For other versions of this document, see http://wikileaks.org/wiki/CRS-RL34442 ------------------------------------------------------------------------------ Order Code RL34442 HUD Proposes Administrative Modifications to the Real Estate Settlement Procedures Act April 10, 2008 Darryl E. Getter Specialist in Financial Economics Government and Finance Division HUD Proposes Administrative Modifications to the Real Estate Settlement Procedures Act Summary Recent unsettling developments in the subprime home loan market have triggered concern in Congress and among the public as to whether borrowers were fully informed about the terms of their mortgage loans. Some observers have suggested that numerous borrowers in the subprime market may have been victims of predatory lending practices or other discriminatory activity. Several bills have been introduced in the 110th Congress that would seek to remedy these perceived abuses. Senate bills S. 1222 (Senator Barack Obama et al.), S. 1299 (Senator Charles Schumer et al.), S. 1386 (Senator Jack Reed et al.), and House bill H.R. 2061 (Representative Stephanie Tubbs Jones et al.) include both suitability and disclosure approaches for addressing these concerns. This report focuses on borrower disclosure, in particular with respect to making all pertinent information about loan terms and settlement costs transparent, so consumers can make well-informed financial decisions when choosing mortgage products. The Real Estate Settlement Procedures Act (RESPA) of 1974 requires standardized disclosures about the settlement or closing costs of residential mortgages. The Department of Housing and Urban Development (HUD) has proposed changes to RESPA designed to facilitate better understanding of mortgage terms as well as to enhance the ability of borrowers to shop for better terms. These changes include (1) a new, standardized good faith estimate (GFE) form; (2) changes in how the yield spread premium (YSP) or broker compensation would be disclosed to the borrower; (3) modifications to the HUD-1 settlement statement; (4) a reading of the mortgage terms to the borrower at the closing table; and (5) allowing for discount pricing of settlement services that would potentially benefit borrowers. HUD plans also to seek legislative changes to enhance its enforcement authority of RESPA. After reviewing specific regulatory reforms, this report provides some survey evidence on the extent of consumer shopping behavior prior to entering into such expensive financial transactions. While there has been concern about lenders not doing enough to provide affordable mortgages, it is also arguably in the best interest of borrowers to shop diligently for the best credit terms. Even if borrowers are made aware of the costs, they may still elect to obtain less affordable mortgages over the entire loan term. Hence, disclosure reform as a way to influence affordability still depends upon the judgement of consumers. Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 HUD-Proposed RESPA Modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 The Good Faith Estimate (GFE) Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Disclosure of Mortgage Broker Compensation . . . . . . . . . . . . . . . . . . . . . . 4 Disclosures Occurring at the Closing Table . . . . . . . . . . . . . . . . . . . . . . . . . 5 Revisions to HUD-1 Settlement Statement . . . . . . . . . . . . . . . . . . . . . . 5 Closing Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Average Cost Pricing and Volume Discounts . . . . . . . . . . . . . . . . . . . . . . . . 6 Proposed Legislative Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Examination of Mortgage Loan Price Differentials and Shopping Behavior Among Consumers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 List of Tables Table 1. Percentages of Homeowners Who Aggressively Shop for Best Credit Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Table 2. Do Interest Rate Differentials Provide Incentive to Shop? . . . . . . . . . . . 8 Table 3. Rates Paid by High Credit Quality Borrowers on Fixed-Rate Mortgages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 HUD Proposes Administrative Modifications to the Real Estate Settlement Procedures Act Introduction Uncertainty generated by recent subprime mortgage repayment problems and subsequent foreclosures has generated concern in Congress and among the public as to whether borrowers are taking on loans they can not afford.1 Borrowers may have obtained expensive, unaffordable loans for various reasons. Perhaps the actual costs of the mortgage were hidden or simply not transparent when borrowers entered into these transactions. Hidden costs can act as a payment shock to a borrower, causing financial distress which could possibly lead to rising foreclosure rates. Borrowers may have entered into high cost loans as a result of discrimination. Recent mortgage repayment problems may reflect a rise in various forms of predatory