Valelly v. Merrill Lynch, Pierce, Fenner & Smith Incorporated

Opinion

UNITED STATES DISTRICT COURT #: SOUTHERN DISTRICT OF NEW YORK DATE FILED: 10/11/24 -------------------------------------------------------------- X SARAH VALELLY, on behalf of herself, : individually, and on behalf of all others similarly- : situated, : : Plaintiff, : 19-CV-7998 (VEC) : -against- : OPINION & ORDER1 : : MERRILL LYNCH, PIERCE, FENNER & : SMITH INC., : : Defendant. : -------------------------------------------------------------- X VALERIE CAPRONI, United States District Judge: This case involves the “sweep” feature of Merrill Edge Self-Directed Investing Accounts. The sweep feature allows Defendant to move automatically or “sweep” Plaintiff’s uninvested cash into a Bank of America money market account.2 On December 28, 2023, the parties filed Daubert motions to exclude the testimony of the opposing party’s expert. Pl. Mot., Dkt. 223; 1 This Opinion & Order will be filed entirely under seal with viewing limited to the parties. The Court notes, however, that although it has previously approved filing certain materials under seal in this case, the redactions applied to material filed in connection with the instant motion appear to have been inconsistently applied. Moreover, the Court is skeptical that this Opinion & Order contains information that overcomes the presumption of public access. See Lugosch v. Pyramid Co. of Onondaga, 435 F.3d 110, 126 (2d Cir. 2006). Accordingly, Defendant must show cause by no later than October 11, 2024, why any portion of this Opinion & Order should be sealed given the Lugosch presumption of access. 2 Plaintiff’s initial complaint asserted claims for quasi contract, breach of contract, breach of suitability standards, and breach of the Massachusetts Consumer Protection Law on behalf of herself and three putative classes. Compl., Dkt. 1. On June 3, 2020, the Court granted Defendant’s motion to dismiss the complaint, but allowed Plaintiff to move for leave to amend her breach of contract claim. June 3, 2020 Opinion & Order, Dkt. 31. On January 25, 2021, the Court granted Plaintiff leave to file an amended complaint, January 25, 2021 Opinion & Order, Dkt. 54, which included additional allegations in support of her claim that Defendant breached the “reasonable rate” provision of the Client Relationship Agreement. First Am. Compl., Dkt. 55 (the “FAC”) ¶¶ 232– 355. The FAC also alleges a “new claim” regarding the interest rate Defendant paid on Plaintiff’s “linked” retirement accounts, id. ¶¶ 2, 356–71, and renewed Plaintiff’s claim for unfair and deceptive trade practices under the Massachusetts Consumer Protection Law, id. ¶¶ 384–93. Def. Mot., Dkt. 222. Defendant also moves to strike portions of Plaintiff’s expert’s rebuttal report and other supplemental email submissions. Defs. Mem. at 1, Dkt. 225. For the following reasons, the Court: (1) DENIES Plaintiff’s motion to exclude the testimony of Defendant’s expert, Dr. Andrea Eisfeldt; (2) GRANTS in part and DENIES in part

Defendant’s motion to exclude the testimony of Plaintiff’s expert, Dr. Darius Palia, and GRANTS Defendant’s motion to strike Dr. Palia’s supplemental “reports.” BACKGROUND3 The Court assumes familiarity with its prior opinions issued over the course of this litigation and will summarize only the most pertinent facts. In August 2017, Plaintiff Sarah Valelly opened three accounts at Merrill Lynch: (i) a Cash Management Account (“CMA”); (ii) a Roth Individual Retirement Account (“Roth IRA”); and (iii) a Traditional Individual Retirement Account (“Traditional IRA”). First Am. Compl. (“FAC”) ¶¶ 26, 43–45, 112, Dkt. 55. The Client Relationship Agreement (“CRA”), which governs all three accounts, contains a so-called “reasonable rate” provision. Id. ¶ 36. Pursuant to the reasonable rate provision, Defendant is

obligated to pay no less than a “reasonable rate” of interest on cash held in Plaintiff’s retirement accounts. Id. ¶ 88. Plaintiff alleges that Defendant breached that contract by failing to pay a “reasonable” interest rate (the “Reasonable Rate Claim”) on her swept cash. Id. ¶¶ 36, 88–89, 226–34. Plaintiff also alleges that Defendant breached the implied covenant of good faith and fair dealing by failing to consider her retirement accounts “linked,” which would have resulted in

