Targeting of African Americans For High Cost Mortgage Grew Worse in 2005, While Fed Downplays Its Own Findings

Byline: Matthew Lee of Inner City Press

   NEW YORK, September 8 -- The targeting of African Americans for higher cost mortgage loans grew more pronounced from 2004 to 2005, data released Friday by the Federal Reserve show.

   The disparities between the mortgage industry pricing for African Americans and whites worsened, even controlling as the industry argues for the change in overall interest rate environment. However, given that the Federal Reserve has yet to take any enforcement action on disparities in lenders' 2004 lending, it is unclear if this new even more disparate data set for 2005 will end what many consumer advocates view as the Federal Reserve's laxity in regulation.

   The report issued by the Federal Reserve on Friday waits until its 39th page to disclose, in the intentionally opaque style of former Fed chairman Alan Greenspan, that "the fact that both spread-adjusted gaps are lower than the comparable unadjusted figures suggests that to the extent that the yield curve changes affected the measurement of racial and ethnic pricing differences, they widen gaps rather than narrow them." Translation: even using the industry's main defense, the yield curse, the disparities grew worse.

   The non-governmental organization Fair Finance Watch, which has raised lending discrimination as a human rights issues, including to United Nations Habitat director Anna Tibaijuka (pictured below, video of Q&A on U.S. Community Reinvestment Act and discrimination here). Where a nation does not act on known discrimination within its borders, FFW argues, it violates treaties it has signed.

   Mortgage lenders were required to release their raw Home Mortgage Disclosure Act data for 2005 on April 1 of this year. 2005 is the second year in which the data distinguishes which loans are higher cost, over the federally-defined rate spread of three percent over the yield on Treasury securities of comparable duration on first lien loans, five percent on subordinate liens. While the Federal Reserve waited six months to compile and analyze the data, a study by Inner City Press of the largest U.S. banks, beginning with Citigroup reached the following findings:

            Citigroup in 2005, in its headquarters Metropolitan Statistical Area of New York City, confined African Americans to higher-cost loans above this rate spread over seven times more frequently than whites, worse than in 2004. Nationwide for conventional, first-lien home purchase loans, Citigroup denied the applications of African Americans 2.69 times more frequently than those of whites, and denied the applications of Latinos 2.02 times more frequently than whites, both disparities worse even than in 2004. Bank of America in 2005 was more disparate to Latinos, denying their applications 2.38 times more frequently than whites, and denying African Americans 2.27 times more frequently than whites.

            Fair Finance Watch designed a way to consider income correlations, by calculating upper and lower income tranches based on each lenders own customers. Nationwide at Citigroup for conventional first-lien loans, 37.73% of upper income African Americans were confined to higher cost loans over the rate spread, versus only 11.46% of upper income whites. Income does not explain the disparities at Citigroup. Nor at HSBC, where less than half of upper income white borrowers were confined to rate spread loans, versus 61.87% of upper income African Americans and an even higher percentage of Latinos, 62.82%. HSBC, which bought Household International in 2002 just after its predatory lending settlement, has increased the interest rates changed by its former Household units. Over eighty percent of HSBC's home purchase loans to African Americans and Latinos were higher-cost loans over the rate spread, much higher than in 2004 at these ex-Household units. In Buffalo, HSBC's long-time headquarters, HSBC in 2005 confined African Americans to higher cost rate spread loans 2.15 times more frequently than whites. 

            In 2005, HSBC made over five thousand super high-cost loans subject to the Home Ownership and Equity Protection Act (HOEPA) -- that is, at least eight percent over comparable Treasury securities.  Wells Fargo made 795 HOEPA loans in 2005. Keycorp, which has said it had discontinued HOEPA loans, made 755 such loans in 2005.

  National City Corporation's  First Franklin made 177,526 higher cost loans over the rate spread in 2005. Merrill Lynch has recently announced a proposal to acquire First Franklin, in order to be able to pool and sell its higher cost loans on Wall Street.

