Adkins v. Stanley

307 F.R.D. 119, 2015 WL 2258231

• Court: District Court, S.D. New York
• Decided: May 14, 2015
• Docket: No. 12-CV-7667(VEC)
• Status: Published

Opinion

OPINION & ORDER

VALERIE CAPRONI, District Judge:

This is one of many cases arising out of the collapse of the housing market. This one comes with a twist: homeowners in Detroit who received subprime loans seek to hold a single investment bank responsible under the Fair Housing Act (“FHA”) for discriminating against African-American borrowers, based on their claim that African-Americans were more likely than similarly-situated white borrowers to receive so-called “Combined-Risk loans.” Plaintiffs allege that Morgan Stanley¹ so infected the market for residential mortgages—and for mortgages written by the New Century Mortgage Company, a now-defunct loan originator, in particular— that it bears responsibility for the disparate impact of New Century’s lending practices. Although Plaintiffs advance creative theories, their class action lawsuit founders on the requirements of Federal Rule of Civil Procedure 23.

Plaintiffs seek to certify a class of “[a]ll African-American individuals who, between 2004 and 2007, resided in the Detroit² area ... and received Combined-Risk Loans from New Century.” Compl. ¶ 229. Plaintiffs define “Combined-Risk loans”³ as loans that are “high-cost” as defined by the Home Mortgage Disclosure Act (“HMDA”), 12 U.S.C. § 2801 et seg.,⁴ and contain two or more of eight risk factors that, they allege, increase the risk of default. Compl. ¶ 34. Defendants oppose class certification, arguing that individual questions will predominate over questions common to the class.

The Court concludes that this class action lawsuit is an inappropriate vehicle to rectify the wrong that Plaintiffs allege Morgan Stanley perpetrated. The subprime mortgage crisis undoubtedly damaged our economy and may have—as Plaintiffs contend—exacerbated preexisting racial disparities in socioeconomic status. While the Court is not unsympathetic to Plaintiffs’ claims, the harmfulness of the terms that Plaintiffs claim that Morgan Stanley caused New Century to include in loans and the role that Morgan Stanley played in causing the terms of specific Plaintiffs’ loans differ considerably within the proposed class; accordingly, Plaintiffs’ proposed class is unworkable.⁵ Plaintiffs’ motion for class certification is therefore DENIED. Defendants’ motion to preclude some opinions in Plaintiffs’ experts’ reports is DISMISSED as moot.

BACKGROUND⁶

I. The Parties

A. Plaintiffs

Beverly Adkins, Charmaine Williams, Rebecca Pettway, Rubbie McCoy, and William Young are African-Americans who purchased or refinanced homes with loans written by New Century, a non-party entity. For example, using an independent broker, Adkins refinanced her and her husband’s home via a 30-year, adjustable rate loan with a substantial prepayment penalty and a 90 percent loan-to-value ratio (“LTV”) (based on an inflated appraisal). Compl. ¶ 129-34; see also Sugnet Deck Ex. 74, Dkt. 129; Reardon Deck Ex. 28, Dkt. 169; Adkins Dep. at 63-65.⁷ When New Century attempted to sell Adkins’ loan to Morgan Stanley, Morgan Stanley “kicked [the loan] out” of the bundle of loans it would buy—meaning that its due diligence efforts flagged the loan and rejected it as undesirable. Gilly Deck ¶ 10 & Ex. E, Dkt. 175. New Century ultimately sold the Adkins’s loan to Credit Suisse First Boston. Expert Report of Timothy J. Ridd-iough, Ph.D. (“Riddiough”), Dkt. 206, ¶ 124. The other individual named plaintiffs have similar stories. Michigan Legal Services, the final plaintiff, alleges that it has been serving low-income communities in Michigan—primarily African-American residents of Detroit—by engaging “in impact-oriented litigation, legislative and administrative advocacy, and client community education.” Compl. ¶¶ 205-20.

B. New Century

New Century, a California-based lender, originated approximately 250,000 subprime mortgage loans per year during the period from 2004 through 2006 (although the number of loans it originated dropped precipitously in the months before its 2007 bankruptcy). Riddiough ¶ 31, FINANCIAL CRISIS INQUIRY COMM’N, FINANCIAL CRISIS INQUIRY REPORT (Jan.2011) (“FCIC”) 71. Plaintiffs allege that New Century was “the second largest originator of subprime residential loans (in terms of loan amounts) each year from 2003 to 2006.” Expert Op. of Ian Ayres in Support of Class Certification (“Ayres”) ¶ 21.

New Century was an “especially aggressive” independent mortgage company that ranked among the leaders in subprime loan originations. FCIC 89. Historically loan originators “avoided making unsound loans because they would be stuck with them in their loan portfolios.” Id. at 7. But like other originators, by the mid-2000s New Century regularly made subprime loans based on questionable underwriting and then sold those loans to investment banks and other seeozzdazy market pzzrchasers, including Morgan Stanley, which securitized them. Licata Dep. at 47, Lindsay Dep. at 93. Like many loan originators, New Century relied on thz’ee “channels” to originate loans—the retail, correspondent, and broker channels. Riddiough ¶ 32.⁸ The vast majority of New Century’s loans were originated through the broker channel, which “originated loans through a network of independent mortgage brokers.” Id. In 2005, the broker channel accounted for approximately 71 percent of New Century’s loans. Id. ¶ 33; see also Ayres ¶¶ 39-40.

Brokers who originated loans were truly “independent,” meaning that after they had written a loan with particular terms, they could shop the loan around to find an originator willing to fund it. See Reardon Deck Ex. 9, McKay Dep. at 99. Brokers could determine whether a loan would likely meet New Century’s guidelines by entering data regarding the borrower and the loan into a computer program known as “FastQual,” which would apply “automated underwriting rules” set by New Century. Id. at 51; see also id. at 131 (“It was really designed to make sure that the underwriting guidelines were being applied consistently.”). When brokers had a loan with terms that were not available on FastQual—for example, when they wanted to qualify a borrower for a loan with an 85 percent LTV but FastQual only provided options up to 80 percent—they could seek an exception from a New Century account executive. Id. at 128-29. While brokers and internal account managers (who generated direct loans) could be “aggressive” with respect to underwriting guidelines, an underwriter from New Century approved every loan that New Century originated, id. at 123-24; the New Century underwriter’s review ranged from ensuring that the documentation matched the data input into the FastQual system, id. at 124, to determining whether a particular loan was worth the risk associated with its terms, id. at 129. Defendants allege that approximately 10 to 25 percent of New Century loans received some sort of an exception to New Century’s underwriting guidelines. Tr. at 60.⁹

In order to originate a loan—through any channel—New Century had to be able to fund the loan. Like most originators, New Century relied on “short-term lines of credit, or ‘warehouse lines,’ from commercial or investment banks.” FCIC 68; see Reardon Deck Ex. 1 (identifying the banks that provided warehouse loans to New Century, including Bank of America, Barclays Bank, Bear Stearns, Citigroup, Credit Suisse First Boston, Deutsche Bank, Morgan Stanley, and UBS).¹⁰ The terms of its warehouse loans permitted New Century to make a loan and assign it to the credit facility, which would then free assets on the facility for New Century’s use. After New Century had sold the loan (either through a whole loan sale or a securitization), it would be removed from the ledger of the warehouse line against which it had been assigned. Tr. at 11314. The parties appear to dispute the extent to which New Century depended on Morgan Stanley’s warehouse line in particular, but in general New Century relied on the existence of warehouse facilities to fund most of the loans that it originated. Cf Riddiough ¶¶ 40-41, Expert Op. of Patricia A. McCoy in Support of Class Certification (“McCoy”) 22.

C. Morgan Stanley

Morgan Stanley, like many other large investment banks, was heavily involved in the securitization and sale of residential mortgage-backed securities (“RMBSs”), both as an investor and as a seller.¹¹ Morgan Stanley was one of a number of banks that had extensive dealings with New Century, as a lender, underwriter of the New Century sec-uritizations, and as purchaser of New Century’s loans.

The parties sharply dispute the relative significance to New Century’s lending practices of Morgan Stanley as compared to the other banks. Plaintiffs’ expert, Patricia McCoy, asserted that Morgan Stanley exerted “singular influence” over New Century. McCoy 22. Defendants’ expert, Timothy Riddiough, disagreed, noting that Professor McCoy did not supply a standard against which to evaluate the claim of “singular influence,” Riddiough ¶¶ 55-56, and that even she conceded that other banks were the “cause” or “principal cause” of some of the so-called Combined-Risk loans that New Century made during the class period, id. ¶ 56 (citing Reardon Deck Ex. 17, McCoy Dep. at 50, 100-05).

The relationship between Morgan Stanley and New Century was indisputably close, although it is not clear how it compares to New Century’s relationships with similar banks. Morgan Stanley officials wrote at the time that “[w]hile they don’t keep specific metrics, we are clearly [New Century’s] largest and most important counterparty.” Sug-net Deck Ex. 7. Other Morgan Stanley-authored materials described Morgan Stanley’s goal “to continue its relationship with New Century in 2005 by maintaining its status as the # 1 whole loan purchaser, # 1 warehouse lender, and # 1 underwriter on a market share basis.” Sugnet Deck Ex. 2 at MS00834840; see also Ex. 3 at MS02685210 (“New Century has approached Morgan Stanley because we are their number one relationship and they would like to keep us their number one relationship.”).

The parties dispute the significance of these self-congratulatory documents, but re-gar dless of whether other banks enjoyed similar relationships with New Century, the record makes clear that at least some Morgan Stanley officials believed that their preferences had a significant impact on New Century’s practices. Morgan Stanley officials described New Century as “extremely open to our advice and involvement in all elements of their operation,” Sugnet Deck Ex. 9, and internal Morgan Stanley documents asserted that “Morgan Stanley is involved in almost every strategic decision that New Century makes in securitized products” and described in detail the bank’s “mutually beneficial” relationship with New Century, Sugnet Deck Ex. 11 at 6. The synergistic relationship between Morgan Stanley and New Century included the placement of Morgan Stanley due diligence staff onsite at New Century. Sugnet Deck Ex. 13 at 387-99; Ex. 14 at 23-24. Contemporaneous Morgan Stanley documents demonstrated some employees’ understanding that New Century “incorporated many of Morgan Stanley’s best practices into [its] origination practices,” apparently “[b]ecause Morgan Stanley is such a large purchaser of loans from New Century.” Sugnet Deck Ex. 12 at 4.

New Century was receptive to the advice of “investors” writ large, see, e.g., Sugnet Deck Ex. 28 at MJM-001152, and was under pressure from the market “to make sales and fund loans,” id. at MJM-001156. New Century had regular meetings “with Wall Street ... to obtain their feedback: what kind of products they wanted, what things New Century was doing that they did or did not like[,] etc.” Id. at MJM-001158; see also id. at MJM-001161 (New Century officials would “go to Wall Street and match their products and their loan pools with Wall Street’s expectations. They would not make any changes internally without making sure [they] complied with what Wall Street wanted.”). The role of Morgan Stanley in particular, as opposed to investor demand in general, is a hotly-contested area not suitable for resolution at this stage of the case.

II. Residential Mortgage-Backed Securities and Subprime Loans

A. RMBSs

By 2015, most readers are surely familiar with the acronym “RMBS” for “residential mortgage-backed security.”

RMBS[s] are “a type of asset-backed security—that is, a security whose value is derived from a specified pool of underlying assets. Typically, an entity (such as a bank) will buy up a large number of mortgages from other banks, assemble those mortgages into pools, securitize the pools (i.e., split them into shares that can be sold off), and then sell them, usually as bonds, to banks or other investors.”

City of Pontiac Policemen’s & Firemen’s Ret. Sys. v. UBS AG, 752 F.3d 173, 177 n. 7 (2d Cir.2014) (quoting Litwin v. Blackstone Grp., L.P., 634 F.3d 706, 710 n. 3 (2d Cir.2011)).

RMBSs can be aggregated into [collateral-ized debt obligations (“CDOs”) ] which are sold in “tranches” based on priority of entitlement to the cash flow. Each tranche of a given RMBS is exposed to the same pool of mortgages, but lower tranches sustain losses before higher tranches in the event that mortgages in the pool default or do not meet payment deadlines. CDOs are similarly divided into higher and lower tranches.

Stratte-McClure v. Morgan Stanley, 776 F.3d 94, 97 n. 2 (2d Cir.2015).

During the so-called “housing bubble,” Wall Street banks exhibited a “quenchless appetite for high-priced, [risky] loans for use in residential mortgage-backed securitization.” McCoy 17; see also Riddiough ¶ 14 (“[T]he aggregate value of non-Agency residential mortgage loan securitizations increased from $50 million in 1995 to $1.2 billion in 2005”). Loan originators—the entities that made mortgage loans to homeowners—sold residential mortgage loans to investment banks (such as Morgan Stanley); the investment banks in turn bundled the loans into RMBSs that were sold on the secondary market. McCoy 17-18.

Plaintiffs claim that “Wall Street’s insatiable demand for high-priced loans caused lenders to cut corners to qualify borrowers however they could.” Id. at 19 (citing Clifford V. Rossi, Anatomy of Risk Management Practices in the Mortgage Industry: Lessons for the Future 36 (Research Institute for Housing America 2010)). Regardless of whether the originators “cut corners,” the demand for RMBSs, among other factors, “led to a significant expansion in the U.S. mortgage market generally, and the sub-prime mortgage loan market specifically.” Riddiough ¶ 26. Plaintiffs postulate that satisfying the appetite for RMBSs “required expanding mortgage lending to borrowers who could not repay.” McCoy 19; see also id. at 24 (“During the housing bubble, investors including Morgan Stanley also pressed [loan originators] to deliver increasingly higher volumes of subprime loans.”).¹²

At least in part to accommodate prospective homeowners who could less clearly afford the homes that they wished to buy, loan originators increased the number of offerings with terms designed to lower the initial cost of a home; these offerings typically offset their lower initial cost by passing the risk of increases in interest rates onto the homeowners. Id. at 24-25. Although they would not make loans to people whom the originators knew to be unable to repay, loan originators “evaluated [people’s] ability to repay based solely on the initial payment, without regard to subsequent payment shock,” “used stated-income and other types of reduced documentation underwriting to mask weak income or assets,” “relied on inflated appraisals as a way to artificially inflate loan-to-value ratios,” “stretch[ed] [their] underwriting guidelines, ... approved] exceptions to [their] underwriting guidelines and ... approved] loans that did not make ‘sense.’ ” Id. at 25; see also Reardon Decl. Ex. 3, Shane M. Sherlund, The Past, Present, and Future of Subprime Mortgages 2 (Fin. & Econ. Discussion Series Divs. of Research & Stats, and Monetary Affairs, Fed. Reserve Bd., Washington, D.C., Working Paper No. 2008-63, Nov. 2008) (“Sherlund”).

B. “Combined-Risk” Loans

The Plaintiffs seek to certify a class of borrowers who received what they describe as “Combined-Risk” loans. They further define these loans as:

loans that meet the definition of high-cost loan under [the] HMDA and also contain two or more of the following high-risk terms: (a) the loan was issued based upon the “stated income,” rather than the verified income, of the borrower; (b) the debt-to-income ratio exceeds 55%; (c) the loan-to-value ratio is at least 90%; (d) the loan has an adjustable interest rate; (e) the loan has “interest only” payment features; (f) the loan has negative loan amortization features; (g) the loan has “balloon” payment features; and/or (h) the loan imposes prepayment penalties.

Compl. ¶ 34. Each of the potential components of a Combined-Risk loan bears some explanation.

1. High-Cost Loans

During the relevant time period, the HMDA defined “high-cost loans” as loans whose annual percentage rate was at least 3 percentage points (for loans secured by a first lien on a dwelling) or 5 percentage points (for loans secured by a subordinate lien on a dwelling) higher than the yield on Treasury securities having comparable maturity periods. Home Mortgage Disclosure, 67 Fed.Reg. 43218, 43223 (June 27, 2002) (amending 12 C.F.R. § 203.4(a)).¹³ Essentially, the “high-cost” component of a “combined-risk loan” means that the interest rate on the loan is substantially higher than the market interest rate on loans made to well-qualified borrowers.

In isolation, high “interest rates increase the risk of default and foreclosure because they raise the borrowers’ monthly payments, putting added strain on often tight family budgets.” McCoy 8. While borrowers who were deemed poor risks qualified only for very high interest rates, a substantial portion of homebuyers who obtained subprime loans were actually eligible for prime rates but were steered to more expensive subprime loans. Id. at 8 n. 15 (collecting sources identifying the percentage of subprime borrowers who qualified for prime loans at somewhere between 10 and 55 percent). The value to lenders of high interest rate loans is obvious—controlling for other factors (such as the risk that the borrower would default), the higher the interest rate, the higher the return for risking the same capital.

2. Stated-Income Loans

“Stated-income” loans were “known to the knowing as ‘liars’ loans’ because in a stated-income loan the lender accepts the borrower’s statement of his income without trying to verify it.” United States v. Phillips, 731 F.3d 649, 651 (7th Cir.2013) (en banc); see also Black’s Law Dictionary 1079 (10th ed.2014). Requiring little or no documentation to support a borrower’s claimed income “opens the mortgage window to large numbers of borrowers who would not qualify ordinarily.” McCoy 10 (quoting Michael LaCour-Little and Jing Yang, Taking the Lie out of Liar Loans: The Effect of Reduced Documentation on the Performance and Pricing of AIL-A and Subprime Mortgages 26 (working paper, Annual AREUEA Conference Paper, 2010)). Not requiring a borrower to supply documentation verifying his or her income has no effect on the risk of default if the borrower is truthful, but the practice permits borrowers who exaggerate or lie about their income to obtain mortgages that exceed their means (and for which they might otherwise be ineligible). Accordingly, Plaintiffs contend that “low-documentation loans substantially raised default rates during the housing bubble.” Id. (collecting sources); but see Reardon Deck Ex. 5, Morgan J. Rose, Predatory Lending Practices and Subprime Foreclosures: Distinguishing Impacts by Loan Category, 60 J. ECON. & BUS. 13, 28 (2008) (“Low-or no-documentation for refinances is generally associated with significantly greater probabilities of foreclosure. In contrast, low—or no-documentation is associated with lesser probabilities of foreclosure for purchase [Fixed Rate Mortgages (‘FRMs’) ], and has no significant effects for purchase [Adjustable Rate Mortgages (‘ARMs’) ]”).

Unlike high interest rates, which have an intuitive appeal to the lenders and to the secondary market, stated income loans lack any features that recommend them to lenders or the secondary market, and the risks associated with such loans make them generally less desirable than a fully documented loan. See Sherlund 16. Accordingly, some investment banks sought to minimize the percentage of no-or low-documentation loans that could be included in a pool of loans being purchased. Riddiough ¶ 97. Still, “the share of fully documented subprime variable-rate mortgages declined from around 75 percent in 2000 to around 60 percent in 2005-2006.” Sherlund 2.

3. Debt-to-income Ratios over 55 Percent

The debt-to-income (“DTI”) ratio “is the ratio of the borrower’s monthly debt obligations to his or her monthly income.” McCoy 9. Plaintiffs offer strong evidence that “higher DTI ratios are positively correlated with higher defaults.” Id. The 55 percent threshold is quite high—in 2013, in the wake of the housing crisis, the federal government imposed a rule that, for qualified mortgages, a “consumer’s total monthly debt payments [including not only mortgage debt, but all debt] cannot exceed 43 percent of the consumer’s total monthly income.” Ability-to-Repay and Qualified Mortgage Standards under the Truth in Lending Act, 78 Fed.Reg. 6408, 6505 (Jan. 30, 2013);¹⁴ see also McCoy 9-10 (citing, inter alia, the Federal Reserve Board of Governors for the proposition “that loans exceeding a debt-to-income ratio of 55% are inherently risky”). There is no intuitive reason for a lender or the secondary market to prefer high-DTI loans to the same loan with a lower DTI; in fact, a rational lender or investor would be expected to prefer a low-DTI loan to an identical loan with a higher DTI ratio.

4. Loan-to-Value Ratios over 90 Percent

The LTV ratio of a residential mortgage loan is the ratio of the principle amount of the loan to the value of the home. When LTV ratios are low, borrowers “usually do not default ... because they can retire their mortgages by selling their houses.” McCoy 11. “But as their LTV ratios mount and approach 100%, their ability to pay off their mortgage by selling their house diminishes once transaction costs are taken into account.” Id.; see also Sherlund 16 (“Assuming the mortgage had a higher loan-to-value ratio ... default rates would be higher.”). Insofar as LTV ratios affect default rates, it is not because the LTV ratio, standing in isolation, affects borrowers’ ability to make monthly payments; for the subset of borrowers who ultimately want to sell or refinance their mortgages, however, a high LTV ratio can prevent refinance and can make a sale difficult or impossible. When the mortgages on a particular home exceed the value of the home (i.e., the combined LTV ratio exceeds 100 percent), the borrower is said to be “underwater.” “Starting in 2007, virtually no lender was willing to refinance an underwater mortgage.” McCoy 11. Nevertheless, “average combined loan-to-value (CLTV) ratios on subprime variable-rate mortgages rose from less than 80 percent in 2000 to over 85 percent in 2005-2006, partly as a result of the more widespread use of piggyback mortgages.” Sherlund 2.¹⁵

As with high DTI ratios, there is no intuitive reason for a lender or the secondary market to prefer high-LTV loans to the same loan with a lower LTV. Because LTV ratios depend on the value of the collateral held by the lender—the more valuable the home, the lower the ratio given the same loan amount— given a choice between otherwise-identical loans, lenders and the secondary market would choose a loan with a lower LTV ratio. See McCoy 11; Riddiough ¶ 97.

5. Adjustable Interest Rates

Whereas FRMs maintain the same interest rate throughout the term of the loan, ARMs adjust the interest rate based on a pre-specified index. “During the period at issue in this case, the most common type of [ARM] was a hybrid ARM,” such as a so-called 2/28 or 3/27 loan. McCoy 12. In a 2/28 loan, the interest rate would be fixed for the first two years and periodically adjustable for the remaining 28. Id. “Hybrid ARMs put the interest rate risk on the borrower, with the attendant hidden risk of payment shock—the risk that monthly payments will rise dramatically upon rate reset. During the housing bubble, many subprime hybrid ARMs had initial rate resets of three percentage points, resulting in increased monthly payments of as much as fifty percent.” Id. As long as housing prices were rising, borrowers could offset any “payment shock” by selling or refinancing their home. Id. But “[i]n an environment of stagnant to falling house prices and stricter underwriting standards, households facing potentially higher mortgage payments due to a mortgage rate reset [found] prepayment [to be] more difficult, thereby increasing the ultimate chances of default.” Sherlund 10. In general, “[borrowers with variable-rate loans have riskier characteristics than those with fixed-rate loans, and riskier loans are more likely to default than to prepay.” Id. at 8.

Perhaps because ARMs placed the risk of rising interest rates on the borrowers, they were a preferred product for Wall Street; investment banks purchasing loans from originators would include conditions requiring a minimum percentage of loans in a pool to be ARMs. Riddiough ¶ 100 (a Barclays bid sheet specified that “at least 77.66% of the loans were be ARMs,” while Morgan Stanley’s specified that “77.91% of the loans were to be ARMs,” and DLJ Mortgage Capital’s provided that “at least 78% of the loans were to be ARMs.”).

6. “Interest Only” Payment Features

Loans with “interest only” features do not fully amortize the principal over the term of the loan; borrowers pay only interest during an initial period (usually six months to three years). McCoy 12. “Interest only” features were pitched as “affordability features” that lowered the initial cost of a loan so that it would be accessible to people who might otherwise not be able to obtain a loan. FCIC 111. Of course, the “interest only” period ultimately ended, and at that point the payments would increase because amortized portions of the principal were then included. McCoy 13. Borrowers who calculated whether they could afford to obtain a particular loan were at risk of agreeing to a loan with manageable payments initially, only to be unable to make the larger payments required once the initial interest-only period expired. This problem was exacerbated when an ARM with an artificially low rate for the initial period took on “interest only” attributes. Id.; FCIC 111. Accordingly, Plaintiffs contend that “interest-only loans originated during the housing bubble had much higher default propensities following recast than comparably seasoned traditional fixed-rate and adjustable-rate loans.” McCoy 13.

“Interest only” loans decreased a secondary market purchaser’s liquidity; accordingly, investment banks and secondary market purchasers sought to limit the percentage of “interest only” loans that they purchased. See Sugnet Decl. Ex. 46.

7. Negative Amortization

In a negative amortization loan, the borrower makes no payments towards the principal (and sometimes part of the interest) during a specified introductory period. After the conclusion of the introductory period, the unpaid balance would be re-capitalized, “the amortization schedule [would be] reset[,] and the borrower [could be] left with potentially unaffordable minimum payments that reflected] the capitalization of unpaid interest, as well as a principal balance that potentially exeeed[ed] the value of the real estate used as security.” Wallace v. Midwest Fin. & Mortg. Servs., Inc., 714 F.3d 414, 417 (6th Cir.2013); see also Wyo. State Treasurer v. Moody’s Investors Serv., Inc. (In re Lehman Bros. Mortg.-Backed Sec. Litig.), 650 F.3d 167, 173 n. 2 (2d Cir.2011). Because of the re-capitalization after the introductory period, the principal could exceed the value of the property at the time of the initial purchase; even steady housing prices could leave a borrower underwater. Plaintiffs’ expert found that “[Researchers who have studied the question agree that nonamortiz-ing and negative amortization features increase the chance of default.” McCoy 13 (collecting sources). The parties do not discuss whether investment banks and secondary market purchasers sought to purchase or to avoid negative amortization loans.

8. “Balloon” Payment Features

In so-called “balloon loans,” a borrower does not fully amortize the principal while making regular payments; instead, either at the end of the term or periodically throughout the loan, the borrower must make a “balloon payment” that is “much larger than earlier [monthly] payments” to cover the portions of the principal not amortized up to that point. Black’s Law Dictionary 1078 (10th ed.2014).¹⁶ Plaintiffs allege that these loans were “used to qualify cash-strapped borrowers during the subprime boom due to their lower payments.” McCoy 13. Borrowers who were only able to afford a loan because of the artificially low initial cost could be expected to struggle when faced with making a one-time payment of the difference between the apparent cost of the loan (as reflected in the monthly payments) and the actual cost of the loan. Id. Accordingly, “[b]alloon loans are also associated with a higher risk of default and foreclosure.” Id. Balloon payments are viewed as an “abusive feature” of subprime loans. Reardon Decl. Ex. 4, Wei Li & Keith S. Ernst, The Best Value in the Subprime Market: State Predatory Lending Reforms 2 (Ctr. for Responsible Lending Feb. 23, 2006) (“Li & Ernst”). Balloon payments were typically due ten years after a loan’s origination. Rose 28. While there is a clear incentive for borrowers facing a balloon payment to refinance the loan, borrowers who could not refinance— because of declining housing prices, their lower viability as a borrower, or for any other reason—would face a large payment shock. On the other hand, banks preferred balloon loans to traditional loans with similar terms, believing that balloon loans led to greater liquidity. Sugnet Decl. Ex. 37.

9. Prepayment Penalties

One of the better-known features of many subprime loans written during the housing bubble, a prepayment penalty is a “charge assessed against a borrower who elects to pay off a loan before it is due.” Black’s Law Dictionary 1314 (10th ed.2014). With a prepayment penalty, the lender was guaranteed a period of time—typically one to five years—during which the borrower could not refinance or pay off the loan without incurring a one-time cost. Borrowers would be “lock[ed] into [such] loans.” McCoy 14. Defendants claim that this feature was present in approximately 99.5 percent of the loans in the putative class. Tr. at 148. The parties dispute the effects of prepayment penalties on an individual borrower’s likelihood of default. Plaintiffs assert that prepayment penalties increased the risk of default by preventing refinancing, which was a common borrower maneuver to avoid default, id., while Defendants point to authority indicating that—perhaps because such penalties were associated with more favorable interest rates—prepayment penalties did not cause defaults, see, e.g., Sherlund 10; Rose 28. Defendants also note that even Plaintiffs’ expert found only that large prepayment penalties increased the likelihood of default, Report of Martha Courchane (“Courchane”), Dkt. 205, ¶ 60 (citing McCoy Dep. at 171-72), and that Michigan law forbade prepayment penalties that were “large” under McCoy’s definition, id.; see also Tr. at 150.

Because prepayment penalties provided a degree of stability and certainty as to the schedule of repayment, lenders sought to ensure that any pool of loans that they purchased would include a high percentage of loans—often around 75 percent—that included prepayment penalties. See Riddiough ¶100.

C. Secondary Market for Subprime Loans

During the period leading up to the collapse of the housing market, low interest rates and low inflation led to an increasing demand for mortgage loans. Although loan originators mostly sold whole loans, they also securitized a rapidly-increasing percentage of the loans that they originated. Riddiough ¶ 14. Investment banks—including (but not exclusively) Morgan Stanley—would underwrite these offerings. McCoy 28, Shapiro Dep. at 167. Whether securitized by the originators or sold to investment banks that securitized them, most of the subprime loans that were originated during the housing boom were sold to third party financial institutions. Cf. City of Ann Arbor Emps. Ret. Sys. v. Citigroup Mortg. Loan Trust Inc., 703 F.Supp.2d 253, 255 (E.D.N.Y.2010). Banks that bought whole loans typically securitized the loans and sold the components to investors that “rang[ed] from small cities in Norway to large Chinese banks.” McCoy 17.

Plaintiffs assert that Morgan Stanley “was the largest purchaser of whole loans originated by New Century” during the relevant period. Id. at 27 (citation omitted); see also Sugnet Decl. Ex. 23, Shapiro Dep. at 167. While the parties do not agree as to the best method of measuring influence, it is clear that Morgan Stanley purchased many fewer loans in 2005 than in the immediately preceding years (and the following year). See Riddiough ¶¶ 43-47. Defendants ascribe this downturn to competition in the marketplace, asserting that Morgan Stanley was unwilling to pay what its competitors paid for New Century’s loans during that time period. Shapiro Dep. at 48-50, Tr. at 67-69.

“Forward sales” accounted for over 90 percent of the whole loans that New Century sold. Sugnet Decl. Ex. 4, Licata Dep. at 24. In a “forward sale,” an investment bank would agree with New Century on a price for a pool of whole loans (many of which may not yet have been written) that would contain specific characteristics as set forth in a “bid sheet.” Id. at 23-29. The parties would reach an agreement on price either through “open bids,” pursuant to which New Century would reveal the characteristics of a loan pool and allow the banks to bid on the pool, or through “reverse inquiries,” pursuant to which an investment bank would indicate what sort of loans it wanted and how much it was willing to pay for such a pool. Id. at 28-29. The purchasing investment bank and New Century would finalize an agreement and terms; those terms generally set ceilings or floors on certain terms being represented in the loans in the pool (e.g., at least 85 percent of the loans must be adjustable rate loans; no more than 42.5 percent could be no-doeumentation loans). See Sugnet Decl. Exs. 46, 48; Kaplan Dep. at 127.

Because forward sales accounted for such a high proportion of New Century’s transactions, New Century had an incentive to originate loans the terms of which were desirable to the purchasing investment banks. Cf. FCIC 105 (“The definition of a good loan changed from ‘one that pays’ to ‘one that could be sold,’ Patricia Lindsay, formerly a fraud specialist at New Century, told the FCIC.”).

III. Disparate Impact of New Century’s Practices

While it is not the focus of the pending dispute, the parties also disagree as to the effects of the lending practices in which New Century engaged. In a comprehensive regression analysis controlling for borrower characteristics, Plaintiffs’ expert, Professor Ian Ayres, found that “[i]n the Detroit region, the [likelihood] that an African-American borrower would receive a Combined-Risk Loan was 1.347 times greater than that of a non-Hispanic white borrower in the Detroit region with similar characteristics;” this conclusion was “statistically significant at the 99% confidence level.” Ayres ¶ 12. Moreover, Plaintiffs assert that “[t]hese disparities persist when measured only for New Century loans purchased by Morgan Stanley.” Id. ¶ 14. Ayres controlled for fifteen variables “that might provide business justified, nondiscriminatory explanations for the product placement (such as credit score, loan-to-value ratio, loan purpose, and the occupancy and property type),” id. ¶ 73, and determined that—as to New Century-originated loans purchased by Morgan Stanley and as to all New Century-originated loans—African-American borrowers were statistically more likely to receive a Combined-Risk loan.

IV. Procedural History of this Case

Plaintiffs initiated this action in October 2012, and it was assigned to Judge Harold Baer, Jr. In December 2012 Defendants moved to dismiss the case, asserting that the case was barred by the statute of limitations and challenging the viability of Plaintiffs’ theory of the case. Judge Baer granted the motion in part but permitted Plaintiffs’ FHA claims to proceed on the theory that Morgan Stanley’s policies caused New Century to make Combined-Risk loans. Adkins v. Morgan Stanley, No. 12-CV-7667, 2013 WL 3835198, at *3 (S.D.N.Y. July 25, 2013) ("Adkins I”). The case was reassigned to the Undersigned following Judge Baer’s death in 2014.

DISCUSSION

In June 2014, Plaintiffs moved for class certification, seeking to certify a class of “all African-American individuals who, between 2004 and 2007, resided in the Detroit region ... and received Combined-Risk Loans from New Century,” Compl. ¶229, for the purposes of determining liability and “crafting appropriate injunctive and declaratory relief,” id. ¶ 226. Although the Complaint left doubt as to what type of class the Plaintiffs sought to certify, Plaintiffs’ briefing makes clear that they seek to certify a class under Rule 23(b)(3).

I. Class Certification under Rule 23(b)(3)

A. Legal Standard

“ ‘The class action is an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only.’ ” Sykes v. Mel S. Harris & Assocs. LLC, 780 F.3d 70, 79 (2d Cir.2015) (quoting Wal-Mart Stores, Inc. v. Dukes, 564 U.S. -, -, 131 S.Ct. 2541, 2550, 180 L.Ed.2d 374 (2011) (other quotation marks omitted)). “A district court may only certify a class if it determines that each Rule 23 requirement is met.” Levitt v. J.P. Morgan Sec., Inc., 710 F.3d 454, 464 (2d Cir.2013). “The party seeking class certification bears the burden of establishing by a preponderance of the evidence that each of Rule 23’s requirements have been met.” Johnson v. Nextel Commc’ns Inc., 780 F.3d 128, 137 (2d Cir.2015).

A district court may certify a class only if the class meets all of the requirements of Rule 23(a) and the relevant requirements of Rule 23(b)—in this case, the requirements of predominance and superiority set out in Rule 23(b)(3). “Regardless of whether class certification is contested, a court may not certify a putative class unless it has performed a ‘rigorous analysis’ and determined that each of Rule 23’s requirements has been met.” Animal Science Prods. v. Hebei Welcome Pharm. Co. Ltd. (In re Vitamin C Antitrust Litig.), 279 F.R.D. 90, 98 (E.D.N.Y.2012) (quoting Gen. Tel. Co. of SW v. Falcon, 457 U.S. 147, 161, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982)). This “rigorous analysis” sets a higher standard than “a mere pleading standard,” Ohio Public Emps. Ret. Sys. v. Gen. Reinsurance Corp. (In re AIG Sec. Litig.), 689 F.3d 229, 237 (2d Cir.2012) (quotation marks omitted).

A class action is appropriate:

only if (1) the class is so numerous that joinder of all members is impracticable; (2) there are questions of law or fact common to the class; (3) the claims or defenses of the representative parties are typical of the claims or defenses of the class; and (4) the representative parties will fairly and adequately protect the interests of the class.

Fed.R.Civ.P. 23(a). In addition to meeting the requirements of Rule 23(a), to certify a class pursuant to Rule 23(b)(3), a plaintiff must establish that “both (1) ‘questions of law or fact common to class members predominate over any questions affecting only individual members,’ and (2) ‘a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.’ ” Roach v. T.L. Cannon Corp., 778 F.3d 401, 405 (2d Cir.2015) (quoting Fed.R.Civ.P. 23(b)(3)). “ ‘While the text of Rule 23(b)(3) does not exclude from certification cases in which individual damages run high, the Advisory Committee had dominantly in mind vindication of the rights of groups of people who individually would be without effective strength to bring their opponents into court at all.’ ” Sykes, 780 F.3d at 81 (quoting Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997)). Although Rule 23(b)(3) “ ‘does not require a plaintiff seeking class certification to prove that each element of her claim is susceptible to elasswide proof,’ ” id. (quoting Amgen Inc. v. Conn. Ret. Plans & Trust Funds, 568 U.S. -,-, 133 S.Ct. 1184, 1196, 185 L.Ed.2d 308 (2013) (alterations omitted, emphasis in original)), it “imposes a ‘far more demanding’ inquiry into the common issues which serve as the basis for class certification” than does Rule 23(a), id. (quoting Amchem, 521 U.S. at 623-24, 117 S.Ct. 2231).

