Case Study Two—Leveraging Income
The Ortegas need 100 percent financing to purchase their first home. Both have strong job histories and make good income. Javier managed a local computer repair company for four years until he was recruited away last month by another firm. His promotion to regional manager resulted in a substantial pay increase from $55,000 to $100,000.
For five years, Wendy has managed the local chapter of a large non-profit, and she currently makes $95,000 a year. Javier’s credit score is 615, and Wendy’s is 555.
Option A—Guideline versus Practice
The industry guideline for calculating income is to average the most recent two years’ W-2s and two pay stubs within the last 30 days. Using this approach, Javier’s income is $77,500. Since Wendy makes more money, the lender uses her 555 score to qualify them. With only $3,000 in savings, they can’t meet the required 10 percent down payment.
The industry practice, however, is to use the most recent W-2 and one pay stub within the last 60 days or use the borrower’s current salary. In this case, Javier’s income is $100,000, which makes him the primary wage earner. Since the lender will use his 615 credit score to qualify the loan, they’re now eligible for 100 percent financing.
Why Is This an Issue?
This practice is common to both prime and subprime lending. When a salaried borrower’s income rises, the new amount is used for qualification purposes. On the surface, this approach seems to make sense—if a person has improved his position in life he should be allowed to use the higher figure.
The guideline, however, exists for a reason. Part of the decision to lend money is based on how a borrower performs over time. Whether it’s paying bills, making money, or holding down a job, each of these factors into a lender’s decision. Though Javier is credited with making $100,000 a year, there’s no track record to support this income.
Using the industry practice has different implications depending on the type of borrower. A consumer with excellent credit has a history of paying on time. This record helps offset the increased risk from giving someone greater buying power. The subprime borrower hasn’t shown this kind of discipline. In many cases, there are no compensating factors to offset the increased risk.
What does this new buying power mean to the Ortegas? With a combined annual income of $195,000, they can qualify for a $750,000 mortgage. On a two-year ARM at 8 percent, the monthly principal and interest payment alone is $5,466. With taxes and insurance it goes to $6,700. If the interest rate adjusts to 10 percent a few years from now, the payment increases another $1,000 a month.
As an industry, mortgage lending is responsible for two things: effectively managing risk and minimizing the chance that consumers do something stupid. Using this method to calculate a subprime borrower’s income doesn’t serve either purpose. Of course, the Ortegas are responsible for putting themselves in this position. No one put a gun to their heads and told them to sign the loan documents. But the industry did nothing to stop it. In fact, the opposite happened. Applying whichever method was needed to increase a borrower’s buying power, the industry practice or the guideline, was a reckless form of risk management.
Option B—Remove the Borrower with the Lower Credit Score
Let’s assume Javier’s income only increased to $85,000, making Wendy the primary wage earner. Using her 555 score, the lender requires a 10 percent down payment. Assuming the Ortegas were willing to buy a less expensive home, they could still qualify for 100 percent financing by just using his income. This means removing Wendy from the application so the lender can use his score. It gives them less buying power but enables them to purchase a home with no money down.
The deal is still being massaged to fit their needs, but at least it’s a better lending decision. By using only his income, the maximum they can borrow is $300,000. Even though Wendy was removed from the loan application, the deal has a significant compensating factor—her income. She may not be signing on the note, but when her salary is factored into the equation, the payment becomes more reasonable. Compared to the initial scenario, which allowed them to maximize their buying power, it’s a better risk for the borrower and the lender.
Why Is This an Issue?
The practice of removing a borrower from the loan application to get better terms happens frequently in subprime. These two options are contrasted to show how the business changed over the last seven years. In the early part of the decade, consumers seemed less inclined to push deals to the absolute limit. But when rates dropped and housing mania replaced all rational thinking, subprime lenders became the crack dealers of the financial industry—pedaling easy money to anyone who needed it. With a lending environment that promoted irresponsible behavior, borrowers threw caution to the wind and structured more loans that resembled the first scenario discussed for the Ortegas.
Case Study Three—It’s All About What You Don’t Tell Them . . .
Jenny Griffin owns a local bakery that makes unique custom cakes and pastries. Like most self-employed business owners, she uses her expenses to reduce her taxable income. Even though her business does well, her tax returns show very little income. Fortunately, the subprime industry has a 24-month bank-statement program that counts her deposits and treats them like income. Totaling the deposits for the last two years and dividing by 24 months will determine her average monthly income. This gives her the same buying power as qualifying with two years of tax returns.
Looking at the bank statements, the broker notices some problems. First, Jenny bounced a couple of checks last year when business slowed. Guidelines dictate that NSF (non-sufficient funds) or overdrafts are cause for immediate decline under the bank statement program. Second, she made a $25,000 deposit 14 months ago after winning a small Texas Hold’em tournament in Las Vegas. Anomalies like large deposits or balance transfers are excluded when calculating average income. If the underwriter doesn’t count this deposit, Jenny’s income will be too low to qualify.