            Considering all conventional first-lien loans, among the most disparate was Washington Mutual and its higher-cost affiliate, Long Beach Mortgage -- together they confined African Americans to rate spread loans 3.70 times more frequently than whites.  Wells Fargo was nearly as disparate, confining African Americans to rate spread loans 3.31 times more frequently than whites. Royal Bank of Scotland and its Citizens Bank units came in at 3.11, and JP Morgan Chase at 2.98.  The disparity at Wachovia was 2.58, and at Atlanta-based SunTrust it was 2.40. The disparity at GMAC, a stake in which Citigroup and others are seeking to buy, was 2.92, while at Countrywide it was 2.86.

            Countrywide's disparity between pricing to African Americans and whites was even worse when considering conventional first lien home purchase loans: Countrywide confined African Americans to rate spread loans 3.53 times more frequently than whites. Countrywide was topped, however, by Milwaukee-based M&I, with a disparity of 3.78, and by Bank of America's MBNA unit, with a disparity of 4.23.

            Bank of America also enabled other subprime lenders in 2005 by securitizing loans through its generically-named Asset-Backed Funding Corporation unit for, among others, Ameriquest, which earlier this year settled predatory lending charges with state attorneys general for $325 million. The settlement only required reforms at Ameriquest Mortgage and two affiliates, but not its largest affiliate, Argent Mortgage. The 2005 data show that Argent made 220,069 higher cost loans over the rate spread, while Ameriquest Mortgage made 122,868 such loans. The reforms announced in support of the predatory lending settlement with the attorneys general cover barely 35% of ACC's high-cost lending. 

            Like ACC / Ameriquest, Citigroup and HSBC, other large subprime lenders also increased the percentage of their loans that were over the rate spread, from 2004 to 2005. At New Century in 2005, fully 215,579 of the company's 268,101 loans were over the rate spread.  Countrywide in 2005 made 190,621 loans over the rate spread. 199,249 of 237,700 loans were over the rate spread at H&R Block, which also in this season offers problematic high-cost tax refund anticipation loans. Further on fringe finance, the study notes that Citigroup helped Dollar Financial to go public, and since continued to lend to and assist this pawn and payday lender.

            The nation's largest bank, Citigroup, was disparate in Metropolitan Statistical Areas all over the country in 2005. In Los Angeles, Citigroup confined African Americans to higher cost rate spread loans 2.13 times more frequently than whites; its disparity for Latinos was 2.02. Citigroup's African American to white disparity was 2.27 in the Washington DC MSA, and 2.72 in Chicago.  In Philadelphia, Citigroup confined African Americans to higher cost rate spread loans 3.43 times more frequently than whites; its disparity for Latinos was 2.50.

            Another of the top four banks which enables predatory lenders is North Carolina-based Wachovia, whose pending application to merge with Golden West has been protested on mortgage discrimination and other grounds. Most recently, the U.S. District Court for the Southern District of New York denied a motion by the Federal Reserve Board to get reconsideration of a decision won by Inner City Press, requiring the disclosure of Wachovia's connections with a range of subprime lenders, including payday as well as mortgage lenders.  Inner City Press v. Federal Reserve Board, 380 F. Supp. 2d 211. On the Federal Reserve Board's motion, the Court ruled that:

"The Board made absolutely no showing in its summary judgment submissions, however, that the disclosure of data regarding Wachovia’s aggregate exposure and loan outstandings to the [subprime lending] clients listed in Exhibit 3 would cause competitive harm to Wachovia or that the public disclosure of this information would make it difficult for the Board to elicit similar information in the future... The Board points to portions of a document entitled 'Subprime Lending and Related Activities' that Wachovia submitted in the public portion of the Merger Application as a ‘glimpse into the conclusory statements [regarding due diligence practices] defendant can expect in future filings’ if merger applicants know such information is to be released to the public. This argument was not made in the Board’s original submission. In any event, without more specific testimony from Wachovia’s representative regarding why Wachovia would not wish its due diligence practices with regard to its subprime lending clients to be made public, it cannot be said that this document represents the limits of what Wachovia would willingly reveal at the Board’s request."

