Mirkin v. XOOM Energy, LLC

Opinion

UNITED STATES DISTRICT COURT EASTERN DISTRICT OF NEW YORK

SUSANNA MIRKIN, Individually and on Behalf of All

Others Similarly Situated, 18-CV-2949 (ARR) (JAM)

Plaintiffs, OPINION & ORDER -against-

XOOM ENERGY, LLC, and XOOM ENERGY NEW YORK, LLC,

Defendants.

ROSS, United States District Judge: Susanna Mirkin (“plaintiff”), a former residential electricity customer of defendants XOOM Energy, LLC and XOOM Energy New York, LLC (collectively “XOOM”), is the lead plaintiff in the present class action, which alleges that she and other similarly situated customers of XOOM were charged exorbitant energy rates in breach of the pricing terms contained in their contracts with XOOM. In August 2023, I denied XOOM’s motion for summary judgment on plaintiff’s breach of contract claim and granted her motion to certify the case as a class action. Op. & Order, ECF No. 151 (“Summ. J. Op.”); Op. & Order, ECF No. 152 (“Class Cert. Op.”). XOOM unsuccessfully petitioned the Second Circuit for interlocutory review of the class certification order under Rule 23(f), see ECF Nos. 166, 174, after which XOOM moved to decertify the class, see ECF No. 183. I denied that motion. Op. & Order, ECF No. 246 (“Op. Denying Decert.”).

In the midst of briefing the decertification motion, plaintiff moved on April 1, 2024, to prepare and serve an “Amended Expert Report” (the “Amended Report”), as a supplement to her original expert report (the “Original Report”). See Joint Mot. for Discovery Conf. Related to a Dispute Regarding Amended Expert Report, ECF No. 187 (“Joint Mot.”). I permitted plaintiff to supplement her expert report by May 10, 2024, pursuant to Rule 26(e)(2). April 15, 2024, Electronic Order (citing Fed. R. Civ. P. 26(e)(2)). In May 2024, XOOM moved to exclude plaintiff’s Original Report as inadmissible under Fed. R. Evid. 702 and Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993). Mem. in Supp. of XOOM’s Mot. to Exclude Plaintiff’s Timely Expert Disclosures, ECF No. 215 (“Mot. to Exclude O.R.”). XOOM separately moved to

exclude plaintiff’s Amended Report as untimely and inadmissible under Fed. R. Evid. 702 and Daubert. Mem. in Supp. of XOOM’s Mot. to Exclude Plaintiff’s Untimely Expert Disclosures, ECF No. 229 (“Mot. to Exclude A.R.”).1 For the following reasons, I GRANT in part and DENY in part XOOM’s motion to exclude plaintiff’s Original Report because its damages models do not conduct a reliable measure of damages. Furthermore, because Plaintiff’s failure to timely disclose her Amended

1 The Amended Report and Original Report were filed as attachments to XOOM’s motion to exclude the Amended Report. See Matthews Declaration – CRA Orig. Rep., Ex. A-2, ECF No. 229-4 (Original Report); Matthews Declaration – CRA New Rep., Ex. A-3, ECF No. 229-5 (Amended Report). Report was neither substantially justified nor harmless, and the factors guiding this court’s discretion favor exclusion, I GRANT XOOM’s motion to exclude plaintiff’s Amended Report as untimely. Background I. Factual Background2

XOOM is an independent energy service company (“ESCO”) that purchases energy from producers on the wholesale market and sells that energy to consumers as an alternative to local utilities. See Summ. J. Op. at 1–2. During the period relevant to this lawsuit, XOOM sold both electricity and natural gas service at either a variable or a fixed monthly rate.3 See Class Cert. Op. at 2. In the spring of 2013, Mirkin contracted with XOOM to purchase residential electricity service under one of its variable rate plans. See id. According to the contract, the “monthly variable rate” would be “based on XOOM’s actual and estimated supply costs which may include but not be limited to prior period adjustments, inventory and balancing costs.” Decl. of Steven L. Wittels in Opp’n to Mot. Summ. J., Ex. 1 (“Enrollment Email & Contract”) at 4, ECF No. 147-2.

Between 2013 and 2016, XOOM used a standard contract that contained identical language to Mirkin’s contract for its New York residential and small business variable-rate customers who purchased its electricity or natural gas services. See Class Cert. Op. at 2. Plaintiff’s variable rate from XOOM fluctuated significantly, and she canceled her service with XOOM after six months. See Summ. J. Op at 3. In the present action, plaintiff

2 The facts of this case are detailed in my prior opinion denying XOOM’s motion for summary judgment, familiarity with which is assumed. I recount only the facts necessary to resolve the present motion. 3 As self-evident from the naming conventions of those products, the variable-rate plans charged a kilowatt-hour (kWh) rate set by XOOM that fluctuated each month, whereas the fixed-rate plans set a single rate over the contract term which did not vary each month. See Op. Denying Decert. at 13–14. alleges that XOOM failed to set its rates “based on” its “actual and estimated supply costs,” as required by its contracts, and instead set prices by considering impermissible factors, such as profit and revenue goals, that were unmentioned in its contracts. See id. at 3, 5–6. As explained in my prior orders, the parties contest the process by which XOOM set its rates. Op. Denying Decert. at 3. The parties agree that XOOM used spreadsheets to estimate and

calculate certain costs of procuring electricity, which were aggregated in its “rate-setting workbooks” and described as “Total Cost.” Id. In various meetings, XOOM then considered the projected “Total Cost” in setting its final rates to be charged to customers. Id. The difference between XOOM’s rates and its Total Cost—the profit—was XOOM’s “margin.” Id. XOOM asserts that, when setting rates, it incorporated certain supply cost components referenced in its contracts, such as prior period adjustments,4 into its margins. Id. Meanwhile, plaintiff contends that the evidence demonstrates that XOOM set its rates based on its “pricing strategies” untethered to its “actual and estimated supply costs” and calculated the margin retroactively after it set its rates, and therefore did not in practice factor its supply costs as a component of its

margin. Id.; see also Amended Report at 20–21. As I have repeatedly explained, resolution of this disputed factual question will be determined by the jury at trial. See Op. Denying Decert. at 11–13. II. Procedural History On August 14, 2023, I denied summary judgment on plaintiff’s breach of contract claim. Summ. J. Op. at 8–19. I began by adopting plaintiff’s construction of the contract to mean that

4 “Prior period adjustments” refers to XOOM’s practice of raising the monthly rate for a given month to offset unanticipated costs in prior months, such as where XOOM’s actual cost of procuring energy to sell exceeded the estimated costs in its workbooks. Op. Denying Decert. at 3 n.3. XOOM was required to set its monthly variable rates using only its actual and estimated supply costs, subject to a reasonable and proportionate margin. Id. at 12–13. I then identified several genuine fact disputes that would need to be resolved by a jury. Although the central dispute was whether XOOM indeed considered factors beyond its actual and estimated supply costs in setting variable rates, id. at 14–19, smaller sub-issues included whether the Total Cost reflected

XOOM’s actual and estimated supply costs (or whether some supply costs were also captured in the margin), id. at 16, and whether the margin itself was reasonable and proportionate to XOOM’s supply costs, id. at 13–14. Two weeks later, I certified plaintiff’s proposed class. Class Cert. Op. at 7–19. As relevant here, I concluded that plaintiff’s original expert report proposed a mathematical damages model capable of offering common evidence of breach and damages across the class. Id. at 9. That model, referred to by the Original Report as “Model 2,” calculated damages by determining the difference between XOOM’s variable rates and its Total Cost—i.e., the actual margin charged by XOOM—as compared to what the model purported was a reasonable margin.

See Original Report at 25. The report opined that the maximum commercially reasonable margin for XOOM’s variable-rate products was equal to the margin it collected for its fixed-rate products (the “Fixed-Rate Margin”), see id., but also noted that should the court or jury determine a different margin was the maximum which XOOM could charge in compliance with the contract, the model could be “easily . . . update[d] . . . to reflect a different margin,” id. at 24 n.51. The difference between the actual margin for a given monthly rate and a reasonable margin represented the purported “overcharge” to the class for that rate, which, when multiplied by the amount of electricity sold under that rate, yielded the experts’ damages calculations. Op. Denying Decert. at 8, 10. In granting certification, I noted that XOOM disputed whether plaintiff could prove the necessary inputs for her model, namely XOOM’s “actual and estimated supply costs” and the reasonable margin that it could permissibly charge, but I concluded that those were issues of fact for the factfinder to resolve. Class Cert. Op. at 13–14; Summ. J. Op. at 13–17. XOOM filed a Rule 23(f) petition with the Second Circuit, arguing that an appeal was warranted because commonality, predominance, and typicality were not met. See Decl. of

Michael D. Matthews, Jr. Supp. Mot. Stay, Ex. C, at 14–21, ECF No. 169-3. The Second Circuit denied the petition in December 2023, concluding that an immediate appeal was not warranted. Mandate, ECF No. 174. Shortly thereafter, XOOM stated its intent to file a motion for decertification on the basis that plaintiff’s expert damages models were faulty, see ECF Nos. 178, 180, and the motion was fully briefed as of May 10, 2024, see Reply Supp. Mot. Decertify, ECF No. 212. In the midst of that briefing, plaintiff moved to serve an amended expert report, arguing that supplementation was necessary in light of (1) my summary judgment and class certification orders’ interpretation of the contract terms, (2) the Second Circuit’s decision in Martinez v.

Agway Energy Services, LLC, 88 F.4th 401 (2d Cir. 2023), and (3) XOOM’s production of updated data on March 1, 2024, regarding the rates that it charged its customers. See Joint Mot. at 1. I permitted plaintiff to supplement her expert report by May 10, 2024, pursuant to Rule 26(e)(2). April 15, 2024, Electronic Order (citing Fed. R. Civ. P. 26(e)(2)). Plaintiff served her expert’s Amended Report with updated damages calculations on that date. Mot. to Exclude A.R. at 7. On May 20, after service of the Amended Report, XOOM moved to exclude the Original Report as irrelevant and as inadmissible under Fed. R. Evid. 702 and Daubert. Mot. to Exclude O.R at 9–25. On May 30, 2024, XOOM moved to exclude the Amended Report as untimely under Fed. R. Civ. P. 26(a), Mot. to Exclude A.R at 8–21, but also asserted similar relevancy and Fed. R. Evid. 702 arguments to those raised in its motion to exclude the Original Report, id. at 21–30. Those motions were fully briefed as of July 26, 2024. See Opp. to XOOM’s Mots. in Limine to Exclude Plaintiff’s Untimely and Timely Expert Disclosures, ECF No. 258 (“Plaintiff’s Opp.”); Reply in Supp. of XOOM’s Mots. in Limine to Exclude Plaintiff’s Untimely

and Timely Expert Disclosures, ECF No. 262 (“XOOM Reply”). Discussion I. The Original Report Is Inadmissible to Prove Damages under Comcast. According to the Supreme Court's decision in Daubert, Fed. R. Evid. 702 requires district courts to exercise a “gatekeeper function,” Restivo v. Hessemann, 846 F.3d 547, 575 (2d Cir. 2017). Under that analysis, I must “focus on the principles and methodology employed by the expert, without regard to the conclusions the expert has reached[,] . . . [to] assess[] whether the expert appl[ied] his stated methodology reliably to the facts of the case” In re Pfizer Inc. Sec. Litig., 819 F.3d 642, 662 (2d Cir. 2016) (quotation marks omitted). Although the Daubert test is a

“liberal” and “permissive” standard in favor of admissibility, Nimely v. City of New York, 414 F.3d 381, 395–96 (2d Cir. 2005), an expert report that offers an “apples and oranges comparison” warrants exclusion, Restivo, 846 F.3d at 577. The Daubert analysis applies with special force to mathematical “models purporting to serve as evidence of damages in [a] class action,” as such a model must “measure only those damages attributable to t[he plaintiff’s] theory” of injury. Comcast Corp. v. Behrend, 569 U.S. 27, 35 (2013). Thus, “at trial[], any model supporting a plaintiff’s damages case must be consistent with its liability case.” Id. (quotation marks omitted); see also Roach v. T.L. Cannon Corp., 778 F.3d 401, 407 (2d Cir. 2015) (“[A] model for determining classwide damages . . . must actually measure damages that result from the class’s asserted theory of injury.”). Plaintiff’s theory of liability is straightforward. According to plaintiff, XOOM’s contracts required XOOM to charge prices that were “directly proportional” to its supply costs, and therefore the contract effectively “capped” XOOM’s prices at a reasonable percentage over its

costs. Plaintiff’s Opp. to Summ. J. at 9, ECF No. 147. Plaintiff asserts that XOOM charged more than that reasonable percentage over cost, thereby violating its contracts. Id. As plaintiff acknowledges, for her damages to be consistent with that theory of liability, her model must “calculate the amount that XOOM overcharged” its customers. Plaintiff’s Mot. for Class Cert. at 21, ECF No. 132; see also Op. Denying Decert. at 10 (noting that plaintiff’s “theory of liability” was that she was “overcharged”). As also acknowledged by plaintiff, to calculate the overcharge her model must estimate the difference between the rate XOOM actually charged its customers and some other rate that it should have charged its customers. Id.; see also Claridge v. N. Am. Power & Gas, LLC, No. 15-CV-1261, 2016 WL 7009062, at *3 (S.D.N.Y. Nov. 30, 2016) (noting

