Semester Schedule
Risk Seminar Fall 2009
Seminars are on Wednesdays
Time: 4:00 - 5:00 PM
Location: 1255 Amsterdam Avenue, Room 903 SSW, 9th Floor
Wednesday, October 28th,
Speaker: Wolfgang Hardle, Humboldt-Universitat, Berlin
Title: CDO and HAC, Collateralized Debt Obligations and Hierarchical Archimedean Copulae
------------------------
Wednesday, September 30th
Speaker: Steven Kou, IEOR, Columbia University
Title: Asian Options
------------------------
Wednesday October 7th
Speaker: Jan Vecer, Dept of Statistics, Columbia University
Title: Change of Numeraire with Perspective Mapping
------------------------
Wednesday, October 21st
Speaker: Professor Rama Cont, Department of Statistics
Columbia University
Title: Systemic Risk in Banking Networks
Abstract:
While systemic risk and financial stability have been the focus of many discussions during the recent financial crisis, no rigorous methodology has been available for studying these sensitive issues. A prominent role is played by the network of links among financial institutions, often neglected in risk management. Through a statistical analysis of a dataset of interbank exposures we show this network structure to be a directed scale-free graph with heavy-tailed degree distributions. The analysis is shown to be compatible with a random graph model with preferential attachment. We then propose a network-based approach to the study of systemicrisk, in which a financial system is modeled as a weighted graph of counterparty relations, and propose a quantitative measure - the Systemic Risk Index - for assessing the systemic risk posed by (a group of) financial institutions in such a network. We discuss some theoretical properties of the Systemic Risk Index and show how it can be estimated in practice using a database of interbank exposures obtained from the Brazilian central bank.
Joint work with Amal Moussa (Columbia University Dept of Statistics) and Edson Bastos (Banco Central do Brasil)
------------------------
Wednesday, October 28th
Professor Wolfgang Härdle
CASE - Center for Applied Statistics & Economics
Institut für Statistik und Ökonometrie
Wirtschaftswissenschaftliche Fakultät
Humboldt-Universität zu Berlin
Title: CDO and HAC
Abstract:
Modelling portfolio credit risk is one of the crucial challenges
faced by financial services industry in the last few years. We propose
the valuation model of collateralized debt obligations (CDO) based on
copula functions with up to three parameters, with default intensities
estimated from market data and with a random loss given default that is
correlated with default times. The methods presented are used to
reproduce the spreads of the iTraxx Europe tranches. We apply
hierarchical Archimedean copulae (HAC) whose construction allows for
the fact that the risky assets of the CDO pool are chosen from six
different industry sectors. The dependence among the assets from the
same group is specified with the higher value of the copula parameter,
otherwise the lower value of the parameter is ascribed. The copula with
two and three parameters models the relation between the loss given
default and the default times. Our approach describes the market prices
better than the standard pricing procedure based on the Gaussian
distribution.
------------------------
Wednesday, November 11
Eugene Xu, Deutsche Bank Securities
Title: The Ghost of Financial Crisis
Abstract:
A preliminary attempt to discuss about the causes of financial crisis that started in 2007, including government policy, effects of globalization, Wall Street and investor behavior, shortcomings rating agencies procedures, etc. The discussion will be less statistical than financial.
------------------------
Wednesday, November 18
Randy Cogill
Assistant Professor
Systems and Information Engineering
University of Virginia
TITLE: Randomized Coding and Queue Dynamics
ABSTRACT: In queuing models arising in communication systems, the choice of when and how to encode data can have a significant impact on queueing dynamics. In this talk we focus on a specific, deceptively difficult queueing model of a single transmitter broadcasting data packets to multiple receivers. Finding a policy that minimizes the average queue length for this system is currently an open problem. We show that a certain randomized coding strategy incurs an average queue backlog that is order-optimal with respect to the number of receivers.
In our analysis, we utilize some useful inequalities for extreme values, steady state performance of Markov chains, and achievable performance of queueing systems. Communication theoretic details will be kept to a minimum.