ABOUT THE AUTHOR

Donald R. Van Deventer, Ph.D.

Don founded Kamakura Corporation in April 1990 and currently serves as Co-Chair, Center for Applied Quantitative Finance, Risk Research and Quantitative Solutions at SAS. Don’s focus at SAS is quantitative finance, credit risk, asset and liability management, and portfolio management for the most sophisticated financial services firms in the world.

Read More

ARCHIVES

Implications of DNA for Retail Default Risk and Consumer Credit Scoring

07/14/2009 11:36 AM

The Broad Institute, which Dr. Lander heads, has been funded with a massive $600 million contribution by philanthropist and entrepreneur Eli Broad and his wife.  The institute is a joint project of Harvard, MIT, the Harvard affiliated hospitals, and the Whitehead Institute.  The institute, described at www.broadinstitute.org , seeks to make a huge DNA data base available to the scientific community, usually with no charge.

Dr. Lander brings a unique background to his role.  Trained as a mathematician, he came to the Broad Institute after 11 years teaching primarily computer science at the Harvard Business School. Boredom and curiosity got him involved in biomedical applications of computer science.  For more on Dr. Lander, please see

http://www.broadinstitute.org/about/bios/bio-lander.html

Why mention this on a finance blog?  The financial implications of this research are just beginning to be understood.  In the insurance industry, DNA is effectively collected when visiting nurses (paid by the insurance company) visit the home of a potential insured and do everything to them one can possibly do, including the collection of blood samples (and therefore DNA).  If DNA is being used to price insurance policies along with other medical data collected by the nurse, it wasn’t disclosed to me when I took out my policy. On the other hand, if 1984 Big Brother concerns were ignored, DNA information would provide an insurance company with an unbelievable competitive advantage because of the importance of genetics in many fatal diseases.

Outside of insurance policy pricing, why would DNA be relevant?  The night I e-mailed my colleagues, my initial thought was that borrower behavior could be better predicted with an understanding of the borrower’s DNA.  For example, was the borrower someone with the DNA of National Economic Council head Lawrence Summers, son of two Ph.Ds and nephew of Nobel Prize winners Kenneth Arrow and Paul Samuelson?  Or was the borrower someone ordinary like me?  I won’t summarize the massive internet presence of debates on intelligence and genetics except to say that today on google there are 6.8 million hits on the search +genetics +intelligence. My hypothesis was that the smartest individuals would be likely to prepay more rationally and more quickly, which should be reflected in efficient market pricing of mortgages.

It turns out my hypothesis was too simple, as usual.

On June 4, 2009 a new study was announced by a team of Harvard-affiliated researchers that found that more than 60% of consumer bankruptcies in 2007 were caused by medical expenses.  The article was scheduled to be published in the American Journal of Medicine in August 2009.

(For a summary of the article see

http://www.consumeraffairs.com/news04/2009/06/bankruptcy_medical_costs.html )

This study found an even higher percentage of bankruptcies were due to medical expenses than an earlier study of bankruptcies in the year 2001.

What this means is that existing default models for consumer credit and for retail credit scoring are dramatically misspecified, because they systematically omit a variable that explains 60% of bankruptcies: the medical condition of the borrower and his or her immediate family.  My initial hypothesis ignored the genetic predispositions toward serious disease that have been steadily uncovered in the last decade, and my hypothesis failed to recognize that these diseases could lead to financial difficulties and bankruptcy.  All of this could come about in spite of consumer credit scores that were sterling at the time the loan was granted.

For many reasons, it may be that governments ban the use of DNA for credit risk assessment on the grounds that it’s “not fair” that someone with bad DNA should be punished again—in the form of higher borrowing costs that reflect the possibility that a higher chance of medical problems meant a higher probability of bankruptcy.  There’s lots of precedence for such a move—Japan banned discrimination against smokers in insurance policy pricing for decades, arguing for “fairness” with a happy by-product of higher profits at government owned Japan Tobacco.

From a purely financial point of view, it is very highly likely that DNA will substantially enhance the predictive capability of consumer default models.  From a public policy point of view, the U.S. government’s proposed new Consumer Financial Protection Agency will surely consider banning the collection of DNA from borrowers, in spite of the fact that insurance companies collect DNA now in the form of blood samples.  Ah yes, yet another conflict between high quality financial economics and public policy!

Comments welcome at info@kamakuraco.com.

Donald R. van Deventer
Kamakura Corporation
Honolulu, July 14, 2009

ABOUT THE AUTHOR

Donald R. Van Deventer, Ph.D.

Don founded Kamakura Corporation in April 1990 and currently serves as Co-Chair, Center for Applied Quantitative Finance, Risk Research and Quantitative Solutions at SAS. Don’s focus at SAS is quantitative finance, credit risk, asset and liability management, and portfolio management for the most sophisticated financial services firms in the world.

Read More

ARCHIVES