lending. This report focuses on current disclosure legislation, which requires the reporting of pertinent loan information to consumers. The Real Estate Settlement Procedures Act (RESPA) of 1974 requires standardized disclosures about the settlement or closing costs, which are costs associated with the acquisition of residential mortgages.2 Examples of such costs include loan origination fees or points, credit report fees, property appraisal fees, mortgage insurance fees, title insurance fees, home and flood insurance fees, recording fees, attorney fees, and escrow account deposits. RESPA currently includes the following provisions: (1) providers of settlement services are required to provide a good faith estimate (GFE) of the settlement service costs borrowers should expect at the closing of their mortgage loans; (2) a list of the actual closing costs must be provided to borrowers at the time of closing, which are typically listed on the HUD-1 settlement statement; and (3) RESPA prohibits "referral fees" or "kickbacks" among settlement service providers to prevent settlement fees from increasing unnecessarily. The Department of Housing and Urban Development (HUD) currently implements RESPA. The Truth-In-Lending Act (TILA) of 1968 is another element of the consumer disclosure regime.3 TILA requires lenders to disclose the cost of credit and repayment terms of mortgage loans before borrowers enter into any transactions. The 1 For more detailed information on subprime lending, see CRS Report RL33930, Subprime Mortgages: Primer on Current Lending and Foreclosure Issues, by Edward Vincent Murphy. 2 P.L. 93-533, 88 Stat. 1724, 12 U.S.C. Sections 2601-2617. 3 TILA is contained in Title I of the Consumer Credit Protection Act, P.L. 90-301, 81 Stat. 146, as amended by 15 U.S.C. Section 1601 et seq. CRS-2 Federal Reserve Board implements TILA through Regulation Z. When homes are purchased, both a real estate and a financial transaction take place simultaneously. TILA governs the disclosure of the credit costs and terms, and RESPA governs disclosure of the closing costs. RESPA may be a reasonable place to implement disclosure improvements specific to the homebuying process, but such reforms arguably could occur under TILA, since they pertain to GFE disclosures that cover lending costs and terms.4 HUD has again proposed changes to RESPA designed to enhance the ability of homebuyers to understand mortgage terms and associated costs as well as enhance their ability to shop for the best deals.5 More transparent information could enhance consumer shopping and discourage predatory, discriminatory, and fraudulent lending practices. Various legislative disclosure proposals, however, arguably depend upon consumers to make prudent choices. Consumers may be fully aware that a particular mortgage loan and associated costs may be expensive yet decide to assume the loan. Disclosure reforms, therefore, may not ensure that consumers will choose more affordable mortgage loans. Evidence on the shopping behavior of consumers will also be presented, which may provide some indication of the extent to which consumers shop to obtain the best terms. HUD-Proposed RESPA Modifications The sections below highlight what may be considered the major items in the recent proposed RESPA revisions, published in March 2008. HUD proposes (1) a new, standardized GFE form; (2) changes in how the yield spread premium (YSP) or broker compensation would be disclosed to the borrower; (3) modifications to the HUD-1 settlement statement; (4) a reading of the mortgage terms to the borrower at the closing table; and (5) allowing for discount pricing of settlement services that would potentially benefit borrowers. HUD plans also to seek legislative changes that are discussed below. The Good Faith Estimate (GFE) Form Although current GFEs contain disclosures that are required by TILA and RESPA, evidence suggests that borrowers still have difficulty understanding and locating relevant mortgage information. A Federal Trade Commission (FTC) study tested 819 consumers to document their understanding of current mortgage cost disclosures and loan terms, as well as their ability to avoid deceptive lending 4 See Patricia A. McCoy, "Rethinking Disclosure in a World of Risk-Based Pricing," Harvard Journal on Legislation, vol. 44, 2007. 