It’s a story we hear from customers all the time: A financial services company wants to move to the cloud for scaling and security reasons but doesn’t because multicast – essential for market data distribution – isn’t available. The company assumes that’s the end of the road and either invests in impossibly expensive solutions or simply gives up and settles with using an on-premises system.
The story shouldn’t end there. cloudSwXtch offers multicast solutions in the cloud for the financial markets sector, without expensive infrastructure rebuilds or complicated code changes. Although market data distribution gets top billing when it comes to multicast use cases for financial services, there are additional benefits for bringing this feature to the cloud.
Let’s look more closely at what the financial sector gains when multicast is in the cloud.
Market Data Distribution
Market data distribution is usually the first impetus for getting multicast into the cloud. Exchanges, pricing systems and banks want to make that real-time data available to more users in more places, monetized in new ways.
Let’s look a little closer at why location matters: High-performance financial services systems like exchanges, ATSs, or pricing systems that exist in other markets, like futures and options, all generate tick data.
Tick data is generally constrained by geography, especially for more distant exchanges outside the U.S. If they wanted to monetize, they’d need to use a data management tool like Refinitiv or software with a global infrastructure for networking. Or, worse, they would have to build their own global network, which is expensive and time consuming
But, to put it plainly, cloud networks are examples of cheap global networks! They’re easy to get to, they can get your data anywhere, to anyone, and they are not costly.
If you can get your multicast data into the cloud without having to convert it, and then fan it out, you can be efficient in how you distribute the data globally. Plus, you' using cloud networks for what they're good at and finding new revenue streams. A win-win.
For financial services, moving the entire trading stack into the cloud for testing QA purposes, is key for keeping production unburdened and not changing the development system while testing is being carried out.
In the not-too-distant past, developers in the financial markets sector had to wait until the stock market was closed and then connect to a simulation in a data center to do all the testing, evaluation and development.
When IEX Exchange decided to move into Azure in 2020, they used cloudSwXtch. Now all of its development engineers can use Azure as their development environment. When a customer wants to test something on their trading system, they can do it in Azure instead of having to connect to IEX Exchange’s Weehawken data center.
Regulatory bodies require disaster recovery for different trading systems assessments, or a backup in case your data center goes down.
swXtch.io typically hears from financial services clients that have disaster recovery needs but don't think the cloud is an option because the performance requirements are lower or the time duration would be shorter. Financial services companies have a separate replicated data center and technology stack somewhere else for this reason
But it’s possible to have a sufficient level of performance just for disaster recovery in the cloud, eliminating unused capital expenditure from infrastructure that sits untouched for most of the year.
Moving Primary Trading Systems to the Cloud
Currently, we’re seeing digital assets trading systems moving to the cloud, because digital assets have lower performance requirements in terms of latencies. As cloud networks continue to improve, however, primary trading systems will certainly move.
Moving your primary trading system to the cloud would require nanosecond level performance. There are enclaves in clouds with specialized networking architecture called InfiniBand. These enclaves can provide nanosecond level performance on the cloud and it's possible to use that high-performance compute technology to build a running system. This isn’t a widespread solution, yet.
As those systems spread out inside the clouds, however, there will be more opportunity for people to use high-performance compute to truly move primary trading systems with nanosecond level networking and latency performance in the cloud. cloudSwXtch will be there to help when this happens by enabling our IP multicast on top of InfiniBand.
Financial services market shouldn’t be locked out of these features when it comes to the cloud SwXtch.io and our cloudSwXtch tool are here to help you cross the bridge to scale and growth. To request more information about how swXtch.io brings on-premises network feature parity to the cloud, please visit www.swxtch.io/contact.