Mortgage Credit Ratings and the Financial Crisis: The Need for a State-Run Mortgage Security Credit Rating Agency MORTGAGE CREDIT RATINGS AND THE FINANCIAL CRISIS: THE NEED FOR A STATE-RUN MORTGAGE SECURITY CREDIT RATING AGENCY Milosz Gudzowski* I. Introduction .................................. 246 II. Background ................................. 249 A. Scope of the Agency's Functions ................ 249 B. History of Mortgage Securitization ..... ..... 251 III. CRA Incentives and Industrial Structure ... ..... 253 A. Dominance of the Issuer Pays Model.................. 254 B. Issuer Pays Model and Reputational Capital .... 255 1. Reputation and Novel Products ..... ..... 256 2. Close Contact and Influence ................. 256 3. Reputation and the Bubble Economy............ 258 4. Reputation and Regulatory License .............. 259 C. Difficulty of Entry. ................. ..... 259 1. Regulatory Barrier .............. ..... 260 2. Economic Barriers .............. ..... 261 D. Conclusion ............................... 262 IV. The 2007-2009 Credit Crisis: Evidence of CRA Failure ................................. 263 V. Proposal .......................... ...... 264 A. Incentive Advantages .............. ..... 265 B. Informational Advantages ................ 266 C. Economic Regulation and NRSROs ... ...... 268 D. Avoiding Externalities........... ......... 271 VI. Public-Private Debate ................. ..... 271 A. Conventional Criticisms of Government Provision of Goods and Services ...... ....... 272 * J.D. Candidate 2010, Columbia University School of Law; M.Phil. Development Studies 2007, Cambridge University; B.A. Economics 2006, Yale University. The author thanks Professor Merritt Fox, Gerald Moody, and all staff and board members of the Columbia Business Law Review for their insightful feedback and editorial assistance with this note. COLUMBIA BUSINESS LAW REVIEW 1. The Quality and Efficiency of the Rating Process ............................ 272 2. Accountability in the Rating Process ............ 274 3. A Costly and Difficult Start .... ......... 275 B. Precedent for Government Involvement ............ 275 C. Conclusion ............................... 277 VII. The Liability Alternative ............... ..... 277 A. Proposals ................................ 277 B. Failure of Liability......... .............. 279 VIII. Conclusion .......................... .... 280 I. INTRODUCTION With the exception of the investment banks, there probably has been no other financial actor with as great a hand in causing the current financial crisis as the major credit rating agencies (CRAs).' CRAs gave residential mortgage backed securities (RMBSs) and RMBS- collateralized debt obligations (R-CDOs) inflated credit ratings,2 enabling financial institutions and other investors to buy trillions of dollars worth of these securities.' Eventually, it became apparent that these securities were ' The following statement by Alan Greenspan, former head of the Federal Reserve Bank, is illustrative of this sentiment on Wall Street: "The consequent surge in global demand for U.S. subprime securities by banks, hedge and pension funds supported by unrealistically positive rating designations by credit agencies was, in my judgment, the core of the problem." The Financial Crisis and the Role of Federal Regulators: Hearing Before the H. Comm. on Oversight and Government Reform, 110th Cong. 18 (2008) [hereinafter Hearing]; accord TECHNICAL COMM. OF THE INT'L ORG. OF SEC. COMM'NS, THE ROLE OF CREDIT. RATING AGENCIES IN STRUCTURED FINANCE MARKETS 3 (2008), http://www.iosco.org/library/pub docs/pdflIOSCOPD270.pdf [hereinafter IOSCO]; Michael J. de la Merced, 3 Financiers on Wall Street's Future, DEALBOOK, Feb. 19, 2009, http://dealbo ok.blogs.nytimes.com/2009/02/19/where-will-finance-go-3-veterans-are-uns ure/ (quoting the former CEOs of Lehman Brothers and AIG and the Chairman of Blackstone Group as blaming CRAs for the market crisis). 2 See Hearing, supra note 1, at 18; infra Parts III & IV. Edward Carr, A Special Report on the Future of Finance: Wild Animal Spirits, ECONOMIST, Jan. 24, 2009, at 8. 246 [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGSAND FINANCIAL CRISIS overvalued, which-because of mark-to-market accounting rules'-forced huge write-downs for those who owned them.' Consequently, the balance sheets of many financial institutions were decimated, some became insolvent, and the whole financial system was pushed to the brink of collapse. Today the private mortgage securitization market is virtually non-existent, and only mortgage securities guaranteed by the Government Sponsored Enterprises (GSEs), Fannie Mae, Freddie Mac, and Ginnie Mae, are bought by investors.' Responsibility for the frozen private mortgage securitization market has been widely assigned to the CRAs.' Because mortgage securitization is so important to reviving the economy, the federal government, through its acquisition and subsidization of the GSEs, has become an underwriter and purchaser of RMBSs and R-CDOs.' It is not a desirable solution, however, for the government to act as a long-term mortgage security investor.'o This Note's proposal ' Mark-to-market means that assets are valued on the balance sheet at market values. Franklin Allen & Elena Carletti, Mark-to-Market Accounting and Liquidity Pricing, 45 J. AcCT. & ECON. 358, 359 (2008). ' Carr, supra note 3, at 8. 6 Id.; see also Michael Lewis & David Einhorn, The End of the Financial World as We Know It, N.Y. TIMES, Jan. 3, 2009, at WK9. See SECURITIES INDUSTRY AND FINANCIAL MARKETS ASSOCIATION, GLOBAL CDO MARKET ISSUANCE DATA 2000-2009, http://www.sifma.org/ uploadedFiles/Research/Statistics/SIFMAGlobalCDOData.pdf [hereinaf- ter SIFMA CDO]; SECURITIES INDUSTRY AND FINANCIAL MARKETS ASSOCIATION, RESEARCH QUARTERLY AUGUST 2009, at 9, http://www.sifma. org/research/ pdflRRVol4-8.pdf [hereinafter SIFMA RMBS]. 8 See infra Part IV. * Dashed Expectations; America's Bank Bail-Out, ECONOMIST, Feb. 14, 2009, at 83; see also Ben S. Bernanke, Chairman, Bd. of Governors of the Fed. Reserve Sys., Address at the UC Berkley/UCLA Symposium: The Mortgage Meltdown, the Economy, and Public Policy (Oct. 31, 2008) (transcript available at http://www.federalreserve.gov/newsevents/speech/ bernanke20081031a.htm); SIFMA RMBS, supra note 7, at 9. 1o The recent history of the GSEs, namely the federal government's implicit backing of their ruinous risk-taking and growth, serves as the prime example of why the government should minimize its involvement in insuring and purchasing these securities. See Parts V and VI for an explanation of why government rating of securities does not lead to the 247 hopes to revive private mortgage securitization with significantly less expansive government intervention. To restart the non-GSE RMBS and R-CDO market and to ensure its continued functioning, this Note argues that a federal government entity ("the Agency") needs to take the place of the CRAs in rating RMBSs and R-CDOs. Inaccurate RMBS and R-CDO ratings are caused in large part by the "issuer pays" system of financing, whereby issuers of securities pay the CRAs for credit ratings, thus creating an obvious conflict of interests.u However, some academics and the CRAs claim that the CRAs' incentive to maintain a strong reputation ("reputational capital") outweighs their desire in any individual case to profit from inflated ratings.12 The current financial crisis, however, disproves this claim and instead indicates that the conflict of interests inherent in issuer financing have overcome the reputational capital constraint and resulted in inflated RMBS and R-CDO ratings.13 Since the issuer pays model is unlikely to change," the CRAs will continue to be incentivized to inflate RMBS and R-CDO ratings, and investors and regulators will probably distrust CRA ratings of R-CDOs and RMBSs for the foreseeable future.'" On the other hand, this Note will show that a government entity can effectively rate R-CDOs and RMBSs. same problems as the government, whether by itself or through implicit backing of the GSEs, purchasing and insuring the securities. n See infra Part III. 1 See Claire A. Hill, Regulating the Rating Agencies, 82 WASH. U. L.Q. 43, 50-51 (2004); Steven L. Schwarcz, Private Ordering of Public Markets: The Rating Agency Paradox, 2002 U. ILL. L. REv. 1, 14 (2002); Fabian Dittrich, The Credit Rating Industry: Competition and Regulation 21-52 (June 4, 2007) (unpublished Ph.D. dissertation, University of Cologne), available at http://papers.ssrn.com/sol3/papers.cfm?abstractLid=991821. 13 See infra Part IV. " See infra Part III.A. " See John Patrick Hunt, Credit Rating Agencies and the "Worldwide Credit Crisis": The Limits of Reputation, the Insufficiency of Reform, and a Proposal for Improvement, 2009 COLUM. Bus. L. REv. 109, 155, 170-76. 