     On January 5 Inner City Press made a request under the Freedom of Information Act request to determine what the Federal Reserve had done on the previous year's data. More than six months later, the Federal Reserve issued a letter stating that is was withholding five linear feet of documents, and would only provide a garbled portion of a single document, described as "a description of the methodology used in generating the HMDA lenders list." The document states:

"The purpose of the Federal Reserve's matched-pair analysis is to compute lender-specific racial or gender disparities in denial rates, high rate pricing incidences and average APR spreads for loans above the threshold controlling for other factors including, market, income and loan amount. Each minority (or female) is matched to as many non-minority (or male) applicants (or borrowers) as meet the matching criteria. The outcomes of the minority (female) is compared with the average outcome of the non-minority (males) matched to it. The difference is the individual minority's (female's) 'matched pair disparity.' The disparities of all matches minorities (females) are averaged by product area or for sub areas such as MSAs...

  "Optionally, the matched pair procedures can be used to test for 'steering' within an organization such as a holding company. The outcome variable is the selection of a particular subsidiary of an organization (say a subprime lender) over another (say a prime lender) and the analysis tests whether this choice is related to the race of gender controlling for other factors including, market, income and loan amount. The user needs to specify how to classify lenders into the 'subprime' and 'prime' groups."

            While Inner City Press will have more once it receives the required mailed version of this document, we now we note Citigroup's recent announcement that it will merge its subprime CitiFinancial into its mostly-prime CitiMortgage, thereby evading this "optional" steering analysis.

            There is a need for more information, including the credit score information that the lending industry opposed being included in Home Mortgage Disclosure Act data. In fact, some lenders resist providing even the data required by law, at least in an analyzable form.

            Fair Finance Watch is demanding action on all of these issues from the relevant regulatory agencies, including the Office of Thrift Supervision (responsible for AIG and Lehman Brothers Bank, among others), the FDIC (still considering giving a bank charter to Wal-Mart), the Office of the Comptroller of the Currency (which since suing to New York last year to block fair lending enforcement has done little to none of its own) and also the Federal Reserve Board.

            Fair Finance Watch responded, "Now that a second year of data is out, with worsening disparities at the largest bank in the nation and many of its peers, there is no more time for the Federal Reserve and other regulatory agencies to equivocate. The time for enforcement actions to combat this discriminatory and predatory lending is now."

Other samples of related or relevant articles:

Stop Bank Branch Closings and Monopolies in the Katrina Zone, Group Says, Challenging Regions- AmSouth Merger

 Birmingham, Alabama, August 20 -- A year after Hurricane Katrina ravaged Louisiana, Mississippi and Alabama, two of the largest banks in the Katrina Zone have applied to merge and save $400 million, in part by closing branches. With the Federal Reserve's comment period on the application by Regions Financial Corporation to acquire AmSouth running through September 14, and the two banks' shareholders' votes set for October 3, consumers and human rights group Fair Finance Watch has filed a fifteen page protest to the deal, requesting public hearings including on what it calls the Katrina Zone issues.

            The challenge represents the first analysis of the 2005 data of Regions Financial's Home Mortgage Disclosure Act data-reporting affiliates, including the subprime specialist Equifirst, cumulating these lenders as Regions and calculating the distribution of loans over the Federally-defined rate spread of 3% over comparable Treasury securities on first lien loans, 5% on subordinate liens (calling these high cost loans).

            The Fair Finance Watch analysis shows that in its home state of Alabama in 2005, Regions confined 51.66% of its African American borrowers to higher cost loans over the rate spread, versus only 23.15% of its white borrowers. That is, Regions confined African Americans to high cost loans 2.23 times more frequently than whites, while denying 30.69% African Americans' applications for loans, versus only 21.29% of whites' applications.