that damages could be calculated for the class “according to the difference between [defendant’s] rates and those of other market participants” or “the difference between [defendant’s] actual charged rate and a hypothetical rate calculated pursuant to the factors described in [its contracts]”). In her motion for class certification, plaintiff purported that her Original Report’s damages model “reflect[ed]” her liability case and calculated the “overcharge” to customers, Plaintiff’s Mot. for Class Cert. at 21, ECF No. 132 (describing the damages model as an “overcharge formula”). Plaintiff purported that her damages model did so by (1) generating the reasonable prices that “[XOOM] should have charged had XOOM based the rates on its actual reported supply costs” and (2) then comparing those “reasonable price[s]” to XOOM’s actual prices. Id. Faced with no contrary argument from XOOM as to the workings of the Original Report’s damages model, I adopted plaintiff’s construction of her Original Report in my prior orders. See Op. Denying Decert. at 4 (describing plaintiff’s Original Report as “calculat[ing] damages by looking at the differences between the variable rates” and “XOOM’s . . . supply

costs . . . plus a reasonable and proportionate margin”). But upon closer review, the Original Report’s damages model does not perform as plaintiff has advertised. Instead of generating a maximum reasonable price based on XOOM’s actual cost, the Original Report conducts the inverse calculation and generates a minimum cost that XOOM should have paid to its suppliers based on XOOM’s actual price. As a result, the Original Report does not calculate a measurement of damages—the class’s overpayment to XOOM—at all. The Original Report’s methodological error flows from its formula, which her expert describes using the following mathematical notation:

“PriceVar – CostVar – (PriceVar * MarginFix ÷ PriceFix) = Original Report” Decl. of Seabron Adamson, ECF No. 248-1, at 5 (“Adamson Decl.”). As a preliminary matter, Model 2 in the Original Report applies plaintiff’s expert’s opinion that, to be reasonable, the margins on XOOM’s variable-rate products should be less than or equal to the margins that XOOM charged on its fixed-rate products, which the Original Report concludes “averag[ed] approximately 19%.” Original Report at 25.5 In other words, XOOM’s variable-rate prices were

5 The Original Report includes two models. Model 2 assumes that the maximum permissible margin for XOOM’s variable-rate products was XOOM’s margin on its corresponding fixed-rate products. Model 1 assumes that XOOM was not permitted to charge any margin under the contract, and that the reasonable price that XOOM was required to charge was equal to its costs. Original Report at 20. Accordingly, Model 1’s output measures the total amount of margin that reasonable only if at most 19% of that price was margin and the remainder of that price (at least 81%) was cost. The model’s math can then be broken down into three steps, as illustrated in the figure below.

Figure |: Original Report Damages Calculation

Actual | Subtract Price I3tertre) 118) (a Reasonabl AWETesst @ bee, eye ease ft © . NCLB eae PNSIIEY ete) . Margin from | Oiiirettons - Actual Price!

AY Cote ra oe eearnmiel al a bas COLO) )iterni □□□ PACA eee GB sexys sioon baa ene (Oley) ISaer storie: (a eerie □ Actual eles) Cost | | Subtract Actual | . Cost . | | | [dS ee! | Actual Price Reasonable Margin Actual Price Minus Model Output Reasonable Margin First, as represented in the blue bar in Figure 1, the model generates a dollar value for XOOM’s reasonable margin from “a percentage of its variable[-]rate” price, XOOM Reply at 13,

XOOM charged the class. /d. Plaintiff does not contest that XOOM was permitted to charge some margin over its costs and that Model 1 therefore fails to accurately estimate damages.

by multiplying the 19% reasonable margin value by XOOM’s actual price.6 For example, if XOOM charged 10¢ per kWh, the model determines that XOOM’s permissible margin at that price was 1.9¢ (19% of 10¢). Second, the model subtracts that permissible margin from the actual price, yielding a dollar value equal to 81% of the actual price, as represented in the red bar in Figure 1.7 Using the above hypothetical numbers, the damages model yields 8.1¢ (10¢ – 1.9¢).

Critically, that number does not represent the reasonable price that XOOM should have charged its consumers. Instead, that number represents the cost at which XOOM’s actual price was reasonable.8 Third, and finally, the model subtracts XOOM’s actual costs from that value to create its final output, represented as the green bar in Figure 1: the difference between (a) what XOOM’s cost should have been and (b) XOOM’s actual costs.9 The dissection of the Original Report’s formula demonstrates that its output measures XOOM’s costs, not plaintiff’s damages. Plaintiff’s claimed injury is that she was overcharged for energy because she paid prices disproportionate to XOOM’s costs to procure that energy, and the proper measure of damages for that injury must estimate how much less XOOM should have

charged than what it actually charged. However, the Original Report instead measures that injury

6 That calculation is performed through the variables in parentheses in the model’s formula: “(PriceVar * MarginFix ÷ PriceFix).” 7 That calculation is performed through the first and last sets of variables in the model’s formula: “PriceVar – (PriceVar * MarginFix ÷ PriceFix).” 8 The damages model’s calculation of “reasonable cost” rather than a “reasonable price” is represented in its formula. Since the fixed margin is simply the fixed-rate price minus the fixed- rate cost, the first two sets of variables in the formula can be expanded as follows: “PriceVar – (PriceVar * [PriceFix – CostFix] ÷ PriceFix).” When multiplied out, those variables can be expressed as follows: PriceVar * (CostFix÷ PriceFix). CostFix ÷ PriceFix is the percentage of XOOM’s fixed-rate prices that was cost (81%). Multiplying that percentage by XOOM’s actual variable-rate prices thus yields what XOOM’s costs should have been for each of those prices. 9 Or at least, what plaintiff purports are XOOM’s actual costs, as whether the model’s inputs accurately reflect XOOM’s costs is a disputed issue of fact. as if XOOM underpaid the utilities because XOOM bought energy too cheaply, and estimates how much more XOOM should have paid than what it actually paid. The failure of plaintiff’s damages model extends beyond a purely mathematical flaw in its methodology. As XOOM points out, because the Original Report’s damages model determines XOOM’s “reasonable margin” solely by reference to XOOM’s actual prices, the model’s

calculation of the “reasonable margin” is inconsistent with plaintiff’s theory of liability that XOOM’s margins and prices must track its costs. First, the damages model’s “reasonable margin” is wholly disconnected from changes in XOOM’s costs. For example, if XOOM charged a price of 10¢ per kWh in months A and B, the model determines that XOOM’s permissible margin in both months was 1.9¢ per kWh (19% of the price) even if XOOM incurred radically different costs to procure energy in those months. Second, under the model’s method of calculation, XOOM was permitted to increase the size of its “reasonable margin” simply by increasing its actual prices, without any change in its costs. To illustrate, if XOOM charged 10¢ per kWh in month A and 20¢ per kWh in month B, the