248 COL UMBIA B USINESS LA W RE VIE W [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS This Note makes the case for the nationalization of RMBS and R-CDO ratings in the following manner. Part II presents a short background of the mortgage security and CRA industry. Part III analyzes the three dominant factors of CRA behavior: (1) the dominance of the issuer pays model, (2) the role of reputational capital, and (3) the oligopolistic structure of the industry. This Part restates the current scholarly view that the incentive and regulatory structure guiding the CRAs is suboptimal. Part IV of this Note examines how the CRAs proximately caused the meltdown of the RMBS and R-CDO market. Part V demonstrates how the Agency can remedy the problems presented in Part III and Part IV. Part VI explains why a government intervention is neither foolish nor unprecedented. Part VII presents variations of the most commonly proposed solution to bad CRA performance: civil liability for bad ratings. This Note concludes that liability systems are overly complicated and ineffective in deterring bad ratings. II. BACKGROUND A. Scope of the Agency's Functions This Note will focus on only one method by which mortgage originators can sell mortgages into the secondary market: RMBS and R-CDO securitization.16 Another method, the covered bond, requires that the issuer guarantee the bond, which means that the bond remains on the issuer's balance sheet." This limitation will prevent covered bonds "6 This Note's proposal seeks to address problems in the "private label" R-CDO and RMBS market. Private label denotes that the GSEs are not involved as buyers or insurers of these securities. Such government involvement, discussed in greater detail in Parts V and VI, is not a viable long-term solution. " See Renzo G. Avesani, Antonio Garcia Pascual & Elina Ribakova, The Use of Mortgage Covered Bonds 15 (IMF Working Paper WP/07/20, 2007), available at http://papers.ssrn.com/sol3/papers.cfm?abstractid=96 1094. 249 from replacing RMBSs and R-CDOs in the United States. 18 There are two dominant forms of mortgage securitization. One type is the relatively uncomplicated RMBS, which is the securitization and tranching'9 of the payments, principal, and interest, from a spectrum of mortgages, with the investor receiving a pro rata share of a particular tranche.2 0 The other type is the more complicated RMBS-based CDO, which consists of the tranched cash flows from several RMBSs.21 Like the RMBSs, investors purchase shares of a particular tranche in an R-CDO.22 Ratings for other types of securities are excluded from the Agency's mandate for the following reasons. First, R-CDOs and RMBSs dwarf other forms of asset securitizations,2 3 so a solution that addresses only R-CDO and RMBS ratings is sufficient to address the bulk of the problem in asset based security ratings. Second, as discussed in Parts III, IV, and V, the Agency would be relatively more qualified to rate these securities and would be relatively less qualified to rate 1s Id. 19 Tranching orders the legal priority to a stream of payments. For example, for a CDO with three tranches, Tranche 1, the most senior, would have to be paid in full before Tranche 2 is paid; Tranche 2 would have to be paid in full before Tranche 3 is paid. Each tranche is rated separately based on the likelihoods that the tranche above it will be paid in full and that there is enough capital left over to satisfy the tranche's obligations. See Leon T. Kendall, Securitization: A New Era in American Finance, in A PRIMER ON SECURITIZATION 11 (Leon T. Kendall & Michael J. Fishman eds., 2000); see also Adam B. Ashcraft & Til Schuermann, Understanding the Securitization of Subprime Mortgage Credit, FED. RES. BANK N.Y. STAFF REP. No. 318, at 20 (2008), available at http://www.newyorkfed.org/research/staff reports/sr318.pdf. 20 See Joseph Mason & Joshua Rosner, Where did the Risk Go? How Misapplied Bond Ratings Cause Mortgage Backed Securities and Collateralized Debt Obligation Market Disruptions 52 (May 14, 2007) (unpublished manuscript), available at http://papers.ssrn.com/sol3/papers. cfm?abstractid=1027475. 21 See Mark L. Korell, The Workings of Private Mortgage Bankers and Securitization Conduits, in A PRIMER ON SECURITIZATION 100 (Leon T. Kendall & Michael J. Fishman eds., 2000). 22 See id. 23 See SIFMA RMBS, supra note 7, at 9-10. 250 COLUMIA BUSINESS LAW REVIEW [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS other securities, particularly "plain-vanilla"2 4 bonds. Third, as explained in Parts III and IV, the current financial crisis is primarily a consequence of bad R-CDO and RMBS ratings, and it is not clear that inaccurate CRA ratings of other securities could lead to a similar financial crisis. Together, these factors strongly suggest limiting the Agency's mandate to R-CDOs and RMBSs. 25 B. History of Mortgage Securitization The modern era of mortgage securitization began in the early 1980s when Freddie Mac started buying from private banks prime mortgages, packaging them, and selling them as RMBSs.26 Until 2001, RMBSs generally took only two forms. One form consisted of prime mortgages which conformed to GSE underwriting requirements.2 7 The other form was the "jumbo" prime mortgage, which was a prime mortgage that exceeded the GSE's cap on mortgage size.28 Between 2001 and 2007, the volume of private label securitizations skyrocketed as issuers packaged greater amounts of non-GSE conforming mortgages into R-CDOs, tranching the cash flows so that all but a few tranches received investment grade29 certification.30 Investment grade 2 "Plain vanilla" will be used throughout this Note to denote bonds that are not structured; that is, corporate bonds, municipal bonds, treasury securities, etc. 25 Throughout this Note, differences between RMBSs/R-CDOs and plain-vanilla bonds will be noted. 26 Lewis S. Ranieri, The Origins of Securitization, Sources of Its Growth, and Its Future Potential, in A PRIMER ON SECURITIZATION 34-37 (Leon T. Kendall & Michael J. Fishman eds., 2000); cf. Kendall, supra note 19, at 6 (discussing government-sponsored entities' ability to provide investors with standardized, consumer-loan-backed securities that carried quasi-government guarantees). 27 Ashcraft, supra note 19, at 2. 28 Id. " Investment grade is a BBB- or higher rating. The BBB- level is set by the individual CRAs. Turmoil in U.S. Credit Market: The Role of the Credit Rating Agencies: Hearing Before the S. Comm. on Banking, Housing, and Urban Affairs, 110th Cong. 18 (2008) (statement of Arturo Cifuentes, Managing Director, R.W. Pressprich & Co.). A BBB- rating 251 COLUMBIA BUSINESS LA WREVIEW certification greatly expands the market for these securities because many institutional investors are required, or have substantial incentives, to buy only investment grade securities." The purpose of these regulations is to restrain institutional investor risk-taking. For example, the Basel I banking regulations heavily incentivize commercial banks to invest in securities given the highest rating, AAA, by the CRAs.32 Since there are only a few corporate issuers with that rating, investment banks met the excess demand by structuring R-CDOs so that the most senior tranche received a AAA rating." SEC regulations also mandate that only Nationally Recognized Statistical Rating Organizations (NRSROs)34 can determine whether a security is investment denotes approximately a 0.75% annual probability of default. A AAA rating denotes a zero-percent chance of default. These categories differ only stylistically as between the three CRAs. Joshua Coval, Jakub Jurek & Erik Stafford, The Economics of Structured Finance, 23 J. ECON. PERSP. 3, 9 (2009). 30 Mason & Rosner, supra note 20, at 8-10. An analysis of why the growth occurred now, and not in some previous period, is beyond the scope of this Note; see generally CHARLES P. KINDLEBERGER & ROBERT ALIBER, MANIAS, PANICS, AND CRASHES: A HISTORY OF FINANCIAL CRISES (5th ed. 2005) (analyzing previous financial bubbles). 31 Mason & Rosner, supra note 20, at 8-10. " See Coval et al., supra note 29, at 22. The following are other examples of such regulation: Net Capital Requirements for Broker- Dealers, 17 C.F.R. § 240.15c3-1 (2008) (defining capital adequacy requirements for broker-dealers under the Securities Exchange Act of 1934); Net Capital Requirements for Money Market Funds, 17 C.F.R. § 270.2a-7 (2008) (defining capital adequacy requirements for money market funds under the Investment Company Act of 1940). " See Coval et al., supra note 29, at 4, 22. 