On the marketing flyers, all top Wall Street firms have claimed to have the best portfolio algo. By checking out the quantitative foundations, we see that the existing solutions are far from being mature. Some major challenges include, for example, solid modeling of market impacts, and intraday risks and correlations, as well as conquering the universal “curse of dimensionality” in dynamic solutions. In this introductory talk, we first review all the major components needed to design and engineer a portfolio algo, and then report a work in progress on building an optimal (static) portfolio algo under generic price dynamics and volume measures. In combination, we demonstrate abundant opportunities for the financial engineering professionals in this emerging area.
Dr. Shen is currently a Vice President in the Algos and Quant Analytics team of the Barclays Investment Bank at New York (formerly Lehman Brothers). Before that, he was also a VP in the Electronic Quant Solutions team at J.P. Morgan. Before pursuing a Wall Street career on financial engineering, Dr. Shen had been an applied mathematician working on information and pattern theory, signal processing, and imaging sciences. He earned his Ph.D. in Applied Mathematics at MIT (1994-1998, under Gil Strang), and spent the subsequent two years as a Computational and Applied Math (CAM) Assistant Professor at UCLA. He had been a tenure-track assistant professor in Applied and Industrial Mathematics at the University of Minnesota, Minneapolis from 2000 to 2007. He has published more than 50 papers on wavelets, signal processing, and imaging and vision sciences, and coauthored (with Tony Chan, President of the Hong Kong UST; formerly Dean/Prof UCLA) the monograph published by the Applied Math Society –“Image Analysis and Processing – Variational, PDE, Wavelet, and Stochastic Methods”. His current research interest mainly focuses on the information and pattern theory aspects of modern electronic markets, and various optimal decision making processes involved.