Featured Image

Real-Time Market Data in the Public Cloud: Adapting to New Business Models amid Legacy Challenges

Data Science and AI

By FactSet Insight  |  April 26, 2023

Eugene Kim, Director of Client Solutions, and Anant Singh, Principal Software Architect, are co-authors of this article.

As demand increases for substantial volumes of real-time financial information, cloud technology services are becoming ubiquitous.

In addition, many exchanges and data providers now provide connectivity solutions in the cloud. This enables financial services firms to consume market data from consolidated/direct feeds and desktop terminal products that supply real-time, pricing, and derived data.

A few recent cloud-enabled developments include more:

  • Exchange- and trading-system delivery via the public cloud

  • Consumption of data feeds via cloud providers and cloud shares

  • Software-as-a-service (SaaS)-based solutions

High Cost of Legacy Technology

Market-data dissemination and consumption—which involve on-premises infrastructure in client data centers—have mostly stayed the same over the past several decades. The centers are costly, require long delivery cycles, and are unable to keep up with increasing business demands.

For example, data volumes continue to rise, currently exceeding one million messages per second or more than 100 billion messages per day. Volumes will increase further over time, given globalization of capital markets and the expansion of cryptocurrencies.

In addition, compliance, regulatory, and artificial intelligence (AI)/machine-learning (ML) initiatives are increasing demand for historical time series data. In a conventional approach to technology, horizontal scaling is deployed (e.g., adding more servers and networking equipment).

For example, implementing the US Options Price Reporting Authority (OPRA) requires 32 individual servers spanning multiple network channels. Factoring resiliency within a location and redundancy with another data center, an OPRA deployment will require 128 servers, plus networking equipment and connectivity.

The Promise of Cloud-Delivered Market Data

Accordingly, financial institutions are starting the transition toward cloud-first operating models with best practices and shared services. They are beginning to see the cloud as an enabler for monetizing proprietary data. Exchange venues have also embraced the public cloud. Recently, the market has seen a wave of strategic partnerships to offer cloud-native solutions, such as the CME Group with Google Cloud and Nasdaq with AWS. Other commercial drivers of cloud adoption in the market data industry include:

  • Automation capabilities for low-value operations and routine tasks

  • Faster time-to-market for deployment/delivery

  • Reduced dependency on large-scale on-premises infrastructure

  • Regional access with minimal Information Technology (IT) staffing

  • Sophisticated "as a service" capabilities

Cloud technologies have also proliferated across adjacent workflows, offering easy integration with third-party suppliers that can provide scale-managed platform services. From a market data infrastructure perspective, on-premises distribution technologies have invested decades in latency, resilience, robustness, and volume handling fine-tuned to bespoke hardware.

The public cloud enables the benefits of this research and development, without physically tethering to customized equipment. As the industry continues to evolve with more cloud solutions for new and specific use cases, the public cloud offers other key technical advantages:

  • Ability to quickly spin-up or unwind products/projects

  • Faster deployment over on-premises data centers

  • Fostering environments for experimentation and innovation

Challenges with Cloud-Delivered Market Data

As with any strategic business decisions, there are challenges to consider along with strengths. For example:

  • Latency. Other than niche direct access trading applications, most business applications do not require ultra-low latency market data (measured in microseconds) and can rely on cloud-distributed data to serve their business needs. Additionally, public cloud technology will continue to evolve with more efficient delivery methods like multicast networking. However, until that time, applications based entirely in the cloud should expect latency ranges between 10 and 50 milliseconds.

  • Shared physical infrastructure. Additionally, the shared nature of cloud environments can sometimes cause variations in performance.

  • Metering. As cloud providers meter by message size, running a cloud environment with an egress to on-premises applications could drive higher operating costs.

The development of any cloud strategy should balance the challenges, opportunity costs, and client retention concerns with the potential benefits from deprecating on-premises equipment and reducing internal support costs.

Cloud providers are also working to reduce friction to adoption. For example, many exchanges offer cloud-delivered data through direct connections or consolidated feed aggregators—all at various price points and service provisions to meet a range of needs across prospective clients.

To learn more, read our white paper, Real-Time Market Data over Public Cloud.

Mr. Anant Singh is the principal software market data architect at FactSet, based in Norwalk, CT. He brings 20 years of experience in data and enterprise architecture strategy to his role in the Content Engineering division at FactSet. Recently, he led the migration of the real-time market data ticker plant to the Amazon Web Services public cloud and enjoys combining long-term business strategy and technical innovation to help financial institutions more efficiently access information. Mr. Singh earned a Master of Science in Engineering for Computer Science from the University of Michigan.

Mr. Eugene Kim is Director of Client Solutions at FactSet, based in New York. In this role, he is responsible for partnering with clients to design technology solutions. He also develops innovative market data offerings and chairs a FactSet Client Advisory Board, which helps steer the overall market data content and product strategy. Mr. Kim earned a Bachelor of Engineering from Cooper Union, and a Master of Science and Master of Business Administration from Columbia University.

This blog post is for informational purposes only. The information contained in this blog post is not legal, tax, or investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.

New Call-to-action


The information contained in this article is not investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.