model determines that XOOM’s reasonable margin was 1.9¢ in month A and 3.8¢ in month B, even if XOOM incurred identical costs of 5¢ in both months. For that same reason, the model cannot calculate XOOM’s reasonable prices consistent with her liability theory that the commercially reasonable price was XOOM’s costs plus a reasonable margin. When that theory is applied to her damages model, the model’s calculations permit XOOM to increase the reasonable price by increasing its actual prices, regardless of its procurement costs. Using the example numbers above, the damages model determines that XOOM’s “reasonable price” was 6.9¢ in month A and 8.8¢ in month B, even though XOOM incurred identical costs of 5¢ in both months. Due to the above discrepancies, the damages model cannot estimate the reasonable margins and prices proportionate to XOOM’s costs that plaintiff claims she should have been charged. Nor can the model reliably serve as a proxy for those margins and prices. Plaintiff’s basic contention is that XOOM’s prices were not determined by its costs, and therefore the model’s calculation of a reasonable margin from XOOM’s prices cannot be translated into one

created from XOOM’s costs. As a result, the damages model cannot calculate a damages number consistent with plaintiff’s theory of liability. In conclusion, the Original Report contains a fundamental mismatch: it identifies the portion of XOOM’s price that XOOM should have spent on supply costs, whereas a proper model would have identified the portion of XOOM’s price that it should never have charged its customers in the first place. Because plaintiff’s Original Report fails to “measure damages that result from the class’s theory of injury,” Waggoner v. Barclays PLC, 875 F.3d 79, 106 (2d Cir. 2017) (quotation marks omitted), it must be excluded as evidence of damages under Comcast, 569 U.S. at 35 (“[A]t trial[], any model supporting a plaintiff’s damages case must be consistent

with its liability case.” (quotation marks omitted)). II. The Original Report Is Admissible to Prove Breach and Liability under Rule 702 and Daubert

Although plaintiff’s Original Report is fatally flawed as a measure of damages, it nonetheless remains admissible as evidence of liability. Model 2 demonstrates that XOOM’s costs should have been higher than its actual costs in order for its prices to have contained only a reasonable profit, and, by the same token, demonstrates that XOOM’s prices were unjustified by its actual costs. As a result, the Original Report shows that XOOM’s actual prices were excessive in comparison to its costs and is therefore admissible as to breach, even if the Original Report cannot determine how excessive those prices were in comparison to the prices that the class should have been charged and is therefore incapable of reliably measuring plaintiff’s injury. XOOM’s remaining arguments in favor of excluding the Original Report as to plaintiff’s liability case are without merit. Indeed, I have already rejected many of those arguments in my prior order denying XOOM’s motion for class decertification. See Op. Denying Decert. at 12–16.

I address below only those arguments that I have not previously rejected. XOOM first objects that Model 2 calculates the permissible variable-rate margin for each month at the corresponding fixed-rate margin during that same month—and that plaintiff’s theory of liability instead requires her expert to calculate the permissible variable-rate margin as a single percentage value equal to the average fixed-rate margin across the entire class period. Although I agree that plaintiff’s model must use a consistent reasonable margin across the class period, the model can easily accommodate the average fixed-rate margin, and it is well settled that “[a] minor flaw in an expert’s reasoning or a slight modification of an otherwise reliable method will not render an expert’s opinion per se inadmissible.” Amorgianos v. Nat’l R.R.

Passenger Corp., 303 F.3d 256, 267 (2d Cir. 2002). XOOM further argues that plaintiff’s model “improperly ignore[s] undercharges,” in that it does not include a setoff for those months in which XOOM charged a price that was less than the maximum reasonable price. XOOM Reply at 12. XOOM does not contend that the months in which it charged less than what its contracts permitted negates its breach during the months in which it charged more than what its contracts permitted. Id. XOOM’s objection is directed solely at the final damages amount calculated by the model. Id., citing Troitino v. Goodman, 225 N.C. 406, 413 (1945) (the injured party “is entitled to the pecuniary difference between his position upon breach of the contract and what it would have been, had the contract been performed. But he is not entitled to be enriched by the breach.” (emphasis added)). Because I have already concluded that the Original Report does not measure damages and is inadmissible as evidence of damages, XOOM’s objection is moot. Taking the above points together, plaintiff’s Original Report is admissible to demonstrate liability, but is inadmissible to prove damages.

III. XOOM’s Motion in Limine to Exclude Plaintiff’s Amended Expert Report As mentioned above, plaintiff served an Amended Report after the close of expert discovery that purported to “supplement” her Original Report in light of XOOM’s expanded data production on March 1, 2024 and various court orders that “clarified” the construction of XOOM’s contracts. Plaintiff’s Opp. at 30. The Original Report advanced two damages models. Model 1 assumed that XOOM was not permitted to charge a margin under its contracts and therefore calculated damages as if XOOM’s permissible margin was 0%, whereas Model 2 assumed that XOOM’s reasonable margin on its variable-rate contracts was equal to XOOM’s margin on its fixed-rate contracts. Original Report at 20–25. The Amended Report contains three models, A, B, and C, all of which rely on the same formula to calculate damages but are each

based on a distinct opinion, to be asserted by plaintiff’s expert, as to XOOM’s reasonable margin. XOOM argues that plaintiff’s Amended Report is untimely under Rule 26(a)(2)(D), and that the Amended Report should therefore be excluded under Rule 37(c)(1). Plaintiff responds that her Amended Report is a timely supplement under Rule 26(e). Under those rules, a party seeking to use an expert witness’s testimony at trial must disclose the expert’s “written report” that contains, inter alia, “a complete statement of all opinions the witness will express and the basis and the reasons for them,” Fed. R. Civ. P. 26(a)(2)(B), and a failure to make such a timely disclosure prevents the party from using that witness at trial “unless the failure was substantially justified or is harmless,” Fed. R. Civ. P. 37(c)(1). However, a party who has previously offered an expert report must “supplement or correct its” report in a “timely manner if the party learns that in some material respect the disclosure or response is incomplete or incorrect.” Fed. R. Civ. P. 26(e)(1)(A). A. The Amended Report Offers New Opinions.

