| BRUSSELS 2007! Workshop 4: "Scoring and Social Discrimination in Credit Contracts" - Abstract from Mr Van Tol |
W4: SCORING AND SOCIAL DISCRIMINATION IN CREDIT CONTRACTS
The issue of discrimination and credit scores has several aspects, each raising different problems:
(1) Minorities and immigrants are disproportionately targeted by sellers of risky financial services products with a high failure rate, thus putting downward pressure on their credit scores.
(2) Credit scoring can overestimate the credit risk of individuals such as immigrants with “thin credit files” because the length of the credit history is a significant component of the credit score.
(3) Institutional/structural racism (i.e., significantly disproportionate impacts of policy and practice) skews the demographic rates of job loss, unemployment and coverage by medical insurance, and these are among the most significant factors leading to failure to pay off bills.
Below are two federal studies on discrimination in credit scoring. The Federal Reserve study, which focuses on lending issues, has some helpful parts related to immigrant credit scores. The FTC study is most notable for the dissenting opinion of one board member.
----------------
CREDIT SCORING AND CREDIT DISCRIMINATION IN THE U.S.
The recent dramatic events in the subprime housing market in the United States have exposed the stunning failures of those who analyse the risk in mortgage-backed securities. If in fact, the credit scoring models used by mortgage lenders are as robust in predicting future performance as their proponents claim, the bond rating agencies should have been able to apply that science to the packages of mortgages that were being securitized and then rated them accordingly and accurately priced the risk. While the bond rating agencies have been exposed to some scrutiny and criticism, those who provide the credit scores largely have not; and we should be asking why. If we are ever to have a time to demand some transparency and proof of socially fair performance from these very proprietary and poorly understood systems, that time is now.
In the United States the stories about this crisis have largely been told in terms of a media framework of “risky borrowers.” That is to say that the current failures in the market are viewed as the result of the natural credit cycle in which, during a period of strong growth in the housing market and low interest rates, the lenders have competed by lowering their lending standards to take in more “risky borrowers” until they got into trouble; and then a backlash began. But if the “risky borrower” view of the problem is correct and the credit scoring system worked properly, then it also should also be correct that the lenders knew that up to 20% of the subprime loans they were originating were going to default and that the bond rating agencies knew that these defaults were going to occur in proportion to how these loans were sliced and diced and distributed among the various mortgage-backed securities. They either knowingly engaged in extremely socially destructive behavior or more likely their credit scoring models were very flawed, because they didn’t adequately capture the risk created by the loan products themselves.
Those of us who were working in the middle of the low income parts of this mortgage market know that the “risky borrower” story is only part of the story of this crisis. The most important part of this story is the “risky loan product” part of the story. And this is where the crux of the problem with discrimination and credit scoring lies. While credit scores may have predictive value about future performance across a class of loans, they have less predictive value about who gets “risky loan products” and who does not, and they certainly have no predictive value across differing types of loans products.
For instance is someone with a credit score of 620 more likely to default on a subprime adjustable rate mortgage than on a fixed rate prime loan and by how much?
According to the Federal Deposit Insurance Corporation (FDIC) 11% of all mortgage loans originated in the United States in 2003 were subprime while this number jumped to 33% in 2005. It’s important to know that in this definition subprime refers to the interest rates and other features of the loan, not necessarily the credit profile of the borrower. It defies belief that the credit profile of people seeking mortgages in the United States changed so dramatically in two short years in a time of relatively normal economic growth and moderate unemployment. What did happen during that time period was that a group of subprime loan products, that were seen as being very profitable, were unleashed on the market and sold heavily, and they were sold especially heavily to African Americans, but also to Latinos and other minority groups.
In 2005 54% of home purchase conventional loans originated by African American borrowers had subprime characteristics while only 21% of the loans to white borrowers did. Among Hispanic borrowers 45% of the home purchase loans had subprime characteristics. For conventional mortgage refinance loans in 2005 those percentages were 49%, 22% and 33% respectively.
The new loan products that swamped the market in 2003 were interest only loans, which according to the FDIC jumped from being 3% of non-prime securitized loans to 30% during this time period. Low documentation/no documentation loans jumped from 25% to 40% of non-prime securitized loans during this time period. Pay option ARMs (negative amortization loans) became popular in some markets and piggy-back loans (loans in which the down payment necessary to avoid having to pay private mortgage insurance was replaced by a second lien loan or a home equity line of credit) also were used widely. Loans referred to by the short hand terms of “2/28s” and “3/27s” rose to prominence during this time period. These risky loan products were sold by brokers who were given more financial incentives to sell these products than to sell good quality fixed rate prime products, and they found that marketing them to demographic groups who traditionally had been denied access to credit or who found mainstream financial institutions unwelcoming was easier than selling them to those who were more comfortable and had a longer history with mainstream financial institutions.
After the elimination of slavery in the United States, African Americans were largely denied access to mainstream financial institutions and were often targeted by those who sell inferior financial services products. It has only been since the passage of the housing civil rights laws in the 1960s and the Community Reinvestment Act in 1977 that tools for fighting a history of discrimination have established a basis in law, but we still live with that legacy of bad financial products being marketed heavily to African Americans.
Whatever predictive value credit scores have, it s important to understand that in the context of a history of discrimination, lowering credit scores can become a self-fulfilling prophecy as financial services access is often limited to bad products that strip wealth and place financial stress on the customer and in turn put downward pressure on their credit score. If there is inadequate legal and regulatory framework to prevent members of a group subject to discrimination from being placed in high cost products regardless of their credit score, then the reputed benefits of a credit scoring system –removing the prejudicial human element from the credit approval process -rings hollow.
Hubert Van Tol, Director for Economic Justice
Rural Opportunities Inc., Rochester, New York
hvantol@ruralinc.org |
| ID: |
40237 |
| Author(s): |
iff |
| Publication date: |
14/09/07 |
|
|
|
|
| | |
Created: 12/09/07. Last changed: 12/09/07.
Information concerning property and copy right of the content will be given by the Institut For Financial Services (IFF) on demand. A lack of explicit information on this web site does not imply any right for free usage of any content. |
