Feedback effects of FICO on consumer credit quality, a statistical solution

March 24, 2009

According to this msNBC video and this CBS video report, with default rates on the rise, credit card companies are desperate to cut costs and reduce risk.  It’s called ‘balance chasing’ and it involves banks cutting credit lines, in one reported case from 19,000$ to 300$.  They can also unilaterally closing accounts to reduce open lines and cut managerial expenses.  By some estimates, 2 trillion dollars worth of consumer credit will disappear by 2010.

In this context, many actions that were once considered good credit practices have now become a burden. Merely having a card you don’t use, for example, has become a ‘risk’ to the individual’s credit security.  Here’s how: If a bank closes a card, a consumer’s overall credit limit is lowered.  This in turn lowers their FICO credit score, sometimes by more than 50 points.  Given the fine print in credit card contracts that allows companies to adjust their terms, the pitch down in score can trigger increased interest rates from as much as 7.99% to 28%.

What results is a severe disruption to a household budgetary routine.  Suddenly there is less credit available, increased payments on multiple cards, dramatically increased debt burden, and a reduced ability to command fresh credit to compensate because of a lowered score.

The downward spiral is caused by the feedback loop at the heart of the the credit scoring system which is designed to allow all lenders to simultaneously monitor a consumer’s behavior with credit. The problem is that although the scoring system is supposed to monitor the consumer, it is also responsive to actions taken unilaterally by creditors on consumer accounts.  The score does not only reflect the changes in the consumer’s behaviour.  It also reflects changes in bank policy. This means that one bank’s internal decision can trigger automated managerial responses in other banks that degrades the consumer’s credit rating, even though the individual’s behaviour has not changed.

In a crisis environment where banks are cutting back on credit lines, and the question of sustaining credit liquidity is of the utmost importance, the personal as well as economic results of this looping effect are devastating.

As CBS reports, new legislation preventing some of these card company practices, was passed in December 2008, but won’t come into effect until 2010.  In the mean time, credit counselors are suggesting that consumers change their user strategies in all kinds of creative ways.  Where consumers were once told to keep (the very) lines (that are getting them into so much trouble) open, in response to the sensitivity to line limits built into the FICO, they are now being encouraged to complexify their card use to make sure these cards get used each month.

Placing responsibility on the public’s shoulders to adjust to this flux of changing demands is an inefficient and disaggregated solution to what is a systemic problem. It also severely undermines the idea that the credit scoring system reflects consumer behaviour, when it is clearly shaping it.  The statistics of FICO have built into them them the rules of the system that generate the spiral.

There are elegant statistical solutions to prevent this problem from happening.  Statistical redesign of the score’s underlying algorithm could prevent unilateral decisions by creditors from affecting scores, at least so dramatically.  The problem could be avoided if FICO could distinguish between a line limit cut by blanket bank policy, and a line limit cut caused by a deleterious consumer action such as a default that triggers a behaviour responsive bank policy.  Once treated as separate events in the score’s underlying statistics the feedback loop that erodes credit quality would be greatly mitigated.

In this time of crisis, when will the ‘political will’ to stabilize the credit system be turned towards the design of hidden financial technologies underlying it, and not only towards the visible actions of people and institutions? This is something that we who study the social effects of financial technologies, sincerely wonder.

One Response to “Feedback effects of FICO on consumer credit quality, a statistical solution”


  1. […] particularly volatile in economic conditions where unemployment and credit line retractions [see previous post] can dramatically affect consumer performance.  This would potentially erode FICO’s credibility […]


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s

%d bloggers like this: