ResCap’s Mortgage Fraud Damages Expert Dr. Karl Snow in the Universal Mortgage Case

Residential Funding Co. v. Universal Am. Mortg. Co.,

Civ. No. 13-3519 (PAM/HB), at *7-10

(D. Minn. Oct. 12, 2018)

REPUBLISHED BY LIT: DEC 27, 2021

a. Breaching Loss approach

The Breaching Loss approach calculates the amount of loss RFC suffered from each bank’s breaching loans. So, for example, if a breaching loan went into default with a balance of $150,000 including interest and fees, RFC’s loss under this approach would be $150,000 minus whatever proceeds were realized from the sale of the property. This approach is based on the repurchase remedy from the Client Guide, which requires a lender to repurchase defective loans within 30 days of RFC’s demand that it do so. (Client Guide § A210(A).) Universal argues that the Breaching Loss approach’s reliance on this remedy is inappropriate because RFC no longer owned any of the loans at issue—they were all sold into the RMBS trusts. Thus, there was nothing to repurchase. Moreover, any loss was incurred by the trusts and their insurers, not RFC, making this damages model inapplicable in any event.

The flaw with the Breaching Loss approach, however, is more simple: it does not account for any discount RFC negotiated in the bankruptcy settlements. Dr . Snow determined that RFC settled its claims for less than 30 cents on the dollar. (Snow Rpt. ¶ 91.) But the Breaching Loss approach measures “each [lender’s] share of the indemnification liability as equal to 100% of the losses on the [lender’s] allegedly breaching loans, despite the steep discount agreed to in the Settlements.” In re RFC SJ Order at 159. This windfall alone makes the Breach Loss approach excludable. Universal’s Motion as to this approach is granted.

b. Allocated Loss approach

The Allocated Loss approach assesses damages based on each bank’s share of total losses on all loans, not just breaching loans. The total losses are capped at the amount RFC agreed to pay in the bankruptcy settlements. Thus, unlike the Breaching Loss approach, this approach provides no windfall to RFC.

Universal challenges this approach because it is not based on Universal’s alleged breaches of the Client Guide, but rather is based only on Universal’s share of the total population of at-issue loans. The Court agrees. The Client Guide does not permit RFC to seek indemnity for any loss a loan experiences, but only for losses caused by Universal’s failure to abide by the Client Guide. Thus, for example, a loan that falls into default because the homeowner became ill would sustain a loss, but unless Universal’s representations regarding the loan were faulty would not constitute a breaching loan for which indemnification was required. Universal’s Motion as to this approach is granted.

c. Allocated Breaching Loss approach

The Allocated Breaching Loss approach allocates a portion of the settlement based on a ratio of Universal’s breaching losses on its at-issue loans to the overall breaching losses on all at-issue loans. This approach does not hold Universal responsible for all losses RFC suffered, but only for its portion of RFC’s settlement obligations.

Universal attacks the Allocated Breaching Loss approach as too speculative to allow a jury to correctly calculate damages. But this method is sufficiently specific to “provide[] the factfinder with a non-speculative basis to assess the value of the claims that are identifiable by Defendants.” In re RFC SJ order at 175. As Judge Nelson found,

The model assigns damages to a [lender] based on the number of loans it sold to RFC, which is a concrete and verifiable number.

The model also assesses economic losses to at-issue mortgages based on reliable loan data.

Further, the model targets its assessment of damages toward only those loans that it estimates contained breaches of the Client Guide and Trust Agreement based on a sampling protocol.

Id. at 175-76.

This approach complies with Minnesota’s requirement that a plaintiff prove damages to a reasonable certainty that need not be mathematically precise.

Poppler v. Wright Hennepin Co-Op Elec. Ass’n, 834 N.W.2d 527, 546 (Minn. Ct. App. 2013).

The Allocated Breaching Loss approach is an appropriate method for calculating damages,

and Universal’s Motion to exclude this approach is denied.

Who is Expert Dr. Karl Snow?

SELECTED EXPERIENCE

Residential Funding Co. v. Universal Am. Mortg. Co.,

Civ. No. 13-3519 (PAM/HB), at *22-24

(D. Minn. Oct. 12, 2018)

Universal’s Daubert Motions

REPUBLISHED BY LIT: DEC 27, 2021

Dr . Karl Snow

Dr . Snow is RFC’s principal damages expert. The Court has already ruled on the merits of some of Universal’s objections to Dr . Snow’s proposed testimony, finding that Dr . Snow’s sampling methodology is sufficiently reliable, and finding that one of the three loss approaches he advocates may be presented to the jury. Universal’s Motion on sampling and the Allocated Breaching Loss approach is denied, but is granted with respect to the two loss approaches the Court previously excluded.

Universal also argues that Dr . Snow’s calculations are fatally flawed because he relied on the analysis of Steven Butler,5 who conducted a reunderwriting as to a sample of loans.

According to Universal, Butler’s reunderwriting analysis is fundamentally unreliable because it focused on whether anything was amiss at the time of the loan’s origination, rather than at the time RFC received notice of the alleged breach with regard to the loan. Universal maintains that a breach at the loan’s origination, such as the borrower’s failure to produce a required document at closing, might not be material later in the life of the loan, and only material breaches could trigger the cure-and-repurchase remedy in the agreements that pooled RFC’s loans into the RMBS trusts.

The question the jury must answer is whether Universal breached a duty to RFC that caused RFC to breach its agreements with the RMBS trusts. Thus, a breach at the time of origination is relevant in the first instance, because such a breach likely triggered Universal’s duties to RFC. Universal can argue to the jury that its alleged breaches did not have a materially adverse effect on the loans and could not have caused RFC’s breaches of its duties to the RMBS trusts. This is not a reason to exclude Dr . Snow’s opinions.

Universal also contends that, even if sampling is allowed, Dr . Snow’s sampling is untrustworthy because he allegedly used the wrong loan population from which to draw his sample. Dr . Snow used a population of approximately 460,000 at-issue loans, which are loans that as of a certain date had actual losses of at least $500 or were 90 or more days delinquent, in foreclosure, or real-estate owned. From this population, he estimated breach rates for Universal’s loans. But more than 2 million loans were involved in the bankruptcy settlements and according to Universal, Dr . Snow could not have correctly calculated a breach rate by sampling only the subset of breaching loans. Moreover, Universal maintains that he cannot reliably estimate the percentage of the settlements that can be attributed to Universal’s share of the at-issue loans as opposed to the other 1.5 million loans.

RFC points out that the remaining 1.5 million loans caused less than $900 million in losses to the trusts, while the 463,000 at-issue loans caused $42 billion in losses.

Indeed, nearly 1.3 million loans caused no loss whatsoever, because they were not in foreclosure or otherwise delinquent. Thus, Dr . Snow’s failure to include those loans in his sampling protocol was correct. Regardless, the inclusion of the total loan population would not materially change Dr . Snow’s conclusions regarding the losses for which Universal is responsible. Universal’s Motion on this point is denied.

Finally, Universal argues that Dr . Snow’s opinions on the breach rates in the insurer settlements is flawed because he allegedly allocates each settlement to specific loan pools according to the breach rate of those pools. Universal contends that Snow did not conduct any pool-specific sampling, so he impermissibly extrapolates breach rates from samples drawn from all the pools and insurers to specific pools and specific insurers. Any issues with Snow’s methods, however, are a matter for cross-examination, and do not render his opinions excludable.

Universal’s Motion to Exclude Dr . Snow is denied, except as to the two loss approaches previously excluded.