5 See HUD's RESPA home page at [http://www.hud.gov/offices/hsg/sfh/res/respa_hm.cfm], which links to the proposed ruling that appeared in the Federal Register, vol. 73, no. 51, March 14, 2008. HUD proposed RESPA revisions on July 29, 2002, but withdrew its recommendations on March 22, 2004 for further consideration and analysis. See letter from Office of Management and Budget addressed to Acting HUD Secretary dated March 22, 2004, at [http://www.reginfo.gov/public/postreview/hud_response9.pdf]. CRS-3 practices.6 The authors found both prime and subprime borrowers did not understand important mortgage costs after viewing mortgage cost disclosures. Some borrowers had difficulty identifying the annual percentage rate (APR) of the loan and loan amounts. Many borrowers did not understand why the interest rate and APR of a loan would differ.7 In addition, borrowers had the most trouble understanding loan terms for the more complicated mortgage products such as those with optional credit insurance, interest-only payments, balloon payments, and prepayment penalties. More borrowers were unable to determine whether balloon payments, prepayment penalties or up-front loan charges were part of the loan. HUD has proposed a new, standardized GFE with recommended changes that it argues will help consumers better understand and locate relevant information about their mortgage products.8 For example, the proposed GFE conveys information about the mortgage terms, whether the interest rate can rise, whether the overall loan balance can rise, whether the loan has a prepayment penalty, whether the loan has a balloon payment, and whether the quoted monthly payment includes a monthly escrow payment for taxes. All of this information about the loan appears on the first page of the GFE. The first page also includes a lump-sum total amount of all estimated settlement fees. Furthermore, a separate GFE will be required for each loan product. For example, a borrower may wish to compare a traditional fixed rate mortgage (FRM) loan with an adjustable rate mortgage (ARM) loan. Both mortgage products must have separate GFEs to ensure that the information provided is unique to each product. HUD argues that these changes to the GFE will lead to less confusion about loan and settlement costs, help the borrower better determine product affordability, and improve comparison shopping.9 In addition to providing a shopping tool for consumers, the HUD proposal also seeks to provide reliable GFEs in the sense that estimated costs do not change substantially by the time consumers are ready to close on their loans. Shopping for the best deal or an affordable loan would be meaningless if the costs were to change when borrowers arrived at closing. Consequently, page 3 of the proposed GFE lists charges that can not increase, charges that are allowed to increase up to 10%, and charges that may change at settlement. For specific charges that should not change or exceed the 10% limit, borrowers would have the option to withdraw their applications and receive full refund of all loan related fees. This change would make it difficult for lenders to generate "costs" or fees that could not be easily justified. 6 See James M. Lacko and Janis K. Pappalardo, Improving Consumer Mortgage Disclosures: An Empirical Assessment of Current and Prototype Disclosure Forms, Bureau of Economics Staff Report, Federal Trade Commission, June 2007, [http://www.ftc.gov/os/2007/06/P025505MortgageDisclosureReport.pdf]. 7 The APR is the annual cost of a loan, which includes the interest cost of the principal loan amount, insurance, and other fees expressed as a percentage. The mortgage interest rate only includes the interest cost of the principal loan amount expressed as a percentage. 8 See the proposed GFE at [http://www.hud.gov/offices/hsg/sfh/res/200803/5180GFE.pdf]. 9 HUD has tested this proposed GFE form among various consumer groups; details about these tests may be found at [http://www.hud.gov/offices/hsg/sfh/res/200803/summary.pdf]. CRS-4 These GFE disclosure requirements may generate controversy. After an earlier proposed ruling concerning GFE modifications and requirements, loan originators argued the GFE would essentially become a binding or final quote, which would be extremely difficult to provide without underwriting. One reason for this criticism stemmed from the fact that loan pricing depends upon having information about borrower credit history and ability to pay. Loan originators need to assess borrower risk (and perhaps the riskiness of the collateral) to generate a binding quote, in particular for high-risk borrowers, a process which takes time and money. The HUD proposal defines two stages in the overall mortgage seeking process, which could be interpreted as a response to the earlier criticism.10 The consumer receives a GFE in stage 1, which occurs prior to the proceeding with the official mortgage application in stage 2. In the first stage, the lender is not expected to have performed any underwriting, and the GFE need only consist of information obtained from the borrower without any verification of borrower statements. Final underwriting is expected to begin in stage 2 after the borrower has expressed a willingness to proceed with an official mortgage application. The GFE becomes binding only if the underwriting process confirms borrower statements and loan qualifications. If the underwriting process reveals that the borrower is unable to qualify for the specific loan product, then the lender may reject the borrower or propose a new GFE for another loan product in which the borrower is more likely to