April 18, 2005
Inner City Press and Fair Finance
Watch have now reviewed the 2004 Home Mortgage Disclosure Act data of three more of the
largest subprime mortgage lenders in the nation – Washington Mutual, AIG / American
General, and Ameriquest – and a similar number of regional banks, including KeyCorp,
SunTrust, Royal Bank of Canada / Centura and ABN Amro, comparing them to corrected data
that Citigroup has released, including the new information concerning which loans are
subject to a rate spread (3% higher than comparable Treasuries on a first lien, and 5% on
a subordinated lien), and have found the following:
Royal Bank of Canada, which in the U.S. owns Centura bank and an Illinois-based
mortgage company, imposes higher-cost rate spread loans nearly four times more frequently
on African Americans than on whites. ABN Amro,
which owns a mortgage company as well as Standard Federal and LaSalle Bank, imposes
higher-cost rate spread loans 4.19 times more frequently on African Americans than on
whites, while denying African Americans’ applications 3.54 times more frequently than
those of whites, and denying Hispanics’ applications 1.84 times more frequently than
those of whites.
Among the banks, KeyCorp in 2004 made 972 so-called HOEPA loans, at costs much
higher than the new rate spread threshold. In fact, 456 of these loans were at rates more
than 10% over the Treasury bill rate baseline – that is, home-secured loans at
interest rates of 12% and up, in a low interest rate environment. ICP is studying the demographics of these HOEPA loans.
The large non-bank subprime lender Ameriquest made more than 45,000 loans at rates
5% or more over Treasuries, and over 270,000 rate spread loans overall. Washington Mutual made 71 HOEPA loans,
and imposed higher-cost rate spread loans 3.26 times more frequently on African Americans
than on whites. AIG FSB, the savings bank owned by the insurance company American International Group, imposed
higher-cost rate spread loans 2.27 times more frequently on African Americans than on
whites in its home state of Delaware in 2004.
Atlanta-based SunTrust, when cumulated with the Memphis-based bank it acquired in
2004, imposed higher-cost rate spread loans 1.92 times more frequently on African
Americans than on whites, while denying African Americans’ applications 2.55 times
more frequently than those of whites, and denying Hispanics’ applications 1.55 times
more frequently than those of whites. There are other issues are SunTrust. In response to
ICP’s comments on its Memphis acquisition, showing that SunTrust was funding dozens
of payday lenders and car title lenders, SunTrust sent a letter to the Federal Reserve,
copied to ICP, stating that "[a]fter consider the potential reputational risks and
consumer harm that could result from lending to such a company, STI is revising its credit
policies to prohibit future loans to all businesses that engage in payday or title
lending." See, e.g., the July 28, 2004
Memphis Commercial Appeal, “NCF, SunTrust Ditch Payday Lenders - Answer
Activists’ Challenge Ahead of Bank Merger,” and Orlando Sentinel, “Bank
Shuns Payday - SunTrust Halts Loans to Fast-Cash Industry.”
In monitoring SunTrust’s compliance with this commitment, ICP has come upon
evidence of a January 2005 loan from SunTrust secured by “all proceeds” of Cash
Advance, Inc. of Jacksonville, Florida. ICP raised this to SunTrust last week, in
connection with obtaining the 2004 mortgage data, and SunTrust refused to address the
seeming violation of the commitment it made, citing its “confidentiality
obligations” but stating that this was a “banking relationship that pre-dated
our representations to the Fed last July.” But the loan to Cash Advance was filed as
an “initial” Uniform Commercial Code lien on January 25, 2005, more than six
months after SunTrust’s commitment to cease such lending. Another defense being
offered is that while the loan is secured by all proceeds of Cash Advance, Inc., it is
somehow not a business loan. ICP has now
raised this SunTrust issue to
federal and state regulators (in Georgia and Tennessee) for their action; ICP is committed
to independent verification and monitoring of commitments.
In further monitoring, ICP has raised to the federal Office of the Comptroller of
the Currency the fact that the mortgage lending data filed by HSBC for its ex-Household
units HFC, Beneficial and Decision One, all point to the OCC as the regulator of these
companies. Each has been state-regulated; HFC
and Beneficial are subject to a $486 million predatory lending settlement with attorneys
general and regulators in 46 states. When HSBC
applied to convert its New York State-charter bank to a national charter with the OCC in
mid-2004, ICP submitted timely comment opposing any shift of HFC and Beneficial from
regulation by the states, at which level HFC and Beneficial are still subject to the
predatory lending settlement. The OCC’s
June 23, 2004 ruling, still on the agency’s web site as Community Reinvestment Act
Decision #122, at www.occ.treas.gov/interp/jul04/crad122.doc,
noted ICP’s concern that
"HSBC’s intermediate parent company, will try to move its
subprime operations from Household International, Inc. (HII), to HUNA in order to preempt
the application of state consumer protection laws.Many of the concerns raised by the
commenter related to HII and its non-bank subsidiaries.
However, as noted above, the OCC currently has no regulatory authority over these
entities, nor would the OCC have any regulatory authority over those entities as a result
of the proposed transaction. Accordingly, the
OCC was unable to address those concerns. However, the applicant has represented that
HII’s branch-based consumer lending business, conducted through Household Finance
Company (HFC) and Beneficial Corporation (“Beneficial”), will continue to be
operated as a state-regulated business."
