Opinion
NOT TO BE PUBLISHED IN THE OFFICIAL REPORTS
APPEAL from an order of the Superior Court of Los Angeles County, Lee Smalley Edmon, Judge, Los Angeles County Super. Ct. No. BC283265
Lerach Coughlin Stoia Geller Rudman & Robbins; Coughlin Stoia Geller Rudman & Robbins, Timothy G. Blood, Pamela M. Parker, Alreen Haeggquist, Leslie E. Hurst, Thomas J. O’Reardon II; Milberg Weiss Bershad & Schulman, Michael Spencer, Susan M. Greenwood; Packard, Packard & Johnson, Craig H. Johnson, Lon D. Packard; Michael Huber; Johnson, Poulson, Coons & Slater and Jonathan E. Johnson for Plaintiffs and Appellants.
Goodwin Procter and Thomas M. Hefferon; Severson & Werson and Michael J. Steiner for Defendants and Respondents.
SUZUKAWA, J.
The trial court denied plaintiffs’ motion for class certification, finding that a well-defined community of interest does not exist because of the individual issues that predominate over common issues. Finding no abuse of discretion, we affirm.
BACKGROUND
Under Code of Civil Procedure section 382, class actions may be brought “when the question is one of a common or general interest, of many persons, or when the parties are numerous, and it is impracticable to bring them all before the court . . . .” “The party seeking certification has the burden to establish the existence of both an ascertainable class and a well-defined community of interest among class members. [Citations.]” (Sav-On Drug Stores, Inc. v. Superior Court (2004) 34 Cal.4th 319, 326.) In order to show a community of interest, the moving party must establish that: (1) common questions of law or fact will predominate over issues requiring separate adjudication; (2) the class representatives possess claims that are typical of the class; and (3) the class representatives can adequately represent the class. (Ibid.)
I. Plaintiffs’ Theory
On October 15, 2002, Utah residents Scott and Erin Pickrell and California resident Barbara Ann Butkus (collectively, plaintiffs) filed this putative class action lawsuit against Countrywide Financial Corp. (sued by its former name, Countrywide Credit Industries, Inc.) and Countrywide Home Loans, Inc., doing business as America’s Wholesale Lender (collectively, defendants or Countrywide). Plaintiffs contended that each class member had been misled to pay underwriting fees that exceeded defendants’ actual computerized underwriting costs by several hundred dollars per loan.
This lawsuit involves the “cream of the crop,” conforming, and conventional loans (class loans) that received a positive computerized score from an automated underwriting system (AUS). The premise underlying this lawsuit is that following the federal government’s 1995 mandate that only conforming, conventional loans with positive AUS scores may be resold to the Federal National Mortgage Association (Fannie Mae) or the Federal Home Loan Mortgage Corporation (Freddie Mac), the underwriting process for class loans was transformed from an expensive, labor intensive process to an inexpensive, computerized process, with a corresponding reduction of $400 to $650 in actual underwriting costs for each class loan. The complaint alleged that once a loan receives a positive AUS score, “underwriting is complete” and “[a]ny additional validation is de minimus” because “FANNIE MAE and FREDDIE MAC ‘guarantee’ that they will purchase the loan on the secondary market, thereby eliminating Countrywide’s risk in the underwriting process.”
Plaintiffs filed this lawsuit seeking to force defendants to pass on to consumers the savings generated by the computerized underwriting of class loans that received positive AUS scores. Plaintiffs alleged that although defendants participate widely in the Fannie Mae and Freddie Mac programs, defendants have failed to adjust their underwriting fees to reflect the reduced costs of computerized underwriting. Plaintiffs contended that Fannie Mae and Freddie Mac charge, at most, $20 for the use of their AUS and that others charge between $30 and $50, but that defendants continue to charge “hundreds of dollars for the same computerized underwriting service. While it may disclose the amount of the underwriting fee it is charging, Countrywide does not disclose that its fee represents a huge mark-up over its actual or reasonable underwriting costs.”