Determining whether to certify a class under Rule 23 will frequently require some evaluation of the merits of the plaintiffs’ underlying claim. Comcast Corp. v. Behrend, 569 U.S.-,-, 133 S.Ct. 1426, 1432, 185 L.Ed.2d 515 (2013). “But the office of a Rule 23(b)(3) certification ruling is not to adjudicate the ease; rather, it is to select the ‘method’ best suited to adjudication of the controversy ‘fairly and efficiently.’ ” Amgen, 133 S.Ct. at 1191; see also Fezzani v. Bear, Stearns & Co., 777 F.3d 566, 570 (2d Cir. 2015) (“ ‘Merits questions may be considered to the extent—but only to the extent-that they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied.’ ”) (quoting Amgen, 133 S.Ct. at 1194-95).

B. Plaintiffs’ Motion for Class Certification Is Denied

Plaintiffs’ motion to certify a class of African-American borrowers in the Detroit area who received Combined-Risk loans suffers from Plaintiffs’ delineation of such loans. Plaintiffs’ definition of a Combined-Risk loan is essentially a fast-food-menu approach (requiring a high-cost mortgage as defined in the HMDA from Column A and two or more risk factors from Column B). This definition yields 247 different potential combinations of the factors identified by Plaintiffs (the “risk factors”), 33 of which actually existed in at least one loan made by New Century to a putative class member, see Tr. at 55. The risk factors on which Plaintiff has chosen to focus are clearly distinct and each affects borrowers differently—and the manner in which each risk factor affects a borrower is context-dependent. Moreover, the “context” that informs the harm (or benefit) caused by a particular risk factor includes the presence or absence of other risk factors. The reality of how the risk factors combine means that each of the 33 combinations of risk factors that actually appeared in a Combined-Risk loan requires a separate analysis with respect to the merits of Plaintiffs’ action. The multiplicity of categories of loans presents a number of obstacles to class certification, not least of which is the difficulty in identifying plaintiffs whose claims are “typical” of the 33 differently-situated subsets of borrowers that exist in Plaintiffs’ class and the difficulty in identifying common questions that predominate over individualized inquiries.¹⁷ The Court also doubts “whether ‘law or fact questions common to the class predominate over questions affecting individual members.’ ” Fezzani, 777 F.3d at 569 (quoting Miles v. Merrill Lynch & Co. (In re Initial Public Offerings Sec. Litig.), 471 F.3d 24, 32 (2d Cir.2006)). Ultimately, the Court has concluded that a class action is inappropriate.

1. Rule 23(a) Factors

a. Plaintiffs Are Sufficiently Numerous

“To meet the requirements of Rule 23(a)(1), the class must be so large that joinder of all members would be impracticable.” McIntire v. China MediaExpress Holdings, 38 F.Supp.3d 415, 423 (S.D.N.Y.2014) (quotation marks omitted). “Numerosity is presumed for classes larger than forty members.” Penn. Public Sch. Emps. Ret. Sys. v. Morgan Stanley & Co., 772 F.3d 111, 120 (2d Cir.2014). This requirement, however, “is not strictly mathematical but must take into account the context of the particular case, in particular whether a class is superior to joinder based on other relevant factors including: (i) judicial economy, (ii) geographic dispersion, (iii) the financial resources of class members, (iv) their ability to sue separately, and (v) requests for injunctive relief that would involve future class members.” Id. In this case, Plaintiffs have proven that the class will likely exceed 4,600 individuals with minimal financial resources; the numerosity requirement is clearly satisfied.

b. There May Be Common Questions of Law and Fact

The commonality “element requires the existence of both at least one question common to the class, and also that a class action ‘has the capacity to generate common answers apt to drive the resolution of the litigation.’ ” Gulino v. Bd. of Educ. of City Sch. Dist. of City of N.Y., No. 96-CV-8414(KMW), 2013 WL 4647190, at *6 (S.D.N.Y. Aug. 29, 2013) (quoting Dukes, 131 S.Ct. at 2556) (alteration omitted, emphasis in original), aff'd 555 Fed.Appx. 37 (2d Cir.2014). Commonality requires a “‘common contention [that] must be of such a nature that it is capable of classwide resolution— which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke.’ ” Id. (quoting Dukes, 131 S.Ct. at 2551) (emphasis in Sykes). Nevertheless, “Rule 23(a)’s requirement of commonality is a low bar, and courts have generally given it a ‘permissive application.’” Brown v. Am. Honda (In re New Motor Vehicles Can. Export Antitrust Litig.), 522 F.3d 6, 19 (1st Cir.2008) (quoting 7A Charles Alan Wright, Arthur R. Miller, Mary Kay Kane, Federal Practice and Procedure § 1763, at 221 (3d ed.2005)).

In this case, Plaintiffs have identified a number of material questions that they believe are susceptible to generalized answers. Questions pertaining to Morgan Stanley’s influence in shaping New Century’s lending decisions, for example, may be susceptible to resolution as to the class as a whole.¹⁸ This does not mean that Morgan Stanley will be liable to the entire class, but it means that the questions whether and to what extent Morgan Stanley controlled New Century and therefore is liable for its loans may be susceptible to classwide resolution. Plaintiffs’ theory of Morgan Stanley’s liability is essentially a market-based approach, analogous to the Basic presumption in securities fraud actions; their claims are thus likely to rise or fall as a class. Cf. Halliburton v. Erica P. John Fund, Inc., 573 U.S. -, -, 134 S.Ct. 2398, 2416, 189 L.Ed.2d 339 (2014). Similarly, Plaintiffs believe that any disparate impact of New Century’s lending practices on African-Americans is likely to be subject to one (or several) proofs across the class, and the disparate impact analysis will not hinge on the varying circumstances pertaining to individual loans. See Ayres ¶¶ 12, 82. The existence of such a large-scale disparate impact study—irrespective of whether its results are ultimately accepted by the factfinder-can be a powerful argument in support of commonality. See, e.g., Stockwell v. City & Cnty. of S.F., 749 F.3d 1107, 1115 (9th Cir.2014) (“If those effects amount to a disparate impact ... it will be so for all class members or for none; their claims rise and fall together.”).

Defendants do not address the question of commonality head-on; they argue instead that any common questions do not “predominate” over the individual questions at issue in the case; the Court agrees, as discussed in Part II.B.2.a. The Court addresses commonality further in the context of its predominance discussion.

c. Plaintiffs’ Claims Are Not Typical of the Putative Class ¹⁹

“Typicality requires that the claims or defenses of the class representatives be typical of the claims or defenses of the class members.” Brown v. Kelly, 609 F.3d 467, 475 (2d Cir.2010) (citing Fed. R.Civ.P. 23(a)(3)). “Rule 23(a)(3) is satisfied when ‘the claims of the class representatives are typical of those of the class, and when each class member’s claim arises from the same course of events, and each class member makes similar legal arguments to prove the defendant’s liability.’ ” Vincent v. Money Store, 304 F.R.D. 446, 455 (S.D.N.Y.2015) (quoting Marisol A. v. Giuliani, 126 F.3d 372, 376 (2d Cir.1997)) (alteration omitted). “The central feature for typicality is that plaintiffs assert ‘that defendants committed the same wrongful acts in the same manner, against all members of the class,’ and the court looks ‘not at the plaintiffs’ behavior, but rather at the defendant’s actions.’ ” Fort Worth Emps. Ret. Fund v. J.P. Morgan Chase & Co., 301 F.R.D. 116, 132 (S.D.N.Y.2014) (quoting Tsereteli v. Residential Asset Securitization Trust 2006-A8, 283 F.R.D. 199, 208 (S.D.N.Y.2012)).

Defendants assert two bases for their claim that the five representative plaintiffs are not typical of the class. First, they assert that each of the plaintiffs is subject to a particularly strong defense.²⁰ The Court is unpersuaded. Certainly the typicality requirement is intended to “ensure that ‘the class representative is not subject to a unique defense which could potentially become the focus of the litigation.’ ” Kelen v. World Fin. Network Nat’l Bank, 302 F.R.D. 56, 65 (S.D.N.Y.2014) (quoting Steinberg v. Nationwide Mut. Ins. Co., 224 F.R.D. 67, 72 (E.D.N.Y.2004)). But “[t]here is little risk here that ‘[the] putative class representative[s] [are] subject to unique defenses which threaten to become the focus of the litigation.’ ” In re Smith Barney Transfer Agent Litig., 290 F.R.D. 42, 46 (S.D.N.Y.2013) (quoting Baffa v. Donaldson, Lufkin & Jenrette Sec. Corp., 222 F.3d 52, 59 (2d Cir.2000)); see, e.g., Fort Worth, 301 F.R.D. at 133-34 (rejecting defendants’ arguments that the class representatives were subject to unique defenses because of their exposure “to public information that would have had the same effect on all class members”).

More persuasively, Defendants argue that the representative plaintiffs reflect only a small subset of the risk factors. Plaintiffs have not established at this stage that Defendants’ conduct “harmed each class member in the same way.” Wallace v. IntraLinks, 302 F.R.D. 310, 315 (S.D.N.Y.2014). Plaintiffs, who had 3/27 ARM loans with prepayment penalties (both characteristics of loans that Morgan Stanley affirmatively sought, see Riddiough ¶ 100), face drastically different proof challenges than Plaintiffs who received stated income loans with a 55 percent DTI (features that Morgan Stanley would have preferred to avoid). The proof required for the first group might focus largely on whether making a loan with favorable terms for three years constitutes a sufficient harm to confer standing, particularly because the class is defined without regard to the outcome of the mortgage (i.e., whether the borrower defaulted, refinanced or sold before the expiration of the three year introductory rate, or paid the loan on schedule despite the more costly payments after the first interest rate adjustment). The second group, conversely, would likely seek to show that Morgan Stanley’s allegedly insatiable appetite for subprime loans caused New Century to originate no-documentation loans with high DTI ratios. The challenges to proving such a claim are different (both in scale and in type) from those plaguing the first group. Morgan Stanley might have purchased loans that led in some convoluted way to New Century’s making such loans, but Morgan Stanley did not affirmatively seek them. Moreover, the harmfulness of the stated income risk factor in particular hinges entirely on whether the borrower misrepresented his or her income. It is therefore difficult to conclude that this is a case in which ‘“each class member’s claim arises from the same course of events and each class member makes similar legal arguments to prove the defendant[s’] liability.’ ” Fort Worth, 301 F.R.D. at 132 (quoting Loftin v. Bande (In re Flag Telecom Holdings, Ltd. Sec. Litig.), 574 F.3d 29, 35 (2d Cir.2009)).

Even if the complicated class described by Plaintiffs could be appropriately certified, the putative class representatives in this case do not represent a broad enough swath of the possible risk-factor permutations to be characterized as “typical” of the putative class. The representative plaintiffs “all had ARM loans, with prepayment penalties, which may have lowered the interest rate on their loans from what they might otherwise have received.” Courchane ¶ 64. McCoy had a stated income loan, and Young had a balloon payment. Id. Pettway had no other combined risk factors. Id. Adkins and Williams had LTV ratios above 90 percent. Compl. ¶¶ 134, 148. “None of the named Plaintiffs had interest only [loans], Option ARM loans, or loans with DTI ratios above 55%.” Cour-chane ¶ 64. Put more succinctly, the representative plaintiffs’ loans contain only four of the 33 permutations of risk factors that are known to exist within the putative class. Two or three²¹ of the risk factors are entirely unaccounted-for within the representative plaintiffs. Plaintiffs have not established that, under these circumstances, they are “typical” of a putative class that includes borrowers with interest-only loans with high DTI ratios but none of the other risk factors.

