Macro Factors Drive Bank Default Probabilities in Hazard Modeling ApproachHONOLULU, December 16, 2003: Kamakura Corporation reported today that the Federal Deposit Insurance Corporation has released “a state of the art approach to measure risks posed to the bank and savings and loan insurance funds” with Kamakura’s Director of Research Robert A. Jarrow as lead author.
Macro Factors Drive Bank Default Probabilities in Hazard Modeling Approach
HONOLULU, December 16, 2003: Kamakura Corporation reported today that the Federal Deposit Insurance Corporation has released “a state of the art approach to measure risks posed to the bank and savings and loan insurance funds” with Kamakura’s Director of Research Robert A. Jarrow as lead author. Professor Jarrow, who is also on the faculty at the Johnson School of Management at Cornell University, co-authored the study with Michael Fu and Huiji Zhang of the University of Maryland and Rosalind Bennett and Daniel Nuxoll from the FDIC.
“The FDIC has taken a very big step forward in its new Loss Distribution Model,” said Donald R. van Deventer, Kamakura Chairman and Chief Executive Officer. “The FDIC study is consistent with the arguments that Kenji Imai and I made in our book Credit Risk Models and the Basel Accords. First, advanced hazard rate modeling produces the most accurate estimates of the probability of failure, consistent with Professor Jarrow’s work on reduced form models. Second, macro-economic factors are the key variables causing correlated failure among corporations and banks. Thirdly, multiple inputs from the stock market, not just the current stock price alone, add powerful explanatory power. And finally, the hazard rate/reduced form modeling approach provides a much closer calibration of actual and expected portfolio losses than a credit modeling approach based on the assumption of normally distributed losses. Using monte carlo simulation to simulate movements of these macroeconomic risk drivers and the default probabilities of every counterparty is the same approach Kamakura has taken in both our Kamakura Risk Manager enterprise-wide risk management software package and our Kamakura Risk Information Services default probabilities.”
The FDIC paper is available for download as research paper number 107 on the research page of the Kamakura web site. The full press release of the FDIC announcement is available at FDIC web site.
The FDIC paper stresses the need to compare the consistency of actual and expected defaults under the model. Kamakura recently released such a paper on credit model testing procedures by Donald R. van Deventer and Xiaoming Wang entitled “Advanced Credit Model Performance Testing to Meet Basel Requirements,” forthcoming as a chapter in The Basel Handbook — A Guide for Financial Practitioners, slated for publication this month from RISK Publications with Michael Ong as editor. The paper (see working paper 106) is also available for download from the research section of Kamakura’s web site www.kamakuraco.com, along with detailed information on Kamakura’s default probability service.