Вы находитесь на странице: 1из 3

My Response to Statement of David Stevens Regarding MBAs HMDA Analysis

Over the past several days I have been engaged in a Twitter discussion regarding claims, made by
Mortgage Bankers Associations President David Stevens, about Fannie Maes and Freddie Macs denial
rates of black mortgage applicants.
In a speech before the National Secondary Market Conference held by the Mortgage Bankers
Association, the President of that trade association, David Stevens, argued for mortgage market reforms
that would replace the GSEs, Fannie Mae and Freddie Mac, with a system that would likely benefit his
members. In support of his argument he stated: The 2012 HMDA data shows a 56% denial rate on GSE
purchase loan applications for African Americans.
After reading Mr. Stevens statement I was stuck by that figure. It appeared quite high and greatly
differed, by orders of magnitude, from my own analysis, the raw numbers put forward by HMDA,
analysis performed by the Federal Reserve
1
or numbers provided by Zillow
2
. As a result, I sent a tweet to
Mr. Stevens asking him to explain where his numbers were from and to publicly disclose the MBAs
methodology.
Rather than answering or providing a link, Mr. Stevens responded with go to our website under
research. When I explained that I could not find their methodology on their website, Mr. Stevens
responded: Its direct from HMDA Josh - really?? This is not even debated by any analyst - its in HMDA -
go online and pull the tables. I was not asking him to direct me to the HMDA data, I was -- as I would do
almost a dozen more times -- asking him to show how he came to that number. Once again Mr. Stevens
suggested that the data was clean, clear and publicly available and went further to insinuate that I must
be an idiot because you r the only one challenging public data.
I explained You guys must have added manuf housing loan denials, GSEs don't do chattel loans. Your
numbers are sketchy, to which Mr. Stevens stated that I was chasing dragons. Again I suggested that
I think you added manuf housing (chattel), included incomplete files, not closed...manipulating #'s
doesn't make them real. Mr. Stevens then stated dude you don't understand hmda - twit to death -
you are simply wrong and even the GSE's don't argue this.
Here we are, days later, and the Mortgage Bankers Association continues to refuse to provide a full
explanation, with actual numbers and assumptions, detailing how they came to a number that is so
wildly different than anyone elses number. Moreover, as we have now heard, Fannie Mae actually does
dispute his numbers and sees serious flaws
3
in the approach employed by the MBA. Lastly, it is
important to remember that neither Fannie Mae nor Freddie Mac are primary lenders and that loans are
originated by mortgage bankers. Sometimes they are held in the portfolios of those lenders, sometimes
they are securitized and sometimes they are sold to the GSEs. The choice is the lenders and not the
GSEs and these lenders often use their own credit overlays to determine how they will handle the

1
http://www.federalreserve.gov/pubs/bulletin/2013/pdf/2012_HMDA.pdf
2
http://t.co/uwWrXPdI4Z
3
http://www.businessweek.com/news/2014-05-22/fannie-sees-flaws-in-groups-black-mortgage-denial-claims
disposition of those loans. As we saw during the crisis, banks frequently steered low-income and
minority borrowers away from conforming and conventional GSE mortgages, for which they could
qualify, toward higher cost mortgages. Even now, it appears many of these firms keep the best loans for
themselves and direct only the less qualified borrowers to the GSEs. In other words, they appear to be
adversely selecting the weakest loans for the government-sponsored enterprises
4
. This raises further
questions about the intelligence of designing a new mortgage-finance system that places the big banks
at its center.
Last evening, David Stevens released a statement on MBAs HMDA analysis
5
and, without providing any
numeric detailing of the methodology or assumptions used, he continued to make assertions that do not
appear plausible or supportable.
In this statement he asserts:
MBA includes manufactured housing (MH) data in its calculation because MBA members lend on this
collateral, and both GSEs have MH loan programs and tout these programs as means to provide
affordable homeownership to lower and moderate income homebuyers. MBA effectively eliminates
chattel, a type of MH property that the GSEs do not fund, by ensuring that only loans which are deemed
conventional in the loan type field are counted.
Unless he presents the methodology, so that we can see that he is correct and I am incorrect, it appears
that Mr. Stevens claim is flatly untrue. My basis for stating so can be found in a February 2014
Factsheet released by the CFPB
6
titled CFPB Takes Steps to Improve Information About Access to
Credit in the Mortgage Market. That release appears to make it clear that there is no way, currently, to
parse the manufactured housing data in the way the Mr. Stevens asserts the MBA has. The CFPB wrote:
Lenders are currently required to report whether a loan will be for a manufactured home. The market
for credit to finance manufactured home purchases is different from the market for credit to finance site-
built home purchases. Additional information on manufactured home loans, including the type of
financing and whether the borrower will own or lease the land where the home is sited, will make it
easier to identify the sources of differences in denial rates, and will improve understanding of
manufactured home financing.
As a result, and without any more transparency about the MBAs approach, it seems fair to conclude
that they merely included all of the manufactured housing applications in their numbers. Given
limitations that prevent the GSEs from non-real estate lending, inclusion of chattel loans is questionable.
In his statement, Mr. Stevens further states:

4
http://www.americanbanker.com/issues/179_99/big-banks-keeping-choicest-mortgages-rather-than-sell-to-
gses-1067663-1.html
5
http://mortgagebankers.org/NewsandMedia/PressCenter/88296.htm
6
http://files.consumerfinance.gov/f/201402_cfpb_factsheet_sbrefa.pdf
MBA includes preapproval denials in our calculation. Preapprovals represent a credit check of a
borrower who has not yet identified a property. Being denied at this stage is an important signal of
availability of credit. A denial is a denial, regardless of where in the loan process it occurs.
This methodology is also seriously flawed for several reasons. First, if a borrower is working with a
mortgage broker, that broker may submit a pre-approval request to several banks. If one bank approves
the loan but the others do not, then the others are each required to report a denial. As a result, using
the methodology the MBA seems to use, one borrower may appear in the data as multiple denials.
Clearly, the inclusion of pre-approval denials, for borrowers who have not even identified a specific
property they would like to purchase or the actual price of that property, is questionable and the
multiple denials for a single borrower should not be included in the MBAs calculation. Moreover, if the
borrower does not provide all necessary information in a pre-approval request they would still be listed
as a denial by the MBAs methodology. This is quite different from the approach included in the actual
application process in which the lender has to flag specific reasons for the denial. These could include
the borrowers credit history, income data, an incomplete application, a withdrawn application or a
closed file.
I am not seeking to prolong this dispute but these issues are important. There is no question that
minority borrowers have less access to credit than whites. This is result of serious historic and current
societal inequalities. We should, each in our own fields, seek the means to address and reduce those
inequalities in all ways possible. The only way we can begin to do so is to get an accurate understanding
of actual instances of inequality and whether those were caused by institutional bias, wealth
differentials, or other factors. In the mortgage finance world that would suggest we ensure that
underwriting approaches are applied consistently, regardless of the borrowers race or gender, so that
we might have a better understanding of whether outcomes are tied to racial bias. Armed with this
information, policy-makers would be in a position to better determine the policy approaches that should
be employed to support minority borrowers.
To that end, I hope that Mr. Stevens and the Mortgage Bankers Association either explain their
methodology in detail and with facts and figures, or retract their assertions.

Вам также может понравиться