More On Underwriting.

By Shari Storm

Editor’s note:  I’m giving this post a plutonium disclaimer.  We are not advocating doing underwriting using the studies that Shari describes below.  She is calling these studies to our attention in the underwriting context being fully aware that to use them would probably be illegal.  It’s just a thought provoking blog post on the internet, folks.

A few months ago, I wrote a post about Dr. Robert Manning’s suggestion that financial institutions consider cash flow, along with credit score, when granting credit.

Since then, I have come across two fascinating articles on this topic.*

The first reports on a study in which researchers showed photos of faces to a group of people (workers from the Mechanical Turks, to be precise).  The participants were asked to answer the question “Would you lend money to this person?” The participants looked at the faces in quick succession and pushed a button for yes or a button for no, depending on their knee jerk reaction to whether or not they would lend them money.

Two incredible things happened. The first is that most of the participants rated the pictures of the faces the same. Secondly, the people behind the faces’ scores were statistically similar to the actual credit scores of the people who were being rated.  So, for example, if the study had a photo of a man with big ears, most of the people who saw the photo of the man with the big ears said they would not lend him money. Turns out, the man with the big ears also had bad credit in real life.  The study also controlled for other variables – from beauty to race to obesity.

The second article discussed a study that suggested that people who don’t have a firm grasp of mathematics are more likely to default on their loans. In this study, they gave a pool of people a math test and over time, the people who did well on the test, paid back more of their debts whereas the people who did poorly on the test had higher default rates.  Want to take the test?

While visually sizing people up for shiftiness or giving quizzes for math comprehensive before granting credit breaks all fair lending regulations, it is something interesting to think about.  It makes me wonder what credit granting in the far future will hold. Perhaps some day they will find a DNA strand associated with creditworthiness and we will all give saliva samples instead of filling out loan applications.

* Articles:  “Physiognomy and economics: About face”, The Economist, March 7th, 2009 (page 86)
“Subprime borrowing and innumeracy: The fear of all sums”, The Economist, May 15th, 2010 (page 84)

Shari Storm is Senior Vice President and Chief Marketing Officer of Verity Federal Credit Union and is the author of the book ‘Motherhood is the New MBA”, available here.


3 thoughts on “More On Underwriting.

  1. Sounds very similar to a study that looked at P2P lending on (can’t find the study as the link I had is no longer working, thanks to Wharton).

    The study analyzed all loan applications listed during a one-year period on They found that listings submitted with a picture of a African-American person — or no picture at all — were significantly less likely to get funded than listings from whites with similar financial information. If funded, African-Americans were subject to a slightly higher rate.

    They also discovered some discrimination against older borrowers, overweight borrowers, and borrowers that they considered to be unattractive. They did find, however, that lenders discriminated in favor of members of the military and women (especially single women).

    Here’s the (ok, one) problem, though: the site’s African-American borrowers were more likely to default (by 36%) than the white borrowers with similar financial information. In addition, members of the military were 49% more likely than non-military borrowers to default.

  2. Ron – The Economist study used Prosper as well, but they found that age, gender, ethnicity and weight did not influence the rater’s reaction. I’ve talked to many a loan officer that say they just have a “sixth sense” about someone’s character. The Economist study was not able to pinpoint what it was that made the raters feel as they did.

  3. Huh…I got 4 out of 5 on the math test. Did one of the problems backwards, so that would be a reading comprehension fail vs a calculation fail. Makes me wonder if the test is also testing reading comprehension, not just the ability to calculate.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s