« Banks and the Right of Setoff - Lehman Brothers Bankruptcy Illustrates a Potential Pitfall | Main | The Cambridge University Chest - Better Than Keeping Your Money Under the Mattress »

July 7, 2009

Clark Abrahams and the Comprehensive Credit Assessment Framework (CCAF)

Clark_abrahams This week, I've been talking by e-mail with Clark Abrahams, long-time financial institutions compliance advisor and author of CREDIT RISK ASSESSMENT:  THE NEW LENDING SYSTEM (2009).  In his July 2, 2009, blog post, he discusses the proposed Consumer Financial Protection Agency ("CFPA").  He sees the administration's push for "plain vanilla" loan products as bringing with it the need for different means of evaluating a borrower's creditworthiness.  He is working on a Comprehensive Credit Assessment Framework ("CCAF")

Credit scores alone are an insufficient indicator of true risk.  Abrahams recognizes the need to include additional "alternative data to qualify borrowers who fall outside of the mainstream"  and to avoid the “one size fits all aspect of credit scoring by conditionally assessing each borrower and deciding what factors are most relevant in a particular situation and what, if any, secondary factors need to be considered in order to render a final decision and, if approved, appropriately price the loan."

Here are some more of his thoughts about evaluating loan applications in the current economic situation and soon-to-be-changing regulatory framework:

" Reading the news, it is obvious that folks like to come up with generalizations, or rationalizations, regarding just about anything.  Perhaps it is the human mind's desire to try to organize information and create rules around observed behavior.   

So, how does this idea relate to the financial crisis, new lending systems, and borrowers, lenders, and investors?  Well, in my travels and discussions with lenders, I have found that they like to generalize a good bit, and their policies reflect that tendency.  In this day and age, it is routine for people to get associated with a number – like their credit score for example.  But people are people, not numbers, and when we start treating people like numbers we end up in a rut like we have today. 

There has been far too much reliance put on credit scores, for starters.  In addition, when lenders decide to tighten or loosen credit standards, they adjust thresholds on individual credit factors in a fragmented fashion – e.g. lower the cap on debt-to-income ratio from 40 to 36 percent, or lowering of the loan-to-value ratio from 90 to 80 percent, or raise the minimum FICO score from 620 to 640, etc.  These sort of adjustments mow down a lot of qualified applicants along with the more marginal ones. 

So what would be a good way to manage risk and adjust credit standards?  The answer is adoption of a comprehensive framework that covers all of the bases at once.  So, what is a proper adjustment of the debt-to-income ratio if credit policy needs tightening?  The answer is: “That Depends!”  For some, the debt ratio does not need to be lowered.  For others it may need to be lowered more than for others.  Still, for others, it may actually make sense to raise it.  I am trying to get folks to imagine a world that avoids sweeping generalizations and treats people more like individuals."

--

My comments:

Long ago and far away -- let's say 30 years ago -- our remembrance is that lenders did sit across the desk from prospective borrowers from the same community, whom they saw every day in the grocery store, the feed store, and in church.  Just like in "It's a Wonderful Life," right?  Well, maybe. . .

Even if this were an accurate picture of lending in the past, we can't go back.  We live in a world of computer models and greater risk tolerance as well as greater consumer appetite for debt.  We've seen the result of carrying risk tolerance and high debt levels come crashing down around us.  We're still struggling to figure out what comes next. 

Objective lending criteria like credit scores make it easy for lenders who lack the skill to make subjective loan decisions.  And they avoid charges of impermissible discrimination.  If it's all quantifiable, it can't be discriminatory.  Can it?  

This is the tension.  How to more accurately evaluate each loan application, without tripping over fair lending laws. 

Other problems are inherent in the credit scores themselves.  Of course, their creators want to maintain the proprietary advantage of keeping confidential the information and algorithms that spit out the scores.  But as long as the scores remain a "black box," we will have trouble on this front.  More study and reform of the consumer credit rating agencies should be on our radar screen.

(ag) July 7, 2009, in Lending Issues




 

July 7, 2009 in Lending Issues | Permalink

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341bfae553ef011571d27d83970b

Listed below are links to weblogs that reference Clark Abrahams and the Comprehensive Credit Assessment Framework (CCAF):

Comments

Great observations Ann! Transparency is a center stage issue here. In Chapter 1 of our latest book, "Credit Risk Assessment, the New lending System for Borrowers, Lenders, and Investors" Sunny Zhang and I explain why transparencey is one of the three root causes of the financial crisis. Just as consumers are able to read ingredients on food before they decide to swallow it, they should also be able to know the make-up of their credit scores before they accept a lender's assessment of their creditworthiness. The argument that knowing the scoring details will allow people to game the system actually holds little substance. First of all, the information that is supplied directly by consumers can, and should, be verified. Second, information not supplied directly by consumers, but that is a by-product of thier usage of credit and performance on timely repayment of thier loans, is most definitely impacted by consumer choices. On the second point, clearly consumers would benefit from having the power to know how their choices for financing their needs impacts their credit score. It may be the case that improving the credit score requires less saving and more spending (and higher credit usage to finance those additional discretionary purchases). In fact, credit scores may encourage people to over-extend themselves by rewarding this type of behavior with a higher score that can further qualify them for additional credit. If this is the case, then instead of keeping the formulas secret, perhaps the formulas need to be changed so that they encourage responsible borrowing, affordable product choices, and are transparent to encourage precisely those habits. This is exactly what the new lending system we are proposing will deliver - greater transparency, and not only to borrowers, but also lenders, investors, rating agencies and regulators. I think Justice Louis Brandeis' quote relating to transparency and honesty in public affairs applies here, namely “Sunlight is the best disinfectant.”

The commonplace practice of generalizing about borrower behavior and loan performance relative to isolated factors in lender credit models leads to flawed credit policies and bad outcomes for both lenders and borrowers! I will elaborate on this more in a future post.

Posted by: Clark Abrahams | Jul 9, 2009 7:58:13 AM

Post a comment