Note on Subprime Lending.

The subprime mortgage industry was built around the concept of risk-based pricing, which meant charging different yields for loans based on the borrowers’ perceived credit quality. In practice, this meant that a subprime lender charged the borrower a higher interest rate for a subprime loan than the rate for prime loans (even if the borrower might have qualified for a prime loan). Golden West rejected the concept of risk-based pricing, believing it would invariably be discriminatory. At Golden West, any borrower who qualified for a loan would receive a prime rate, irrespective of the borrower’s financial information or other characteristics.

Virtually all of the subprime lending in the early- and mid- 2000s used a “2/28” loan product. The 2/28 loan refers to a loan that was structured to trigger a significantly higher interest rate and monthly payment after two years. For example, a 2/28 might have a “teaser” rate for two years of 7%, but the rate could jump to 12% or more after two years, creating a significant risk of payment shock to the borrower. Another popular subprime loan was a “3/27”, which had similar features and would recast after three years. Golden West made no 2/28 or 3/27 loans.

Subprime lending was a small proportion of the overall mortgage market until the late 1990s and early 2000s, when its growth was fueled by the combination of technological advances, investor demand for higher yields, and the purchase of subprime operations by major mortgage banks. In order to generate high volumes of loans, subprime lenders looked for ways to shortcut traditional underwriting. A principal tool they used to expedite underwriting decisions (and to justify charging higher rates to borrowers) were FICO credit scores, which have never been fully validated for residential mortgage lending (they were initially adopted for consumer credit). There are many factors that call the veracity of FICO scoring into question: (a) three different agencies can give widely different FICO scores for the same borrower at the same point in time, and lenders can play games with which FICO scores they choose to use; (b) FICO scores can change quickly for reasons unrelated to true credit risk or, alternatively, the scores can move much too slowly to capture actual risk; and (c) FICO scores can be manipulated; companies exist to help borrowers improve their credit scores in ways that do not meaningfully alter the borrowers’ real risk profile. Golden West, unlike most other lenders, continued to do traditional, holistic underwriting (looking at the borrowers’ actual credit history and the appraisal of the property), rather than relying on shortcut methods like FICO credit scores as a sole or primary determinant for underwriting.

Back to top

a. Manual Underwriting of Each Loan.

The company evaluated the creditworthiness of potential borrowers based primarily on credit history and an evaluation of the potential borrower’s ability to repay the loan. When evaluating a borrower’s ability to pay, Golden West assessed the ability to make fully amortizing monthly payments, even if the borrower had the option to make a lower initial monthly payment. By contrast, many mortgage banking operations only assessed the borrower’s ability to make a minimum, non-amortizing payment. In its underwriting decisions, Golden West also evaluated the characteristics of the property and the loan transaction, including whether the borrower was purchasing or refinancing the property and would occupy the property. Golden West used systems developed internally based on decades of experience evaluating credit risk. Although Golden West used credit scores and technological tools to help with underwriting evaluations, its trained underwriting personnel reviewed each file and analyzed a wide range of relevant factors when making final judgments. Higher-level approvals within the underwriting organization were obtained when circumstances warranted.

b. Internal Appraisal Staff.

Golden West appraised the property that secured loans by assessing its market value and marketability. Golden West maintained an internal staff to conduct and review property appraisals. Any external appraisers used for loans that were originated and retained in portfolio were required to go through a training program with the company, and each external appraisal was reviewed by Golden West’s internal appraisal staff. Golden West did not rely on external automated valuation models (AVMs) in its appraisal process, which is what mortgage banking operations commonly did.

c. Lending on Moderately Priced Properties with Low Loan-to-Value Ratios.

Golden West focused on originating high-quality loans on moderately priced properties because, historically, these properties tended to hold their values better than high-priced properties, particularly in weak housing markets. Golden West did not emphasize lending on higher-priced properties because of concerns about greater price volatility and the larger potential loss if these loans did not perform.

d. Low Loan-to-Value (LTV) Ratios.

The loan-to-value ratio, or LTV, is the loan balance of a first mortgage expressed as a percentage of the appraised value of the property at the time of origination. A combined loan-to-value, or CLTV, refers to the sum of the first and second mortgage loan balances as a percentage of the total appraised value at the time of origination. Golden West focused its lending activity on loans with original LTVs or CLTVs at or below 80%. Golden West’s average LTV at origination was 71%. In the early-to-mid-2000s, many other lenders were regularly making loans equal to the value of the house (100% LTVs).

a. Conducted periodic loan reviews;
b. Analyzed market trends in lending territories and adjusted loan terms, such as required original LTV or CLTV ratios;
c. Reviewed loans that became nonperforming assets to evaluate if there were detectable signs that the company should incorporate into the training of underwriting and appraisal staff;
d. Identified segments of the portfolio that might have more vulnerability to credit risk, either because of geography, LTV or CLTV ratio, credit score, or a combination of these and other factors; and
e. Worked with customers who may present potential risks and offer them counseling or other programs to try to reduce the potential for future problems