The New Lending System - Putting common sense into compliance and risk assessment

In last Thursday's front page article in The Wall Street Journal, Tom Lauricella, Jason Zweig, and Conor Daugherty said that "A year after the U.S. economy was brought to its knees by the bursting of the housing bubble, credit fo consumers is still being ratcheted back." The Federal Reserve reported on Wednesday that total consumer credit outstanding contracted for the seventh straight month (which has not occurred since 1991), falling $12 billion in August. Lauricella, Zweig, and Daugherty go on to point out that some of the decline is due to lenders' exposure on thier real estate loans in particular, and to the reluctance of consumers to borrower given the tightening job market and the loss of their home equity. So, what does this imply? Well, given that historically consumers spending has fueled 70% of the nation's economic growth, coupled with the fact that consumers financed a large portion of thier purchaes through borrowed money and the unemployment rate is headed north of 10% going into 2010, we cannot expect the consumer sector to lead an economic rebound anytime soon. If anything, I would expect consumers to continue to curtail spending and to attempt to increase their savings while de-leveraging to the greatest extent possible. Many find themselves in a larger home than they can afford, or stuck with a second home in a real estate market where liquidity at an acceptable price point has dried up due to the number of foreclosed properties that continue to flood the market (I have heard reports of an additional 3.5 million homes that will go into foreclosure by the end of 2010).

Let's consider the question "What can lenders do to help get the economy back on track?"

As lenders asssess lending opportunities and risks in today's troubled economy, they must consider many factors, and unavoidably all sorts of questions pop into their minds. For example:



Item: Congress is poised to enact new laws and regulations in the wake of the financial crisis and the banking industry is experiencing a great deal of FUD anticipating the outcome.

Item: Banks are struggling to properly assess credit risk so that they can make profitable loans – the demand is there, but they are hesitant to lend.

Item: It’s now the end of another quarter and lenders are looking at one or more of their product lines and finding that they've come up short in terms of revenue, and have mounting losses. “What happened?”

Clearly, the FICO score is too blunt and and too opaque a method with which to manage credit nowadays. With CCAF you can find the answers and take immediate and appropriate action.

Take, for example, the question “How should we tighten (loosen) credit?” which I highlighted in red in the above figure. Let's contrast the curent "state-of-the-art" with the CCAF approach for addressing this question, which is both a natural, and a recurrent, one in the world of lending operations.

In his recent article “How Did Economists Get It So Wrong?” New York Times journalist Paul Krugman takes the reader on a journey from the birth of economics with the publication of Adam Smith’s “The Wealth of Nations” in 1776 to the theories of John Maynard Keynes and finally to today’s salt water versus fresh water schools of economic thought. He concludes that the Keynesian economic theory remains the best framework for making sense of recessions and depressions. Advancing the notion that markets are not perfect, people are not always rational, Krugman quotes H. L. Mencken: “There is always an easy solution to every human problem — neat, plausible and wrong.” Krugman offers his opinion that the vision of the economic profession therefore will likely not be neat, may seem a bit unclear, but will hopefully be partially right!

Mainstream economists were not the only ones who did not see the freight train coming. We can add many more to that list! Someone quipped that the regulators just needed a dashboard with dials that measure greed. I would offer that if the nation's credit system had been able to leverage thechnology like the Comprehensive Credit Assessment Framework (CCAF, pronounced See-Caf) then we would have seen the handwriting on the wall long before the problem reached such giagantic proportions. CCAF provides early warning on non-delinquent trends and concentrations that precede performance declines. Furthermore, it surfaces loans where the loan products, the collateral, and the borrower are all associated with high risk. Needless to point out, that is a very bad combination!

Even more important, CCAF would not have allowed the risky loans to be made in the first place because it does not put riskier borrowers purchasing over-priced homes into the riskier loan products. This is because CCAF integrates borrower payment history with borrower capacity, borrower capital, borrower equity in the property being financed, and borrower and property vulnerability to future market and economic scenarios.

In the first chapter of our latest book that describes a new lending system for borrowers, lenders, and investors, we trace through the suspected causes and perpetrator of the financial crisis to 3 roots as depicted below:




Credit Risk Assessment: The New Lending System for Borrowers, Lenders & Investors, Clark Abrahams & Mingyuan Zhang, copyright 2009, SAS Institute, Inc. Reprinted with permission of John Wiley & Sons, Inc

CCAF addresses all three root causes in the following ways:

1. It is fully transparent, unlike credit scoring models. Its classification and qualification of borrowers is open and immediately verifiable by logical inspection, i.e. the categorization of the borrowers by the 5 C's of Credit correlates with their risk score. No one can look at a scorecard and tell if it is correct -- not even FICO !!

2. It is comprehensive and uses common sense -- CCAF does not rely on models to determine what is important in granting a loan and it does not have static weights like a scorecard -- in fact it adjusts its score based on all relevant factors viewed simultaneously! It is not a one-size fits all approach. The expression "To the boy with a hammer, the world is a nail" comes to mind when I think of the FICO credit brueau score and how much it contributes to the problem of inadequate risk assessment.

3. CCAF promotes a balanced approach to lending that emphasizes affordability, rather than:

A) the lowest monthly payment for a loan amount sufficiently large to put the borrower at considerable risk if everything doesn't go according to plan, or

B) the highest return for the lender associated with somewhat less suitable loan products for the borrower's core need.

Sadly, we are now witnessing the fallout from unchecked greed, namely:

i) borrower's homes in foreclosure
ii) lenders having higher loan loss reserves and owning more property in their OREO portfolio
iii) investors holding ill-liquid investments with accompanying negative returns.