Rule 26(e) does “not . . . permit a party to supplement an expert report at any time it wishes,” nor does that rule permit a party to disclose new expert opinions by simply labeling those opinions as a “supplement” to a prior report. Coene v. 3M Co., 303 F.R.D. 32, 42 (W.D.N.Y. 2014). Moreover, courts in this Circuit have repeatedly ruled that Rule 26(e) “is not . . . a vehicle to permit a party to serve a deficient opening report and then remedy the deficiency through the expedient of a ‘supplemental’ report.” Anthem, Inc. v. Express Scripts, Inc., 660 F. Supp. 3d 169, 185 (S.D.N.Y. 2023) (internal quotation marks omitted) (collecting cases). 1. The Amended Report Offers Two New Opinions as to the “Reasonable Margin.” Two of the models contained within the Amended Report undeniably rely on opinions that were not asserted in the Original Report but could have been asserted at that time. First, Model A opines that the maximum reasonable margin that XOOM could permissibly charge is 7.33%: the average difference between XOOM’s documented costs in its rate setting workbooks and the price at which its supplier utilities sold directly to their own customers. Amended Report at 30–31. Similarly, Model B opines that XOOM’s maximum permissible margin is 5%: based

on a 2019 New York regulation that capped fixed-rate energy prices at 105% of the average price at which utilities sold energy directly to their customers. Amended Report at 35. Plaintiff’s Original Report did not assert those opinions, much less contain her expert’s basis for doing so. Accordingly, those opinions do not “supplement” her Original Report. 2. The Amended Report’s Calculation of the Fixed-Rate Margin Differs from the Original Report’s Calculation of the Fixed-Rate Margin, but the Significance of that Difference is Immaterial. Unlike Models A and B, Model C in the Amended Report relies upon the same underlying opinion as Model 2 in the Original Report, namely, that the fixed-rate margin was the maximum “reasonable” variable-rate margin. Original Report at 24–25; Amended Report at 39– 41. Although the Original and Amended Reports calculate XOOM’s fixed-rate margins from different data sets, XOOM concedes that, insofar as that difference is concerned, the Amended Report is a permissible supplement given XOOM’s production of further data after the Original Report was prepared. Mot. to Exclude O.R. at 10. XOOM instead raises a different objection. The Amended Report opines that XOOM’s actual fixed-rate margin averaged 21.2% of XOOM’s fixed-rate costs.10 Amended Report at 41. Meanwhile, the Original Report opined that XOOM’s fixed-rate margin averaged 19% of XOOM’s fixed-rate prices.11 Adamson Decl. at 5. According to XOOM, that difference between the two reports’ calculations of the fixed-rate margin means that the Amended Report’s calculation is a new opinion that does not supplement the Original Report.

However, I need not decide whether the Amended Report’s calculation of the fixed-rate margin in Model C is a permissible supplement to the Original Report’s calculation of the fixed- rate margin in Model 2. As will be discussed below, the two reports rely on different formulas, and while both formulas incorporate the average fixed-rate margin as an input, the formulas utilize that input differently. That difference between the reports’ formulas is determinative of whether Model C in the Amended Report advances a new opinion.

10 Put another way, the Amended Report expresses the margin as a cost percentage: Margin ÷ Cost 11 Put another way, the Original Report expresses the margin as a price percentage: Margin ÷ Price At present, it is sufficient to note that the Amended Report’s fixed-rate margin-over-cost percentage (Model C’s input for XOOM’s reasonable margin) can be converted into a margin- over-price percentage (Model 2’s input for XOOM’s reasonable margin) according to a set equation.12 Adamson Decl. at 6–7. Thus, the statement that (A) XOOM’s reasonable margin was at most 21.2% of its actual costs is equivalent to the statement that (B) XOOM’s price was

reasonable if at most 17.5% of that price was margin (21.2% ÷ 121%) and the remaining 82.5% was cost (100% ÷ 121%). 3. The Amended Report’s Calculation of Damages from the Fixed-Rate Margin Relies on a Different Formula than does the Original Report. Although both Model 2 in the Original Report and Model C in the Amended Report use the fixed-rate margin as an input into their respective damages formula, the manner in which the two reports’ formulas employ that input are meaningfully different. Both the Original and Amended Reports purport to calculate the overcharge to the class: the difference between the price XOOM’s customers paid and the “reasonable” price that the customers should have paid. Original Report at 24–25; Amended Report at 39–41. That overcharge, when multiplied by XOOM’s sales figures, and summed across XOOM’s products and the class period, yields the total damages to the class. Original Report at 24–25; Amended Report at 39–41. But, as explained above, the Original Report did not conduct that calculation, and was not consistent with plaintiff’s theory of liability that XOOM’s reasonable price was XOOM’s variable-rate costs plus a margin proportionate to those costs. Accordingly, if the

12 Specifically, price-based and cost-based percentages can be converted into one another using the following equations: PricePercentage = CostPercentage ÷ (CostPercentage + 1) CostPercentage = PricePercentage ÷ (1 – PricePercentage) Amended Report had continued to use the same methodology as the Original Report, the Amended Report would fail for the same reason as the original. However, plaintiff has altered the formula underlying the Original Report in her Amended Report. As explained below, the Amended Report now properly measures “the difference between the variable rate[ price]” and “XOOM’s . . . supply costs . . . plus a

reasonable and proportionate margin.” Op. Denying Decert. at 4. Unfortunately for plaintiff, however, those changes also mean that her Amended Report now advances an untimely damages model that is wholly distinct from the one within her Original Report. Despite plaintiff’s efforts to mask the flaws in her Original Report, and its distinguishability from the Amended Report, by arguing that the two reports’ damages models measure the “same thing,” Plaintiff’s Opp. at 24, an examination of the Amended Report demonstrates the contrary. While plaintiff’s Amended Report may now advance a damages formula “consistent with [her] liability case,” Comcast Corp. v. Behrend, 569 U.S. 27, 35 (2013), the alterations to the Original Report are too great for the former to “supplement” the latter under Rule 26.

As explained by plaintiff’s expert, her Amended Report calculates damages from the following formula: “PriceVar – CostVar – (CostVar * MarginFix ÷ CostFix) = Output” Adamson Decl. at 5. Figure 2: Amended Report Damages Model

Actual Price NY Cele Olay □□□ Ooi [roe ETS)

Servet] 8) Cs : : NY Evecare BAZ Cnaty | fel PANe (Oe) 8) ONCE tele) (a : : □ Ese | Actual PaXe Oe)l oe : : Cost weoeters| | FromActual | ere) : Price, subtract | | Reasonable | " Price : | FN Ores : : | | | Actual Cost Reasonable Margin Reasonable Price Model Output

As shown in Figure 2, the Amended Report’s damages models can be broken into three steps. First, as represented in the blue bar, the Amended Report generates a dollar value for XOOM’s reasonable margin by multiplying XOOM’s costs by 21.2% (XOOM’s average fixed- rate margin as a percentage of its fixed-rate costs).!? For example, if XOOM’s cost was 5¢ per kWh, the model determines that XOOM’s reasonable margin was 1.06¢. That first step introduces the critical difference between the Original and Amended Reports: the Original

13 That calculation is performed by the variables in parentheses in the model’s formula: “(Costvar * Marginrix + CostFix).”