The complaint defined the proposed class as consisting of the thousands of borrowers who, on or after January 1, 1995, had paid an underwriting fee to defendants and received a positive AUS score for their loan. Plaintiffs alleged that because the underwriting fees exceeded the actual computerized underwriting costs, the fees were unconscionable, unfair, unlawful, and fraudulent, and were prohibited by: (1) the Unfair Competition Law (Bus. & Prof. Code, § 17200 [UCL or section 17200]) (first cause of action); (2) the California Consumers Legal Remedies Act (Civ. Code, § 1750 et seq. [CLRA]) (third cause of action); and (3) the Utah Consumer Sales Practices Act (Utah Code, § 13-11-1 et seq. [UCSPA]) (fourth cause of action). The complaint sought restitution, injunctive relief, attorney fees, costs, and other relief.
An October 25, 2004 summary adjudication order disposed of the second amended complaint’s three common law claims for unjust enrichment and constructive trust (fifth cause of action), breach of bailment agreement (sixth cause of action), and conversion (seventh cause of action). In addition, plaintiffs are no longer pursuing their second cause of action for the alleged overcharging of flood certification fees in violation of section 17200.
Plaintiffs moved to certify the class on the theory that it is possible to prove, on a classwide basis, that each member had paid an underwriting fee that exceeded by several hundred dollars the actual AUS costs incurred to obtain the positive AUS score, which is all that is needed to underwrite a class loan. According to plaintiffs’ expert witness J. F. Morrow, a mortgage company director, “Underwriting the subject loans is performed by an AUS. The AUS analyzes the data input by the Originator according to the FANNIE MAE/FREDDIE MAC guidelines that are programmed into the AUS to determine whether the borrower qualifies for the loan. [¶] . . . Before the development of the AUS, lenders used human underwriters to underwrite loans. Now, lenders limit the underwriter’s role for the subject loans to reviewing documentation. [¶] . . . The AUS automatically underwrites the data input by the Originator by calculating the housing expense-to-income and total debt-to-income ratios and provides a list of documents that need to be included in the loan package. No Countrywide underwriting employee is involved in this process because it is all done automatically by computer. Residential real estate loans rated ‘Accept,’ ‘Accept-Plus,’ ‘Approve’ or a similar code by the AUS are underwritten in a uniform manner that is virtually indistinguishable from loan to loan.” Morrow further stated, “Automated underwriting allows underwriters to focus on loans that are not rated with an ‘Accept’ or similar code, where the underwriting judgment is critical as a manual review of credit and risk is required.”
Of relevance to this appeal, plaintiffs sought to establish that a well-defined community of interest exists among the class members who were damaged by paying allegedly excessive underwriting fees for class loans. Plaintiffs argued that common issues would predominate over individual questions because the class members could show, using common evidence, that: (1) defendants’ actual underwriting costs were limited to AUS costs; (2) defendants’ actual AUS costs could be determined by using cost accounting methods; and (3) defendants’ reasonable AUS costs could be determined by referring to what other lenders charge for automated underwriting.
In their opening brief, plaintiffs stated that “[g]athering, reviewing and verifying the completeness and accuracy of the information put into the computer system are additional steps in the loan process, but they are not underwriting. . . . Indeed, the evidence showed that Countrywide charges separate fees for these functions, in addition to what it charges for ‘underwriting.’”
II. Defendants’ Opposition
Defendants argued that class certification should be denied because of plaintiffs’ failure to establish the existence of a well-defined community of interest among the class members. Specifically, defendants contended that individual issues would predominate over common issues for the following reasons.
First, according to defendants’ expert witness Middleton Thompson, an accredited underwriter, the AUS is simply “a tool used by underwriters, it does not underwrite loans.” According to Thompson, “Underwriters have to perform many functions that an AUS simply cannot do. An AUS cannot perform tasks that involve the physical inspection of documents or data, for example. Thus, it cannot review a W-2 to determine whether it has been altered. Nor can it look at the pictures on an appraisal report to evaluate the comparables selected. An AUS can’t clear conditions. It can’t read a title report. It can’t insure that the income information input into it was properly seasoned or qualified. In short, it can’t replace the human judgment required for underwriting. Indeed, even though I consider myself to be an experienced underwriter, it sometimes takes me five to seven hours of work just to reach an underwriting decision on an ‘Accept’ loan.”
Second, the underwriting process for each loan differs with the type of borrower, property, and loan involved. According to Thompson, “[t]here are hundreds of conforming loan programs. The work necessary to underwrite loans in these programs varies significantly based on the individual characteristics of each loan product [e.g., fixed rate versus adjustable rate], each prospective borrower [e.g., salaried versus commissioned income], and each property proposed to secure the mortgage [e.g., owner-occupied versus investment].”
