Lehman Brothers Powers POINT with Kdb

2 Dec 1999 | , , , ,
Share on:

Palo Alto (2 Dec 1999) – Kx Systems, a world-renowned pioneer in high performance vector-based technology, announced today that Lehman Brothers (NYSE:LEH) has selected Kdb as the underlying database for an enterprise decision support platform for fixed income portfolio and index analysis. The system, called POINT, contains over ten years of portfolio data on Lehman’s global family of bond indices. The underlying database, called Kdb, allows Lehman’s internal and external end users to perform sophisticated time series analyses on the portfolio data in real time.

Lehman chose Kdb over several leading database alternatives because of its speed and unique capabilities. Rich Wagner, Senior Vice President of Lehman Brother’s Fixed Income Research Group, said, “The most important database requirement for POINT was functionality. We wanted to be able to ask new types of complex time-based queries that we couldn’t with our existing relational database system. With Kdb, POINT can now handle comparisons of portfolios versus indices over time. This capability, in addition, to Kdb’s speed and simplicity, has made an enormous contribution to POINT’s success.”

Before selecting Kdb as the underlying database for POINT, Lehman rigorously tested its capabilities, requiring Kdb to meet, or exceed, exacting standards for a number of criteria:

  1. Processing Speed — Kdb’s special treatment of time, combined with the fact that the data is ordered, resulted in very fast processing speeds. This allowed POINT to perform historical analyses on portfolios of securities across indices and sectors in a fraction of the time it takes with other systems.
  2. Programming overhead — Programming time was dramatically reduced with Kdb because of its efficient syntax. Pages of old code were literally reduced to lines with Kdb
  3. .Scalability — Scaling up proved easy with Kdb. During the initial implementation period, the database grew from 12 to 33 GB without any loss in performance, largely because of Kdb’s unique efficiencies in table structure.
  4. Perform complex financial functions — These are native to Kdb’s query language, KSQL, so it easily performs all financial functions.
  5. Require minimal training — Kdb’s query language, KSQL is so similar to SQL92 that it requires very little training.
  6. Perform under peak conditions — Lehman found that Kdb’s speed made it better at handling high levels of demand for compute-intensive queries than other database products.
  7. Integrate easily — POINT’s architecture relies on a number of technology layers. Connections to Kdb were easy using either its interprocess communication facility or industry standard interfaces.

The system currently has hundreds of thousands of securities, with hundreds of fields of data for each security, and will continue to grow.

© 2018 Kx Systems
Kx® and kdb+ are registered trademarks of Kx Systems, Inc., a subsidiary of First Derivatives plc.

SUGGESTED ARTICLES

kdb+ breaks STAC benchmark record

Kdb+ and Intel® Optane™ SSDs break more performance records in latest Intel STAC benchmarks

1 Nov 2017 | , , ,

FD (AIM:FDP.L, ESM:FDP.I) announces that its Kx Systems division is proud to be part of Intel’s latest record-breaking tick analytics benchmark run on the Intel Skylake processor in conjunction with the brand new Intel ® Optane™ SSD DC P4800X. The tests broke 11 out of 17 of the existing response time records for performance on a single server.

kdb+ for Industrial internet of things

Kx Systems Joins the Industrial Internet Consortium

5 Sep 2017 | , , , ,

Kx Systems, a subsidiary of First Derivatives plc (FD), and provider of the industry-leading kdb+ time series database, announces that it has joined the Industrial Internet Consortium (IIC), the world’s leading organization transforming business and society by accelerating the adoption of the Industrial Internet of Things (IIoT). Kx will work with other IIC members to support digital transformation through the adoption of best practices in big and fast data.