Report’s damages models calculate the reasonable margin by taking a percentage of XOOM’s actual prices, whereas the Amended Report’s models calculate the reasonable margin by taking a percentage of XOOM’s actual costs. Second, the Amended Report’s model adds its calculation of XOOM’s reasonable margin to XOOM’s actual costs, thereby determining a dollar value for XOOM’s reasonable price—

XOOM’s costs plus a reasonable margin proportionate those costs.14 Using the above hypothetical numbers, the model determines that the maximum reasonable price permitted by the contract was 6.06¢ (5¢ + 1.06¢). Third, and finally, the model subtracts its “reasonable price” from XOOM’s actual price to create its final output, represented as the green bar in Figure 2: the amount XOOM overcharged its customers. Thus, assuming that XOOM had charged 10¢ per kWh in my hypothetical scenario, the Amended Report would determine that the class was overcharged by 3.94¢ per kWh (10¢ – 6.06¢). The Amended Report’s measurement—the overcharge to XOOM’s customers—is certainly a more accurate reflection of plaintiff’s theory of injury than that advanced in her

Original Report. However, the question at hand is not the accuracy or admissibility of plaintiff’s Amended Report on its merits. Instead, I must determine whether the Amended Report supplements the Original Report or is a replacement of the latter with undisclosed and untimely theories. See Engler v. MTD Prods., Inc., 304 F.R.D. 349, 356 (N.D.N.Y. 2015) (“[C]ommon sense suggests (and numerous decisions confirm) that an expert report that discloses new opinions is in no way a mere supplement to a prior report.” (internal quotation marks omitted)); Lidle v. Cirrus Design Corp., No. 08-cv-1253, 2009 WL 4907201, at *5 (S.D.N.Y. Dec. 18,

14 That calculation is performed by the latter two sets of variables in the model’s formula: “– CostVar – (CostVar * MarginFix ÷ CostFix)” 2009) (Rule 26 “is not . . . a vehicle to permit a party to serve a deficient opening report and then remedy the deficiency through the expedient of a ‘supplemental’ report.”). The alterations between the two reports’ formulas are too significant for the Amended Report to “supplement” the Original Report. As shown in Exhibit 1: Figure 3 below, a comparison of the Original and Amended Reports’ models demonstrates that the reports use

different formulas and that the outputs created from those formulas measure different concepts.15 Unlike the Original Report, the Amended Report determines a “reasonable margin” that is proportionate to, i.e., is based on, XOOM’s actual costs. As a result, it avoids both of the erroneous results in the Original Report’s model that I have explained above. See supra p. 12. First, unlike the Original Report, the Amended Report’s calculation of XOOM’s reasonable margin is affected by increases or decreases in XOOM’s actual costs. To illustrate, if XOOM’s costs were 5¢ in month A and 10¢ in month B, the model determines that XOOM’s reasonable margin was 1.06¢ and 2.12¢, respectively. By contrast, the Original Report’s model would determine that XOOM’s reasonable margin was 17.5% of its actual prices in those months,

regardless of what costs XOOM had incurred. Second, the Amended Report yields identical dollar values for XOOM’s reasonable margin (and reasonable price) whenever XOOM incurred the same costs. For example, whenever XOOM incurred 10¢ of costs, the Amended Report’s formula always determines that XOOM’s reasonable margin was 2.12¢ and its reasonable price was 12.12¢. By contrast, under the Original Report’s formula, XOOM could increase the

15 Because the Original Report uses price percentages, the Amended Report’s cost-based “reasonable margin” must first be converted to a price-based percentage. As explained above, a 21.2% reasonable margin over cost percentage is equivalent to a 17.5% reasonable margin over price percentage (21.2% ÷ 121.2%). See supra p.18. “reasonable” margin and price by simply increasing its actual prices, regardless of what costs it had incurred. See id. Based on the analysis above, plaintiff’s Amended Report is not a permissible supplement, as it uses different methodology and measures different concepts than does her Original Report. Compare Whole Woman’s Health All. v. Hill, No. 18-cv-1904, 2020 WL 7129727, at *3 (S.D.

Ind. Dec. 3, 2020) (finding new report was not proper supplement where expert “substituted a different comparison between . . . data [sets]” rather than “correcting the report to reflect accurate data” (alterations adopted)); with Sitts v. Dairy Farmers of Am., Inc., No. 16-CV-287, 2020 WL 3467993, at *3 (D. Vt. June 24, 2020) (allowing supplementation where expert used the “same [mathematical formula] he previously employed to reanalyze the data”). The Original Report simply does not measure what plaintiff’s expert claimed that it measured, and as numerous courts in this Circuit have concluded, “mistakes in arithmetic can be corrected (unlike the mistake of using an unreliable methodology).” Malletier v. Dooney & Bourke, Inc., 525 F. Supp. 2d 558, 657 (S.D.N.Y. 2007) (emphasis added). Plaintiff purported that the Original

Report measured class damages and did so by comparing XOOM’s prices to what prices it should have charged, when that report in fact accomplished neither task. Plaintiff now belatedly offers an Amended Report that attempts to correct both errors by substituting a different formula. Rule 26 simply does not permit such gamesmanship, nor does it give “a license to the disclosing party to offer a new theory out of time when the responding party has demonstrated that the disclosing party’s initial theory is incorrect.” Bozick v. Conagra Foods, Inc., No. 19-cv-4045, 2021 WL 1198320, at *3 (S.D.N.Y. Mar. 30, 2021).16 B. The Amended Report is Untimely. Plaintiff’s Amended Report offers new opinions and relies on an entirely new methodology, and is plainly untimely. The parties’ deadline to serve expert reports was October 3, 2022. ECF 115-1. Plaintiff served her Amended Report May 10, 2024. Mot. to Exclude A.R. at 7.

Although I recently permitted plaintiff to serve a supplemental expert report, that order was limited to supplementation of her Original Report. April 15, 2024, Electronic Order (citing Fed. R. Civ. P. 26(e)(2)). It did not, as plaintiff contends, give her carte blanche to offer new expert opinions. Nor did I have occasion to determine whether plaintiff could go further, as plaintiff’s expert had not even written a new report at the time that I issued my order. Joint Mot. at 1. As set forth above, the Amended Report is not a supplemental disclosure under Rule 26(e), see Est. of Jackson by Jackson v. County of Suffolk, No. 12-CV-1455, 2019 WL 1676000, at *2 (E.D.N.Y. Apr. 17, 2019), and therefore remains untimely notwithstanding my order permitting supplementation of the Original Report under that rule.

