.jpg)
The Rutgers University Master of Financial Statistics and Risk Management (FSRM) program is pleased to proclaim its first annual Rutgers Statistics for Financial Risk Management Conference a success. On November 7, 2013 more than 150 practitioners, academics and students gathered to hear distinguished experts from around the world share their insights on the statistical facets of risk management. Rutgers also inaugurated the FSRM Center of Excellence.
"The conference's success illustrates the growing recognition that statistics plays a vital role in measuring, monitoring and managing all types of risk that affect financial institutions," says Neville O'Reilly, Associate Director of the FSRM Master's Degree program. "Rutgers is committed to being a leader in financial statistics, which we will demonstrate by delivering a high quality educational program, conducting collaborative research in our Center of Excellence and facilitating conferences."
In his presentation entitled Cascading Defaults and Systemic Risk of a Banking System, Jin-Chuan Duan of the National University of Singapore explained the different between systematic and systemic risk. He talked about asset and liability dynamics in stressed scenarios and the impact of netting. Citing a study of 15 British banks, he explained why a shock to the system is exogenous, while a cascading default is endogenous.
Wolfgang Härdle of Humboldt-Universität zu Berlin presented CoVaR in Very High Dimensions. CoVaR determines the value at risk of the financial system conditional on other risk factors being under distress. Härdle used several bank examples to explain risk analytics and determination in the face of many variables. He also explained the mathematics behind an iterative technique referred to as Linearize, Average, Estimate, Penalize (LAEP).
James Hodson of Bloomberg's Machine Learning and Statistical Inference group talked about Cross-lingual Information Arbitrage: Mitigating Global Market Inefficiencies. Bloomberg publishes about 1.2 million news stories daily, and it builds big data models that help market participants digest this information. Hodson explained the constraints of the models and the risk associated with them. He also presented ways that the available streams of data can be harnessed to provide a fuller picture in a timely manner through Machine Learning and Natural Language Processing techniques.
University of Chicago's Ruey Tsay spoke about Market-based Credit Ratings. He explained the difference between absolute and relative credit ratings. Then he demonstrated a method for classifying a company and constructing new credit ratings directly from asset prices, specifically credit default swaps. He also discussed the best methods for matching non-traded firms to traded firms.
Holger Rootzén of Chalmers University of Technology in Göteberg, Sweden explained how to manage the risk of extreme and rare events in his presentation entitled Taming Black Swans with Statistics. He cited examples of how statistics can be applied in the realm of financial crises, climate change, pandemics, natural disasters and traffic accidents.
The final presenter was Nassim Taleb, a former derivatives trader turned scholar, currently distinguished professor of risk engineering at New York University's Polytechnic Institute, and the author of The Black Swan and Antifragile. Taleb discussed the unreliability of measuring tail events in fat-tailed domains, the technical issues associated with the Law of Large Numbers and Extreme Value Theory. He also proposed fragility as a non-linear response to random events and model error and offered heuristics as a way of dealing with model and estimation errors.
"The feedback from conference participants was overwhelmingly positive," notes O'Reilly. "We look forward to staging the Rutgers Statistics for Financial Risk Management Conference again in 2014."
