Kx in Academia: Carnegie Mellon University teaching with kdb+

7 Feb 2017 | , ,
Share on:

One of the ways the Kx community grows is through the Kx academic program, which provides free kdb+ licenses for teaching and research to select universities. The global list of institutions with academic licenses includes Harvard Business School, the Haas School of Business at UC Berkeley, the University of Chicago and Carnegie Mellon University’s Tepper School of Business.

Kdb+ user Robert Almgren, president and co-founder of Quantitative Brokers, a broker-dealer and provider of agency algorithms, is also an adjunct professor of Finance at Carnegie Mellon. He uses Carnegie Mellon’s kdb+ academic license for teaching.

Over the past few years, Robert has jointly taught “Market Microstructure and Algorithmic Trading,” with Duane Seppi in the Master of Science in Computational Finance program at Carnegie Mellon’s Tepper School of Business.

This course presents foundational concepts and current developments relating to trading in financial markets including algorithmic and high-frequency strategies, optimal order execution, execution quality analysis, the operation of limit order markets, the regulatory and institutional landscape, programming and IT infrastructure, and the economics of market microstructure.

The class is heavily data-oriented. The students get to work on a month of futures tick data from the CME Foundation, with over 3 billion total data records. For this amount of data, a sophisticated product like kdb+ is required. Carnegie Mellon hosts the data on a kdb+ server, and lets the students access it either from the q command line, or using the interface to the R statistical language.

The instructors give the students some basic instruction in kdb+ and show them how to do useful queries directly in the data server. Then they use the data throughout the course to explore empirical aspects of market microstructure.

Robert and Duane’s students last semester were very much up to the challenge. Kx engineer Ruairí Devlin gave a presentation to the class in October about kdb+. During an animated session of live coding the class put Ruairí through his paces, demonstrating significant knowledge of trading systems, and kdb+.

If you are an alumni or professor at a university which could also benefit from a kdb+ academic license, drop us a line at academic@kx.com.


Java API for kdb+

31 May 2018 | , , ,

In the latest in an ongoing series of kdb+ technical white papers published on the Kx Developer’s site, Kx engineer Peter Lyness has written about how the Java API for kdb+ can be used to enable a Java program to interact with a kdb+ process.

kdb+ tick database disaster recovery

Kx Technical Whitepaper: Disaster recovery planning for kdb+ tick systems

2 Nov 2017 | , , , ,

This whitepaper discusses disaster recovery and failover concepts from the perspective of the gateway layer accessing a typical kdb+ tick system used in capital markets applications. The end goal of constructing this plan is to ensure high availability of the application via the gateway where possible, considering all conceivable failure scenarios and outlining any actions required to prevent data loss, minimize any downtime and keep the application accessible.

Collaboration with Kx and Thomson Reuters using kdb+

The Power of Collaboration: Kx and Thomson Reuters

2 May 2017 | , , , , , ,

What defines a great collaboration? Kx’s Chief Revenue Officer Pat Brazel writes in this blog about our current Kx and Thomson Reuters partnership. Over the past 12 months Kx and Thomson Reuters’ domain experts, and engineers have jumped into the Velocity Analytics 8 product development process with good faith and real enthusiasm. Out of that has come a product delivery, roadmap and sales strategy that both companies are proud of. To both firms credit, we seem to have avoided the dreaded “not-invented-here” syndrome. Both companies have brought the best of their technologies together to create a product which builds on, and exceeds, the foundation on which it is built.