qualify. If the proposed two-stage process is implemented, then the final loan rates initially stated on the GFE forms would likely become the actual ones borrowers would receive after underwriting, since the majority of borrowers are considered prime or high credit-quality. Lenders currently advertise the interest rates that prime borrowers are likely to be charged, and high credit quality borrowers are arguably already able to shop for loans.11 Subprime or high risk borrowers, however, encounter difficulties shopping for loan rates and may continue to do so under this proposed system. Lenders typically charge higher rates to riskier borrowers to compensate for the additional risk, and such rates are typically determined after underwriting has occurred. Hence, low credit quality borrowers may be less likely to obtain estimates of loan rates prior to final underwriting that would not change afterwards. Assuming no substantial shifts in the current proportion of prime relative to subprime borrowers, or that the share of prime borrowers diminishes as a result of further borrower risk gradations, underwriting at the GFE stage might not be necessary for the vast majority of consumers to obtain fairly reliable pricing information of mortgage products. Disclosure of Mortgage Broker Compensation HUD is proposing a new way to disclose to the borrower how the lending institution or a mortgage broker, who is an agent that works for a lending institution, is paid. Loan charges may be collected either through points (upfront fees), or via the 10 See Federal Register, vol. 73, no. 51, March 14, 2008, p. 14035. 11 See Patricia A. McCoy, "Rethinking Disclosure in a World of Risk-Based Pricing," Harvard Journal on Legislation, vol. 44, no. 1, winter 2007. CRS-5 interest rate mechanism, which is referred to as the yield spread premium (YSP), or some combination of these two pricing mechanisms.12 On page two of the proposed GSE form, the total origination costs are disclosed in item 1. The division of these costs into points and YSP is disclosed in item 2. A "credit" that represents the dollar value of loan origination costs not paid at settlement would appear in item 2. In this context, "credit" does not mean the borrower would receive a refund from the loan originator. Instead, credit refers to the loan origination costs that the borrower still pays, not upfront at settlement, but in the form of a higher interest rate, or the YSP. Conversely, the dollar value of fees paid upfront at settlement would appear as a "charge" in item 2. For a given interest rate, both credit and charge amounts in item 2 should add up to the total loan origination costs, which appears in item 1. The adjusted origination costs, which appear in box A, are the difference between the total loan origination costs and the YSP; the adjusted origination costs refers to the amount of total upfront fees that will be paid at settlement. If borrowers realize that mortgage loan origination costs may be collected by some combination of upfront fees and YSP, then they may also realize that it is possible to choose between paying higher upfront fees for a lower interest rate or lower upfront fees for a higher interest rate. In order to facilitate the understanding of the tradeoff between interest rates and points, HUD has also proposed including a comparison table on page three of the new GFE form. For any given loan arrangement, the proposed GFE discloses how a loan with the same principal face value and a lower interest rate results in higher upfront settlement costs; it discloses how the same loan with a higher interest rate results in lower upfront settlement costs.13 Disclosures Occurring at the Closing Table Revisions to HUD-1 Settlement Statement. As stated earlier, the use of a standardized HUD-1 settlement statement is required at all settlements or closings involving mortgage loans. The HUD-1 lists all settlement charges paid at closing, the seller's net proceeds, and the buyer's net payment. HUD proposes modifying the HUD-1 to link GFE and HUD-1 charges. The itemized charges listed on the HUD-1 would include references to the same charges originally listed on the proposed GFE. With these references, it may become more apparent to borrowers what charges remained the same or changed from the estimation stage to the closing stage. Closing Script. HUD has proposed a closing script that settlement agents must read to borrowers at closing. The closing script, which, like the GFE, is prepared for each loan type, restates the loan terms and highlights any changes in 12 The mortgage interest rate and the YSP are not identical. The YSP is defined as the difference between the total coupon interest rate and the actual wholesale interest rate of the loan. For example, a loan with a market or wholesale rate of 6% may have a total coupon rate of 6.5%, and 0.5% is the compensation going to the mortgage broker, which is the YSP. Some mortgage lenders may pay brokers up