I encourage you to check out the latest edition of SAS.com magazine.

In the Industry Outlook Section (pages 4-5) Sunny and I share our thoughts on the current financial crisis -- its cause, what can be done about it, and how we can prevent reoccurrences.



Essentially, we propose that what is needed is a new form of lending system (Comprehensive Credit Assessment Framework, or CCAF -- pronounced See-Caf). "WHY?" you might ask. The answer, in a nutshell, is that there a number of flaws in the current lending systems that need to be addressed, such as:

1. Current risk assessment models are inaccurate

a. Failure to consider all relevant factors simultaneously
b. Over-reliance on past performance

2. Modeling approach is unrealistic

a. Purely data driven and overly relies on quantitative methods
b. A black box and hard to interpret
c. One size fits all
d. Risk rating is performed prior to segmentation

3. Models are static with limited life

a. They become less accurate over time
b. Model re-development and model validation processes are very tme-consuming

CCAF is a brand new system, based on a holistic analytic framework that addresses the entire lending value chain , which is depicted in a figure from our recent book below:



CCAF creates loan-level transparency for securitization investors, which provides them with the ability to have a 360 degree view of all risk components of each loan which relate to borrower payment performance, stability, capacity, and capital, in addition to loan collateral and the vulnerability of both borrower and collateral to future market and economic conditions.

CCAF affords:

1. A far broader range of outcomes being considered
2. A transparent process that captures the business reality for borrowers, lenders, and investors
3. An adaptive system that actually becomes more predictive over time

Please do check out our latest published article and let us know what you think!

In her recent blog, Law Professor Ann Graham shared with her readers some of my thoughts on this subject of "generalizing"and provided some good comments of her own on the credit granting process. I contend that in the area of credit granting, most generalizations are no good. An example of a generalization in the judgmental days of credit granting (from page 187 of our first book) was "Never lend to beauticians, bartenders, or barbers." In the age of credit scoring, generalizations have emerged among credit professionals in various lines of business. A couple of examples, say in automobile lending, are:

1. The notion that the FICO score captures 40% of the information value for a credit applicant in determining the likelihood that they will pay as agreed on their loan. As a consequence, a lot of weight may be put on the FICO score for risk-based pricing whereby a customer who normally pays cash for his/her car and who does not maintain a credit balance on thier credit card may be charged more than someone who uses credit to the extreme and lives on thre edge--but has not failed to make a payment yet.

2. The notion that a ten percent increase in the payment-to-income ratio above some pre-determine threshold will make an applicant twice as likely to default on their car loan. Would you think a higher payment-to-income ratio would affect a borrower who lives paycheck-to-paycheck with little savings more than a borrower who has little other debt and a high savings rate?

These generalizations lead to flawed credit policies, bad loan decisions, and alienated customers. As these examples point out, the problem is not the use of judgment or the use of credit scoring or statistical analysis. The problem is the fact that sweeping gereralizations are just plain inaccurate, whether they are rooted in opinion or from a model that was developed on historical loan performance data. Our new lending system avoids the pitfals of generalizations due to its unique handle-based segmentation feature which first seeks to classify borrowers comprehensively before passing judgment on them.

In the absence of complete information and consideration of all relevant factors, credit granting practiced today is, in effect, generalizing. We assert that this was a major contributory factor to the financial crisis, economic downturn and credit crunch that consumers are feeling the brunt of these days. We cover this topic in great depth in our second book.

Sweeping financial reform legislation, backed by President Obama and his Administration, is moving through Congress. It is currently being reviewed and revised by Chairman Frank’s Financial Services Committee and is expected to be put before the full House of Representatives by September. It is expected that the Senate will probably take a different approach and move a large, more comprehensive package later in the year. If enacted in its current form, which is highly likely, the House version will have a broad impact including, the creation of a new agency (Consumer Financial Protection Agency) that will have a very extensive mandate and powers, including the ability to write regulations to implement consumer financial protection laws, such as ECOA, TILA, FCRA, HMDA, HOEPA, RESPA, and so on. The new legislation will require lenders to report additional data to the government under pre-existing laws and regulations, such as HMDA. Furthermore, lenders will be compelled to offer “plain vanilla” products to consumers, in response to a perception that a lack of understanding of loans on the consumer’s part, including the implications of loan terms and conditions, contributed significantly to the current financial crisis.

Sunny and I see the handwriting on the wall for further standardization, not only of loan products, but the process by which consumers are qualified for loans.

Our comprehensive credit assessment framework (CCAF) perfectly fulfills this need in three key areas, namely:

1. The new law seeks to ensure that consumers have, and understand, information that will help them make their financial decisions.

--> CCAF’s transparency, and its ability to portray all primary factor ratings for consumers with a single, interpretable number, is precisely what’s needed to satisfy this requirement.

2. Further, there is some additional emphasis on ensuring fair lending laws are doing the job they were intended to do, especially relative to the means by which consumers are classified and qualified for a loan.

--> Proper classification lies at the heart of that issue of fairly qualifying consumers for loans prior to determining the riskiness of the transaction. Our new lending framework would automatically identify similarly situated borrowers across protected classes, e.g. specific minority groups, women, and so on, and would facilitate comparisons relative to loan denials, above threshold pricing, market penetration, and other measures that characterize lending practices.

3. Finally, providing the underserved consumer segment greater access to financial services and products is another end goal of the legislation.

--> CCAF can help on that front as well, since it leverages alternative data to qualify borrowers who fall outside of the mainstream and it avoids 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.