The following post comes to us from Brent J. Horton, assistant professor at Fordham University Gabelli School of Business.

In my recent Article, Toward a More Perfect Substitute: How Pressure on the Issuers of Private-Label Mortgage-Backed Securities Can Improve the Accuracy of Ratings,[1] which is scheduled to be published in Volume 93 of the Boston University Law Review this winter,[2] I propose a burden shifting procedure that will force issuers of private-label MBS to take ownership of the ratings incorporated into their registration statement (e.g., Aaa, Baa3)—specifically, the accuracy of those ratings. The issuers will take ownership because they (the issuers) will be forced to prove that the ratings are accurate, or face liability pursuant to Section 11 of the Securities Act of 1933.

My burden shifting proposal works as follows: where a harmed investor brings a cause of action pursuant to Section 11 against the issuer, alleging that a MBS registration statement incorporated an inaccurate rating, the burden will shift to the issuer to establish (1) that the loss model used by the rating agency was state-of-the-art, and (2) all inputs used by the rating agency were correct and up-to-date. If the issuer cannot meet this burden, the rating will be deemed a material misstatement, satisfying the most difficult element of a Section 11 claim.

Properly incentivized, the issuer is in the best position to place reputational pressure on the rating agency to use a state-of-the-art rating model. Even more importantly, the issuer—which manages the flow of information from originator to rating agency—is in the best position to remedy inaccuracies as to loan-level data (those inaccuracies can play havoc on even the best rating models). In short, forcing the issuer to take ownership of the rating is the best way to make ratings a more perfect substitute. As the title of my Article suggests, the goal is to make ratings more perfect; as heuristics,[3] they can never be absolutely perfect.

Whether the rating agencies were affirmatively manipulating ratings to keep their client (the MBS issuer) happy, or the explanation is more innocent (one executive claimed they “drank the Kool-Aid” and truly believed that the structure of the MBS shielded the investor from risk),[4] it is evident that the ratings process was deeply flawed.

To rate private-label MBS, rating agencies calculate the likely cash flow from the mortgages underlying a MBS issuance, and then apply that likely cash flow to the multiple tranches. Each MBS tranche is given a rating. Because the junior tranches receive payment only after the senior tranche, a reduction in cash flow (because of mortgage defaults) is most likely to impact the junior tranches. After the senior tranches drink up the cash flow, it is possible that there will be nothing left for the junior-most tranche. This process is referred to as a “waterfall”.[5] Generally speaking, depending on the likely cash flow, the most senior tranche received an Aaa rating; the lowest publicly traded tranche may receive a Baa3 rating. Both are considered investment grade, but a Baa3 rating is considered more speculative.

However, if the rating agency overestimates the likely cash flow—i.e., underestimates defaults—then the lowest tranche may not deserve an investment grade rating. Accurately calculating the number of defaults that will occur is therefore essential to the rating process; it requires that the rating agency consider data about the underlying mortgages provided by the issuer (observed loan-level data), as well as data about national trends in housing prices, defaults, and how those defaults correlate (assumed macro-economic trends). However, it turns out that the observed loan-level data was distorted by information asymmetries between borrower and mortgage originator, and again between mortgage originator and issuer, and again between issuer and investor.[6] Further, the assumed macro-economic trends were wildly optimistic, for example, the rating agencies assumed a 4% increase in home prices every year for the foreseeable future, when past experience tells us that it is not uncommon to have a 30% reduction in the event of a recession.[7]

My proposal avoids the problems that arise from placing liability on the rating agency: the rating agency may find the threat of liability too great, and decide to withdraw from the offering process altogether.[8] The issuer is less likely to withdraw from offering MBS. Further, my proposal mirrors the legislative scheme of Section 11 of the Securities Act, holding the issuer primarily (and strictly) liable for material misstatements in the registration statement.

[1] Available at SSRN: http://ssrn.com/abstract=2318136. The title of this Article was inspired by Professor Steven L. Scwartz, who wrote that prior to the 2008 economic crisis, investors relied on “such imperfect substitutes as rating-agency ratings and the results of mathematical models.” Steven L. Schwarcz, The Future of Securitization, 41 Conn. L. Rev. 1313, 1323 n.51. (2009) (emphasis added).
[2] Boston University Law Review publications are available at http://www.bu.edu/bulawreview/.
[3] Steven L. Schwarcz, Regulating Complexity in Financial Markets, 87 Wash. U. L. Rev. 211, 222 (2009) (“investors often resort to simplifying heuristics, such as credit ratings, as substitutes for attempting to fully understand the investment being analyzed.”).
[4] Permanent Subcomm. On Investigations, U.S. Senate, Wall Street and the Financial Crisis: Anatomy of a Financial Collapse 278 (2011) (quoting Moody’s Chief Credit Officer: “we drank the Kool-Aid”).
[5] See Kathryn Judge, Fragmentation Nodes: A study in Financial Innovation, Complexity and Systemic Risk, 64 Stan. L. Rev. 657, 681 (2012) (explaining the financial structure of MBS, and resulting waterfall, in detail).
[6] See Adam J. Levitin & Susan M. Wachter, Explaining the Housing Bubble, 100 Geo. L.J. 1177, 1230 (2012) (“The unregulated, private securitization market is rife with information asymmetries between financial institutions and investors.”).
[7] An excellent source for comparing housing price trends over the past century is Robert J. Shiller, Data for Figure 2.1 in Irrational Exuberance, available at http://www.econ.yale.edu/~shiller/data.htm.
[8] See Jason W. Parsont, NRSRO Nullification: Why Rating Reform May Be In Peril, 77 Brooklyn L. Rev. 1015, 1019-20 (2012) (discussing the possibility of rating agencies refusing to provide ratings in the face of direct liability).