Page 497 - untitled

468 Dissenting Statement

the Lehman analysts—into believing that Fannie and Freddie did not hold or had
not guaranteed substantial numbers of high risk loans, and thus that there were
many fewer such loans in the fi nancial system than in fact existed.
Of course, in the early 2000s there was no generally understood defi nition
of the term “subprime,” so Fannie and Freddie could defi ne it as they liked, and the
assumption that the GSEs only made prime loans continued to be supported by their
public disclosures. So when Fannie and Freddie reported their loan acquisitions to
various mortgage information aggregators they did not report those mortgages as
subprime or Alt-A, and the aggregators continued to follow industry practice by
placing virtually all the GSEs’ loans in the “prime” category. Without understanding
Fannie and Freddie’s peculiar and self-serving loan classifi cation methods, the
recipients of information about the GSEs’ mortgage positions simply seemed to
assume that all these mortgages were prime loans, as they had always been in the
past, and added them to the number of prime loans outstanding. Accordingly, by
2008 there were approximately 12 million more NTMs in the fi nancial system—and
12 million fewer prime loans—than most market participants realized.
Appendix 1 shows that the levels of delinquency and default would be 86
percent higher than expected if there were 12 million NTMs in the fi nancial system
instead of 12 million prime loans. Appendix 2 shows that the levels of delinquency
would be 150 percent higher than expected if the feedback eff ect of mortgage
delinquencies—causing lower housing prices, in a downward spiral—were taken
into account. Th ese diff erences in projected losses could have misled the rating
agencies into believing that, even if the bubble were to defl ate, the losses on mortgage
failures would not be so substantial as to have a more than local eff ect and would not
adversely aff ect the AAA tranches in MBS securitizations.
Th e Commission never looked into this issue, or attempted to determine
what market participants believed to be the number of subprime and other NTMs
outstanding in the system immediately before the fi nancial crisis. Whenever
possible in the Commission’s public hearings, I asked analysts and other market
participants how many NTMs they believed were outstanding before the fi nancial
crisis occurred. It was clear from the responses that none of the witnesses had ever
considered that question, and it appeared that none suspected that the number was
large enough to substantially aff ect losses aft er the collapse of the bubble.
It was only on November 10, 2008, aft er Fannie had been taken over by the
federal government, that the company admitted in its 10-Q report for the third
quarter of 2008 that it had classifi ed as subprime or Alt-A loans only those loans
that it purchased from self-denominated subprime or Alt-A originators, and not
loans that were subprime or Alt-A because of their risk characteristics. Even then
Fannie wasn’t fully candid. Aft er describing its classifi cation criteria, Fannie stated,
“[H]owever, we have other loans with some features that are similar to Alt-A and
subprime loans that we have not classifi ed as Alt-A or subprime because they do not
43
meet our classifi cation criteria.” Th is hardly described the true nature of Fannie’s
obligations.
On the issue of the number of NTMs outstanding before the crisis the
Commission studiously averted its eyes, and the Commission majority’s report
43 Fannie Mae, 2008 3rd quarter 10-Q. p.115, http://www.fanniemae.com/ir/pdf/earnings/2008/q32008.
pdf.