This workshop will bring together researchers in machine learning from computer science, statistics, and electrical engineering; financial engineers and economists; applied mathematicians; legal scholars; econometricians; practitioners; and regulators to address the challenging questions raised by the post-mortem analysis of financial crisis data. In light of recent theoretical, empirical, and technological progress (e.g. big data analytics, better understanding of fire sales and liquidity shocks, and secure multi-party computation), the participants will revisit recent market anomalies to find, in hindsight, what could have been done to predict, prepare for, and/or prevent them given the current technology.
The tools and methods for bank supervision and the implementation of the Dodd-Frank Act will be addressed, and experience from “forensic agencies” such as NTSB and OFR will be central to these discussions.
The workshop will promote research on diagnostics for online detection of market manipulation and the identification of data patterns foreshadowing crashes. Examples of specific topics to be covered include the dynamic nature of the data at current trading speeds; the influence of social networks on trading; analysis of the Flash Crash of 2010 and the mini flash crashes since then; forensic analysis of OTC trading; real-time risk-monitoring of exchange traded securities; and the role of complexity in creating new forms of systemic risk in the financial industry.
This workshop will feature the 2015 Green Family Lectures by Andrew Lo. This series consists of a public lecture (Monday, 5/18), a research lecture aimed at a general math and computer science audience (Tuesday, 5/19), and a technical lecture that is part of the workshop (Monday, 5/18).
Marco Avellaneda
(New York University)
Jean-Philippe Bouchaud
(Capital Fund Management)
René Carmona
(Princeton University)
Andrew Lo
(Massachusetts Institute of Technology)
George Papanicolaou
(Stanford University)
Thaleia Zariphopoulou
(University of Texas at Austin)
Copyright. All Rights Reserved.