Back to Basics: A Wishlist for Cloud Providers from the Financial Exchange Industry
|
Greater investment in hardware and networking needed to make cloud-based colocation and multicast a reality
As some of the world’s largest financial industry participants gathered in Boca Raton last week, one fact that was hard not to miss was the increasing presence (and sponsorship) of the major cloud providers. These cloud providers are investing billions into the financial services industry with the ultimate promise of delivering AI in all its glory. AI might eventually succeed in providing some great solutions, but to the rightly risk-averse audience that is the global financial exchange industry, it is still limited in what it can provide.
As we have yet to realise the promise of AI, cloud providers in the meantime should concentrate on getting back to basics. Many hardware development issues still need to be resolved, for example, such asx cloud-based colocation and multicast protocols. Only then can exchanges fully benefit from the cloud’s flexibility, scalability, cost-effectiveness and efficiency.
“Dear Cloud Providers”: More investment needed for colocation and multicast
High frequency traders (HFTs) have long relied on trading strategies involving colocation, the practice of locating the trading servers in close proximity to the exchange data. Colocation plays a crucial role in modern capital markets by facilitating faster and more efficient trading. For exchanges, it offers a revenue stream through fees charged for access to colocation facilities, as well as enhancing market quality and efficiency by attracting HFT firms and other high-volume traders.
The promise of moving colocation to the cloud, however, is simply not there yet. Much more investment by cloud providers needs to take place in terms of hardware and networking, all aimed at optimising low-latency performance, compliance and security. Once the market for cloud-based colocation becomes more efficient, costs will hopefully fall to a more reasonable level.
Likewise for multicast, the protocol that simultaneously delivers data to multiple clients ensuring low-latency updates with real-time market information. But here again, the ability to reach low latency levels remains a serious concern. The solution would be for cloud providers to manage connections between participants within the same cloud so they can achieve low latency speeds granted by proximity. This could be achieved by cloud providers selling a hosting solution, although this has not yet taken place. The solution in this case, pertains to a business model, rather than a technical solution.
Fully cloud-based trading
In five to ten years’ time we will inevitably get to a much more comprehensive level of cloud-based trading that also incorporates colocation and multicast protocols. Given the exponential changes we have seen occurring across other industries, exchanges should prepare themselves for this eventual transition.
We are beginning to see a few developments in this move towards more cloud-based trading. For example, Exberry built a cloud-native matching engine on AWS that can process 1 million trades per second, with 20 microseconds latency. We need to remember, however, that not all markets are sensitive to latency issues. These exchanges could feasibly transition to the cloud today, forgoing expensive on-prem data centres and forever lead-times on software development, delivery and testing.
Acceleration towards cloud-based trading is inevitable
Until more investment by cloud providers is made into hardware and networking around colocation and business models around multicast, exchanges will stick to their colocation revenue streams. We hope to see cloud providers addressing these issues so there is an alternative to the colo business model. Once these performance issues are solved, however, a rapid acceleration towards change will take place. It might take a few years, but it is going to happen.
The Evolution of Exchange Data: Transforming Big Data into Relevant Insights
This decade marks a groundbreaking period in financial markets. Annual data generation is now exceeding all information created throughout human history. How can this exponential growth in data be harnessed to serve the interests of financial exchanges?
GPU-Edge Trading and Exchange-as-a-Service: New Frontiers in Exchange Technology?
The landscape of cloud hosting for trading technology has undergone a sizable shift in recent years. Major exchanges like Nasdaq, TMX and Aquis are now embracing cloud technology, signalling an evolution from scepticism to acceptance.
AFM & Exberry Tech Day: Accelerating the Exchange Revolution
The recent AFM & Exberry Tech Day in Amsterdam proved to be an exceptional gathering of minds and a catalyst for innovation.