34 If anyone's "investment grade" determination were deemed sufficient for regulatory purposes, then there would not be many securities without that label. Thus, the SEC, when first incorporating ratings into regulations, limited qualifying ratings to those made by NRSROs. However, it did leave it up to the NRSROs to determine the qualifications for an investment grade rating. See JOHN C. COFFEE, JR., GATEKEEPERS: THE PROFESSIONS AND CORPORATE GOVERNANCE 289 (2006). In Part V.C, this Note explains why ratings based purely on default risk mask important differences between the default of a tranche in an R-CDO or RMBS and the default of a regular corporate bond. 252 [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS grade or AAA." The focus of this Note is the three NRSROs, Moody's, S&P, and Fitch, that dominate credit ratings, particularly the ratings of RMBSs and R-CDOs.3 6 This Note will refer collectively to these three CRAs as "CRAs." Another historical feature of the CRA industry is that the CRAs are protected from negligence liability by the First Amendment." They have beat back non-fraud based liability because they have convinced courts that their ratings are constitutionally protected opinions." If, however, Congress or the courts could somehow undo this immunity, this Note finds, in Part V.C., that any sort of liability scheme for poor ratings would be unlikely to solve the problems afflicting CRA ratings of RMBSs and R-CDOs. III. CRA INCENTIVES AND INDUSTRIAL STRUCTURE This Part summarizes scholarly work on the incentive and industrial structure faced by the CRAs, with a focus on the issues that are most relevant to ratings of RMBSs and R- CDOs. There are three main factors explaining the poor performance of CRAs. First, the issuer pays model is inevitable-the alternative, the subscriber pays model, is unlikely to work and may not be desirable. Second, reputational capital is an insufficient incentive for CRAs to achieve accurate ratings of RMBSs and R-CDOs.39 Third, " Ashcraft, supra note 19, at 38. 36 SEC, ANN. REP. ON NATIONALLY RECOGNIZED STAT. RATING ORGS., June 2008, at 35 [hereinafter NRSRO Report]. 3 See generally Hunt, supra note 15, at 54-57. 31 Id. However, the edifice of the First Amendment defense is cracking. See Erik Larson, Moody's, S&P Subprime Suit Narrowed by Judge's Ruling, BLOOMBERG, Oct. 19, 2009, http://www.bloomberg.com/ap ps/news?pid=20601103&sid=akVbwpqs-QJO. " See COFFEE, supra note 34, for an extensive discussion of gatekeepers and reputational capital. "Typically, the term [gatekeepers] connotes some form of outside or independent watchdog or monitor- someone who screens out flaws or defects or who verifies compliance with standards or procedures." Id. For Coffee, these are CRAs, auditors, securities analysts, corporate attorneys, and investment bankers. Id. at 1-2. 253 COLUMBIA BUSINESS LAW REVIEW even if the reputational market worked in theory, its significance would be constrained by the near-monopolistic structure of the RMBS and R-CDO rating industry.40 A. Dominance of the Issuer Pays Model In the 1970s, the major CRAs switched from a subscriber- pays41 system to an issuer pays system.42 One reason for this switch was that the issuer pays model was the only means of financing that would enable CRAs to issue ratings across a broad spectrum of securities." Thus, when mutual funds demanded broader ratings coverage in the 1970s, the CRAs accommodated them by switching to the issuer pays model.44 Another possible reason is that advances in photocopying in the 1970s made it easier to disseminate ratings.45 Thus, the greater ease and frequency of theft rendered the collection of subscription fees increasingly difficult for the CRAs.46 With dissemination technology improving and the numbers of rated securities increasing, the subscriber pays model cannot realistically become economically feasible.4 7 Additionally, the subscriber pays model may not be desirable from a fair disclosure perspective. Jeffrey Manns notes that smaller investors may not be able to afford the 40 See infra Part III.C. 41 Subscribers were investors who paid for the CRAs' ratings. They received books from the CRAs' listing companies and their respective three-letter credit ratings. See COFFEE, supra note 34, at 289. 4 Jeffrey Manns, Rating Risk after the Subprime Mortgage Crisis: A User Fee Approach for Rating Agency Accountability, 87 N.C. L. REV. 1011, 1056 (2009); Dittrich, supra note 12, at 18; Hunt, supra note 15, at 32 n.144. ' See Dittrich, supra note 12, at 19. 4 See id. See id. See id. 17 There are several small subscription-based credit rating agencies. Their business model is to sell detailed analyses of a limited number of securities. Although theft is still obviously an issue, their reports are not as easy to disseminate as simple three letter ratings. Additionally, their ratings coverage is miniscule compared to the CRAs. See COFFEE, supra note 34, at 288-89. 254 [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS subscription fees." The subscriber pays system would also have to overcome the SEC requirement that CRAs that use material nonpublic information must make their ratings publicly available.49 Either the SEC changes the rules, thereby condoning selective disclosure by the issuers, 0 or CRAs would be relegated to using only publicly available information in their ratings determinations. Either scenario is undesirable."' B. Issuer Pays Model and Reputational Capital With the seeming inevitability, and desirability from a disclosure perspective, of the issuer pays model, there has to be some other mechanism with which investors can ensure that CRAs are accountable to them and not to the issuers. The reputational capital mechanism (in conjunction with competition),5 2 according to many scholars and the CRAs, provides that check on performance." The theory is that investors value a Moody's, S&P, or Fitch rating because these companies have reputations for producing accurate ratings." Without the investor's confidence, the argument goes, issuers would not pay the CRAs to have their securities rated. 5 Why would a CRA risk compromising its reputation with investors by "cheating" on a rating, particularly when * See Manns, supra note 42, at 49. * See SEC Regulation FD, 17 C.F.R. § 243.100 (2009). '0 The SEC is now considering changing this rule to allow subscription-based CRAs access to material nonpublic information. See Re-Proposed Rules for Nationally Recognized Statistical Rating Organizations, 74 Fed. Reg. 6485 (proposed Feb. 9, 2009) (to be codified at 17 C.F.R. pts. 240 & 243). See, e.g., Floyd Norris, Trading One Risk for Another, N.Y. TiMES, Sept. 25, 2009, at Bl. Disallowing CRAs from using nonpublic information would invariably reduce the quality of the ratings. A detailed discussion of why selective disclosure is undesirable is beyond the scope of this Note. 52 See infra Part III.C. " See sources cited supra note 12. 5 Dittrich, supra note 12, at 21-56. 55 Id. 255 no single issuer accounts for more than a few percentage points of a CRA's revenue? 1. Reputation and Novel Products First, it may not be the case that there is reputational capital at risk for particular ratings. John Hunt argues that a CRA can leverage its prestige in rating other products- e.g. corporate bonds-to credibly rate novel debt products.56 But the reputational spillover flows in only one direction.7 Because CRAs have been rating RMBSs and CDOs for a short period of time relative to other debt instruments, investors do not associate poor RMBS and R-CDO ratings performance with the ratings performance of traditional securities.8 Second, because investors generally lack the requisite breadth of information to value RMBSs, it is more difficult to assess CRA performance, potentially allowing bad performance to pass undetected (the current crisis being an exception)."9 Logically, investors should therefore infer that the CRAs are not actually putting their reputation at risk for novel products. Hunt argues that investors did not make this inference, allowing the CRAs to abuse the investors' trust with faulty RMBS and R-CDO ratings.60 2. Close Contact and Influence Second, R-CDO and RMBS ratings involve close contact between the CRA analyst and the issuer. . With regular corporate bonds, the debtor's ability to affect its initial credit rating is comparatively limited.' However, with R-CDOs 56 Hunt, supra note 15, at 45. 57 Id. 58 Id. 5 IOSCO, supra note 1, at 2. 60 Id. 61 Mason & Rosner, supra note 20, at 13. A corporation usually cannot, without great cost, change its capital structure or projected cash flows so that a particular debt issuance achieves a desired credit rating. See id. 256 COL UMBIA B USINESS LA W RE VIE W [Vol. 20 10 No. 1:245] MORTGAGE CREDIT RATINGSAND FINANCIAL CRISIS and RMBSs, assets can be added and tranches restructured, all easily within the control of the issuer who desires a particular rating.62 Thus, issuers "engage in a process that is 'iterative and interactive,'" with the CRAs informing the issuer of the "requirements . . . to achieve target ratings."63 At first glance, intimate communication does not seem to be troublesome; it could help CRAs make more accurate ratings and allow for greater issuer satisfaction. But, as with a hostage that eventually sympathizes with his or her captors, with close contact comes the. potential for an issuer to cloud the judgment of a CRA. A recent article in the New York Times alleges that Countrywide was able to persuade Moody's to increase a rating simply by calling and complaining, without offering new information about the quality of the assets.6 4 Although Countrywide may have also employed other levers besides persuasion, this example demonstrates how close contact may be problematic. To address this problem (and several other problems described later in this Note), the SEC, in early 2009, adopted new rules restricting communication between CRA rating analysts and executives that negotiate fees for the issuers.6" However, as long as issuers pay for their ratings, their ability to "persuade" the CRAs will probably be quite strong. Additionally, CRAs sell consulting services to issuers.66 The accounting industry was criticized in the wake of Enron 62 Id. 63 Id. 64 See Gretchen Morgenson, Debt Watchdogs: Tamed or Caught Napping?, N.Y. TIMES, Dec. 7, 2008, at Al. 65 See Amendments to Rules for Nationally Recognized Statistical Rating Organizations, 74 Fed. Reg. 6456 (proposed Feb. 9, 2009) (to be codified at 17 C.F.R. pts. 240 & 249b); Annette L. Nazareth, SEC Publishes Final Rules for Credit Rating Agencies, THE HARVARD LAW SCHOOL FORUM ON CORPORATE GOVERNANCE AND FINANCIAL REGULATION, Feb. 5, 2009, http://blogs.law.harvard.edulcorpgov/2009/02/05/sec-publishe s-final-rules-and-proposes-additional-rules-for-credit-rating-agencies/. 