to 2% in YSP. 13 The GFE never discusses loan details using the term annual percentage rate (APR), which expresses the total interest and settlement costs as a percentage. CRS-6 charges or fees that occurred from preparation of the initial GFE to settlement. The borrower must sign and be given a copy of the closing script after it has been read. Average Cost Pricing and Volume Discounts In another element of the RESPA revisions package, HUD proposes permitting average cost pricing and volume discounts on settlement service fees. While still making it unlawful for lenders or settlement agents to benefit from kickbacks that increase borrower fees, the proposal would allow pricing arrangements that may financially benefit borrowers. Average cost pricing refers to dividing the total costs associated with a specific service (e.g. appraisal service) by the total number of services (appraisals) provided, and charging borrowers for the average cost of the services. Average cost pricing may apply to all services. Under this system, some borrowers will pay less for services that may actually cost more; but the proposed rule prevents some borrowers with actual fees below the average cost to be charged more than average cost. The average cost of various service fees, which will be computed over specified periods, will be based upon methods established by HUD. Settlement service agents will also be permitted to provide volume discounts that may be passed on to borrowers. Proposed Legislative Actions HUD also seeks legislative changes that will give the agency greater statutory authority to enforce specific sections of RESPA, in particular sections dealing with key disclosure provisions.14 The Department of HUD requests that the HUD Secretary be granted the authority to impose penalties for RESPA violations. Legislation may also be necessary to authorize procedures requiring that borrowers receive the HUD-1 closing statement three days before settlement and establishing a uniform statute of limitations for RESPA violations. Examination of Mortgage Loan Price Differentials and Shopping Behavior Among Consumers Assuming implementation issues such as costs associated with reforming RESPA are addressed, consumers still need to shop actively for the best loan terms to obtain more affordable mortgage products.15 In this section, the Federal Reserve Board's 2004 Survey of Consumer Finances (SCF) is used to examine consumer behavior with regards to shopping for affordable credit terms.16 Assuming 14 See Federal Register, vol. 73, no. 51, March 14, 2008, p. 14033. 15 In a press release, the Mortgage Bankers Association alludes to the amount paperwork, which translates into costs, that would increase if the RESPA proposal were implemented at [http://www.mortgagebankers.org/NewsandMedia/PressCenter/60847.htm]. 16 The SCF, which is compiled every three years by the Federal Reserve Board, provides information on the assets, liabilities, and demographic characteristics of approximately 4000 U.S. families. For more information about the 2004 SCF, see Brian K. Bucks, Arthur B. (continued...) CRS-7 competitive market behavior, borrowers having similar risk characteristics should pay similar rates for their loans. Evidence pertaining to mortgage pricing differentials and shopping behavior will be presented.17 All SCF respondents were asked if they were more likely to "shop around for the best credit terms." Mortgages originating between 1996 through 2004 were retained for this analysis. The responses to the survey question make it possible to identify whether homeowners exhibited behavior consistent with obtaining an affordable mortgage loan rate. The results in Table 1 suggest that less than a third of the homeowners in the sample admit to shopping aggressively for the best credit terms. There is little behavioral difference between borrowers with fixed rate or adjustable rate mortgages (ARMs). Homeowners obtaining ARMs may not aggressively shop if they had planned either not to stay in their residences for a long period of time or to refinance in the near future. Fixed rate borrowers, however, would be expected to have significantly higher shopping rates than borrowers with ARMs, since they would be locking in a rate for a longer period of time. The results indicate that relatively few borrowers shop aggressively to acquire the best credit terms. Table 1. Percentages of Homeowners Who Aggressively Shop for Best Credit Terms Total Number Total w/ Fixed Rate Total w/ ARMs Total Who Homeowners in Mortgages Who Shop Who Shop Shop Sample Aggressively Aggressively 1809 26.5% 26.9% 23.9% Source: CRS calculations using the 2004 SCF, Federal Reserve Board. Were the differences in rates large enough for shopping to matter? Answering this question requires placing some controls on the data sample and variables. The data set has been limited to include only fixed-rate conforming mortgages.18 Mortgage holders with traditional 30-year fixed-rate conforming mortgages were chosen for this part of the analysis to avoid mixing of terms, adjustable, or jumbo mortgage loan rates. Unlike adjustable rate mortgages (ARMs), borrowers with fixed rate mortgages do not share interest rate risk with the lender. When interest rates 16 (...continued) Kennickell, and Kevin B. Moore, "Recent Changes in U.S. Family Finances: Evidence from the 2001 and 2004 Survey of Consumer Finances," Federal Reserve Bulletin, vol. 92 February 2006, pp. A1-A38. 