Option A—Being Resourceful
Fortunately, the lender’s account executive knows a few things the broker doesn’t. In reviewing the statements, he noticed her deposits were greater during the most recent 12 months than the previous year. After crunching the numbers, he determines her income is high enough to qualify using just one year’s bank statements. The broker didn’t know the lender had a 12-month bank statement program through a different investor. Using this program, Jenny gets around the issue of the large deposit since it happened over a year ago.
The rep also knows how to get around the bounced checks. The guidelines require every page from the bank statements be submitted to underwriting. The common practice is only to provide the first page from each month’s statement. This page displays a summary of her activity including total monthly deposits. To find the NSF, the underwriter has to search through a hundred pages of bank statements, which no one has time for. To expedite the process, most investors accept the front pages and never ask for the others, so the bounced checks go unnoticed.
Why Is This an Issue?
Admittedly, bouncing a check is not a heinous infraction. It’s a mistake that can happen to even the most credit savvy consumers. But if Jenny bounced checks every month, which would make her a greater risk, she still could have been approved for the loan. It raises the question, “What good is a policy if no one adheres to it?”
In many ways this example resembles the last case study. Just as Javier Ortega was given credit for his newly increased salary, Jenny Griffin’s underwriter is only using 12 months of income to determine the likelihood that she’ll make her payments. For a self-employed borrower who poses a higher risk, it’s an aggressive lending policy.
The bank statement program could also be manipulated in other ways. Let’s assume the NSF showed up on the first page of a monthly statement. If Jenny could qualify using the other 11 statements while still dividing the total deposits by 12 months, the broker would submit the loan but omit the statement. As lon
g as the calculations worked, the underwriter would approve the loan.
Case Study Four—Now You See It, Now You Don’t
The mortgage industry uses a standard approach to determine which credit score is used for qualification purposes. First, for a borrower’s score to be valid, a minimum of two credit bureaus must produce scores. In most cases, the bureaus are unable to produce a score if a borrower has little or no credit history. A complete credit report will include scores from all three repositories (Equifax, TransUnion, and Experian), but as long as two scores are produced a borrower can still qualify. Second, lenders will use either the middle of three scores or the lower of two scores as the qualifying number.
Bill and Rita Watson are purchasing a home, but they’ve got a problem. Their car was repossessed nine months ago, resulting in a $25,000 collection account. Since it’s a large balance and occurred recently, investors require it to be paid off as a condition of the mortgage. In most cases, this isn’t feasible, which means the deal gets denied. However, the lender’s account executive notices the collection agency handling the account reports only to Equifax. Some small and midsized creditors will report to only one or two repositories in order to reduce costs. If the broker reissues the credit report but removes Equifax, producing information from only the other two bureaus, the collection disappears. One minute the account is there and the next it’s gone. When the underwriter reviews the loan, she has no idea it ever existed.
Why Is This an Issue?
This is a radical example of how to manipulate credit. It has no redeeming value aside from approving borrowers for a loan they didn’t qualify for, the subprime equivalent of three-card monte.
Dropping borrowers from the application or removing bureaus from a credit report was a manipulative process. Here are some other examples in which this method was used to qualify borrowers:
• The Watsons have a large car payment that’s preventing them from qualifying. Since the debt was reported to only one bureau, the broker drops that bureau from the borrower’s credit report and the debt disappears. The underwriter is unable to accurately determine the borrower’s debt-to-income ratio.
• If a debt or collection account that prevents the Watsons from qualifying is in only one of their names, the lender drops that person from the application. If the remaining borrower doesn’t make enough money to qualify on his or her own, the lender puts that person into a stated income loan.
Using this tactic requires a certain amount of luck to be successful. Whenever a repository or borrower is dropped from a deal, it’s possible the loan will get worse, not better. If the dropped repository has the highest of the three scores, the borrower ends up with a lower score since the lender uses the lower of the two remaining scores versus the middle of the original three scores. With a lower credit score, the borrower’s grade could go down, which would require a larger down payment. It’s fortunate that a large percentage of creditors report to all three repositories; otherwise subprime lending could approve substantially more unqualified borrowers.
Case Study Five—Automated Underwriting
The widespread use of automated underwriting (AU) technology brought subprime lending into the twenty-first century. Although systems were slow to develop, most lenders had some form of AU technology in place by 2004. The more robust systems used risk-based decision-making that went beyond basic underwriting guidelines. If a borrower’s compensating factors warranted a loan exception, the best systems could make that call.