Third, AUS results are only as accurate as the information supplied by the borrower, which the lender must verify. Referring to the named plaintiffs, Thompson stated that “[t]he Pickrells’ loans, for instance, were so-called ‘reduced documentation’ loans. These loans typically do not require the submission of tax returns, or other traditional proof of income. A borrower’s stated income, however, cannot be accepted at face value. Part of the underwriting of such loans is determining whether the stated income is within the reasonable range for the borrower’s occupation and geographic location. That work is performed by a human being, not an AUS.” In addition, Mr. Pickrell was self-employed. For self-employed borrowers, “the stability and health of their business sometimes need to be evaluated. The Pickrells’ loan files demonstrate, for example, that work was done to verify his company’s existence, locations, phone numbers, and licensing status.”
Thompson also pointed out that because Butkus’ loan was secured by a condominium, underwriting her loan required “verification that the condominium either has been approved by Fannie Mae or Freddie Mac, or that it meets the requirements of their spot approval questionnaires. The condominium’s bylaws need to be reviewed for buyback provisions, rights of first refusal, and to ensure that no discriminatory covenants exist. The adequacy of the condominium association’s insurance also needs to be reviewed. Multi-family properties require a consideration of vacancy factors, the adequacy of the borrowers’ reserves, and perhaps, their experience as a landlord. Appraisals for such properties often utilize a different format that looks at both sales and rental values for comparable properties.”
Fourth, the AUS must often be run several times for each loan if the borrower’s circumstances change during the application process. Again referring to the named plaintiffs as an example, defendants pointed out that “many aspects of the Pickrells’ loan applications changed over time. In addition to shopping for both 15 and 30-year terms, the loan amount on the Pickrells’ first loan increased between application and closing in order to finance a prepayment penalty. . . . The appraised value of the property also increased. . . . The loan amount on the Pickrells’ second loan also increased by $10,000 from application to closing. . . . In contrast, the loan amount for Ms. Butkus’ first loan decreased. . . . In addition, prior to closing her second loan, Ms. Butkus had taken out a second mortgage, thereby increasing her combined loan to value ratio. . . . All of these changes required changes in the underwriting process for these borrowers from loan to loan and within loan applications, those changes were reflected in the loan file, documents, closing conditions, and multiple CLUES [the acronym for defendants’ AUS] reports.”
Fifth, the AUS must often be run several times for each loan if there are questionable results before an “accept” recommendation is obtained. Again referring to the named plaintiffs as an example, defendants showed that “the Pickrells received two recommendations of ‘Refer’ in the eight times that CLUES was run in connection with their first loan, three recommendations of ‘Refer’ in the four times it was run on their second loan, and two recommendations of ‘Refer’ and one recommendation of ‘Error’ in the eight times it was run on their third loan. . . . Butkus received one recommendation of ‘Refer’ in the six times that CLUES was run in connection with her first loan.”
CLUES provides underwriters with two other recommendations: “Error,” which indicates that vital information is missing, and that the user has to correct the problem and resubmit the loan; and “Refer,” which indicates that the application needs to be referred to an underwriter for a manual review of the file.
Sixth, it would be difficult to calculate the actual underwriting costs for each loan on a classwide basis, given that defendants do not track or charge underwriting fees based on actual costs but according to what the market will bear in each region. According to defendants, their underwriting fees “were not uniform, but varied from branch to branch and loan-to-loan. The evidence demonstrated that underwriting fees at Countrywide’s California retail branches varied from $100 to $250 and that 31 of the approximately 150 retail branches charged no underwriting fee at all. . . . Underwriting fees in California wholesale branches ranged from $300 to $750. . . . Moreover, like all lender fees, underwriting fees are subject to negotiation, and are sometimes reduced or waived or paid by sellers, brokers or builders.”
Finally, deciding whether to issue a loan requires an exercise of judgment that is not based solely on the AUS score, which is merely a recommendation. Loans that receive positive AUS scores are not automatically approved for sale to Fannie Mae or Freddie Mac. Freddie Mac requires lenders not to rely solely on the AUS score in determining whether to issue a loan. Additionally, according to the declaration of John Kelly, Countrywide’s senior vice president of artificial intelligence, Countrywide remains obligated to indemnify or repurchase failed loans even after they are resold to Fannie Mae or Freddie Mac.