This is not a case in which Plaintiffs’ “ ‘injuries derive from a unitary course of conduct by a single system.’” Smith Barney, 290 F.R.D. at 46 (quoting Marisol A., 126 F.3d at 377). Plaintiffs did not all receive identical letters from a debt collector, cf. Kalkstein v. Collecto, Inc., 304 F.R.D. 114, 121 (E.D.N.Y.2015), or invest in the same security, cf. McIntire, 38 F.Supp.3d at 424. Instead, this is a ease in which differently-situated plaintiffs assert that numerous different Morgan Stanley policies drove the decisions of numerous third parties, leading the plaintiff class to suffer a racially disparate adverse impact. Cf. Bolden v. Walsh Const. Co., 688 F.3d 893, 898 (7th Cir.2012) (declining to certify a class in a race discrimination case against a large company in which “[t]he 12 plaintiffs did not experience the working conditions at all 262 sites either individually or collectively, and a given plaintiffs bad experience with one of the five supervisors ... named does not present any question about the conduct of ... many other superintendents and foremen.”). It is not immediately apparent that “by prosecuting [their] own case[s], the named plaintiff[s] ‘[would] simultaneously advance[ ] the interests of the absent class members.’” Vitamin C Antitrust Litig., 279 F.R.D. at 105 (quoting 1 Joseph M. McLaughlin, McLaughlin on Class Actions § 4:16 (8th ed.2011)). In short, the Court lacks “a concrete, non-speculative basis” on which to conclude that determining defendants’ liability to the representative plaintiffs would resolve defendants’ liability to the broadly-defined class. Stream Sicav v. Wang, No. 12-CV-6682(PAE), 2015 WL 268855, at *5 (S.D.N.Y. Jan. 21, 2015). Accordingly, Plaintiffs have not carried their burden of showing that their claims are typical of the claims of the class.

d. Named Plaintiffs Are Adequate Representatives of the Putative Class

Rule “23(a)(4) requires that in a class action, ‘the interests of the class’ must be ‘fairly and adequately protected.’ ” Charron v. Wiener, 731 F.3d 241, 249 (2d Cir.2013) (quoting Fed.R.Civ.P. 23(a)(4)) (alteration omitted). “Determination of adequacy typically ‘entails inquiry as to whether: 1) plaintiffs interests are antagonistic to the interest of other members of the class and 2) plaintiffs attorneys are qualified, experienced and able to conduct the litigation.’” Cordes & Co. Fin. Servs. v. A.G. Edwards & Sons, Inc., 502 F.3d 91, 99 (2d Cir.2007) (quoting Baffa, 222 F.3d at 60).²² “ ‘A class representative must be part of the class and possess the same interest and suffer the same injury as the class members.’” Amchem, 521 U.S. at 625, 117 S.Ct. 2231 (quoting E. Tex. Motor Freight Sys., Inc. v. Rodriguez, 431 U.S. 395, 403, 97 S.Ct. 1891, 52 L.Ed.2d 453 (1977) (alteration and other quotation marks omitted)). The fact that plaintiffs might seek relief pursuant to different statutes—for example, when some members of the class are time-barred from asserting some otherwise-relevant claims'—does not mean the representative is inadequate if the gestalt of the plaintiffs’ claims is the same. Sykes, 780 F.3d at 89-90. Moreover, Courts are “ ‘wary of [defendants’] efforts to defeat representation of a class on grounds of inadequacy' where, as here, the effect of an inadequacy finding would be to ‘eliminate any class representation.’ ” Vitamin C Antitrust Litig., 279 F.R.D. at 100 (quoting Kline v. Wolf, 702 F.2d 400, 402 (2d Cir.1983)).

In this case Defendants advance two arguments why Plaintiffs are inadequate class representatives. First, they argue that Plaintiffs seek to collect damages on their own behalf while only seeking disgorgement on behalf of the putative class. Defendants allege that res judicata principles that prohibit claim-splitting would foreclose any subsequent recovery by absent class members. See, e.g., In re Methyl Tertiary Butyl Ether (“MTBE”) Prods. Liability Litig., 209 F.R.D. 323, 339 (S.D.N.Y.2002); Reppert v. Manin Lumber & Cedar Co., 359 F.3d 53, 56 (1st Cir.2004). There are, however, “several exceptions to the claim-splitting principle, one of which arises when ‘the court in the first action has expressly reserved the plaintiffs right to maintain the second action.’ ” Vitamin C Antitrust Litig., 279 F.R.D. at 115-16 (quoting Restatement (2d) of Judgments § 26(1)(b) (1982) (alteration omitted)). Moreover, “courts generally allow plaintiffs in class actions to sue for injunctive relief on behalf of the class and then bring damages claims in subsequent individual actions.” MTBE, 209 F.R.D. at 339 (collecting cases).²³ Because disgorgement, like injunctive relief, is an equitable remedy, Plaintiffs contend that the remedy they seek is analogous to the injunctive relief sought in Vitamin C.²⁴

It is not clear that disgorgement is appropriately analogized to injunctive relief in a class action context because disgorgement, like damages, must be sought under Rule 23(b)(3), while injunctive and declaratory relief are both available under Rule 23(b)(2). See Randall v. Rolls-Royce Corp., 637 F.3d 818, 820 (7th Cir.2011). Unlike damages, however, “ ‘the primary purpose of disgorgement orders is to deter violations of the laws by depriving violators of their ill-gotten gains.’” F.T.C. v. Bronson Partners, LLC, 654 F.3d 359, 373 (2d Cir.2011) (quoting S.E.C. v. Fischbach Corp., 133 F.3d 170, 175 (2d Cir.1997) (alteration omitted)). Because disgorgement serves a different purpose from damages, principles of res judicata would probably not bar any plaintiff from asserting a separate claim for damages. Of course, “[a] judgment in favor of either side is conclusive in a subsequent action between them on any issue actually litigated and determined, if its determination was essential to that judgment.” Cooper v. Fed. Reserve Bank of Richmond, 467 U.S. 867, 874, 104 S.Ct. 2794, 81 L.Ed.2d 718 (1984). Accordingly, while absent class members risk losing on the merits and being precluded from separately asserting, for example, a claim that Morgan Stanley is culpable for their loans, if the class were to win on the merits they would be able to pursue separate claims for damages. Because the putative class representatives have not needlessly compromised the ability of the class to secure its relief, Defendants’ argument that they are inadequate class representatives on that basis is unpersuasive.

Defendants’ second argument relative to Plaintiffs’ adequacy hinges on the possibility that disgorgement would be available for some, but not all, members of Plaintiffs’ proposed class. While this dispute might speak to the questions of predominance and superiority of the class format, it does not speak to the adequacy of the putative class representatives. See Amchem, 521 U.S. at 625-26, 117 S.Ct. 2231 (“The adequacy inquiry under Rule 23(a)(4) serves to uncover conflicts of interest between named parties and the class they seek to represent.”). Defendants have not shown that the “ ‘plaintiffs’] interests are antagonistic to the interest of other members of the class,’” even if some plaintiffs may ultimately advance claims that are unavailable to other members of the class. Sykes, 780 F.3d at 90 (quoting Baffa, 222 F.3d at 60). The named representatives’ “interests in maximizing the class recovery” are aligned with those of the class. Kalkstein, 304 F.R.D. at 121. Accordingly, the putative class representatives would be adequate class representatives of the proposed class.

2. Rule 23(b)(3) Factors

“Rule 23(b)(3) imposes two additional burdens on plaintiffs attempting to proceed by class action, namely, predominance and superiority.” Sykes, 780 F.3d at 81. “Predominance is satisfied ‘if resolution of some of the legal or factual questions that qualify each class member’s case as a genuine controversy can be achieved through generalized proof, and if these particular issues are more substantial than the issues subject only to individualized proof.’ ” Roach, 778 F.3d at 405 (quoting Catholic Healthcare W. v. U.S. Foodservice Inc. (In re U.S. Foodservice Inc. Pricing Litig.), 729 F.3d 108, 118 (2d Cir.2013)). “The superiority requirement reflects the goal of class actions to achieve economies of time, effort and expense, and promote uniformity of decision as to persons similarly situated, without sacrificing procedural fairness.” N.J. Carpenters Health Fund v. DLJ Mortg. Cap., Inc., No. 08-CV-5653(PAC), 2014 WL 1013835, at *11 (S.D.N.Y. Mar. 17, 2014) (quotation marks and alterations omitted). Four other factors—individual control of litigation, prior actions involving the parties, the desirability of the forum, and manageability—should also be considered in making these determinations. Fed.R.Civ.P. 23(b)(3); Sykes, 780 F.3d at 82. Although structurally these factors “apply to both predominance and superiority, they more clearly implicate the superiority inquiry.” Sykes, 780 F.3d at 82 (collecting cases); but see AIG Sec. Litig., 689 F.3d at 242 (district court erred by “view[ing] manageability and predominance as two independent inquiries”); Seijas v. Republic of Arg., 606 F.3d 53, 58 (2d Cir.2010) (“[Wjhether the court is likely to face difficulties managing a class action bears on whether the proposed class satisfies the predominance and superiority requirements.”). Of the four factors that inform the Court’s predominance and superiority inquiries, “manageability ‘is, by [ ] far, the most critical concern in determining whether a class action is a superior means of adjudication.’ ” Sykes, 780 F.3d at 82 (quoting 2 William B. Rubenstein, Newberg on Class Actions § 4.72 (5th ed. West 2014)).

a. Common Questions Do Not Predominate over Individual Questions

To determine whether Plaintiffs have met their burden on predominance, a court “must assess (1) the ‘elements of the claims and defenses to be litigated’; and (2) ‘whether generalized evidence could be offered to prove those elements on a class-wide basis or whether individualized proof will be needed to establish each class member’s entitlement to relief.” Johnson, 780 F.3d at 138 (quoting 1 Joseph M. McLaughlin, McLaughlin on Class Actions § 5:23 (11th ed.2014)). The mere existence of individual inquiries does not doom a potential class; Rule 23(b)(3) “anticipates the existence of individual issues.” Sykes, 780 F.3d at 87. What will doom a class, however, is the determination that common issues will be overwhelmed by individual issues. While Rule 23(a)(2) requires commonality, the predominance requirement of Rule 23(b)(3) “imposes a ‘far more demanding’ inquiry.” Id. (quoting Amchem, 521 U.S. at 623-24, 117 S.Ct. 2231). In order to satisfy the predominance requirement, the proposed class must be “‘sufficiently cohesive to warrant adjudication by representation.’ ” AIG Sec. Litig., 689 F.3d at 239-40 (quoting Amchem, 521 U.S. at 623, 117 S.Ct. 2231).

The predominance inquiry frequently hinges on whether elements of each class member’s ease can be proven through generalized proof, and whether the issues that can be so proven are more substantial than the issues subject only to individualized proof. UFCW Local 1776 v. Eli Lilly & Co., 620 F.3d 121, 131 (2d Cir.2010). If, in order to prove causation or liability, a trial will need to address the facts of each individual claim, then the Plaintiffs have not carried their burden. See, e.g., Bolden, 688 F.3d at 896; Mata v. Citimortgage, Inc., No. 10-CV-9167(DSF), 2012 WL 7985175, at *2 (C.D.Cal. July 20, 2012).

In order to determine predominance, the Court must start with the elements of the underlying cause of action. Erica P. John Fund, Inc. v. Halliburton Co., 563 U.S. 804, -, 131 S.Ct. 2179, 2184, 180 L.Ed.2d 24 (2011) (internal quotation marks omitted). To establish a prima facie case of discrimination based on disparate impact under the FHA,²⁵ the plaintiff must prove that the defendant is covered by the FHA,²⁶ that the defendant engaged in an outwardly neutral practice, and that the defendant’s facially neutral act or practice produced “a significantly adverse or disproportionate impact on persons of a particular type.” Tsombanidis v. W. Haven Fire Dep’t, 352 F.3d 565, 574-75 (2d Cir.2003) (quotation marks and emphasis omitted); Keller v. City of Fremont, 719 F.3d 931, 949 (8th Cir.2013) (identifying two other elements of a disparate impact claim under the FHA as whether the defendant can show that the act or practice had a “manifest relationship to a legitimate, non-discriminatory policy objective and was necessary to [attain] that objective, and if so whether plaintiffs can show that a viable alternative means was available to achieve the legitimate policy objective without discriminatory effects”) (quotation marks omitted).

Common issues do not “predominate” over individual issues in this case for a number of reasons. First, although Plaintiffs attempt to cast this case as a straightforward one—did Morgan Stanley exert a “singular influence” over New Century and thereby cause it to make “toxic loans” that had a detrimental effect on borrowers who were disproportionately African-American— no element of their theory is susceptible to one generalized proof. The amount of influence that Morgan Stanley exerted over New Century, for example, requires proof along a number of different axes. The Court need not examine the merits of Plaintiffs’ causation theory to note that the extent to which Morgan Stanley caused New Century’s lending practices could vary based on the risk factors present in a particular loan²⁷ and Morgan Stanley’s financial involvement in the particular loan.²⁸ Plaintiffs’ claim that Morgan Stanley’s lax due diligence affirmatively caused New Century to issue Combined-Risk loans (an illogical proposition in the first place) may be even less persuasive in the context of loans that were caught and rejected by Morgan Stanley’s due diligence process. Similarly, putative class members whose loans were made as “exceptions” to New Century’s underwriting guidelines would need to prove that the basis for the exception was Morgan Stanley’s preference, not the borrowers’; these borrowers would be differently-situated from those who obtained loans pursuant to the guidelines that allegedly “baked in” Morgan Stanley’s preferences.

