Wall Street Letter Volume XLV Issue 13 | Page 3

news 18 – 24 April 2013 Data Published by Pageant Media London Thavies Inn House London EC1N 2HA T [phone] F [phone] EDITORIAL Jeanene Timberlake Managing editor [phone] [email protected] Ryan Gorman Reporter [phone] [email protected] Gwyn Roberts Group head of content [phone] [email protected] Indira Peters-DiDio Data manager [phone] [email protected] PRODUCTION Claudia Honerjager Head of production c.honerjager@ pageantmedia.com Rachel Kurzfield Sub-editor Eleanor Stanley Sub-editor New York 240 W 37th St. Suite 302, NY 10018 T [phone] F [phone] COMMERCIAL Lucy Guest Group commercial manager [phone] [email protected] Stefano Giacoia Content Sales Manager [phone] [email protected] EVENTS Beth Hall Head of events [phone] [email protected] DISTRIBUTION Fay Muddle Circulation manager [phone] [email protected] PAGEANT MEDIA Charlie Kerr Chief executive Printed by The Manson Group Boston Post rebrands, to launch predictive analytics Boston Post Analytics is rebranding its business as it prepares to roll out an adaptive predictive analytics platform, according to founder David Marra. The company is launching Arialytics as it launches Aria, the flagship predictive analytics platform that is set for launch in early May, said Marra in an interview. The predictive engine is aimed at institutional asset managers looking to leverage big data in their research as they make trading decisions, he added. “Big data predictive analytics has taken off in a lot of industries,” Marra said, adding: “It’s about to take off in the investment community.” Aria is able to take data from any source, analyze it and create predictive trading models, Marra explained. The speed of model creation is dependent on a firm’s infrastructure and the system can be scaled almost limitlessly, he said. Regardless of speed, Marra pointed out the offering can build multiple models more quickly than a team of researchers and quants, freeing up resources to take advantage of trading opportunities. “Decreased research time creates five models instead of one,” Marra said. The models are also developed without bias, which helps improve the performance of the model in use, and they are continuously updated to fit an asset manager’s strategy as conditions change, Marra continued. Each model references between 600 and 800 variables to generate results and runs over cloud infrastructure on the same tech as the world’s top 500 super computers, Marra noted. The offering can be used by any 03 01 continued... ISSN# 726-98790 © 2013 Pageant Media Ltd. All rights reserved. COPYRIGHT NOTICE: No part of this publication may be copied, photocopied or duplicated in any form or by any means without Pageant Media’s prior written consent. Copying of this publication is in violation of the Federal Copyright Law (17 USC 101 et seq.). Violators may be subject to criminal penalties as well as liability for substantial monetary damages, including statutory damages up to $100,000 per infringement, costs and attorney’s fees. The next WALL STREET LETTER will be available the week of 04/22/13 The upcoming issue will feature our regular industry news coverag