3 The Court will refer to the relevant submissions as follows: Plaintiff’s memorandum of law in support of her motion, Dkt. 224, as “Pl. Mem.”; Defendant’s memorandum in opposition to Plaintiff’s motion, Dkt. 240, as “Def. Opp.”; Plaintiff’s Reply memorandum, Dkt. 253, as “Pl. Reply”; Defendant’s memorandum of law in support of its motion, Dkt. 225, as “Def. Mem.”; Plaintiff’s memorandum of law in opposition to Defendant’s motion, Dkt. 244, as “Pl. Opp.”; and Defendant’s Reply Memorandum, Dkt. 251, as “Def. Reply.” Citations to alphabetical exhibits (e.g., Ex. A) refer to the exhibits attached to the Declarations of Paul Mishkin, Dkts. 229, 242, 252, the exhibits attached to the Declarations of Emer Burke, Dkts. 228, 231, 255, and the exhibits attached to the Declaration of Antoinette Adesanya, Dkt. 246. a higher interest rate being paid on the swept cash (the “Statement-Linking Claim”). Id. ¶¶ 2, 356–71. The parties identified their respective experts on June 16, 2023. Pl. Mem. at 7, Dkt. 224. Plaintiff disclosed to Defendant that she would rely on an expert report prepared by Dr. Darius

Palia (the “Palia Report”); Defendant disclosed that it would rely on an expert report prepared by Dr. Andrea Eisfeldt (the “Eisfeldt Report”). Id. The parties exchanged written reports on June 23, 2023, and rebuttal reports on July 20, 2023. Id. On September 15, 2023, Defendant moved for summary judgment and moved to exclude the testimony of Dr. Palia.4 Dkts. 191, 201. On the same day, Plaintiff moved for partial summary judgment and moved to exclude the testimony of Dr. Eisfeldt. Dkt. 192. On October 5, 2023, Defendant moved to strike additional analyses of Plaintiff’s proposed expert. Dkt. 211. Thereafter, the Court dismissed the parties’ motions without prejudice, Dkt. 214, and ordered the parties to file new Daubert motions before moving for summary judgment, Dkt. 218. The parties each filed their Daubert motions on December 28, 2023. Dkts. 222, 223. Each party opposed

the other party’s motion, Dkts. 241, 245. I. Dr. Eisfeldt’s Report Because neither party contests the expertise of the other party’s expert, the Court will not discuss their credentials at length. Dr. Eisfeldt is an economist on the faculty of the Anderson School of Management at the University of California, Los Angeles. Eisfeldt Report ¶ 1, Ex. D, Dkt. 231–4. Her research examines fixed income investing, banking, interest rates, over-the-counter markets, and equity-

4 Prior to this stage of the litigation, Defendant had filed a Daubert motion to exclude Plaintiff’s previous expert, Dr. Officer, Dkt. 112; the Court granted Defendant’s motion on March 21, 2023, Dkt. 149. Plaintiff filed a motion for reconsideration, Dkt. 159, which the Court denied on June 28, 2023, Dkt. 182. based compensation. Id. Her research has received multiple prizes.5 Id. Outside of academia, Dr. Eisfeldt has eight years of asset management experience, having been a consultant and then Chief Economist at Structured Portfolio Management, a mortgage hedge fund; she also served as a consultant for AQR Capital Management, an asset management firm. Id. ¶ 3.