These concerns are not akin to the existence of individualized defenses—these are real differences among class members that result in arguments’ being available to some on issues of causation that will be unavailable to others. In short, the question of causation is simply not subject to a elasswide proof. This is itself sufficient to establish that common questions do not predominate. See, e.g., IPO Sec. Litig., 471 F.3d at 43 (reversing class certification because questions related to plaintiffs’ knowledge of allegedly concealed information could not be determined on a classwide basis).

But other aspects of Plaintiffs’ case are also not subject to classwide proof. Plaintiffs’ theory that Morgan Stanley exerted “singular influence” over New Century, for example, is stronger with respect to borrowers who obtained loans in 2004, when Morgan Stanley may well have been the proverbial 800-pound gorilla in the room, rather than with respect to borrowers who obtained loans in the first half of 2005, during which time Morgan Stanley did not win a single bid to purchase bulk loans from New Century. See Shapiro Dep. at 48-49. Insofar as the basis for Plaintiffs’ claim that Morgan Stanley dictated the terms of New Century’s loans is predicated largely on Morgan Stanley’s position in the marketplace, there is ample reason to conclude that the proof will be different relative to a loan written during a five-month period during which New Century sold no loans to Morgan Stanley than it will be relative to a loan written during a period when Morgan Stanley accounted for almost half of New Century’s business. Cf. Ridd-iough Figure 7. Thus, questions of Morgan Stanley’s influence over New Century are not “common” across time.

Nor is the claim that Combined-Risk loans (as opposed to, say, high-cost loans without any of the risk factors) are per se harmful susceptible to class-wide proof. Even assuming, arguendo, that Professor McCoy’s testimony that each risk factor in isolation increases the risk of default is both admissible and persuasive to the fact finder, Defendants have nevertheless established that the effect of any given risk factor depends on the presence vel non of other risk factors,²⁹ and on externalities, such as the general real estate market and oil prices, see Sherlund 11. While this does not mean that Defendants would necessarily prevail on the merits, it does mean that each of the 33 different categories of Combined-Risk loan included in Plaintiffs’ proposed class would need to establish the harmfulness of that particular combination of risk factors.³⁰ Each potential combination of risk factors might also yield different results with respect to a disparate impact analysis.³¹

Furthermore, each combination of factors might need a separate study to determine the likelihood that loans of that type were caused by Morgan Stanley’s policies or practices. See Riddiough 62 Figure 13 (depicting Morgan Stanley’s share of New Century’s loans with various combinations of risk factors, ranging from loans with DTI over 55 percent and LTV ratios over 90 percent, in which Morgan Stanley had a minimal share, to those loans containing balloon payments and prepayment penalties, of which Morgan Stanley had a considerably greater share). Morgan Stanley’s relationship with New Century evolved over time, and the putative class varies wildly with regard to the types of loans and the various roles that Morgan Stanley played in each. A large number of Morgan Stanley’s “policies and practices” affected the putative class in different ways. Cf. D.L. v. Dist. of Columbia, 713 F.3d 120, 127 (D.C.Cir.2013). The need to conduct separate analyses for hundreds of combinations simply to determine whether Morgan Stanley caused harm to the Plaintiff is enough to cause individual issues to predominate over common issues.³²

The Court recognizes that Plaintiffs do not need to be identically situated in order for their claims to be susceptible to generalized proof that predominates over individual questions. See, e.g., Anwar v. Fairfield Greenwich Ltd., 306 F.R.D. 134, 145 (S.D.N.Y.2015) (finding that “common issues predominate^]” “even assuming Defendants’ claims that certain communications to class members may not have been uniform” because they were “uniformly misleading”) (internal quotation marks and citations omitted); Public Emps. Ret. Sys. of Miss. v. Goldman Sachs Grp., 280 F.R.D. 130, 139 (S.D.N.Y.2012) (common issues predominated despite the existence of different certificates because the misrepresentations on which each investor was alleged to rely were the same; “questions of materiality and loss causation [were thus] subject to objective standards and generalized proof’). But Plaintiffs’ case would involve many issues that are “generalized” only as to a small subsection of the proposed class; such inquiries are simply not subject to a common, classwide resolution. The existence of so many different groups of Plaintiffs, differentiated by the nature of their loans and by the role that Morgan Stanley played vis-á-vis those loans is fatal to Plaintiffs’ claim of predominance.

The final blow to Plaintiffs’ claim that common questions predominate is the role of New Century’s brokers in the origination process. When considering whether discretion afforded to third parties defeats the existence of common questions of law or fact, the court’s inquiry must focus on whether “there was a common and unlawful mode by which the [parties] exercised their discretion.” Rodriguez v. Nat’l City Bank, 726 F.3d 372, 385 (3d Cir.2013); see Dukes, 131 S.Ct. at 2554.³³ “On this point, Dukes is clear: class members must unite acts of discretion under a single policy or practice, or through a single mode of exercising discretion, and the mere presence of a range within which acts of discretion take place will not suffice to establish commonality.” Miller v. Countrywide Bank, N.A. (In re Countrywide Fin. Corp. Mortg. Lending Practices Litig.), 708 F.3d 704, 708 (6th Cir.2013).³⁴ Accordingly, when Merrill Lynch maintained policies that permitted brokers in the same office to form “teams” based on whatever criteria they chose and assigned accounts within an office to high performers-thereby permitting them to earn more money and to advance in the company—the Seventh Circuit held that plaintiffs had asserted common questions as to whether those policies caused a disparate impact above and beyond what would exist absent those policies. McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, 672 F.3d 482, 488-90 (7th Cir.2012). The Court found there to be a logical tie between the challenged policy (which exacerbated inequalities existing in the status quo) and the racial disparity that resulted. Id. at 489; see also Calibuso v. Bank of Am. Corp., 893 F.Supp.2d 374, 389 (E.D.N.Y.2012).

On the other hand, when a claim rests on the discretion of a loan officer, even if the loan officer is making loans against the backdrop of facially race-neutral underwriting criteria, courts have held that their discretion prevents a finding of commonality or predominance.³⁵ In Rodriguez, for example, the Third Circuit held:

Even if Plaintiffs had succeeded in controlling for every objective credit-related variable ... the regression analyses do not even purport to control for individual, subjective considerations. A loan officer may have set an individual borrower’s interest rate and fees based on any number of nondiscriminatory reasons, such as whether the mortgage loans were intended to benefit other family members who were not borrowers, whether borrowers misrepresented their income or assets, whether borrowers were seeking or had previously been given favorable loan-to-value terms not warranted by their credit status, whether the loans were part of a beneficial debt consolidation, or even concerns the loan officer may have had at the time for the financial institution irrespective of the borrower.

726 F.3d at 384.

When discretion is concentrated in “upper-level, top-management personnel,” that discretion may not defeat commonality, while “the exercise of discretion by lower-level employees” generally will. Scott v. Family Dollar Stores, Inc., 733 F.3d 105, 114 (4th Cir.2013). In this case, the “discretion” at issue was exercised by mid-level New Century employees and myriad independent brokers who made independent decisions in selecting terms for a particular loan. Finley Decl. ¶4, Dkt. 209. Moreover, every New Century loan (including those written by brokers) was underwritten by a New Century employee who exercised discretion when the loan did not adhere to New Century’s (unchallenged) guidelines.³⁶ Id. ¶ 8; McKay Dep. at 96-97; cf. Rodriguez, 726 F.3d at 384. The connection between Morgan Stanley’s desires (seeking loans with particular features) and the disparate impact on the African-American community in Detroit could—consistent with Plaintiffs’ theory of the case—be entirely explained by brokers’ exercise of discretion in writing loans.³⁷

The existence of some common questions of fact does not establish that such questions “predominate.” In this case, the number of meaningful variations among the putative class would require mini-trials as to many, maybe hundreds, of groups of borrowers; these variations prevent a finding that class-wide issues predominate.

b. A Class Action Is Not a Superior Vehicle for This Action

Plaintiffs seeking to certify a class under Rule 23(b)(3) must also establish “that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” Fed.R.Civ.P. 23(b)(3). Rule 23(b)(3) lists four factors that “clearly implicate the superiority inquiry.” Sykes, 780 F.3d at 82. These factors include:

(A) the class members’ interests in individually controlling the prosecution or defense of separate actions;

(B) the extent and nature of any litigation concerning the controversy already begun by or against class members;

(C) the desirability or undesirability of concentrating the litigation of the claims in the particular forum; and

(D) the likely difficulties in managing a class action.

Id. at 81 (quoting Fed.R.Civ.P. 23(b)(3)). Of the four factors, the most critical issue in determining whether a class action is a superior means of adjudication is manageability. Id. at 82. “[M]anageability is an issue peculiarly within a district court’s discretion.” Seijas, 606 F.3d at 58 (citations omitted).

In this case, no other litigation is pending, and Plaintiffs have sufficiently demonstrated that the Southern District of New York, where Morgan Stanley is based, is a desirable forum. Also on Plaintiffs’ side of the ledger is the impracticability of expecting thousands of class members (who are unlikely to be wealthy or sophisticated) to pursue expensive litigation requiring expert statistical analyses tending to show a disparate impact. “In such circumstances, the class action device is frequently superior to individual actions.” Id.; see also U.S. Foodservice Pricing, 729 F.3d at 130 (“Rule_ 23(b)(3) class actions can be superior ... where the costs of bringing individual actions outweigh the expected recovery.”); Kalkstein, 304 F.R.D. at 123. But the determination of superiority must look to the evidence on the record and to the class that plaintiffs actually seek to certify. See Parker v. Time Warner Entm’t Co., L.P., 331 F.3d 13, 22 (2d Cir.2003).

The class that Plaintiffs seek to certify is unmanageable. As noted previously, Plaintiffs’ proposed algorithm (high cost plus two or more risk factors) yields 33 permutations that actually appeared in at least one borrower’s loan. The various factors have different effects on borrowers and lenders in combination than they have in isolation; that reality would require the factfinder to consider separately the way that each unique permutation affected the borrowers and the extent to which Morgan Stanley caused loans to be written with that particular combination of risk factors. The various roles that Morgan Stanley (and other banks) played with regards to different categories of loans at different periods of time will also require separate factfinding as to the banks’ relative culpability in different contexts. Cf. UFCW Local 1776, 620 F.3d at 135 (describing “the unworkable complexity of joining as a single class of plaintiffs some individuals who [plaintiffs] argue should never have been prescribed [the drug at issue] and some individuals who they argue were properly prescribed [the drug], but paid too much for it.”). When there is no uniform trial that could address the discrete issues presented, a case “fails the predominance and superiority criteria of Rule 23(b)(3).” Johnson, 780 F.3d at 140. Such is the case here; accordingly, Plaintiffs’ Motion for Class Certification is DENIED.

C. No Alternative Class Would Cure the Defects Identified in Plaintiffs’ Motion

During oral argument on the class certification motion, Plaintiffs for the first time asserted that “at minimum ... there are sufficient common questions with common evidence as to the loans that Morgan Stanley bought.” Tr. at 26-27. The Court is not persuaded. Plaintiffs’ case is certainly stronger on the merits for that set of loans, see 42 U.S.C. § 3605(b)—indeed, that is one of the many differences within Plaintiffs’ proposed class that undercuts class certification. But even if Plaintiffs might more easily prove Morgan Stanley’s culpability for disparate impact as to such a plaintiff class, they would still need to overcome many of the other impediments to class certification previously discussed. Notably, Plaintiffs would still need to overcome the challenges presented by the complex Combined-Risk definition, which would still yield class members whose loans do not appear to be inherently harmful and class members whose loans have only Combined-Risk factors that Morgan Stanley did not seek. Such a class would still include individuals whose loans went through vastly different processes (including some whose loans may have initially been placed on a different bank’s warehouse line, or been made with an eye towards a different bank’s forward sale, before being included in a pool that Morgan Stanley purchased).

Finally, in contrast to the late-in-the-game suggestion that the Court consider certifying a different class than the one proposed, Plaintiffs’ theory of the ease has been consistent since it was filed three years ago: Morgan Stanley’s preferences dictated New Century’s behavior; New Century wrote loans based on Morgan Stanley’s preferred terms even when it did not intend to sell the loan to Morgan Stanley; and, therefore, the damage wrought by New Century was caused by Morgan Stanley. That theory is a necessary component of Plaintiffs’ class definition, which was included in their initial pleadings in this action. See Compl. ¶ 231(a)-(g) (identifying the “common question” of “whether Morgan Stanley’s policies with respect to purchasing New Century loans for securitization included requirements for loans with high-risk features”). To change the theory now would unfairly prejudice Defendants, who have spent considerable time and money litigating Plaintiffs’ initial theory. While the evidence on which Plaintiffs would rely may, as they asserted at oral argument, already be in the record, none of the briefing has focused on the alternative theory, and Defendants have not been given an opportunity to develop or produce evidence regarding Plaintiffs’ newly-proposed class.

