At Stanford University, Kay Giesecke, assistant professor in the department of management science and engineering, works at quantifying and measuring credit risk.
His research group, called Cred-itLab, has been funded by grants from JP Morgan Chase & Co., Moody's Corp., Credit Suisse Group, American Express Co. and the Global Association of Risk Professionals.
He also has helped MSCI Barra develop its risk modeling software, which cranks out probabilities of default for thousands of debt issuers on a daily basis.
Professor Giesecke is a leader in credit risk modeling with whom we have developed several credit models and co-authored several academic papers, said Lisa Goldberg, an executive director in research at MSCI Barra, New York.
Explained Mr. Giesecke: I build mathematical and empirical models of credit risk and how it should be measured. Specifically, I estimate the expected loss of holding a corporate bond over an expected time horizon.
From the perspective of an institutional fixed-income investor who holds hundreds or thousands of corporate bonds and is exposed to credit risk, the model helps monitor risk, he said.
Essentially, he takes company daily stock prices over many years and balance sheet and other financial accounting data including debt maturity structure to assess and assign the probability of a corporation going bankrupt within a designated period of time.
The model links stock price and maturity structure of debt to estimate the probability of bankruptcy, Mr. Giesecke said.
This helps you monitor risk in a portfolio and spot names likely to default, early as they begin to deteriorate. You can put the names on a watch list, he said.
Investors, depending on their risk sensitivity, can adjust the model's parameters to raise or lower the default probability for screening corporations.
He calls the model I-squared, for incomplete information model of credit risk.
The model also can be used to estimate relative value of credit default swaps and the equity market. Investors can use it to identify opportunities for trading credit default swaps against the equity market, he said.
If I believe the price of one market doesn't reflect true risk, I can devise a trading strategy to profit from it, he said, speaking hypothetically.
In a loose sense, it is a statistical arbitrage but not a true arbitrage, because an arbitrage is an opportunity to make a risk-free profit, Mr. Giesecke said. So this trade would not be risk free, but a bet on convergence of two markets.
In addition, Mr. Giesecke has worked with JPMorgan officials on a new model that analyzes risk of corporate collateralized debt obligations.
For the past few years, most market participants used a simple model to analyze risk of CDOs, Mr. Giesecke said. That model has been too simplistic and systematically underestimated risk of CDOs. For rating firms, it led to too high ratings and led investors to buy paper that was overoptimistically rated too high. He calls the more complex model the top-down CDO model.
I'm still working on refining the model for corporate CDOs. Mr. Giesecke said. We extended the model to cover sovereign CDOs.Barry B. Burr