6 See Frank Partnoy, How and Why Credit Rating Agencies Are Not Like Other Gatekeepers 60 (Univ. of San Diego Sch. of Law, Legal Stud. Research Paper Series, Research Paper No. 07-46, 2006), available at http://ssrn.com/abstract=900257. 257 COLUMBIA BUSINESS LAW REVIEW for the other services it sold to its clients.67 Sarbanes-Oxley curtailed this practice for accountants," but not for CRAs.6 9 The 2009 SEC CRA rules do not prohibit cross selling of services, but do ban "recommendations," which are pre- rating communications whereby the CRA advises the issuer on how to structure the security to achieve a certain rating.o However, enforcing this ban may be difficult, as the SEC rules are vague about which communications will be allowed and which will not. 3. Reputation and the Bubble Economy The fear of losing reputational capital may succumb to the exuberance of a bubble economy.72 Given their close contact with the issuers of these securities, the CRAs may have become infected by the same irrational exuberance that the issuers experienced. It did not help that Moody's and McGraw-Hill, S&P's parent, were publicly traded companies with inflated pre-crisis stock prices.74 For example, Moody's stock was trading at over twenty times earnings,7 ' and its market capitalization was fifteen billion dollars in 2005.76 Analyst and executive bonuses were paid with stock options7 and with the difficulty of maintaining such a high market valuation, CRA analysts and executives may have felt the pressure, just like their clients, to maximize the short-term 67 COFFEE, supra note 34, at 65, 151, 291. 6 Id. at 166-67; Partnoy, supra note 66, at 60. 69 Nazareth, supra note 65, at 2. 70 See id. 71 Id. 72 See generally COFFEE, supra note 34, at 67-69 (describing market euphoria among gatekeepers). 7 See Gretchen Morgenson, How the Thundering Herd Faltered and Fell, N.Y. TIMEs, Nov. 9, 2008, at BU1 (chronicling the downfall of Merrill Lynch, one of the major RMBS and R-CDO issuers). 74 See Partnoy, supra note 66, at 65, 69 n.33. 7 Id. at 65. 76 Id. " Morgenson, supra note 64. 258 [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS rewards of a bubble economy." This meant keeping the issuers happy. 4. Reputation and Regulatory License Because so many institutional investors are required, or are heavily incentivized, to buy only "investment grade" securities, it is extremely important for an issuer that its securities achieve an investment grade rating." The CRAs police that boundary in their role as NRSROs.o With so much riding on achieving investment grade status, the NRSROs might alter their standards at this boundary. The evidence is mixed as to whether this really happens with respect to corporate bonds. A recent study says yes,"' while two other studies suggest no.82 Regardless, the temptation for CRAs to "cheat" will always be strong for ratings at the investment grade boundary. C. Difficulty of Entry Notwithstanding the discussion above, let us assume that reputational capital is important to the CRAs. The theory is then that the CRAs compete along the axis of reputation." Reputational capital competition forces the CRAs to maintain and improve their ratings; otherwise their reputation declines and investors discount the quality of 7 See generally COFFEE, supra note 34. * Dion Bongaerts, K.J. Martijn Cremers & William Goetzmann, Tiebreaker: Certification and Multiple Credit Ratings 2 (Yale ICF Working Paper No. 08-27, 2009), available at http://ssrn.com/abstract=1307782 (finding that it accounts for a 45 basis-point difference in yield). 8o See supra Part II.B. s Bo Becker & Todd Milbourn, Reputation and Competition: Evidence from the Credit Rating Industry 6 (Harv. Bus. Sch. Working Paper 09-051, 2009), available at http://ssrn.com/abstract=1278150. 82 Bongaerts et al., supra note 79, at 27; Daniel M. Covitz & Paul Harrison, Testing Conflicts of Interest at Bond Ratings Agencies with Market Anticipation: Evidence that Reputation Incentives Dominate 4-6 (FEDS Working Paper No. 2003-68, 2003), available at www.federalres erve.gov/Pubs/feds/2003/200368/200368pap.pdf. " See, e.g., Hill, supra note 12, at 45. 259 COLUMBIA BUSINESS LAW REVIEW their ratings." Uncertainty about ratings would lower the price of the rated securities, thus motivating issuers to seek out CRAs with better reputations." However, even if reputation among investors could serve as an axis along which CRAs compete, the oligopolistic industrial structure of the CRA industry would limit its significance. There is a two-rating norm for securities, so three competitors are needed to avert a "partner monopoly."86 The CRA industry has barely avoided a partner monopoly- there are three competitors, Moody's, S&P, and Fitch, which account for more than 98% of all ratings. In the plain- vanilla rating industry, with changes to the SEC's NRSRO rules, there is a chance of more competition." In the insurance industry, for example, A.M. Best accounts for about 25% of all ratings." However, for asset-backed securities ratings, which include RMBS and R-CDOs, the big three are responsible for almost 98% of all ratings.90 Competition in this class of securities, as opposed to other classes of securities, is probably less likely for the following reasons. 1. Regulatory Barrier Until recently, the SEC aided the CRA's oligopoly by withholding NRSRO designations." Per the 2006 CRA Reform Act mandate, the SEC has increased the number of NRSROs to ten.92 Although the regulatory burden has now been lifted, economic factors create as effective a burden as regulations. 8 See id. at 50-51. 85 See id. * Mason & Rosner, supra note 20, at 9. m NRSRO Report, supra note 36, at 35. * See COFFEE, supra note 34, at 291-92. 8 NRSRO Report, supra note 36, at 35. 90 Id. at 36. 91 COFFEE, supra note 34, at 291. A comprehensive history of the NRSRO designation process is beyond the scope of this Note. 9 NRSRO Report, supra note 36, at 1, 34. 260 [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS 2. Economic Barriers First, for RMBSs and R-CDOs, there is less opportunity for an entrant to establish a reputation. For ordinary corporate bonds, the market can serve as a second rater of the security since much of the information that a CRA uses to rate a corporate bond is in public filings." However, investors of RMBSs and R-CDOs have no other mechanism to judge CRA performance and are therefore comparatively more dependent on the CRA.94 This means a competitor CRA would have a much harder time entering the market for rating RMBSs and R-CDOs because investors can only ascertain reputation, and not performance (unlike ratings for corporate bonds), during a RMBS and R-CDO rater's early years of existence." Second, ratings on regular bonds reflect the idiosyncratic risks of the asset. Thus, a CRA's corporate rating incorporates company and industry specific risks, which can be quite subjective, e.g. quality of management, products, trends, etc.96 But, for RMBSs and R-CDOs, diversification generally eliminates the idiosyncratic risk of the assets underlying the security.97 Since all or most of the risk is systematic, it is necessary to get the whole picture. Therefore, entry cannot be made through a piecemeal approach, but needs to be at the scale of the entire RMBS and R-CDO market, thus preventing access for those potential entrants that do not have enough financing to enter on such a large scale. Third, there is a chance that the CRAs can use their existing dominance in regular bond ratings to push out competitors in the RMBS and R-CDO market. The CRA Reform Act of 2006 is supposed to prevent tying,9" but tying " IOSCO, supra note 1, at 3-5. 9 Id. 9 Dittrich, supra note 12, at 62. * Asheraft, supra note 19, at 43. 9 Id. 98 Hunt, supra note 15, at 17 n.71. Tying is the practice of conditioning the purchase of one product on the purchase of another. 