II. Exclusion of Plaintiffs’ Experts

In support of their motions, Plaintiffs submitted a number of expert reports. Most critically, Professor Patricia McCoy submitted a report describing the effects of the risk factors on borrowers and the “singular influence” that Morgan Stanley exerted over New Century. Sugnet Decl. Ex. 4. Professor Ian Ayres submitted a statistical regression analysis that, he claims, proves that Combined-Risk loans had a disparate impact on the African-American community in Detroit. Sugnet Decl. Ex. 5. Defendants have moved to exclude all of the McCoy Report and limited language in the Ayres Report.³⁸ Defendants also submitted a number of dueling expert reports in opposition to class certification, most notably including the report of Professor Timothy Riddiough, submitted to counter the McCoy Report’s conclusions.

“Under Daubert v. Merrell Dow Pharmaceuticals Inc., expert testimony is admissible if the expert is proposing to testify to (1) scientific knowledge that (2) will assist the trier of fact to understand or determine a fact or issue.” U.S. Foodservice Pricing, 729 F.3d at 129 n. 12 (citing 509 U.S. 579, 592, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993)). “Under [Federal Rule of Evidence] 702, an expert witness, unlike a lay witness, is ‘permitted wide latitude to offer opinions, including those that are not based on firsthand knowledge or observations.’” Major League Baseball Props., Inc. v. Salvino, Inc., 542 F.3d 290, 310 (2d Cir.2008) (quoting Daubert, 509 U.S. at 592, 113 S.Ct. 2786). The role of district courts as gatekeepers of expert testimony under Rule 702 is well-established. See Nimely v. City of N.Y., 414 F.3d 381, 396 (2d Cir.2005).

“Neither the Supreme Court nor the Second Circuit has definitely decided whether the Daubert standard governs the admissibility of expert evidence submitted at the class certification stage.” Chen-Oster v. Goldman, Sachs & Co., No. 10-CV-6950(AT)(JCF), 2015 WL 1035350, at *1 (S.D.N.Y. Mar. 10, 2015); see U.S. Foodservice Pricing, 729 F.3d at 129 (same, but noting that Dukes “offered limited dicta suggesting that a Daubert analysis may be required at least in some circumstances”). In this case, the Court need not determine whether the challenged experts’ reports should be excluded under Fed.R.Evid. 702. The Court has fully considered the opinions of Professors McCoy and Ayres in its class certification analysis and has nonetheless determined that class certification is inappropriate. Accordingly, Defendants’ Motion to Preclude is dismissed without prejudice if the Plaintiffs are ultimately able to certify a class.

The Court nevertheless notes numerous concerns about McCoy’s report. First, in offering her opinion as to Morgan Stanley’s responsibility for New Century’s loans, McCoy offers no expert analysis—instead, she simply marshals evidence unrelated to her expertise in consumer mortgages. Based on her understanding of New Century’s reliance on Wall Street in general, McCoy concludes ipse dixit that Morgan Stanley had a “singular influence” on New Century’s practices. McCoy 22-25. Her analysis lacks any benchmarks or comparators; McCoy refused to admit or deny the possibility that “there were other singular influences as to New Century,” including “other secondary market participants, banks, [or] other institutions.” McCoy Dep. at 40-41. Because she conducted neither a scientific study nor a qualitative comparison with meaningful benchmarks, McCoy’s opinion as to Morgan Stanley’s influence over New Century would not be admissible under Rule 702.³⁹ The same holds true for her related conclusions regarding Morgan Stanley’s role as a purchaser, underwriter, and securitizer; McCoy’s speculation regarding the role that Morgan Stanley played lacks any metrics or benchmarks against which it could be appraised (like, for example, comparing the role that other banks played in generating similar loans). Because her methodology is unreliable,⁴⁰ Rule 702 would not countenance her testimony as to Morgan Stanley’s influence on New Century. This is true irrespective of the validity or invalidity of her conclusion. “ ‘The focus of the admissibility inquiry ... must be solely on principles and methodology, not on the conclusions that they generate.’ ” SR Int’l Bus. Ins. Co. Ltd. v. World Trade Ctr. Props., LLC, 467 F.3d 107, 134 (2d Cir.2006) (quoting Daubert, 509 U.S. at 595, 113 S.Ct. 2786) (alteration omitted). Plaintiffs would, of course, be permitted to summarize the relevant documentary evidence describing the relationship between Morgan Stanley and New Century, but “factual testimony about matters that require[] no specialized knowledge” is not the province of experts. United States v. Mejia, 545 F.3d 179, 196 (2d Cir.2008).

Defendants’ motion likely would be denied, on the other hand, as to at least some of McCoy’s testimony regarding the effects of the risk factors on a borrower’s likelihood of default. McCoy’s report is replete with citations to scholarly reports, and her qualitative reasoning as to causality expresses opinions informed by her significant relevant experience. Cf. United States v. Farhane, 634 F.3d 127, 158-59 (2d Cir.2011). The Court is not persuaded by all of McCoy’s testimony as to the role of the risk factors, but the “gaps or inconsistencies in [her] testimony ... ‘go to the weight of the evidence, not its admissibility.’” SR Int’l, 467 F.3d at 134 (quoting Campbell v. Metro. Prop. & Cas. Ins. Co., 239 F.3d 179, 186 (2d Cir.2001)).

Finally, Defendants’ motion would likely be granted as to the Ayres Report but to almost no effect. The Ayres Report focuses on the disparate impact of the Combined-Risk loans that New Century made. Insofar as one stray clause suggests that Ayres believes these loans to have been caused by Morgan Stanley, that clause would be inadmissible opinion unrelated to the subject of Ayres’ study. The balance of Ayres’ report would be admissible.

Defendants’ motion is DISMISSED without prejudice to renewal.

CONCLUSION

For the foregoing reasons, Plaintiffs’ Motion for Class Certification is DENIED and Defendants’ Motion to Exclude All Opinions Contained in the Report of Patricia A. McCoy and One Opinion Contained in the Report of Ian Ayres is DISMISSED as moot. While the Plaintiffs’ April 16, 2015 letter did not fully address the effect that a denial of class certification would have on the Plaintiffs’ practical ability to pursue this lawsuit (notwithstanding the fact that they would not legally be barred from doing so), the Court recognizes the likelihood that this ruling constitutes a “death knell” for Plaintiffs’ lawsuit, cf. Sumitomo Copper Litig. v. Credit Lyonnais Rouse, Ltd., 262 F.3d 134, 140-41 (2d Cir.2001), and that appellate review pursuant to Rule 23(f) may therefore be appropriate, see Levitt v. PriceWaterhouse-Cooper LLP, No. 07-3334-mv, 2007 WL 4060136 (2d Cir. Nov. 7, 2007). The Clerk of the Court is respectfully directed to terminate Dkt. 127 and Dkt. 188.

SO ORDERED.

1 . Although a number of corporate entities are named as Defendants, the Court refers to all Defendants as "Morgan Stanley.”

2 . Plaintiffs include nine counties in defining "the Detroit metropolitan area: Genesee, Lapeer, Livingston, Macomb, Monroe, Oakland, St. Clair, Washtenaw, and Wayne.” Compl. ¶ 116.

3 . The definition of a "Combined-Risk loan” was created by Plaintiffs for the purpose of this litigation; it is not a generally-accepted term in either the world of residential lending or the world of investment banking.

4 . The HMDA defines “high-cost” loans as "first lien loan[s] with an annual percentage rate and borrowing costs that exceed by more than 3 percentage points Treasury securities of comparable maturity, or [] subordinate lien loan[s] that exceedf] the Treasury benchmark by more than five points.” Compl. ¶ 31.

5 . For ease of reference, this opinion refers to the risk factors the Plaintiffs focus on as loan "terms” even though several of the factors are not loan terms at all (e.g., high loan-to-value ratios or stated-income loans).

6 . Findings of fact are drawn from the depositions, declarations, expert reports, and other exhibits that the parties submitted in connection with the motion for class certification. At this stage, the Court does "not resolve factual assertions related to the merits but state[s] them as the parties’ assertions.” Sykes v. Mel S. Hams & Assocs., LLC, 780 F.3d 70, 75 (2d Cir.2015) (internal quotation marks and alterations omitted).

7 . Both parties included excerpts from depositions in advocating for and against class certification and separately included the full deposition transcripts of each witness. Accordingly, the Court cites to tire full deposition transcripts, which the parties are directed to file (as discs if necessary) on the public docket, in redacted form if necessary pursuant to the Protective Order.

8 . The retail channel, operating under the name Homel23 Corporation, directly lent to consumers and to builders. Riddiough ¶ 32. New Century’s Wholesale Division ran the correspondent and the broker channels. Through the correspondent channel, the Wholesale Division "purchased loans that had already been funded by other entities known as 'correspondent lenders.’ ” Id.

9 . References to "Tr.” refer to the transcript of the April 9, 2015 evidentiary hearing and oral argument.

10 . Both parties’ experts agree that New Century has substantial credit capacity from Wall Street generally and from Morgan Stanley specifically. See Riddiough 1138 (“New Century’s total credit capacity (including its warehouse loans and its asset-backed commercial paper facility) increased during the proposed class period, from $7.4 billion at the beginning of 2004 to $16.9 billion at the time of its final quarterly SEC filing for the period ended September 30, 2006.”); Expert Op. of Patricia A. McCoy in Support of Class Certification ("McCoy”) 22 ("New Century relied so heavily on warehouse loans that it had approximately $8.5 billion in warehouse lines from Morgan Stanley and other Wall Street firms at the end of third quarter 2006.”).

11 . Morgan Stanley’s involvement with RMBSs has led to a bevy of lawsuits (many unsuccessful) asserting claims arising from the collapse of the housing market. See, e.g., Stratte-McClure v. Morgan Stanley, 776 F.3d 94, 96 (2d Cir.2015) (affirming dismissal of a claim that Morgan Stanley misled its investors by understating its exposure to the RMBS market); In re Morgan Stanley Mortg. Pass-Through Certificates Litig., 23 F.Supp.3d 203, 205 (S.D.N.Y.2014) (describing "alleged violations of the Securities Act of 1933 ... in connection with purchases of certain issu-ances, including purchases of certificates from [a] Morgan Stanley Mortgage Loan Trust”); Nat’l Credit Union Admin. Bd. v. Morgan Stanley & Co., No. 13-CV-6705(DLC), 2014 WL 241739, at *1 (S.D.N.Y. Jan. 22, 2014); Emps. Ret. Sys. of Gov't of V.I. v. Morgan Stanley & Co., 814 F.Supp.2d 344, 348-49 (S.D.N.Y.2011) (describing plaintiffs’ allegations that "Morgan Stanley had direct, inside, non-public information regarding the deteriorating quality of New Century's loans” but failed to downgrade the credit ratings of affected collateralized debt obligations ("CDOs”)); In re MBIA, Inc., Sec. Litig., 700 F.Supp.2d 566, 572 (S.D.N.Y.2010); Abu Dhabi Commercial Bank v. Morgan Stanley & Co., 651 F.Supp.2d 155, 165-66 (S.D.N.Y.2009). In connection with allegations regarding the bank’s sale of RMBSs, Morgan Stanley announced in February 2015 "that it had reached a $2.6 billion settlement with the Justice Department over the sale of mortgage securities before the financial crisis.” Nathaniel Popper, Bank Settles Federal Case over Crisis in Mortgages, N.Y. TIMES, Feb. 26, 2015, at B3.

12 . Of course, purchasers of RMBSs would undoubtedly argue that there would not have been such an outsized demand had the purchasers known of the poor quality of the loans that were being bundled into RMBSs.

13 . The regulation currently in force includes loans with rates only 1.5 percentage points higher than the prime rate for first liens and 3.5 percentage points higher for subordinate liens. See 12 C.F.R. § 203.4(a)(12)(i).

14 . This rule actually raised the maximum DTI, which had previously been 36 percent. Ability-to-Repay and Qualified Mortgage Standards under the Truth in Lending Act, 78 Fed.Reg. at 6416.