Defendant engaged Dr. Eisfeldt to compare the interest rates paid by Bank of America, N.A. (“BANA”) on deposits in Defendant’s Retirement Asset Savings Program (“RASP”)6 to interest rates paid on comparable products by other firms, and to discuss issues related to Plaintiff’s allegations that her accounts did not receive the benefits of “statement linking.” Id. ¶ 5. A. Dr. Eisfeldt’s Analysis of RASP Interest Rates A sweep program is a service offered by brokerage firms; it typically involves the automatic transfer of uninvested cash from the brokerage account into a deposit account at a bank that may or may not be affiliated with the broker-dealer. Id. ¶ 14. Brokerage firms offer sweep programs connected to retirement accounts to permit any cash in the account that is sitting

idle to be deposited into a bank account where the cash can accrue interest until it is invested in something else. Id. ¶ 15. Some sweep options transfer cash to banks that are affiliated with the broker, and others transfer cash to banks that are unaffiliated with the broker. Id. ¶ 16. The two

5 Specifically, Dr. Eisfeldt’s research has twice received the Amundi Smith Breeden Prize, awarded to the best papers published in the Journal of Finance each year; she has also received the Jensen Prize in the Journal of Financial Economics, as well as grants from the National Science Foundation Grant and the Banque du France. Eisfeldt Report ¶ 1. 6 Defendant’s RASP automatically transfers cash balances of $1 or more daily to one of its participating institutions – BANA being the primary one. Id. ¶ 33. RASP sweeps cash to the primary depository institution until the cash balance nears a threshold of $246,000; subsequent cash is then swept to a secondary depository institution. Id. ¶ 34. The interest rates are determined by BANA, and the interest rate depends on the asset tier to which an account belongs. Id. ¶ 35. The asset tier is determined by the total asset value in the eligible retirement accounts and sweep accounts that are linked. Id. Because a RASP account is connected to a retirement account, as opposed to being a standalone money market deposit account, Dr. Eisfeldt opined that an appropriate analysis of RASP interest rates should look to other programs that function similarly. Id. ¶ 36. primary benefits to the account holder of using a sweep program are: (i) the ability to earn interest on cash that would otherwise be idle, and (ii) the ability to trade quickly with available cash. Id. ¶¶ 17–18. Sweep programs differ from other interest-bearing products and, as such, they pay interest at different rates than other types of cash investments, including savings

accounts, money market mutual funds, stable value funds, and non-sweep money market deposit accounts. Id. ¶ 20. Dr. Eisfeldt explained why, in her opinion, interest rates paid on funds in non-sweep accounts are not comparable to interest rates paid by sweep programs. First, cash affected by a sweep program is transitory; the cash is automatically on deposit until the customer decides to invest the cash elsewhere. Id. ¶ 21. Second, because sweep programs hold cash for the period between investment decisions, the funds must be readily available so that they can be reinvested without delay. Id. ¶ 22. The balance of sweep programs can be volatile, and banks typically offer higher interest rates to deposit accounts with more stable and growing balances. Id. ¶ 23. Sweep programs tend to receive lower interest rates compared to deposits that have more

predictable balances, like certificates of deposit, because cash from sweep programs is less attractive to banks. Id. Third, sweep programs that are insured by the FDIC, like Defendant’s, can guarantee capital preservation for customers until they are ready to invest the cash; products such as money market mutual funds (“MMFs”) have returns that are based on the performance of the underlying securities, and those returns can decrease when the market dips. Id. ¶ 24. Government MMFs are low-risk investments because government securities are backed by the full faith and credit of the U.S. government. Id. ¶ 26. Although MMFs and sweep programs are both very low-risk, there are key differences. Among other differences, there can be uncertainty regarding the U.S. Treasury’s ability to repay existing debt and issue new debt due to the debt

ceiling; that uncertainty can lead to liquidity strains on government MMFs. Id. ¶¶ 27–28. FDIC insurance, in contrast, is funded by insurance premiums assessed on deposits at insured banks; the premiums are paid to the Deposit Insurance Fund, which reimburses depositors if a bank fails. Because of this fund, the U.S. debt ceiling is not a primary source of risk for FDIC-insured accounts as it is for government MMFs. Id. ¶ 29. Fourth, the sweep programs at issue are