See also, Buffalo News of June 13, 2004, “HSBC Hit on Downstate
Lending Patterns,” reporting that ICP “says the move could let Household avoid
state scrutiny if it became a subsidiary of the new national bank. A national
investigation by multiple state attorneys general led to a settlement in September 2002
with all 50 states. Household agreed to pay $484 million in refunds to customers and to
make dramatic changes in its practices. HSBC officials insist that the bank and Household
are separate and there are no plans to reorganize Household under HSBC Bank USA. They say
the lending offices and practices of subsidiaries Household Finance and Beneficial Finance
will remain under state purview.” But
that is not what is reflected in the 2004 HMDA data filed by HSBC – there, the
ex-Household units are portrayed as regulated by the OCC.
ICP notes, however, that neither company is named or disclosed in the OCC’s
online listing of national bank operating subsidiaries, at <www.occ.treas.gov/OpSublist.pdf>.
This issue of stealth preemption on state consumer protection laws, implicating both HSBC and the OCC, is being raised to
Congress and state officials.
Some lenders continue to throw up obstructions to access to their mortgage and rate
spread rate. Lehman Brothers, which like AIG owns a savings bank in Delaware as well as
two large subprime lenders, is attempting to require ICP to sign a confidentiality /
privacy agreement (as has another large subprime lender, which now states it is
reconsidering its position). ICP has written to Lehman Brothers, and to the Office of
Thrift Supervision and other regulators, reminding them that under the Home Mortgage
Disclosure Act this data must be released, without conditions. There are other
obfuscations: Fifth Third Bank has provided its data in PDF format, which can be viewed
(as printed pages) but not analyzed; Countrywide is claimed to not be able to provide the
data in the format it was submitted to the regulators, despite ICP working for a week
providing Countrywide with information about free software and formatting options.
In a new low, Citigroup on April 13 informed ICP that the data Citigroup had given
it on March 31 was incomplete and incorrect. Based on that data, provided by Citigroup the
full month after ICP’s request, ICP conducted an analysis and found for example that
for home purchase loans at Citigroup in 2004, African Americans were 4.34 times more
likely to receive higher-cost rate spread loans than whites. Citigroup’s spokesman,
asked to respond by the Associated Press and the American Banker newspaper, called
ICP’s findings, and its director, “reckless,” and claimed that the data
showed otherwise. See, e.g.,“U.S. community group alleges Citigroup, Bank of America
discriminate in mortgage lending,” by Eileen Alt Powell, Associated Press, April 4,
2005; “First HMDA Fallout - Activists Hit Citi, B of A,” by Hannah Bergman,
American Banker, April 5, 2005, Pg. 1; and "Groups Make Hay of HMDA Data,"
National Mortgage News, April 11, 2005, Pg. 2.
On April 14, ICP received from Citigroup new compact disks and repeated its
analysis. The number of originated loans and
mortgage records have remained the same – 351, 811 loans and 1,218,402 records. But the number of the loans that are higher-cost
rate spread loans has increased from 11,000 in
the first, incorrect CD, to fully 93,103 rate spread loans in the second set of data. That
is to say, the data Citigroup provided on March 31 underreported its 2004 higher-cost
loans by 82,103 rate spread loans. Based on the new data, fully 26.46 percent of
Citigroup’s originated loans in 2004 were higher-cost rate spread loans.
(This is still lower than at HSBC, where 32.7% of 2004 loans were higher-cost rate
spread loans – but it is much than at Wells Fargo, where 9.13% of 2004 loans were
higher-cost rate spread loans. For
home purchase loans, Wells Fargo
denied the applications of African Americans 2.28 times more frequently than those of
whites, and those of Latinos 2.02 times more frequently than whites. At Citigroup, the disparity for African Americans
is higher (a denial rate for African Americans 2.54 times higher than for whites), while
for Latinos it is slightly lower (a denial rate for Latinos 1.93 times more frequently
than whites). These comparisons are for the holding companies as a whole,
cumulating all of their HMDA-reporting affiliates).
Based on the new data, for home purchase loans at Citigroup in 2004, African
Americans were 3.88 times more likely to receive higher-cost rate spread loans than
whites. While this is slightly lower than the
disparity, 4.34 to one, in ICP’s first study based on the data Citigroup provided, it
is still much higher than for example the lenders reviewed above. Strangely, the Wall
Street Journal’s April 11 report, based on Citigroup’s self-generated
percentages, had Citigroup appearing less disparate than nearly all other lenders. (HSBC
was not included in the Wall Street Journal’s report, despite making more rate spread
loans in 2004 than either Citigroup or Wells Fargo).
While ICP’s analysis of Citigroup’s second, ostensibly correct batch of
data is continuing, ICP stands by its finding, that the disparities by race in high-cost
lending at Citigroup are worse than at its peers.
Citigroup had more than a month to
prepare, but released data that undercounted its high cost loans by a power of seven. The
new data makes Citigroup look even worse and more disparate, and makes it all the more
important that the Federal Reserve stick to and firm up its March 2004 ruling that
Citigroup should not significantly expand until it fixes its compliance woes.
Citigroup’s problems include systemic racial disparities and predatory lending.
Methodology and Scope of Review
As stated above, both for its April 4 report and for this report, ICP Fair Finance
Watch reviewed Citigroup’s 1,218,401 loan mortgage loan applications records for
2004. For the other banks in this report, ICP reviewed Washington Mutual’s 1.3
million mortgage records, including 71 HOEPA loans. KeyCorp reported merely 184,955
mortgage records in 2004, including 972 HOEPA loans. RBC Centura and RBC Mortgage together
reported 92,766 mortgage records. SunTrust Bank, together with its mortgage company and
NCF, reported 293,302 records (only 27,208 of these were from NBC). ABN Amro, cumulating
its mortgage company, Standard Federal and LaSalle Bank, reported 411,341 mortgage records
in 2004.
For further information, click here to contact us
|