15 . Of course, other factors—notably, stagnating or falling housing values—greatly exacerbated the risk inherent in a loan with a high CLTV ratio. Sherlund 3.

16 . One common type of "balloon loan” was the 40/30 loan, in which the loan was amortized over a 40-year period, but the borrower would make regular payments for only the first 30 years, at which point the borrower would owe a “balloon payment” that would extinguish the remaining principal balance of the loan. Adam J. Levitin & Susan M. Wachter, Explaining the Housing Bubble, 100 GEO. L.J. 1177, 1200 n. 70 (2012).

17 . As discussed infra, the number of relevant subsets may substantially exceed 33, particularly if the Court were to view as differently-situated borrowers whose loans were included, for example, in a reverse-bid initiated by Morgan Stanley, in a forward sale pursuant to a standard bid won by Morgan Stanley, or in a sale to Morgan Stanley after the loan was made. Furthermore, each permutation could have different subgroups, depending on the timing of the purchase—Morgan Stanley may have exerted greater influence over New Century’s practices, for example, in 2004 than it did in the first half of 2005 (when it did not win a single bid to purchase mortgages from New Century). Eveiy additional meaningful distinction could yield another 33 subgroups of differently-situated borrowers.

18 . The existence of some individualized issues is irrelevant to the question of commonality, although it weighs heavily in the predominance inquiry.

19 . "[T]he Supreme Court has acknowledged that, in certain 'contexts, the commonality and typicality requirements of Rule 23(a) tend to merge.' ” Sykes, 780 F.3d at 84 n. 2 (quoting Dukes, 131 S.Ct. at 2551 n. 5) (alterations omitted). In this case, as in some other cases, the typicality inquiry is also intertwined with adequacy under Rule 23(a)(4), see, e.g., Bishop v. N.Y. City Dep’t of Hous. Preservation & Dev., 141 F.R.D. 229, 240 (S.D.N.Y.1992), and with predominance under Rule 23(b)(3), see, e.g., Stream Sicav v. Wang, No. 12-CV-6682(PAE), 2015 WL 268855, at *5 (S.D.N.Y. Jan. 21, 2015).

20 . Specifically, Defendants assert that (1) Adkins, McCoy, and Young will face difficulties establishing Morgan Stanley's responsibility for their loans in light of the fact that Morgan Stanley due diligence officials rejected their loans for purchase by Morgan Stanley; (2) Williams faces a similar causation challenge insofar as New Century originated her loan during a period in which Morgan Stanley did not purchase any loans from New Century; and (3) Pettway will face unique difficulties in establishing that equitable tolling is appropriate in her case.

21 . Plaintiffs do not allege that any representative had a negative amortization loan, but Courchane did not list it among the features absent from the putative representatives' loans. See Compl. ¶¶ 123-204, Courchane ¶ 64.

22 . Defendants do not identify any inadequacies of the proposed class counsel, and the Court finds none. Plaintiffs’ proposed counsel would more than adequately represent the class, notwithstanding the Court’s other reservations regarding the suitability of this case for class certification.

23 . See also Vitamin C Antitrust Litig., 279 F.R.D. at 114-15 ("’[E]very federal court of appeals that has considered the question has held that a class action seeking only declaratory or injunctive relief does not bar subsequent individual suits for damages.’ ”) (quoting Hiser v. Franklin, 94 F.3d 1287, 1291 (9th Cir.1996) (emphasis from Vitamin C Antitrust Litig.)).

24 . Defendants, as an aside and without seeking relief, challenge the availability of disgorgement as a remedy in this case. Disgorgement traditionally "is a distinctly public-regarding remedy. available only to government entities seeking to enforce explicit statutory provisions." F.T.C. v. Bronson Partners, LLC, 654 F.3d 359, 372 (2d Cir.2011). The FHA permits courts to award actual and punitive damages as well as injunctive relief. 42 U.S.C. § 3613(c). The Ninth Circuit has held that disgorgement is unavailable to a private party seeking to enforce the FHA. Smith v. Pac. Props. & Dev. Corp., 358 F.3d 1097, 1106 (9th Cir.2004) (”[T]he remedy of disgorgement only arises where a prior relationship between the parties subject to and benefitting from disgorgement originally resulted in unjust enrichment.”). Even if disgorgement is not available at this stage, however, if class treatment were otherwise appropriate, the Court could certify the class for the limited purpose of determining Morgan Stanley’s liability. Fed.R.Civ.P. 23(c)(4); Johnson, 780 F.3d at 138 ("Common issues— such as liability—may be certified, consistent with Rule 23, even where other issues—such as damages—do not lend themselves to classwide proof.”).

25 . The Supreme Court will soon decide whether FHA provides a cause of action for housing discrimination based on disparate impact. See Tex. Dep't of Hous. & Cmty. Affairs v. Inclusive Cmtys. Project, Inc., No. 13-1371.

26 . Loan purchasers and mortgage securitizers are covered by the FHA; they are prohibited from imposing different terms or conditions for those purchases because of race. Adkins I, 2013 WL 3835198, at *9 (quoting 24 C.F.R. § 100.125); see also 42 U.S.C. § 3605(a). Thus, Morgan Stanley is plainly covered by the FHA’s text insofar as it purchased or securitized mortgages. The Court expresses no view whether that coverage extends as far as Plaintiffs’ theory, which seeks to hold Morgan Stanley responsible under the FHA for loans that it neither purchased nor securitized.

27 . For example, although Defendants concede that Morgan Stanley sought loans with prepayment penalties, Riddiough ¶ 100, there is no evidence that Morgan Stanley affirmatively sought no-documentation loans. Accordingly, Plaintiffs' proof as to Morgan Stanley's role in New Century’s making no-documentation loans will depend on their ability to establish that Morgan Stanley’s desires as a secondary market purchaser of loans was driving the larger market for residential mortgage loans. That contention could be stronger in the context of Combined-Risk loans that had risk factors affirmatively sought by Morgan Stanley than it would be in the context of Combined Risk loans that have only risk factors that were not desired by Morgan Stanley (say, no-documentation loans or loans with a high LTV).

28 . For example, when Morgan Stanley made a purchase “based on expected loan characteristics of a pool without necessarily having the actual loans," Vanacker Dep. at 43-44, and before the actual loans were made, the argument that Morgan Stanley "dictated” the terms on which New Centuiy made its loan would be discernably different from the same argument in the context of a loan that was never touched by Morgan Stanley (e.g., a loan originated by New Century, sold to Barclays, and securitized by HSBC).

29 . See, e.g., Rose 24 (prepayment penalties, balloon payments, and stated income loans "each affect the probability of foreclosure differently, depending on [whether a loan’s rate is fixed or adjustable]”), id. at 26 (finding that purchase FRMs with reduced documentation and either a long prepayment penally period or a balloon payment (but not both) are associated with substantially lower probability of foreclosure).

30 . The complexity of the use of multiple risk factors that can be mixed and matched many different ways to define membership in the class is underscored, for example, by Professor Ayres’ decision to control for three of the eight risk factors in his regression analysis assessing the probability that a borrower would obtain a Combined-Risk loan. See Ayres 48 Table 12.

31 . Plaintiffs’ description of the Combined-Risk loan "as a 'proxy for the type of layered-risk loans associated with high rates of default and foreclosure,'” Reply at 11-12 (quoting McCoy 16), is creative but ultimately unpersuasive. The Court recognizes that many Combined-Risk loans were "bad loans” that created a heightened risk of default beyond what was likely given the borrower’s creditworthiness and income. But Plaintiffs have combined a hodge-podge of factors that have different effects on the borrowers and varying desirability to investors. Even if this class were a reasonable proxy for recipients of "bad” loans by some measure—and Plaintiffs have not shown that it is—the proof necessary to establish the elements of the cause of action differs considerably as to Plaintiffs whose participation in the makeshift class was brought about by the presence of different terms of their loans. Accordingly, even if the class were defined in an "accurate” way by some measure, it is not defined in a manner that is susceptible to classwide proof that predominates over individualized questions.

32 . Cf. Johnson, 780 F.3d at 148; UFCW Local 1776, 620 F.3d at 135 ("[S]howing injury by general proof is precluded by uncertainty about what the alternatives ... [to the alleged harm] would have been, and how they would have been distributed amongst the plaintiffs.”); Stream Si-cav, 2015 WL 268855, at *4 ("Absent a concrete presentation of the evidence on which such a finding could be made in this case, the Court cannot find that [] injury to all class members can be established by a common proof.").

33 . The Dukes Court analyzed the question of discretion afforded to managers at Wal-Mart stores throughout the country as a commonality issue; because the predominance inquiry subsumes the commonality inquiry, however, that distinction is of no moment here.

34 . Plaintiffs rely heavily on one case in which brokers' discretion was held not to defeat commonality. See Ramirez v. Greenpoint Mortg. Funding, Inc., 268 F.R.D. 627, 632-35 (N.D.Cal.2010). That case, however, relied heavily on the Ninth Circuit's en banc decision in Dukes v. Wal-Mart, 603 F.3d 571 (9th Cir.2010), which was reversed by the Supreme Court, - U.S. -, 131 S.Ct. 2541, 180 L.Ed.2d 374. The Ramirez decision specifically quoted language from the intermediate court's opinion indicating that a policy of subjective discretion "is a ‘ready mechanism for discrimination' ” bearing particularly heavy scrutiny. 268 F.R.D. at 634 (quoting Dukes v. Wal-Mart Stores Inc., 603 F.3d at 612). This theory was soundly rejected by the Supreme Court. Dukes, 131 S.Ct. at 2554. But even if it did not rely on a reversed Ninth Circuit decision, the Ramirez case involved fewer discretionary decisions because the defendant was the loan originator. Policies established by the originator could be the basis for liability in Ramirez but could be the basis for a defense verdict here, where the originator is not the defendant.

35 . See, e.g., Countrywide, 708 F.3d at 708 ("The plaintiffs claim that the discretion Countrywide has given its sales force is exercised in a common way—by limited variation of the par rate.... [Pjlaintiffs [do not] demonstrate that this range, rather than discretionary decisions made within this range, disparately impacted the proposed class.”) (internal quotation marks omitted); Barrett v. Option One Mortg. Corp., No. 08-CV-10157, 2012 WL 4076465, at *2-3 (D.Mass. Sept. 18, 2012) ("[Although] African-American borrowers on average spent about $134 more per year on their mortgages than similarly situated white borrowers,” plaintiffs did not "point to any common mode of exercising discretion that was shared by all of Option One’s brokers ... [such as] claiming] that Option One’s brokers uniformly exercised their discretion by considering specific attributes that produce disparate impact—such as 'scores on general aptitude tests or educational achievements.’ ”) (quoting Dukes, 131 S.Ct. at 2554); In re Wells Fargo Residential Mortg. Lending Discrimination Litig., No. 08-MD-1930(MMC), 2011 WL 3903117, at *3-5 (N.D.Cal. Sept. 6, 2011) (evidence that discretion afforded to "loan officers and mortgage brokers” yielded a disparate impact was insufficient to demonstrate commonality).

36 . The, FastQual system permitted independent brokers to learn quickly whether specific terms would comply with New Century's underwriting guidelines (although it did not bind New Century to make the loan) but did not eliminate the broker’s discretion in selecting which terms to include in a loan or when to deviate from the guidelines. Finley Deck ¶ 8; cf. Countrywide, 708 F.3d at 708.

37 . Plaintiffs’ argument that brokers’ discretion was minimal is unpersuasive, particularly in light of their discussion of "steering,” a process by which brokers or loan officers would write Combined-Risk or other subprime loans for borrowers who were eligible for prime loans. McCoy 8 &nl5.

38 . Plaintiffs submitted other expert reports that are not the subject of Defendants' motion to preclude, including those of Geoffrey Oliver and Professor Thomas Sugrue. Sugnet Decl. Exs. 6-7.

39 . It is likely that McCoy's analysis regarding Morgan Stanley's relationship with New Century satisfies “none of the four factors identified in Daubert"—she has not tested her methodology, subjected her qualitative descriptions of one company's singular influence over another to peer review or publication, analyzed it through any method susceptible to evaluation for known error rates, or shown a general acceptance of her methodologies. Zaremba v. Gen. Motors Corp., 360 F.3d 355, 358 (2d Cir.2004).

40 . It is a stretch to describe what McCoy did as a "methodology.” Essentially, she summarized a number of internal emails that appear to have been hand-picked by Plaintiffs' counsel and speculatively devised a narrative for the nature of Morgan Stanley's interactions with New Century therefrom.