They also helped remove the guesswork. With AU approval, a broker had something more tangible to tell his borrower and his realtor. Since lenders stood behind their systems, an approval all but guaranteed a loan would fund. As long as the property value could be substantiated and the information on the application could be verified, loans would close.
AU systems not only modernized the subprime industry, they helped address the greatest frustration for lenders and brokers—scores that declined because lenders had reordered credit reports. Before automation, most brokered loans underwent a similar process. The broker sent a loan application and credit report to his account executive for prequalification. Let’s assume the borrower had a 590 credit score and was preapproved for 100 percent financing. When the loan arrived at the lender’s office three weeks later, the lender ordered a new credit report, which is standard operating procedure. In this case, the score dropped to 550. Since the lender used their scores to underwrite the loan, the borrower was no longer eligible for 100 percent financing. With no cash available for a down payment, the deal quickly fell apart.
AU technology solved this problem. When a broker used a lender’s AU system, he could utilize his credit report to get the approval. Once the loan file arrived at the lender’s office, the underwriter would access the system and print out the broker’s credit report in the lender’s name. The issue of falling credit scores became a thing of the past.
Why Is This an Issue?
When a credit score drops, it means one of three things:
1. The borrower was recently late on a payment.
2. He has used more of his total available credit by running up his credit cards.
3. He has applied for new credit.
When any or all of these happen, the borrower becomes a greater credit risk, which causes his score to deteriorate.
Most investors considered credit reports to be valid for 60 days. This meant a loan had two months to close from the date the credit report was issued, otherwise a new report had to be ordered. Some investors were more aggressive. Credit reports issued through Assetwise, RFC’s AU system, were valid for 120 days. Although it provided lenders with a great selling tool, this may have been the single worst risk policy implemented in the history of RFC. The 60-day time limit has a purpose. High-risk borrowers are less responsible than prime borrowers when it comes to managing their credit. Allowing them 120 days between the time credit is ordered and a loan is closed is long enough to have every account go to collection, file for bankruptcy, get divorced, and have time to spare.
This credit policy created a Catch-22. With RFC, when a broker’s credit report being used through Assetwise was more than 60 days old, we usually pulled credit again to be certain nothing drastic had happened to the borrower. When the credit score dropped significantly, it left us with two choices: decline the loan and upset the broker by not standing behind the AU system, or close a mortgage for a borrower who no longer qualified. To save face with our customer and remain competitive, we usually chose the latter.
For all the benefits AU technology brought to subprime lending, one thing is clear—automation helped lenders close loans that should have been declined. Eight years ago the issue of falling credit scores was a common occurrence in subprime lending. Until automation became a standard part of the business, 10 to 15 percent of loans that brokers submitted to underwriting were turned down for this reason. Getting an AU approval for loans that should have been denied didn’t make the borrowers creditworthy—it meant technology had found a way to circumvent the issue.
Appraisals
Property valuation is a highly subjective process. Ask two residential real estate appraisers to assess a property’s value and you’ll likely get two different answers. Pose the same question to the consumer, broker, lender, and investor, and you’ll get four more. When it comes to valuing real estate, everyone has an opinion. While the most important judgment belongs to the appraiser, it’s by no means the most reliable.
As impartial evaluators, appraisers are supposed to remain objective, following a set of rules and guidelines to determine a property’s fair market value. Their opinion shouldn’t be influenced by anything other than the available data in the marketplace. This is a case where theory and reality are seldom in sync.
Appraisers rely on the lenders and brokers who hire them to make a living. Being true to their profession and pleasing the customer is a difficult balancing act. When a broker orders an appraisal, he provid
es an estimate or target value for the property to the appraiser. If the appraiser has problems consistently reaching this number, the broker will hire someone else. Any appraiser who goes strictly by the book can struggle to get repeat business.
Since property valuation is subjective, there’s an acceptable fudge factor for appraising real estate. This is an allowable deviation—an amount or percentage a property’s appraised value can vary from what a lender or investor thinks it is worth.
Allowing brokers to choose the appraiser, combined with the fudge factor, created a system that was vulnerable to abuse. As commissioned salespeople with no vested interest in a loan’s performance, brokers have the means and the motive to influence the final appraised value. Since the influence is very real and subprime lenders know it, they’ll come to view any broker-ordered appraisal with a high degree of skepticism.
Why is value so critical to the broker? If a home is under contract for $250,000 but it appraises for only $240,000, the deal is in jeopardy. The lender uses the lower of the purchase price or the appraised value as the final home value for lending purposes. If the seller isn’t willing to drop the sales price to meet the appraised value, the buyer needs to bring in the difference ($10,000) in addition to whatever funds were already required for the down payment and closing costs. For cash-strapped subprime borrowers, this usually kills the deal.
Confessions of a Subprime Lender Page 9