III. Trial Court’s Ruling
The trial court denied the motion for class certification, stating that although the proposed class is ascertainable, evidence of a well-defined community of interest was lacking because, “after carefully considering all of the evidence, the court does not find that Plaintiffs have met their burden of establishing that common questions predominate.” The trial court stated: “Plaintiffs’ argument for class certification is based on the premise that class loans are underwritten at Countrywide by computers, rather than people. Specifically, Plaintiffs assert that Countrywide’s underwriting consists of a computerized assessment of risk, when the loan information is run through CLUES, Countrywide’s automated underwriting system. However, the court finds that the evidence presented by Countrywide demonstrates that the underwriting of each loan varies depending on the borrower, the property and the loan program and that each involves some level of human underwriting. Specifically, Countrywide has presented evidence that underwriting starts before CLUES is run and continues after it is completed and includes some work that can only be performed by human beings. . . . Countrywide has presented persuasive evidence that even the loans that are ultimately rated ‘Accept/Approve’ by CLUES involve a substantial amount of underwriting that varies depending on the particular borrower, property and loan program. Thus, in addressing the costs of underwriting each loan, Defendants would be entitled to present loan-specific evidence regarding the work done on each loan and how the underwriting compared with any fee collected. In light of that fact, [the] court finds that it is not feasible to determine these issues on a class-wide basis.”
Similarly, regarding the allegations that the underwriting fees were unconscionable and, therefore, prohibited by the CLRA and UCSPA, the trial court stated that common issues would not predominate because numerous individual issues—including the manner in which the fees were calculated, negotiated, and disclosed to each borrower; whether the fees were so unreasonable as to shock the conscience; whether other meaningful choices were available to the borrower; and the sophistication of the borrower—must be determined as to each loan.
DISCUSSION
We review the trial court’s order denying the motion for class certification under the abuse of discretion standard. (Sav-On Drug Stores, Inc. v. Superior Court, supra, 34 Cal.4th at pp. 326-327.) “‘Because trial courts are ideally situated to evaluate the efficiencies and practicalities of permitting group action, they are afforded great discretion in granting or denying certification. . . . [Accordingly,] a trial court ruling supported by substantial evidence generally will not be disturbed “unless (1) improper criteria were used [citation]; or (2) erroneous legal assumptions were made [citation]” [citation]. . . . “Any valid pertinent reason stated will be sufficient to uphold the order.”’ [Citations.]” (Ibid.)
I. Plaintiffs Did Not Establish a Community of Interest
As stated earlier, “[t]he party seeking certification has the burden to establish the existence of both an ascertainable class and a well-defined community of interest among class members. [Citations.]” (Sav-On Drug Stores, Inc. v. Superior Court, supra, 34 Cal.4th at p. 326.) In order to show a community of interest, the moving party must establish that: (1) common questions of law or fact will predominate over individual issues; (2) the class representatives possess claims that are typical of the class; and (3) the class representatives can adequately represent the class. (Ibid.)
In order to maintain a class action, the named plaintiffs must belong to the class they claim to represent. It is only by possessing claims that are typical of the class and by being able to represent the class adequately that the named plaintiffs acquire the legitimacy to bind other class members with notice of the action. (Trotsky v. Los Angeles Fed. Sav. & Loan Assn. (1975) 48 Cal.App.3d 134, 146.)
In this case, the trial court denied class certification because of plaintiffs’ failure to show that common issues would predominate over individual issues regarding the non-computerized steps required to underwrite each loan. Although plaintiffs do not challenge the sufficiency of the evidence to support that determination, they contend that the trial court intruded into the merits of the litigation by rejecting their theory that underwriting is an entirely computerized process. Plaintiffs argue that the trial court inappropriately accepted defendants’ evidence regarding the individualized nature of underwriting and prematurely rejected plaintiffs’ conflicting evidence regarding the computerized nature of underwriting. The trial court, however, did not improperly reach the merits of the lawsuit in denying the class certification motion, but properly based its conclusion on substantial evidence that individual issues would predominate over common issues.