261 COLUMBIA BUSINESS LAW REVIEW rules may not be easily enforced. For example, major RMBS and R-CDO issuers like J.P. Morgan and Bank of America must have their own corporate bonds rated by two of the big three CRAs. It is not difficult to imagine that the CRAs would use this as leverage to demand they also rate these banks' RMBSs and R-CDOs. CRAs engage in the tying-like behavior of "notching."9 Notching is the assignment of a lower rating to a CDO whose underlying assets, e.g. RMBSs, were not originally rated by the CRA rating the CDO.oo Because the CRAs have some justification in adjusting ratings based on who rated the original assets (i.e., the CRA cannot vouch for the quality of the underlying ratings) the SEC has to be selective in prosecuting these violations."o' It remains to be seen whether the SEC can reliably determine whether particular instances of notching are legitimate. D. Conclusion Private competition does not ensure quality RMBS and R- CDO ratings for two reasons: the unimportance of reputational capital and the oligopolistic nature of the industry. Furthermore, the issuer pays model creates a strong temptation for CRAs to issue inflated RMBS and R- CDO ratings, which the next Part shows was the case in this financial crisis.102 These problems are unavoidable. The issuer pays model is the entrenched form of industrial organization,' and it is unlikely that the steps of the RMBS and R-CDO rating process can be changed so that reputation becomes more important0 ' or entry becomes more effective.0 " Therefore, another entity, the Agency, is needed to rate RMBSs and R-CDOs. * Id. 100 Id. 101 Id. 102 See infra Part IV. " See supra Part III.A. '04 See supra Part III.B. 1o0 See supra Part III.C. 262 [Vol. 20 10 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS IV. THE 2007-2009 CREDIT CRISIS: EVIDENCE OF CRA FAILURE Issuance of private label RMBSs and R-CDOs dropped from approximately seven hundred billion dollars in 2007, to forty billion dollars in 2008, to nearly nonexistent in 2009.10 Two reasons for this decline may be that the appetite for risk has declined across the board and that the housing bubble burst.1 7 However, issuance of corporate bonds only dropped about thirty percent between 2007 and 2008,"o8 and equity issues actually increased."' As of October 2009, most capital markets except the securitization market have seen substantial rebounds."0 Scholars,"' the media,"2 the marketplace, "' Congress," 4 and regulators", have concluded that a poor CRA rating process is responsible for a substantial part of the decline in volume. This Part lists the proximate causes of the collapse of RMBS and R-CDO issuances during this financial crisis. As explicated in Part II.B.2, more so than investors in other securities, R-CDO and RMBS investors were dependent on CRA ratings to set prices, and when confidence 106 SIFMA RMBS, supra note 7, at 9. 107 See SECURITIES INDUSTRY AND FINANCIAL MARKETS ASSOCIATION, RESEARCH QUARTERLY NOvEMBER 2008 (2008), at 6, available at http://www.sifma.org/research/pdflRRVol3-10.pdf. 108 Id. at 1. 1o9 Id. "' Aline van Duyn, Battle to Restore Confidence in Securitisation, FIN. TIMES, Oct. 29, 2009, at A4; see also SIFMA RMBS, supra note 7, at 2-7. n. See, e.g., Hunt, supra note 15. 112 See, e.g., Morgenson, supra note 73. .' IOSCO, supra note 1, at 9. 114 Henry Waxman, Chairman of the House Committee on Oversight and Government Reform, accused the CRAs of either fraud or gross incompetence. Credit Rating Agencies and the Financial Crisis: Hearing Before the H. Comm. on Oversight and Gov't Reform, 110th Cong. 15 (2008) (statement of Rep. Waxman, Chairman) [hereinafter Committee Hearings]. ". IOSCO, supra note 1, at 2 (criticizing the CRAs for not incorporating a perceptible decline in the quality of the origination process into ratings). 263 in the rating process rightfully evaporated, so did any idea of how the RMBSs and R-CDOs should be valued.'16 Additionally, the CRAs continually adjusted their methodologies"' and downgraded debt right after an initial rating without explanation."' CRAs have also been critiqued for not downgrading R-CDOs and RMBSs quickly enough when new information came to light."' Additionally, each agency had its own methodology for rating the securities. 20 These methodologies were not internally consistent, as an AAA rating for an RMBS incorporated different, and ultimately incomparable, types of risk than the same rating for a plain-vanilla bond.121 Furthermore, it was difficult to monitor the CRAs as modeling and performance data provided by the CRAs was not specific enough.12 Finally, recent media coverage has been particularly harsh; financial journalists, members of Congress, and Alan Greenspan have all accused the CRAs of either gross breaches of independence or incompetence. 123 At this point, after the CRAs' many failings have come to light and after nearly a year of frozen markets, it is hard to believe that the CRAs can resurrect the market's and regulators' trust in their R-CDO and RMBS ratings. The following Part presents the plan that will replace the CRAs and unfreeze private RMBS and R-CDO issuances. V. PROPOSAL One of the ways the federal government has been trying to stem the current economic contraction is by attempting to 116 IOSCO, supra note 1, at 9. See also A Personal View of the Crisis: Confessions of a Risk Manager, EcoNOMIST, Aug. 9, 2008. 11. IOSCO, supra note 1, at 10. u1 See id. at 11. 119 Id. 120 Id. at 9. 121 See infra Part V.C. 122 IOSCO, supra note 1, at 9. 123 See e.g., Morgenson, supra note 64; Hearing, supra note 1; Hunt, supra note 15, at 48. COL UMBIA B USINESS LA W RE VIE W [Vol. 20 10264 No. 1:2451 MORTGAGE CREDITRATINGSAND FINANCIAL CRISIS resuscitate the housing market.124 It has done so by purchasing mortgage securities through the GSEs and the Federal Reserve Bank.125 However, this is a short-term solution. 12 6 This Note's proposal helps lower mortgage rates without such an expansive government intervention. To regain investor confidence and demand for RMBSs and R- CDOs, thereby increasing securitization and lowering mortgage rates,12 7 this Note proposes that a federal government entity ("the Agency") take over the CRAs' responsibilities in rating RMBSs and R-CDOs. This Part will enumerate the reasons why the Agency is superior to the CRAs in rating RMBSs and R-CDOs. A. Incentive Advantages The Agency will experience less pressure from issuers to inflate ratings because its funding will come from general revenue sources and will not be tied to turnover. 128 There will be no "iterative and interactive" 129 contact between 124 Dashed Expectations, supra note 9. 125 Bernanke, supra note 9, at 2; Deborah Levine, Investors See Mortgage Rates Rising as Fed Wraps Up Buys, MAIUCETWATCH, Nov., 3, 2009, http://www.marketwatch.com/story/story/print?guid=2BOFOC39-C2 25-4837-BA89-EBF4CA6769CA. 126 The federal government, however, might have a hard time extricating itself from its ownership of Fannie Mae and Freddie Mac. Charles Duhigg, U.S. Likely to Keep the Reins On 2 Fallen Mortgage Giants, N.Y. TIMES, Mar. 3, 2009, at Al. Given the criticisms of the federal government's pre-crisis implicit backing of Freddie Mae and Freddie Mac and the subsequent nationalization of these two agencies, it is politically unfeasible that the federal government will continue to be a major purchaser or insurer of mortgage securities beyond its current, temporary involvement with the GSEs. See generally Brent Horton, In Defense of Private-Label Mortgage-Backed Securities, 61 FLA. L. REV. 827 (2009) (describing the failure and consequences of government backing of GSEs). 127 The Obama Rescue, EcONOMIST, Feb. 14, 2009. 128 Issuers will still pay for ratings, but payments will go to a general revenue fund and not to the Agency. As explained in Part VI.A.1, even accounting for inefficiencies of a government rater, total payments by issuers to the general revenue fund should equal the costs of the Agency without a significant increase in rates. 129 Mason & Rosner, supra note 20, at 13. 265 COLUMBIA BUSINESS LAW REVIEW issuers and analysts because analysts will not seek issuers' business. There will be no employee stock options motivating analysts to increase firm profitability. As with many government agencies, there is the concern that the lure of the private sector might influence employee judgment. This problem cannot be completely eliminated, but can be mitigated with reasonable rules prohibiting Agency analysts from seeking employment with the issuers they had recently rated. Unlike any self-imposed restrictions by the CRAs, barriers to a "revolving door" between the Agency and the private sector issuers will be credible. B. Informational Advantages The Agency would regulate all RMBSs and R-CDOs purchased by regulated financial institutions. Since there are economies of scale in rating RMBSs and R-CDOs," 0 the Agency will already have an advantage over the current CRAs that rate most, but not all, of these securities.'3 1 The Agency will enforce best practice standards for the origination of residential mortgages that become part of the securities. The recent emergence of the "Originate to Distribute" model of mortgage lending has caused deterioration in the quality of lending standards.132 The Agency will have the willingness to address this problem and will not be stymied by any coordination issues that the three competing CRAs may experience. Investors will value the uniformity of having one rater and the Agency's transparency." The Agency will have no competitors, so it can make available to the public all its methods, models, and assumptions. Also, with an open system, the Agency can incorporate ideas from a spectrum of interested third parties, particularly academics and financial 130 See IOSCO, supra note 1, at 8. ..1 See NRSRO Report, supra note 36, at 35. 