to be maintained by the broker. Instead, brokers can simply connect to it via the internet, leased lines or a VSAT [very small aperture terminal].
“FOW is already used by a large number of our members to trade in our derivatives segment,” said BSE spokesperson Ketan Mehta. “This is a hosted execution management system that is accessible to anyone with an internet connection and a web browser.”
The Next Wave
The concept of a capital markets community, or financial ecosystem, involves a series of data centers interconnected by networks, through which transactions, data and services can be shared.
“Our core goal for cloud is to provide on-demand capacity ready for use almost immediately,” said Henderson at NYSE Technologies. “For our managed services, the cloud underpins almost everything we will be doing. We will leverage the cloud architecture as part of our growth there going forward. It’s targeted to be a high performance infrastructure that retains the security and reliability that our customers expect.”
The ability for the cloud to provide large amounts of computing power spontaneously is useful for computationally-intensive research tasks, like analyzing market data and back-testing.
“The biggest strength is its ability to have huge amounts of computational power available at the click of a button that’s only utilized (and paid for) as long as the research task requires it,” said Tyler Moeller, chief executive of Broadway Technology, a trading technology company.
The biggest limitation is the fact that this computational power is not physically co-located with exchanges.
“In fact, cloud server farms are usually located far away from any other technology centers to minimize power and property costs, and that virtualization degrades performance for real-time tasks that need to share high volumes of time sensitive data—i.e. market technology,” said Moeller.
For those tasks, dedicated hardware, tightly controlled networks, careful monitoring of each machine’s performance, regular performance tuning, traffic isolation and other optimization techniques are all critical for good trading.
“With all the investments exchanges have made in recent years into building their own platforms, with a higher focus on latency, it will take a while before exchanges move their core businesses like trade matching into the cloud,” said Vladimir Kislenkov, product manager of order routing at CQG, a provider of analytics and order routing software.
However, the “cloud” can also be taken to mean environments hosted by a third party, outside of a firm’s internal IT, under which definition there is use for market technology.
There are vendors today that will provide hosted environments that are highly tuned for low latency performance and are not operated by a trading firm’s internal IT.
These are almost all dedicated machines and networks within a co-located data center, but operated and often owned by a qualified third-party vendor.
“This has proven to be a great way for trading institutions to get systems up and running extremely fast and to have them managed by a highly-qualified third-party team solely responsible for that system,” said Moeller at Broadway.
There are many advantages to this approach over a firm purchasing dedicated co-location space, including having a large highly-qualified team available to manage the machines, hosting environments that are efficiently managed, the ability to scale up and down more easily and cost savings due to shared infrastructure.
The cloud is becoming one of the primary channels of market data distribution,” said Stephane Dubois, chief executive of Xignite, a financial market data cloud provider. “What has been the model for data distribution in the last 20 years has always been technology focused.
“Exchanges gave you a feed that gives you everything like a firehose and you have to take it all in and figure out what to do with it. That worked fine five to 10 years ago, but with increasing volumes, firms said, ‘Do we really need this big fat pipe and infrastructure to get all this information?’”
Mehta of BSE added: “We are early days still in terms of cloud solutions in Indian markets. But it offers obvious benefits for many participants that do not want to carry the burden of investing in their own infrastructure. Many vendors also offer these kinds of services and will increasingly do so in India going forward.”
Firms should use data to gain insight from trade flows and optimize their system performance.
Elections in the Netherlands, France, and Germany could spark volatility.
Platforms are expected to consolidate down to two providers.
IOSCO suggests the good old days weren't all that good.
BIS says efficiency gains will be evolutionary, not revolutionary.