Algorithmic Trading Success: “Hand Off to the Machine”

Combining benchmarked high-performance storage with information and the human element for business advantage.
Topics in this article

Quantitative finance firms are facing exponential growth in daily transactions. Their systems now process terabytes of market data and hundreds of thousands—or millions—of jobs. Their trading algorithms need to drive amazingly rapid decision-making at massive scale.

And that’s the quandary. To gain business advantage in a competitive trading market, how can you generate the required infrastructure performance while eliminating any roadblocks and distractions for your valued quant teams?

For me, it’s about eliminating boundaries between technology and trading strategy—for a simple, virtually instantaneous “handoff to the machine.” And that may be easier than you think. Quant firms around the world rely on our Dell Unstructured Data Storage (UDS) solutions to drive this approach. As a technology provider member of the STAC Benchmark Council, we continue advancing the enterprise-grade, high-performance infrastructure that firms rely on to accelerate their models and decisions.

Our approach brings together three elements to help you achieve real business advantage:

    • Speed – with ease of use and protection
    • Information and knowledge — from a range of unstructured data
    • Technique – the human element

Speed with Ease of Use and Protection

High-frequency trading infrastructure needs to be fast, of course—but also easy to use and protect. Dell PowerScale scale-out NAS data storage is a proven infrastructure on all counts.

Its blazing performance (high concurrency, throughput and IOPS) and massive scale help quant organizations ask more questions and create arbitrage opportunities faster. According to STAC benchmarking research, Dell PowerScale All-Flash platforms provide:

    • Real-time performance on smaller data sets (<10TB) and near real-time performance on large data sets (>10 TB) at ultra-high concurrency.
    • Shorter model development times, with 7.7x faster results vs. NBBO scores.

The latest performance boost begins in July, when PowerScale F900 and F600 all-flash models introduce QLC flash drives, providing incredible economics and double the node density to support quant firms’ high-capacity financial modeling workloads. The next release of the OneFS operating system (version 9.5) planned for this year will unlock streaming read throughput gains of up to 25% or more across our PowerScale F-series all-flash portfolio.

For ease of use, PowerScale easily integrates with existing systems through a broad set of S3, S3a CAS, NFS, NFSoRDMA and SMB protocols—true multi-protocol support that (unlike some alternatives on the market) enable you to integrate the same data via ANY protocol, not just via one or another.

And we never stop deepening the comprehensive range of PowerScale data protection and security capabilities essential for your continued success. Highlights include:

    • Building upon OneFS 9.3 and 9.4 security enhancements through planned submission to the Federal APL in a future release
    • Comprehensive ransomware detection and remediation options (air gap, cyber vault and more)
    • Strong authorization through External KMIP and an upcoming release of MFA
    • Support for more than one Active Directory or LDAP
    • Highly efficient and writeable snapshots perfect for quick dev and test branches
    • Active-active replication with strict data consistency, streamlining data management and operations

Information and Knowledge From a Range of Unstructured Data

Combining disparate and unstructured data sources with market data gives quant firms a competitive edge. However, unstructured data sources like social media, news feeds, weather trends, event data and regulatory submissions can quickly increase your data storage requirements.

Firms can take advantage of information and knowledge that others may not be able to simply by choosing the right storage infrastructure. It is no longer sufficient to pull from third-party data sources over the Internet. The velocity of today’s algorithms requires fast access to on-premise data to take advantage of quickly moving and/or quickly disappearing opportunities.

Dell PowerScale and Dell ECS object storage facilitate rapid insights through their scale-out capabilities, multi-protocol support and easy management of incoming unstructured data.

Technique – Through the All-Important Human Element

We also believe it’s essential to hire the best people you can and give them the tools to succeed. The human element of AI and analytics reigns supreme in quant analysis, and acquiring the best talent is important to accelerating results. With the proper toolsets at their disposal, a good analyst or data scientist can build multiple strategies that generate revenues for the firm.

For example, with the right people and right technique, you’re able to understand:

    • Which algorithm to use
    • Tradeoffs between the speed and skill of algorithms
    • When it makes sense to go with a simpler design

However, even the best talent can be hamstrung by poor infrastructure. Building a fast, easy-to-use data storage infrastructure that will feed the analytics and AI pipelines is paramount to increasing the efficiency of your highly valued, expert staff.

You want your analysts and data scientists working on models and defining new strategies—not searching for data, cleaning it and trying to validate the proper source of truth. These are not value-added operations.

Dell PowerScale enables humans to empower the machines of high-frequency trading and realize their potential. PowerScale’s advances in the ultra-high-performance algorithmic environments can help increase your quantitative, algorithmic and high-frequency trading team’s efficiency and performance with first-in-class data access, delivery and management.

Join us at the STAC Summits

Dell Technologies is a Platinum Sponsor at this spring’s STAC summits: in Chicago on May 10, London on May 19 and New York City on June 1. We look forward to meeting with you at one of the events to explore how you can remove the infrastructure “handoffs” to drive success in algorithmic trading.

About the Author: David Noy

David Noy brings 25 years of experience in the storage and data management industry. He spent nearly a decade leading engineering and product management teams for numerous companies, including Dell Technologies, NetApp, Veritas and Cohesity. Today, David leads two industry-leading technology divisions at Dell Technologies, Unstructured Data Storage and Data Protection, where he is helping to embolden innovation around data management and hybrid cloud; and driving advancement of holistic solutions to help heighten business success for customers worldwide.
Topics in this article