Regions NY self-cheering

            In Louisiana in 2005, Regions confined 54.92% of its African American borrowers to higher cost loans over the rate spread, versus only 27.88% of its white borrowers. Regions confined African Americans to high cost loans 1.97 times more frequently than whites, while denying 30.71% African Americans' applications for loans, versus only 22.27% of whites' applications.

            In neighboring Mississippi, Regions in 2005 confined 38% of its African American borrowers to higher cost loans over the rate spread, versus only 18.38% of its white borrowers. Regions confined African Americans to high cost loans 2.07 times more frequently than whites, while denying 35.87% African Americans' applications for loans, versus only 24.68% of whites' applications.

            Throughout Mississippi and their other footprint states, the banks have been asking community groups and charities to write letters of support, including references to a Community Reinvestment Act pledge the two banks announced.  The Fair Finance Watch comments argue that given the high percentage of Regions' mortgages which are high-cost, the pledge may represent a promise of predatory lending.

            While Fair Finance Watch has focused the regulators on these three Katrina Zone states, nationwide in 2005 Regions confined fully 73.55% of its African American borrowers to higher cost loans over the rate spread, versus only 51.78% of its white borrowers. In Florida in 2005, Regions confined 66.97% of its African American borrowers to higher cost loans over the rate spread, compared to 45.98% of its white borrowers. And in North Carolina, the headquarters of Regions' subprime unit Equifirst, Regions in 2005 confined a whopping 88.76% of its African American borrowers to higher cost loans over the rate spread, versus 71.66% of its white borrowers.

            Regions and AmSouth have continued supporting other subprime lenders.  Uniform Commercial Code filings filed by Fair Finance Watch show for example that Regions on July 18, 2005, made a loan secured by all "accounts and proceeds" to Eagle Title Loans, Inc. of Athens, Alabama. Also in Alabama, Regions lends to Twin States Pawn of Butler and Boaz' Sand Mountain Pawn. In Louisiana, Regions lends to LA Pawn Shop of West Monroe. In Arkansas, Regions lends to A-1 Pawn of Russellville. In the Sunshine State, Regions lends to Deerfield Pawn Brokers of Deerfield, Florida.

            The issue of banks funding such fringe financiers is one that's in evolution. In response to similar comments from Fair Finance Watch, the Atlanta-based bank SunTrust committed to stop lending to auto pawn and payday lenders.

            AmSouth, which Fair Finance Watch says refused to provide its mortgage data in computer analyzable form, lent to Rent to Own Pasco of Pasco, FL, and Pasco Jewelry and Pawn in the same city. The Fair Finance Watch comment conclude that "while the merger should be denied on all of the above grounds, any merger of this size in the still-unrepaired and underbanked zone impacted by last year's hurricanes militates for a required Katrina Zone CRA Lending Plan, and for public hearings." 

            How this call for hearings will fare, in the face of the letters of support solicited by the banks, remains to be seen. But the need to focus on economic justice in the areas hit by Hurricane Katrina is hard to dismiss if one looks at the region, so to speak, in this one-year anniversary of disregard and destruction.

JPMChase to Close Four Branches in Low and Moderate Income Tracts

BRONX, NY, May 18 -- JPMorgan Chase has today for the first time specified that it has identified in low- and moderate-income census tracts four of the Bank of New York branches it seeks to acquire "which are located close to a JPMCB branch." This is essentially code language that these four low-income branches would be closed if the acquisition is approved. JPM Chase's statement, in a May 18 letter responding to Fair Finance Watch's April 17 and May 6 comments to the Office of the Comptroller of the Currency, declines to provide the addresses of these four branches and the 46 other branches, some surely adjacent to low-income tracts, which the letter projects would be closed. The figure "four LMI branches" is qualified by the statement "in New York City." Since many of Bank of New York's branches are outside of the five boroughs, might even more than four low- and moderate-income census tract branches be closed?  Developing...

Other Inner City Press reports are archived on www.InnerCityPress.org -