different from other interest-bearing products because they are part of the retirement brokerage account; accordingly, funds do not need to be withdrawn or shares sold before the cash can be invested. Id. ¶ 31. Dr. Eisfeldt opined that Crane7 provides market data on sweep rates from comparable programs to which RASP interest rates should be compared. Id. ¶ 37. In order to assess whether the sweep programs reflected in the data Crane reports are comparable to RASP, Dr. Eisfeldt reviewed the underlying programs that make up the Crane Brokerage Sweep Index and considered whether the programs are FDIC-insured, available for retirement accounts, and have the ability to automatically transfer cash from the brokerage account to the sweep account. Id. ¶ 41. According to Dr. Eisfeldt, every program in the Crane data satisfied the criteria except for two.8 Id. ¶ 41. Although Dr. Eisfeldt identified differences in how these programs are

7 Crane is recognized as a reputable source for data about money markets by the press and peer-reviewed journals, and its data has been used by regulators and academics. Id. ¶ 38. Fidelity used Crane data on sweep rates as a proxy for the industry average in its advertising. Id. Crane’s “Broker Sweep Intelligence” is a product available by subscription. Id. ¶ 39. That product summarizes the interest rates paid by the sweep programs of 11 of the U.S. broker-dealers that represent the largest and most comparable entities to Defendant that offer sweep programs. Id. Crane collects interest rate data weekly. Id. Dr. Eisfeldt states that the Crane data is sufficiently granular that it provides interest rate data for different asset tiers. Id. ¶ 40. Crane computes its own “Crane Brokerage Sweep Index,” which averages the main sweep options for the 11 brokerages it covers at each asset tier level. Id. 8 The two exceptions are Fidelity’s “Cash Management Account” and E*Trade’s “Extended Insurance Sweep Deposit Program”; both are FDIC-insured sweep programs, but neither is available for retirement accounts. Id. ¶ 41 n.57. These two brokerages offer separate FDIC-insured sweep programs for retirement accounts, which Crane does not track. Id. Dr. Eisfeldt confirmed that the inclusion of these two programs in the Crane index is either conservative or unlikely to affect her analysis. Id. ¶ 41. implemented, she concluded that the differences are not substantive and do not affect their suitability as comparators for RASP.9 Id. ¶ 41. Dr. Eisfeldt noted that BANA monitored the interest rates paid by on a monthly basis and considered those sweep programs to be comparable to

RASP. Id. ¶ 46. In addition to those three brokerages, Dr. Eisfeldt noted that the Crane data includes discount and online brokerages (i.e., E*Trade, TD Ameritrade, and Fidelity), which BANA does not consider to be Id. ¶ 47. Dr. Eisfeldt’s analysis found that, from December 2016 to December 2022 across all asset tiers, the simple average of RASP rates was 0.009 percentage points lower than the median rate of comparable programs. Id. ¶ 49. The average difference ranges from -0.043 to 0.036 percentage points, which includes periods of time when the average RASP rate was higher than, equal to, or lower than the median rate. Id. Dr. Eisfeldt compared the equally-weighted average of RASP rates to the 25th percentile of other comparable programs to assess whether RASP rates were an outlier on the low end of interest rates. Id. ¶ 50. When compared to the 25th percentile, Dr. Eisfeldt found average RASP rates to be 0.043 percentage points higher.10 Id. RASP rates were not an outlier compared to the interest