132 IOSCO, supra note 1, at 5 n.9. .33 See id. at 9 (noting that variety in CRAs' ratings methodologies and lack of disclosure makes evaluation of CRAs' performance difficult). 266 [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS commentators. A more transparent system would have been particularly useful the last few years because warning signs of an impending structured finance collapse were "on the industry side ... dismissed," while academics who were anticipating a collapse were ignored.1 34 Outsiders will be able to critique the Agency's performance, for example, by pointing out where an old model does not fit new data. Additionally, to the extent that this does not conflict with uniformity or disrupt marketability, there can be addendums to a rating that acknowledge differing opinions within the Agency or by outside experts. In its 2009 CRA rule changes, the SEC made some progress in addressing CRA transparency. The SEC now requires CRAs to report material deviations from models.' Also, the SEC added the requirement that CRAs disclose to the public general performance standards and, for ten percent of securities ratings, complete case histories.'3 6 Additionally, the CRAs will be required to disclose their assessments of the quality of the originator's information.' However, these disclosure requirements do not go as far as the Agency's disclosure requirements. Invariably, as profit maximizing entities, the CRAs will hold back on revealing specific models, assumptions, or other information they would want to keep secret from their competitors.'3 8 On the other hand, the Agency's transparency will be designed to allow investors, scholars, and other interested third parties to recreate step-by-step the Agency's credit rating process. Only with that level of candor will investor confidence be resurrected and maintained for R-CDO and RMBS ratings. 134 See Steve Lohr, Wall Street's Extreme Sport, N.Y. TIMES, Nov. 5, 2008, at Bl. 's' Nazareth, supra note 65. 136 Id. 137 Id. "' If the SEC mandated total disclosure, competitors would copy models and there would be no incentive to invest in superior models. Carol Ann Frost, Credit Rating Agencies in Capital Markets: A Review of Research Evidence on Selected Criticisms of the Agencies, 22 J. AccT., AUDITING & FIN. 469, 476 n.26 (2007). 267 COLUMBIA BUSINESS LA WREVIEW Additionally, it should be noted that the SEC is making these rule changes after the CRAs' damage has already been done. With this record of inaction, it becomes a serious question whether the SEC will really be able to enforce these transparency measures or address other future problems. Manns notes that "the advantages or skills that make private gatekeepers serve as attractive complements or substitutes for public enforcement may also equip these gatekeepers with the tools to facilitate illicit activity and to obfuscate their malfeasance."3 9 It seems that only when their malfeasance becomes a crisis, as it has now, does the SEC respond. The Agency, unlike the SEC, will have the advantage of being on the front lines of the securitization process, inspecting each RMBS and R-CDO, and anticipating problems in the securitization process, instead of just reacting to them. C. Economic Regulation and NRSROs The Agency's restrictions on RMBS and R-CDO creativity will protect the interests of investors. The Agency needs to protect investors, even sophisticated ones, because issuers took advantage of investors' trust in the rating process to engage in "ratings arbitrage."' According to the Securities Industry and Financial Markets Association (SIFMA), "arbitrage CDOs attempt to capture the mismatch between the yields of assets [CDO collateral] and the financing costs of the generally higher rated liabilities [CDO tranches]."141 In other words, through the use of ratings, the total-an R- CDO security composed of RMBSs-can be made to look better than the sum of the parts-a portfolio of RMBS securities. Approximately 80% of CDOs are arbitrage CDOs.142 139 Manns, supra note 42, at 1026. 140 Partnoy, supra note 66, at 75--80; see also SIFMA CDO, supra note 7. 141 SIFMA CDO, supra note 7, at 5. 142 Id at 2. 268 [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS Basic finance instructs that a portfolio security, an R- CDO for example, should not be more valuable than the sum of its assets if an investor can do the same thing, namely group RMBSs into a portfolio.'43 The existence of arbitrage CDOs suggests that issuers are attempting to create this inflated valuation, and they are succeeding because of CRA ratings.1' The mechanism for fooling investors, according to Coval, is that the ratings hide the fact that the risk profile of R-CDOs is different from the risk profile of corporate bonds.1 45 Default risks for a regular bond are a reflection of unsystematic and systematic risk.4 6 However, for most R- CDOs, because of diversification in the asset pool, default risk reflects only the systematic risk associated with that type of asset.' Therefore, because a regular bond's unsystematic risk can be eliminated by a diversified investor, a corporate bond's yield should reflect only systematic risks, which, by definition, are less than the sum of the two risks.'4 8 However, with R-CDOs, diversification does not reduce risk, so there should not be a similar reduction of yield.14 9 Nevertheless, if both securities are rated the same (because their overall default risk is the same), investors may have thought the securities should yield the same."5 o Taking a more optimistic view of investors' intelligence, maybe institutional investors themselves were engaging in ratings arbitrage to fool regulators and subvert regulations that restrain risk-taking.' Thus, while the CRAs can claim 14 Partnoy, supra note 74, at 75. 1" Id. 145 Coval, supra note 29, at 20-21. 146 Id. 147 Id. 148 Id. 149 See id. 1 0 Id. Former Federal Reserve Chairman, Alan Greenspan, was confused as well. Cyrus Sanati, Greenspan Says He Was Mystified by Subprime Market, DEALBOOK, Feb. 12, 2009, http://dealbook.blogs.nytime s.com/2009/02/12/greenspan-says-he-was-mystified-by-subprime-market/. 151 Coval, supra note 29, at 22. 269 that the overall risk of default is the same between different types of securities rated investment grade, thereby justifying the same rating, the nature of the risks is certainly not the same (systematic versus unsystematic).15 2 This probably subverts the goal of regulators who must have thought, just like some investors, that ratings reflect comparable risks.'53 It seems imprudent that such important regulatory power be assigned to the profit maximizing CRAs, which, as the current crisis bears out, either tried to subvert these regulations or did not stop others from subverting them. Partnoy suggests eliminating CRA regulatory power by changing financial institution regulations to use other measures of capital adequacy, namely the credit default swap (CDS) prices of the assets.54 This idea, however, is controversial' 55 and Partnoy acknowledges that it would have limited application because the CDS market is not liquid for many securities.'5 6 The 2009 CRA rule changes sought to reduce dependency on CRA ratings by replacing ratings with standards for certain regulations.' However, the SEC can go only so far; ratings are pervasive in financial regulation because there probably is no better way to do it.' Since credit ratings are not going to go away, the rater needs to improve. ,2 Id. at 21. ' Id. at 22. The SEC considered marking novel securities with a special symbol so investors would know they differ from traditional bonds, but eventually ruled against it. John P. Hunt, The SEC's Proposed Rating Agency Rules: Unresolved Conflicts 2 (Berkeley Ctr. for Law, Bus. & Econ. White Paper, 2008), available at http://ssrn.com/abstract=1284709. " Partnoy, supra note 74, at 79. A CDS is like an insurance policy on debt-it pays out to the holder the face value of principal and interest if there is a default on the underlying security. The cost of this insurance can serve as a proxy for the creditworthiness of the underlying assets. Id. 'r" Hunt, supra note 15, at 31. 156 Partnoy, supra note 74, at 93. 1 Hunt, supra note 15, at 30. 118 Id. at 31. 270 COL UM BIA B USINESS LA WRE VIE W [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS D. Avoiding Externalities Another reason the Agency is better suited to rate R- CDOs and RMBSs is that there are significant externalities to poor quality ratings. One such externality is the contagion effect of financial shocks."'9 The current financial crisis, which improper RMBS and R-CDO ratings played a large part in causing, has dried up finance and shaken investor confidence in the capital markets. 