March 27, 2014
Thank you for providing the public with the opportunity to comment on your further proposals regarding ABS loan level data disclosures. Until 2011, I managed a group at Moodys that gathered and analyzed disclosures for Mortgage Backed Security and other ABS deals. In 2012, I co-authored a study entitled Restoring Trust in Mortgage Backed Securities for the Reason Foundation (at http://reason.org/files/study_restoring_trust_in_mbs_final.pdf).
At present, I have no relationship to the ABS market. However, as a policy analyst and concerned citizen, I wanted to share my perspective on this topic. These views are my own disinterested opinions, and do not reflect the perspectives of any former employer or client.
(1) In his statement announcing the reopening of the comment period (at http://www.sec.gov/News/PublicStmt/Detail/PublicStmt/1370540851698#.UzTjS_ldWSo), Commissioner Piwowar notes that Registered offerings by private label residential mortgage-backed securities (MBS) remain almost non-existent, with only a single sponsor making such an offering in 2013.
He later asks commenters to focus on Auto ABS and other ABS asset classes.
Low residential MBS issuance is primarily the result of GSE dominance of new mortgage originations. If Congress and the Administration agree on a plan to scale back Fannie Mae and Freddie Mac, RMBS issuance is likely to rise substantially. In this event, the RMBS market could once again become a source of systemic risk. By encouraging transparency and facilitating deal analysis, the SEC can minimize the chances of a future financial crisis. Loan by loan disclosure is much more important for RMBS than for Auto ABS or Credit Card securitizations, so I suggest the conversation should continue to focus on mortgage-backed deals.
(2) The SECs new memorandum on asset level disclosures (at http://www.sec.gov/comments/s7-08-10/s70810-258.pdf) recommends that loan level data reside on issuer provided web sites rather than on EDGAR. Under the proposal, these issuer provided web sites would be open to investors and potential investors. Access should be extended to rating agencies, analytics firms, data providers and academics. These groups provide data analysis, commentary and deal modeling services that may be useful to investors. So long as members of these groups agree to terms of use restricting public dissemination of raw borrower-identifying data, they should have equal access to the information.
(3) The Commission should seek synergies between this new class of asset level disclosure web site and the rating agency focused sites required under regulation 17g-5. This provision requires hired rating agencies and issuers to provide access to deal data used in the rating process to non-hired rating agencies. Since this data include loan level information, there is likely to be substantial overlap between the 17g-5 sites and those required under the new proposal. Issuers should be able to create a single web site containing all information required by investors, potential investors and non-hired rating agencies as well as the other parties mentioned earlier (academics, data and analytics providers). Issuers should also be able to delegate this responsibility to information providers. Ideally, there would be a single web site (or a small number of sites) that contain asset-level and 17g-5 required information. This would allow users to easily compare deals.
(4) As I discussed in a 2012 Bloomberg commentary (http://www.bloomberg.com/news/2012-06-03/to-help-fix-housing-disclose-mortgage-addresses.html), I favor disclosure of loan level address data. The fact that an individuals property is encumbered by a mortgage is a matter of public record. Some analytics firms combine loan level RMBS data with this public records information. Since this data merging process is an expensive service, the results are not equally available to all investors. The private label RMBS market will be more liquid and less prone to crisis, if investors and those that serve them have access to all material information affecting the bonds they need to evaluate.