Denver / QWAFAFEW

QWAFAFEW Denver Meeting - Feb. 20th

Date/Time:  Wednesday, February 20, 2008 at 6PM
Location:   Via Trattoria (Located on Wynkoop between 18th and 19th in LoDo) 6PM: Networking, hors d’oeuvres (including vegetarian options), and complimentary wine; 6:30PM: Presentation
Admission:  $10 Members; $35 Guests
RSVP:  Secure your seat by sending your admission check to the address below by February 13, 2008.
Address:  QWAFAFEW Denver, PO BOX 19013, Boulder CO 80308
Questions:  Email Denver@qwafafew.org

Presentation: “Portfolio Construction from Ordering”
Speaker: Gioel Molinari 
Mr. Molinari, one of ClariFI's founders, is responsible for the company's technology vision, strategy, and product development and plays a leading role for the company during a time of unprecedented growth. He has extensive experience working at the intersection of mathematics and computer science, having successfully developed leading-edge quantitative technologies to solve novel business problems. Among his many accomplishments, Mr. Molinari spearheaded the design and development of the Sengent Broker Investment Advisor Grid, an Internet "grid"-based quantitative modeling platform; a high throughput Bluetooth testing environment as System Architect for Ericsson; and the world's first visual expression analysis application as Bioinformatics Engineer for CuraGen Corporation. Prior to founding ClariFI, Mr. Molinari was a Software Architect for Xcelerate Corporation where he led the development of several successful e-business projects including E-Exchange.com. He holds a Bachelor of Science degree in chemical engineering from Carnegie Mellon University.

Presentation Summary: Traditional mean variance portfolio construction requires inputs of expected return and asset covariances. In practice however, the production of expected return estimates is a very sensitive and error prone process. In particular, forecasts produced by multifactor models may not be in the form of return estimates at all, which requires ad-hoc transformation for use in portfolio optimization. We present a method for optimal portfolio construction that replaces the requirement for expected returns with ranking information and accomodates the numerous forms of rankings that practioners produce with their forecasting models including sector-relative and multifactor rankings. This presentation is based on research conducted in collaboration with Neil Chriss and Robert Almgren.

QWAFAFEW Denver Steering Committee:
Tommi Johnsen, University of Denver
Elisabetta Basilico, Qwest Asset Management
Chris Gantz, AMG National Trust Bank
Anna Coppola, Flagship Global
J.P. Tremblay, Standard & Poor's Investment Services