Plaintiffs cite only the first half of the established principle that “[a]t a class certification hearing, the court should not make any determination of the merits or validity of the claim. [Citation.]” (Bartold v. Glendale Federal Bank (2000) 81 Cal.App.4th 816, 829.) The other half of the principle, which applies to this case, is that “when the merits of the claim are enmeshed with class action requirements, the trial court must consider evidence bearing on the factual elements necessary to determine whether to certify the class. ‘When the trial court determines the propriety of class action treatment, “the issue of community of interest is determined on the merits and the plaintiff must establish the community as a matter of fact.” [Citation.]’ (Caro v. Procter & Gamble Co. (1993) 18 Cal.App.4th 644, 656.)” (Bartold v. Glendale Federal Bank, supra, 81 Cal.App.4th at p. 829.)
Contrary to plaintiffs’ position, “the fact some evidence relevant to a class action determination may have been relevant also to the merits of the lawsuit did not preclude the court from considering such evidence at the hearing on [plaintiffs’] motion for class certification.” (Caro v. Proctor & Gamble Co., supra, 18 Cal.App.4th at p. 656.) In this case, the trial court correctly considered evidence regarding the underwriting process, even though such evidence is also relevant to the merits of the lawsuit, in determining whether common issues would predominate over those requiring individual adjudication. The trial court’s determination that common issues would not predominate is supported by substantial evidence, including evidence that: (1) the circumstances of each loan and borrower are different, as illustrated by the fact that the named plaintiffs’ loans required numerous changes and AUS runs before positive scores were obtained; and (2) even loans that receive positive AUS scores are not automatically approved because Freddie Mac requires lenders not to rely solely on the AUS score in determining whether to issue a loan and lenders remain obligated to indemnify or repurchase loans that are resold to Fannie Mae or Freddie Mac.
Although plaintiffs’ expert witness Morrow attested that the underwriting of class loans is an entirely automated process, the record does not show that Morrow was familiar with defendants’ underwriting procedures, either in general or with regard to the named plaintiffs’ individual loans. Significantly, Morrow expressed no knowledge or opinion regarding the adjustments and numerous computer runs that were made regarding the named plaintiffs before positive AUS scores were obtained for their loans.
The premise underlying Morrow’s declaration and the complaint was that the cost of underwriting any conforming, conventional loan with a positive AUS score is $20 to $50. Given that eight computer runs were required to obtain a positive AUS score for the named plaintiffs’ loans, plaintiffs argued that “even if a computer program is run eight times, under plaintiffs’ theory, that only amounts to $160 cost (at about $20 per run). . . . That is a far cry from the minimum $450 fee Countrywide routinely charges for this service.” But if, as alleged in the complaint, it costs $50 per run to use an AUS, eight computer runs would cost $400, which is close to the $450 fee that defendants allegedly charged.
Even under plaintiffs’ theory of automated underwriting, loans that initially do not receive positive AUS scores, such as the loans issued to the named plaintiffs, require human judgment. Morrow stated that “[a]utomated underwriting allows underwriters to focus on loans that are not rated with an ‘Accept’ or similar code, where the underwriting judgment is critical as a manual review of credit and risk is required.” Had defendants not exercised human judgment regarding the named plaintiffs’ loans, the AUS would not have been run multiple times with different figures and the loans would simply have been rejected. Accordingly, even under plaintiffs’ theory of automated underwriting, it appears that individual issues would predominate over common issues because of the varying circumstances of each class loan.
II. The Cited Cases Do Not Support Plaintiffs’ Position
In Linder v. Thrifty Oil Co. (2000) 23 Cal.4th 429, the Supreme Court stated: “When the substantive theories and claims of a proposed class suit are alleged to be without legal or factual merit, the interests of fairness and efficiency are furthered when the contention is resolved in the context of a formal pleading (demurrer) or motion (judgment on the pleadings, summary judgment, or summary adjudication) that affords proper notice and employs clear standards. Were we to condone merit-based challenges as part and parcel of the certification process, similar procedural protections would be necessary to ensure that an otherwise certifiable class is not unfairly denied the opportunity to proceed on legitimate claims.” (Id. at pp. 440-441.)
Citing Linder, plaintiffs contend that “a procedural motion for class certification is no place for merits-based arguments,” which should be resolved by means of demurrer, motion for judgment on the pleadings, summary judgment, or summary adjudication. Plaintiffs analogize this case to Linder, claiming that the trial court, after receiving the parties’ competing interpretations of the term “underwriting,” rejected plaintiffs’ interpretation and improperly denied certification on the ground that plaintiffs could not prevail on their claim.