60 Alternatively, in case there is a future crisis similar to the current one, the Agency will be able to accurately gauge and resuscitate the health of investors who hold Agency-rated securities. The Agency, because it has rated all of these securities, would be an immediate source of information on the value of RMBSs and R-CDOs, therefore assisting federal policymakers in assessing the health of banks and other institutional investors who purchased them. This ability would be useful in the current crisis because the Treasury Department and Federal Reserve's inability to value RMBSs and R-CDOs has hindered their attempts to strengthen the banking system.'6 1 Also, the CRAs actually obfuscated government policymaking by failing to downgrade debt when their models showed that it should be downgraded.'6 2 The Agency will be designed to work with other agencies and not contravene policymaking in this manner. VI. PUBLIC-PRIVATE DEBATE This Part analyzes the choice of the Agency versus the CRAs within the broader debate of which activities should be in private hands. A complete framework of when an activity should be privatized, and when it should not, is beyond the scope of this Note. Instead, this Part addresses some of the 15 Id. at 58-59; see also The Obama Rescue, supra note 127; Hearing, supra note 1. 160 Hunt, supra note 15, at 58-59. 161 John C. Coffee, Jr., Address at Columbia Law School Panel Discussion on the Current Financial Crisis (Oct. 8, 2008), available at http://www.law.columbia.edullaw school/educationtech/streaming/live. 162 Manns, supra note 42, at 1088. 271 common criticisms of nationalization, finding that these criticisms are either inapplicable or can be mitigated in the context of RMBS and.R-CDO ratings. Also, this Part looks at other instances of public provision of goods and services. A. Conventional Criticisms of Government Provision of Goods and Services Many of the criticisms of nationalization have their intellectual roots in public choice theory.163 To summarize, public choice theory argues that government actors are influenced by rational self-interest and not some ideal of virtuous public service.1" Therefore, because there is no profit-motive, government actors will not innovate or work hard, but instead, will be motivated by illegitimate political goals.165 However, for the Agency's tasks, several of the commonly stated criticisms of nationalization-lack of innovation, over-formality, cost, and political corruption- can be sufficiently mitigated, or, in fact, may even become virtues. However, one common criticism is still applicable: difficulties in designing and initializing the government takeover. 1. The Quality and Efficiency of the Rating Process One common criticism of government providers is that they suffer from a lack of innovation.1 66 However, this is arguably an Agency virtue, as it is excess innovation that 16 Jane S. Shaw, Public Choice Theory, in THE CONCISE ENCYCLOPEDIA OF ECONOMICS (David R. Henderson ed., 1993); see also Amir Hefetz & Mildred Warner, Privatization and its Reverse: Explaining the Dynamics of the Government Contracting Process, 14 J. PUB. ADMIN. RES. & THEORY 171, 173 (2004). 164 Shaw, supra note 163. ' Steven Cohen, A Strategic Framework for Devolving Responsibility and Functions from Government to the Private Sector, 61 PUB. ADMIN. REV. 432, 433 (2001). 16 Hefetz, supra note 163, at 173. 272 COL UMBIA B USINESS LA W RE VIE W [Vol. 2010 No. 1:245] MORTGAGE CREDITRATINGS AND FINANCIAL CRISIS played a large part in creating the current financial crisis. 167 Innovative debt instruments, as explained in Part IV.C., were often a way of illegitimately leveraging investor trust. Even the standard-bearer of free markets, The Economist, questions the need for such complexity. The magazine criticizes the illiquidity and valuation difficulties complex securities pose, particularly in times of market stress.6 The Agency will therefore limit innovation by limiting the types of securities it is willing to rate to those that serve the purposes of ratings-based regulations and are not simply designed to take advantage of investors through over- complexity. If investors wish to buy non-Agency rated debt, they could still do so, but these securities would not satisfy ratings-based regulations. Another common criticism of government is over- formality; too many bureaucratic procedures create inefficiencies and slow the delivery of goods and services.69 However, strict regimentation is arguably what the rating process needs. Investors desire the knowledge that the security is reviewed in a predictable fashion or that any changes are immediately and thoroughly explained. 70 The CRAs have been accused of not following their own rules when they changed ratings without apparent reason and when they were slow to downgrade."'7 Manns argues that if the CRAs had only followed their ratings models, and downgraded certain securities at an earlier moment, the Federal Reserve and the Treasury Department would have been able to avert the Bear Stearns collapse."" Therefore, although the Agency may be slow and regimented, it would be predictably slow and regimented. Furthermore, it is arguably the lack of predictability, not necessarily the speed 167 In Plato's Cave, ECONOMIST, Jan. 24, 2009, at 13. 168 Id. 169 Cohen, supra note 165, at 433. 170 See IOSCO, supra note 1, at 11. 171 IOSCO, supra note 1, at 11; Manns, supra note 42, at 1087. 172 Manns, supra note 42, at 1088. 273 at which ratings are upgraded or downgraded, that caused problems in this current financial crisis."' Another criticism may be the cost of the Agency. There is a common perception that government accomplishes its tasks in a less efficient manner than private enterprises do.7 First, this notion may be false; Elliot Sclar argues compellingly that this perception is driven more by ideology than by reasoned inferences from empirical evidence." Second, the CRAs are incredibly profitable."' Moody's income before interest and taxes, for example, was one billion dollars in 2007, or fifty percent of revenue.7 Thus, even if, operationally, the Agency were half as efficient as Moody's, the social cost of its services would be equivalent. Third, as explicated in Part V.C., the prevention of crises such as the current one would justify, if necessary, the higher costs that the Agency might. impose on issuers compared to the CRAs. 2. Accountability in the Rating Process Another common criticism is that government agencies are unaccountable and therefore prone to bad decision- making and corruption."8 However, the transparency of the Agency will help mitigate problems of subpar judgment and enable outside accountability. Assumptions and models will be open to scrutiny. For example, interested parties can 173 The 2009 SEC regulations do require that the CRAs announce any changes in rating methodology. However, as argued in Part V.B, it remains to be seen how well the SEC can enforce such a vague requirement. 174 ELLIOTT D. SCLAR, You DON'T ALWAYS GET WHAT YOU PAY FOR: THE EcoNOMIcs OF PRIVATIZATION 4 (2000). "1 Id. at 11-15. 176 Partnoy, supra note 74, at 62-68. 177 Moody's Corp., Current Report (Form 8-K), at 13 (Feb. 5, 2009). 2007 was the height of the market. Though 2008 earnings were lower, operating margin stayed high at forty-two percent. Id. Before 2007, Moody's operating margin and profits were not substantially different. See Partnoy, supra note 74, at 62-68. 178 Cohen, supra note 165, at 433. 274 COL UMBlA B USINESS LA W RE VIE W [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS track the Agency's performance and offer suggestions for more accurate models or better supported assumptions. Additionally, the transparency will enable the public to expose instances in which models or assumptions have been manipulated for illegitimate political ends. As explained in Part V.C, the Agency's transparency will be more comprehensive than that which a profit-motivated entity could ever achieve. Lastly, investors will still have to purchase the securities. By setting a price for the security that is lower than the one predicted by the rating, investors can signal a belief that the rating is not free from illegitimate political interference. 3. A Costly and Difficult Start One probably unavoidable problem is that it will take a considerable amount of effort and funding to get the Agency up and running. An initial obstacle is that the Agency will start off without the historical data and personnel needed to construct the models. Because the CRAs' NRSRO designations will be taken away by this plan, the CRAs may want to exit the R-CDO and RMBS rating business. Therefore, the Agency might be able to purchase models and data relatively cheaply and hire the personnel that used to work for the CRAs (or at one of the now-bankrupt issuers). Nevertheless, this process might be more difficult than it initially appears, and there may be other costs and obstacles that this Note does not anticipate. How to deal with obstacles in the Agency's formation are questions for another Note. B. Precedent for Government Involvement We do not always rely on a profit-motivated entity, notwithstanding any sort of incentive structure that an outside regulator can design. In various contexts, the government acts as an inspector, ensuring that private entities comply with certain regulations. One such setting is meat inspection; a USDA veterinarian inspects every cow carcass before it is processed into beef for human 275 COLUMBIA BUSINESS LAW REVIEW consumption.' Another example is the EPA's unannounced inspections of chemical facilities.s 0 These government agencies could, in theory, contract out their tasks to private agencies that could operate under a regulatory framework, much as the current CRAs operate under a SEC regulatory framework. However, for these tasks, the public recognizes that neither competition nor the profit-motive leads to a better outcome. In the corporate governance world, there was a plan in the early 1900s to have the Treasury Department directly employ accountants to audit publicly traded companies."' In 1898, Congress created the Industrial Commission, a blue ribbon panel of corporate governance experts, to study regulation of large corporations and trusts.' A minority report of the Commission proposed requiring audits of these large corporations, but this measure was not included in the Commission's final report. 