C. Exclusion of the Amended Report’s Damages Models Is Warranted. The Second Circuit has directed courts in this Circuit to consider four factors in exercising their “wide discretion to impose sanctions” for discovery violations under Rule 37(c)(1): (1) the party’s justification for its late disclosure; (2) the importance of the evidence sought to be precluded; (3) the prejudice suffered by the opposing party a result of having to prepare to meet the new testimony; and (4) the possibility of a continuance. Design Strategy, Inc. v. Davis, 469 F.3d 284, 294, 296 (2d Cir. 2006). The Second Circuit has further counseled district

16 I note that Models A and B in the Amended Report use the same formula as Model C. Amended Report at 29–35. Thus, the untimely changes in Model C that I have identified above also apply to Models A and B. courts to “consider less drastic responses” prior to imposing “the extreme sanction of preclusion.” Outley v. City of New York, 837 F.2d 587, 591 (2d Cir. 1988). The first factor heavily favors XOOM, as plaintiff has failed to present a substantial justification for the timing of her disclosure. Plaintiff principally contends that her expert’s alterations to his report were necessary to conform his models to certain post-discovery rulings

that “clarified the constraints on XOOM’s rate-setting,” Plaintiff’s Opp. at 30, and therefore asserts that her late disclosures should be “viewed in a distinct[ly favorable] light.” Anthem, Inc., 660 F. Supp. 3d at 185 (concluding that supplemental expert report containing new opinions should not be excluded because it was made in response to the court’s “conclusions of law as to the contract at issue in [that] case” that “foreclosed [plaintiff’s] prior damages theories”). Specifically, plaintiff contends that my August 2023 orders denying summary judgment and granting class certification clarified that XOOM was permitted to charge a margin as part of its variable rates, but that its margin and (therefore its rates) were required to be “proportionate” to its supply costs. See Summ. J. Op. at 13 (interpreting XOOM’s contract to require its rates and

margin to “remain proportionate, but not equal to the supply costs over time.”). Plaintiff’s argument rings hollow. Although it is true that, prior to those orders, I had not definitively interpreted the contract according to that reading, my orders merely concluded that plaintiff’s reading of the contract was correct. For example, in opposing summary judgment, plaintiff argued that XOOM’s contract language required it to charge a margin that was “consistent” and “proportional” to its costs, and represented that her Original Report “reflect[ed] th[at] reasonable interpretation.” Plaintiff’s Opp. to Summ. J. at 9, ECF No. 147. Plaintiff cannot seriously contend that her experts were surprised by my “novel” contract interpretation and needed to revise their models when that interpretation was plaintiff’s theory all along, especially given that she represented that her expert reports reflected that theory. There is simply no reason why plaintiff’s Original Report could not have complied with plaintiff’s own reading of the contract. Plaintiff was responsible for informing her expert of her theory of liability and damages. See Comcast, 569 U.S. at 38 (“The first step in a damages study

is the translation of the legal theory of the harmful event into an analysis of the economic impact of that event.” (citation, internal quotation marks, and emphasis omitted)). Her failure to inform her expert is no justification whatsoever. See Matthews Declaration – Adamson Dep. at 3, Ex. A- 1, ECF No. 229-3 (“Q. . . . Were you asked to assume a particular interpretation of the pricing provisions? A. No, not really.”); see also 3M Innovative Props. Co. v. Avery-Dennison Corp., No. 01-CV-1781, 2005 WL 6019652, at *1 (D. Minn. Oct. 21, 2005) (striking portions of a “new supplemental expert report” that included “previously undisclosed” materials where court’s construction of the theory of liability was “[]anticipated” by the parties). Indeed, even if it were true that plaintiff needed to update her models in light of my

interpretation of the contract, plaintiff offers no justification for her failure to notify me or XOOM of her intent to alter her expert report to comply with that interpretation. My summary judgment and class certification orders were decided in August 2023, and plaintiff did not amend her report or provide XOOM with notice of her intent to do so until seven months after those decisions. See Joint Mot. at 1. In response, plaintiff contends that her failure to serve an amended report at an earlier date is justified by XOOM’s March 2024 production of supplemental data, and asserts that it would have been a waste of the parties’ and the court’s resources for her to submit successive reports advancing her new opinions and then updating those opinions with XOOM’s new data. Plaintiff’s Opp. at 31. But of course, while that may justify a delay in producing her Amended Report, it does not justify plaintiff’s failure to disclose her intention to advance new opinions and methodologies. See Rekor Sys., Inc. v. Loughlin, No. 19-CV-7767, 2022 WL 2063857, at *9 (S.D.N.Y. June 8, 2022) (noting that a party can hardly complain that their expert reports should not be excluded as untimely where “[t]hey did not ask for [additional] time”).

Plaintiff also contends that the Second Circuit’s decision in Martinez, 88 F.4th at 414, issued after the close of discovery, made “key pronouncement[s]” that “affected [her] expert’s analysis,” Plaintiff’s Opp. at 11–12. Specifically, plaintiff asserts that Martinez “made clear that XOOM’s contract prohibited the use of subjective process or factors to set monthly variable rates.” Id. at 31. However, the relevant passages in Martinez that discussed XOOM’s contracts had previously been stated in the Second Circuit’s prior decision in this case, see Martinez, 88 F.4th at 411–12 (citing Mirkin v. XOOM Energy, LLC, 931 F.3d 173 (2d Cir. 2019)), and there is no reason why plaintiff’s expert could not have looked to that prior decision to inform his analysis. In any event, plaintiff overreads Martinez. The court did not state that XOOM was

prohibited from using “subjective process[es] or factors” in setting its monthly rates, Plaintiff’s Opp. at 31; rather, the court explained that the plaintiff’s expert in that case could not opine on his “subjective and unsupported view of what a ‘reasonable margin’ would have been,” Martinez, 88 F.4th at 413. That proposition should have been self-evident to plaintiff and her expert. After all, an expert cannot support his opinions with mere ipse dixit. See Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997) (noting that a district court is not required “to admit opinion evidence that is connected to existing data only by the ipse dixit of the expert”). Finally, plaintiff’s justifications are undercut by the repeated insistence of her counsel and expert that the models advanced in the Amended Report “are not substantively different” from those in the Original Report, and that the two reports “express[] the same thing.” See, e.g., Plaintiff’s Opp. at 45. As explained above, that is simply not true, and her counsel’s and expert’s attempt to mask the meaningful differences between the two reports’ methodology can mean only one of two things. Either plaintiff’s counsel and her expert do not fully understand their own models, or they know that the Original Report is flawed and that the Amended Report

corrects its flaws, but have attempted to mislead me into believing that the two models are identical. The former is plainly not a “substantial” justification for plaintiff’s untimely disclosure under Rule 37, see Fed. R. Civ. P. 37(c)(1), whereas the latter amply supports the issuance of sanctions. See Design Strategy, Inc., 469 F.3d at 296 (“[A]lthough a ‘bad-faith’ violation of . . . Rule 26 is not required in order to exclude evidence pursuant to Rule 37, it can be taken into account as part of the party’s explanation for its failure to comply.”). By contrast, the second factor (the importance of the evidence) favors the plaintiff. As I have just explained, the Original Report is inadmissible as evidence of class damages, and plaintiff has offered no other method of calculating damages on a classwide basis. See id. at 297