Unlike Linder, however, the court below did not decide the certification question based on the potential merits of plaintiffs’ lawsuit. There were no merit-based challenges to the complaint below. Defendants did not argue, for example, that the complaint failed to state a viable claim under the UCL, CLRA, or UCSPA, nor did the trial court make such a finding. In this case, defendants properly argued that plaintiffs had failed to meet their burden of establishing a community of interest, which, as previously stated, presents a factual issue that must be decided on the merits. (Bartold v. Glendale Federal Bank, supra, 81 Cal.App.4th at p. 829.)
Plaintiffs also cite Lebrilla v. Farmers Group, Inc. (2004) 119 Cal.App.4th 1070 and In re Cipro Cases I & II (2004) 121 Cal.App.4th 402, for the proposition that the trial court erroneously credited defendants’ witnesses over plaintiffs’ witnesses in concluding, prematurely, that underwriting involves more than the use of an AUS. The order, however, when read as a whole and reasonably construed, does not support plaintiffs’ interpretation that the trial court improperly prejudged the merits of the litigation. The trial court correctly limited its ruling to the question whether common issues would predominate over individual issues.
Plaintiffs’ reliance on Lockheed Martin Corp. v. Superior Court (2003) 29 Cal.4th 1096 is misplaced. In affirming the denial of class certification, the Supreme Court stated that the plaintiffs had failed to “place substantial evidence in the record [to show] that common issues predominate. [Citation.]” (Id. at p. 1108.) Significantly, the Supreme Court acknowledged that in order for the moving party to meet its burden on a class certification motion, substantial evidence must be submitted to show that common issues predominate. The moving party may not, as argued by plaintiffs in this case, rely solely on a theory alleged in the complaint to meet their burden.
Finally, Aguiar v. Cintas Corp. No. 2 (2006) 144 Cal.App.4th 121 does not assist plaintiffs because it is distinguishable. Class certification was denied by the trial court in Aguiar because the proposed class was not ascertainable and lacked a well-defined community of interest, and class treatment was not the superior means to resolve the litigation. The appellate court reversed, stating in relevant part that “[t]he case is replete with common questions of law and fact, a point not addressed by the trial court and not effectively disputed by” the employer. (Id. at p. 136.) In this case, on the other hand, the trial court carefully considered the evidence and concluded that plaintiffs had failed to present substantial evidence to establish that common issues predominate.
III. Unconscionability
Based on the same theories mentioned above, plaintiffs seek reversal of the trial court’s determination that common issues concerning the unconscionability of the underwriting fee do not predominate over questions requiring individual adjudication. For the reasons previously stated, we also reject this contention. Given that the trial court’s ruling on this point is adequately supported on the grounds discussed above, we need not reach the remaining issues raised in plaintiffs’ brief.
DISPOSITION
The order is affirmed. Defendants are awarded their costs.
We concur: EPSTEIN, P. J,. WILLHITE, J.
The CLRA’s exclusive class certification requirements are stated in Civil Code section 1781, subdivision (b). As under Code of Civil Procedure section 382, included among the necessary factors for class certification under Civil Code section 1781, subdivision (b), is the requirement that common questions of law or fact must predominate over the questions requiring individual adjudication. (Civ. Code, § 1781, subd. (b)(2).) In this case, with respect to all of plaintiffs’ claims, including the CLRA claims, we conclude the trial court properly determined that individual issues predominated over common issues.
If by this statement plaintiffs are asserting that they were billed twice for the same services, we reject the contention as unsupported by the pleadings and the record. The complaint, which alleged only that plaintiffs were overcharged for underwriting services, made no reference to double billing. The evidence cited by plaintiffs, consisting of two HUD-1 settlement statements regarding loans issued to the named plaintiffs and a “chart prepared by plaintiffs to illustrate separate services and the fees charged,” do not explicitly refer to double billing. Although plaintiffs contend that the HUD-1 settlement statements showed that “separate fees” were charged “for underwriting, loan processing, field review and document preparation,” the settlement statements do not explain the specific services that were covered by the fees, making it impossible to determine from the documents alone whether there was duplicate billing. As for the chart that was prepared by plaintiffs, because it was not evidence, the trial court considered it “only as Plaintiffs’ counsel’s argument about what they believe[d] the evidence shows.”