17 All descriptive statistics presented in this analysis are computed using SCF sample weights. For information about the weights, see Arthur B. Kennickell, Revisions of the SCF Weighting Methodology: Accounting for Race/Ethnicity and Homeownership (Washington: Board of Governors of the Federal Reserve System, December 1990, [http://www.federalreserve.gov/pubs/oss/oss2/method.html]. 18 The definition of a conforming loan is a mortgage loan that meets the guidelines set by Fannie Mae and Freddie Mac. CRS-8 change, borrower cash outlays remain unchanged if they have traditional fixed rate mortgages. Hence, the risk-preference characteristics of the borrower are held constant in this analysis. It is possible to control for some of the interest rate volatility over the time, since the SCF asks respondents for the year they obtained their mortgage.19 The mortgage loan rates were normalized by the average rate for the year in which they were obtained, so borrowers with mortgages in different years could be compared. For example, a borrower might report obtaining a 30-year fixed loan in 2002 with an interest rate of 6%. Given that the average conventional 30-year fixed mortgage rate in 2002 was 6.54%, according to interest rate data provided by Freddie Mac, the normalized loan rate is 6/6.54, or a value of 0.9174. This indicates the borrower received a loan price below average during that year. All loan rates reported by 2004 survey respondents have been normalized by the average mortgage rate by origination year; then a loan price distribution was produced to compare rate pricing differentials.20 Table 2 shows the mean normalized rates for the bottom and top portions of the mortgage interest rate distribution for all homeowners in the sample. Homeowners in the bottom 25% of the loan distribution paid lower interest rates for their loans relative to homeowners in the top 25% of the distribution. The average rate in the bottom quartile was 47% lower than the mean rate in the top quartile, and the mean mortgage rates in the top and bottom quartiles were statistically different from each other at the 99% level of significance.21 These results indicate that an incentive to shop for the best credit terms would have existed, given the wide range of available interest rates. Table 2. Do Interest Rate Differentials Provide Incentive to Shop? Actual Number Mean Standard Borrowers in Normalized Deviation of Quartile Mortgage Rates Normalized Rates Bottom Quartile 173 0.7978 0.0031 Top Quartile 146 1.1731 0.0056 Source: CRS calculations using the 2004 SCF, Federal Reserve Board. Did creditworthy borrowers get better pricing relative to those of low credit quality? Assuming lenders price according to borrower risk, a larger share of high credit quality borrowers would be expected to pay rates in the bottom quartile. Does 19 The SCF does not have information on closing costs, so it is not possible to use APRs. 20 There was very little month-to-month volatility in 1996 through 2002, but volatility increased from 2002 through 2004 as interest rates declined. 21 The median interest rates were 0.8162 for the bottom quartile and 1.1168 for the top quartile. CRS-9 aggressive shopping help high credit quality borrowers receive more favorable pricing? Answers to these questions require first identifying high credit quality mortgage holders. The SCF does not have credit score information, so a proxy for borrower creditworthiness was constructed. A high credit quality borrower, therefore, is defined as one who has reported "not being behind 2 months or more making loan repayments in the last year," "not being turned down completely for credit in the last 5 years," and "never having filed for bankruptcy." After controlling for year of loan origination, degree of creditworthiness, and borrower risk preferences, the results in Table 3 indicate pricing differences did exist among high credit quality borrowers. The results also show that less than one-third of high credit quality borrowers reported aggressively shopping for the best credit terms. According to Table 3, 76% of all borrowers paying rates in the lower 25% percentile of the normalized mortgage loan rate distribution were high credit quality borrowers, and 63.3% of all borrowers paying rates in the upper percentile were of high credit quality. Although the share of high quality borrowers in the top quartile is smaller than the share in the bottom quartile, there is still a large share of borrowers paying