(finding second factor weighed against exclusion of “evidence of lost profits” as that evidence “was essential to proving these damages”). At the same time, however, I find that the importance of the evidence does not weigh heavily in her favor, as its importance “only serves to underscore” that her “delayed submission” is “inexcusable.” Point Prods. A.G. v. Sony Music Ent., Inc., No. 93-CV-4001, 2004 WL 345551, at *11 (S.D.N.Y. Feb. 23, 2004). This is “not a case in which a new or previously unknown event has recently come to light,” but instead an “eleventh hour effort to rescue a deficient expert report that has been the basis for the course of this litigation.” Id. Weighing the third and fourth factors in tandem, plaintiff’s untimely disclosure of the Amended Report has caused XOOM significant prejudice that would not be fully cured by a continuance of discovery. Plaintiff’s Amended Report was served nineteen months after the close of expert discovery, and the extensive length of that delay itself results in prejudice. See Design Strategy, Inc., 469 F.3d at 297 (noting that “the fact that discovery had been closed for

approximately one and a half years” weighed “heavily [in favor of exclusion] on both the prejudice and possibility of continuance factors.” (internal quotation marks omitted)). Moreover, in the period since the close of discovery, I issued several decisions premised on the notion that plaintiff’s Original Report was consistent with her theory of liability and damages. See Class Cert. Op. at 15 (granting class certification based on the finding that plaintiff “has identified a damages model that would apply to each class member”); Op. Denying Decert. at 16 (finding that plaintiff’s “proposed damages model is consistent with her theory of liability”). Critically, in issuing those orders, I had no occasion to examine the latent flaw underlying the Original Report, as XOOM did not argue “that the [formula] itself [wa]s faulty.” Id. (quoting Class Cert. Op. at

14). However, XOOM’s motion to exclude plaintiff’s Original Report has now revealed that the report’s formula is faulty, and plaintiff’s untimely attempt to correct her models in her Amended Report may require that I now revisit the validity of my prior orders. See Advanced Analytics, Inc. v. Citigroup Glob. Mkts., Inc., 301 F.R.D. 31, 41 (S.D.N.Y.), objections overruled, 301 F.R.D. 47 (S.D.N.Y. 2014) (“Absent preclusion, discovery will need to be re-opened in order to provide Defendants with an opportunity to explore [an expert’s] new revelations that were not previously advanced, rendering [Defendants’ summary judgment] motion, addressed in good faith to the theories and evidence that had been disclosed, an expensive waste of effort.” (alterations adopted and internal quotation marks omitted)). XOOM’s prejudice cannot be fully mitigated by simply rewinding the litigation to the class certification stage and concluding that plaintiff’s Amended Report corrects the methodological flaw that I have identified in the Original Report. Critically, plaintiff notified XOOM of her intention to issue a new report just three days before XOOM’s deadline to file a Daubert challenge to the Original Report. That date was also three weeks after XOOM filed its

decertification motion, which relied on a challenge to the validity of plaintiff’s Original Report and presaged XOOM’s arguments to exclude the Original Report. As discussed above, XOOM is correct that the Original Report is fatally flawed, and plaintiff’s attempt to correct that report in light of XOOM’s arguments amounts to an improper effort “to offer a new theory out of time when the responding party has demonstrated that the disclosing party’s initial theory is incorrect.” Anthem, Inc., 660 F. Supp. 3d at 185 (internal quotation marks omitted). The disclosure of expert reports is not an opportunity for the proponent to seek their opponents’ constructive feedback and to untimely reconfigure the expert’s methodology to rebut any errors revealed by the opponent.

Furthermore, although the Amended Report corrects the flaw in the Original Report that I have identified above, the Amended Report may have introduced other flaws that have yet to be clearly raised or identified. Due to plaintiff’s excessive delay in producing her Amended Report, XOOM has not had sufficient time to fully investigate the Amended Report. To challenge it, XOOM must at minimum obtain further discovery from plaintiff’s expert, including his deposition; be given the opportunity to file a new rebuttal report; and resubmit its motion to exclude the Amended Report, all of which will prolong this litigation and sap the parties’ financial resources. See Morritt v. Stryker Corp., No. 07-CV-2319, 2011 WL 3876960, at *7 (E.D.N.Y. Sept. 1, 2011) (finding admission of untimely expert testimony was “severe, as discovery would have to be reopened to permit additional expert reports and depositions” (internal quotation marks omitted)); Rienzi & Sons, Inc. v. N. Puglisi & F. Industria Paste Alientari S.P.A., No. 08-CV-2540, 2011 WL 1239867 (E.D.N.Y. Mar. 30, 2011) (“[P]rejudice to [the moving party] is particularly great because discovery is closed and would have to be reopened for [the moving party] to appropriately respond to the damages calculations.”).

Finally, XOOM has already filed its motions in limine and submitted its trial exihibits, witness lists, proposed jury instructions, and other pretrial order components, and did so prior to plaintiff’s production and service of her Amended Report. Those submissions might also need to be revised now that plaintiff has submitted that report, which wholly replaces her Original Report. A continuance would require XOOM to redo much of its work over the past year, and, while available in the abstract, would not fully remedy the harms to XOOM resulting from the timing of plaintiff’s disclosure. See Kelly v. Beliv LLC, No. 21-CV-8134, 2024 WL 1076217, at *8 (S.D.N.Y. Mar. 12, 2024) (“[N]otwithstanding the availability of a continuance, a continuance is not appropriate here because discovery is closed.”).

IV. Remedy Accordingly, I GRANT in part and DENY in part XOOM’s motion to exclude plaintiff’s Original Report. The Original Report is admissible as part of plaintiff’s liability case under Rule 702 and Daubert, as it is sufficiently reliable evidence that XOOM’s costs did not justify its prices. However, the Original Report is inadmissible to prove damages, as it is inconsistent with plaintiff’s theory of injury under Comcast and fails to “reflect[] a reliable application” of the expert’s “principles and methods to the facts of the case.” Fed. R. Evid. 702(d) (emphasis added). Furthermore, because Plaintiff’s failure to timely disclose her Amended Report was neither substantially justified nor harmless, and the factors guiding this court’s discretion favor exclusion, I GRANT XOOM’s motion to exclude plaintiff’s Amended Report as untimely. Finally, in light of these orders, the parties are ORDERED to provide supplemental briefing on whether or not classwide proceedings remain viable in this case. The parties are directed to confer and jointly propose a briefing schedule.

SO ORDERED.

/s/ Allyne R. Ross United States District Judge

Dated: September 2, 2023 Brooklyn, New York EXHIBIT 1 Figure 3: Demonstrating that Original and Amended Report Calculate Different Outputs Based on Their Conflicting Conceptions of the Reasonable Margin Actual | © Fo Price 1 Reasonable Margin I XOoM ! based on Actual Price Charged toy Charged to Customers I I Price) I PNants □□□ cs Oe OTE CONC ieee Grose)

Oereatiey ROluitt □□□□□□□□□□□□□□□□□□□□□ O18)

1 Reasonable Margin I I based on Actual Cost to | XOOM __ □ (21.2% of Actual Cost) 4! Actual | 1 I \ Cost ! I I to XOOM , | I I | I I i Actual Cost to I I Actual Cost to I | I I | I I | I I | I I 1 I Po tl Po Original Report Amended Report 33