18 However, it showed that at least a minority of corporate governance experts felt that only a government authority could be entrusted to handle the task of large-corporation accounting. Lastly, the government intervenes in instances where the private sector has stopped providing an important good or service. A recent example is the Treasury Department and Federal Reserve's takeover of the GSEs.m' These two government agencies have begun to undertake the activities traditionally reserved for private financial institutions, namely, investing in RMBSs and R-CDOs, because the private sector is not doing so."' Similarly, private investors 17 Cary Coglianese & David Lazar, Management Based Regulation: Prescribing Private Management to Achieve Public Goals, 37 LAW & SOC'Y REV. 691, 696 (2003). 180 Id. at 699. 181 GARY JOHN PREVITS & BARBARA DuBis MERINO, A HISTORY OF ACCOUNTANCY IN THE UNITED STATES 184-86 (1998). 182 Id. 183 Id. '" The Obama Rescue, supra note 127, at 2. 185 Id. 276 [Vol. 20 10 No. 1:245] MORTGAGE CREDITRATINGS AND FINANCIAL CRISIS have lost confidence in CRA ratings of RMBSs and R-CDOs, and thus a government entity should take the CRAs' place. C. Conclusion A detailed exposition of the Agency's form and functions is beyond the scope of this Note. This Part merely argues that the common shortcomings that hobble government interventions elsewhere may not be applicable to the RMBS and R-CDO rating process. Innovation is not really an issue, over-formality may be desirable, and accountability would be greatly facilitated by the Agency's transparency. Furthermore, government involvement is not unprecedented, particularly when the private sector fails or the costs of failure are partially or wholly externalized. VII. THE LIABILITY ALTERNATIVE This Part explains why establishing the Agency is superior to another commonly proposed solution to poor CRA performance: liability for bad ratings. Hunt and Manns have devised two different frameworks for imposing liability in cases of poor CRA ratings. However, both of their proposals are unwieldy and inadequate. This Part will first present these two proposals and offer specific critiques. Then the Note will show why it is difficult, as a general matter, to make the imposition of liability work. A. Proposals Hunt proposes forcing the CRAs to either disclose that they are making low-quality ratings or pay a predetermined penalty if the ratings turn out to be bad."' Hunt recognizes that some minimum standard of quality would have to be established and suggests ways to do so. The SEC, courts, or some outside contractor could enforce this minimum 1 Hunt, supra note 15. 277 COLUMBIA BUSINESS LAW REVIEW quality."' As for the penalty, Hunt suggests disgorgement of profits or some multiple thereof.'8 One weakness of this proposal is the assumption that the CRAs acted rationally. In other words, it presupposes that the cause of the recent CRA ratings failure is the fact that the CRAs knew they could get away with low-quality ratings. This may be one reason why the CRAs acted the way they did, but there are other causes of bad performance that this proposal does not address."8 ' Manns proposes the formation of an investor committee for every RMBS and R-CDO that would finance and monitor the quality of the CRA's rating.9 o Building upon the bankruptcy creditor committee concept, Manns believes that these committees would be effective in advocating for creditor interests in ways that individual, unorganized creditors could not."' The committees would contractually delineate and enforce exactly the role they want the CRA to play.'92 Violations of these rights would be adjudicated by the SEC, which, through a heightened pleading standard, would ensure that the lawsuits not be frivolous.'93 However, creditor committee liability would exacerbate the marketability of these securities. These securities are hard to value because an asymmetry of information exists between their holders and potential buyers; ratings are supposed to ameliorate this disparity.'94 Each creditor committee contract would create further differentiation among securities, thereby exacerbating valuation difficulties. Additionally, bankruptcy case law dealing with creditor committees has taken many years to develop--a difficult 187 Id. at 72. 188 Id. at 54 n.228. Hunt limits liability because full damages liability can be over-deterring. Id. at 62. 1'8 See supra Part III.B.2-3. 190 Marns, supra note 42. 191 Id. at 1055. 192 Id. 193 Id. at 1082 n.292. 194 See Coffee, supra note 161. 278 [Vol. 2010 No. 1:245] MORTGAGE CREDIT RATINGS AND FINANCIAL CRISIS process that non-bankruptcy creditor committee case law would have to go through as well.' B. Failure of Liability Beyond the specific criticisms to the plans offered above, there are several shortcomings with civil liability systems generally. First, it is difficult to determine when exactly a rating is "bad" and therefore difficult to know when a CRA's determination warrants punishment.' 6 Defaults are inevitable; how does a tribunal decide whether the rating was incorrect, or whether the default was just an instance where the CRA correctly assessed the probability of default, and this probability came to fruition? Furthermore, defaults, especially for highly-rated debt, are relatively infrequent.' With such a small sample of defaults by which to judge performance, it would be difficult for a tribunal to decide whether ratings for any given period were low quality."' Second, civil liability requires substantial government oversight, particularly in devising performance standards.' In the creditor committee model, standards are devised by the contracting parties, but the SEC still has to adjudicate complaints.2 00 Generally, any system that requires oversight, whether by a tribunal, the SEC, or investor committees, raises a problem that Manns readily concedes, which is that the "advantages or skills that make private gatekeepers serve as attractive complements or substitutes for public enforcement may also equip these gatekeepers with the tools to facilitate illicit activity and to obfuscate their malfeasance."20' Any overseer would therefore face talented " See Manns, supra note 42, at 1067 (acknowledging the novelty of creditor committees outside the bankruptcy context). 196 Both Hunt and Manns address this in their plans, but acknowledge difficulty in devising the correct standards. See Hunt, supra note 15, at 67; Manns, supra note 42, at 63. ' Ashcraft., supra note 19, at 39. 198 Id. 199 Manns, supra note 42, at 1079. 200 Id. at 1082. 201 Id. at 1079. 279 adversaries who could, as the current crisis has demonstrated, obfuscate their malfeasance quite well.20 2 VIII. CONCLUSION Ultimately, the Agency only addresses one of the many problems affecting the financial services industry. Financial intermediaries will doubtless continue to find ways to sidestep regulations and expose the broader financial system to unwanted risks. Nevertheless, this Note's proposal is an important step in fighting systemic risk because it puts a government actor in the middle of an important risk taking process, allowing developing problems to be prevented ex ante rather than just remedied ex post. One problem that regulators have failed to adequately address is subpar RMBS and R-CDO ratings. As a consequence, the mortgage securitization market is frozen (outside of government purchasing through the GSEs). Unfreezing securitization through massive GSE purchasing is not a plausible long-term solution. In order to bring back private investment, investors need confidence in ratings that only the Agency can provide. Without the Agency, conflicts of interest will continue to plague the issuer pays model and cause bad ratings. Also, proposals predicated on liability are difficult to implement and ultimately ineffective. Therefore, a clean sweep is needed; one that replaces the CRAs with a government entity that truly serves investors and does not subvert financial regulation. 202 See supra Part V.C. 280 COL UMIA B USINESS LA W RE VIE W [Vol. 2010