relatively higher loan rates, given their seemingly low levels of credit risk. Table 3. Rates Paid by High Credit Quality Borrowers on Fixed-Rate Mortgages Share Black and Share High Share Black and Actual Share High Hispanic High Credit Quality Hispanic High Number Credit Quality Credit Quality Who Shop Credit Quality Borrowers Borrowers Who Shop Aggressively Borrowers Aggressively Bottom 173 76.4% 29.7% 16.9% 27.9% Quartile Top 146 63.3% 29.3% 24.6% 41.6% Quartile Source: CRS calculations using the 2004 SCF, Federal Reserve Board. Why did a large share of high credit quality borrowers have loan rates in the upper quartile of the distribution? One explanation may be an increase in lending practices resulting in borrowers only able to obtain high priced loans. Another explanation may be that consumers did not aggressively shop for the best credit terms. The results in Table 3 reflect a low degree of aggressive shopping by consumers. Less than one-third of high credit quality borrowers report aggressive shopping for the best credit terms, and the difference between the percentages of high credit quality shoppers in the quartiles is still relatively small. Shopping may arguably not matter for high credit quality borrowers in the bottom quartile, since more than two-thirds managed to get better pricing without being aggressive shoppers. The results in Table 2, however, suggest a strong incentive to shop did exist, given that the interest rates in the quartiles were significantly different. Hence, it may be surprising that fewer high credit quality households, especially those in the top quartile of the distribution, reported being aggressive when shopping for the best CRS-10 terms. Of course, respondents may have applied various subjective interpretations of "aggressive" when answering the survey question, which could affect the results. To what extent might discrimination explain the pricing differential? The weighted percentages for Black and Hispanic borrowers have also been reported in Table 3. Approximately 17% of the high credit quality borrowers in the bottom quartile were black or Hispanic. Only 27.9% of the minority borrowers reported shopping aggressively for lower terms, which was not much different from the percentage of the entire sample. Almost 25% of the high credit quality borrowers in the top quartile were black or Hispanic, and 41.6% reported shopping aggressively for the best credit terms. Greater information on the credit shopping experiences of minority borrowers would therefore be a useful component to further investigation of this question. Minorities may have established fewer relationships with traditional financial institutions and therefore be more likely to feel intimidated when dealing with them. As a result, these borrowers might be more prone to shopping within the universe of subprime lenders, rather than seeking prime lenders. In addition, a disparate impact, which occurs when a practice or procedure has less favorable consequences on particular groups, may be a factor. If minorities disproportionately earn less income and accumulate less wealth, then loan-to-value and qualifying ratios for these groups would be affected, which would likely affect loan pricing. Hence, it is possible for minority borrowers who aggressively shop to pay higher loan rates if their shopping experiences are vastly different from the majority population.22 The data from the 2004 SCF suggests that less than a third of all homebuyers shop aggressively for the best credit terms, which raises questions about consumer judgment. Would disclosure reform have led to a reduction of subprime mortgage troubles? On the one hand, if the mortgage shopping experience is frustrating under current disclosures, enhanced disclosure may help motivate more consumers to become more active shoppers. On the other hand, consumers who shop in a casual manner for credit terms may not pay much attention to mortgage loan details regardless how they are presented. As a result, it is difficult to determine whether enhanced or expanded disclosure would help more people obtain affordable mortgages, since consumers are not required to shop aggressively when making financial decisions. 22 Discrimination cannot be proven using SCF data. Federal agencies follow Interagency Fair Lending Examination Procedures, provided by the Federal Financial Institutions Examination Council (FFIEC), to evaluate unlawful discrimination. The Department of Justice, which investigates fair lending cases, establishes a pattern of discrimination before charging lenders with violation of federal discrimination laws. The information necessary to establish any patterns of discrimination is not available from SCF data. See [http://www.dallasfed.org/ca/pubs/fair.html] and [http://www.ffiec.gov/PDF/fairlend.pdf]. ------------------------------------------------------------------------------ For other versions of this document, see http://wikileaks.org/wiki/CRS-RL34442