9 Brokerages differ in their definitions of asset tiers and the number of tiers offered, which reflect business decisions about attracting customers; those differences do not affect how the programs function. Id. ¶ 42. Further, the relationships between the brokerages and the depository institutions to which cash is swept vary, as some brokerages sweep cash to affiliated institutions, and some sweep to both affiliated and non-affiliated banks. Id. ¶ 43. Aside from Merrill, three of the 11 institutions only sweep to affiliated institutions; five sweep to both affiliated and non-affiliated banks. Id. Of the five brokerages that sweep to both, customers receive a single interest rate regardless of the institution into which their cash is swept. Id. Programs have differing compensation arrangements between brokerages and depository institutions; some brokerage firms receive a flat fee per account while others receive a percentage fee based on the level of assets deposited. Id. ¶ 45. 10 RASP Tier 1 rates were 0.004 percentage points lower than the 25th percentile; Tier 2 rates were 0.034 percentage points higher than the 25th percentile; Tier 3 rates were 0.030 percentage points higher than the 25th percentile; Tier 4 rates were 0.111 percentage points higher than the 25th percentile. Id. ¶ 50 n.73. rates offered by comparable institutions for the measured time period.11 Id. Dr. Eisfeldt then created benchmark rates by taking the median and 25th percentile interest rates and compared them to the RASP rate. Id. ¶ 52. Dr. Eisfeldt created graphs for each asset tier. Those graphs show that, in 2016, rates were close to zero for both RASP and the benchmark rates; thereafter,

interest rates began to rise with RASP leading the pack. Id. ¶ 53. Interest rates for other programs continued to rise after RASP rates leveled off, with RASP rates falling below the benchmark rates.12 Id. Dr. Eisfeldt ’s conclusion is that RASP rates were comparable to rates offered by competitive sweep programs. Id. ¶¶ 55, 57. Although Dr. Eisfeldt also analyzed Plaintiff’s “statement linking” claim, because Plaintiff’s Daubert Motion does not seek to exclude that portion of her report, it will not be discussed here. II. Dr. Palia’s Report Dr. Palia has a Ph.D. in finance from New York University and is on the faculty of the Rutgers Business School. Palia Report, Ex. A, App’x C at 10; ¶ 1, Dkt. 229–1. Dr. Palia has

published in top finance and economics academic journals, and his work has won academic prizes.13 Id. ¶ 3. Plaintiff engaged Dr. Palia to opine on whether Defendant’s RASP rates were “reasonable,” to develop and implement a damages model on behalf of the putative class (the

11 From December 2016 to December 2022, across all asset tiers, an equally-weighted average of RASP rates was 0.008 percentage points lower than the average rate of sweep programs. Id. ¶ 10 n.74.

12 The RASP rates for Tier 4 were the exception; in 2019 those rates were higher than the interest rates paid by other programs. Id. ¶ 53 n.75.

13 He has published papers in the American Economic Review, Journal of Finance, Journal of Financial Economics, Journal of Law and Economic, Journal of Financial and Quantitative Analysis, Review of Financial Studies, and RAND Journal of Economics. Id. ¶ 3. “Reasonable Rate Class” in the report includes persons with Merrill Edge retirement accounts with cash balances swept under RASP between December 15, 2016 through March 15, 2020), and to analyze the interest rates paid by online retirement sweep accounts and calculate damages for all persons with Merrill Edge retirement accounts with cash balances swept under RASP

from January 1, 2022, through May 32, 2023. Id. ¶¶ 8–9. A. Dr. Palia’s Analysis Whether RASP Rates Were “Reasonable”14 Dr. Palia’s report opines that Defendant set its sweep rates in a manner inconsistent with the fair market value standard15 and that the RASP rates paid were not reasonable. Id. ¶ 20. According to Dr. Palia, there are widespread conflicts of interest across the retail brokerage industry. Accordingly, brokerage firms that sweep funds to an affiliated bank cannot be valid comparators against which to assess whether RASP rates were reasonable. Id. ¶ 15. Dr. Palia opines that Defendant’s methodology for setting interest rates yielded unreasonably low rates for swept cash in RASP accounts because the methodology was designed to maximize net interest income to Defendant by reducing and delaying interest rate increases paid to the customer on

swept cash. Id. ¶¶ 17–18. BANA sets the interest rate it pays on Cash Management Accounts (“CMA”) based on “the client’s overall assets under management”; “[h]igher tiered accounts receive a higher-pass- through of market rates.” Id. ¶ 58. The rates of interest paid are based on a pricing index called

14 Although the Court has repeatedly held that a government MMF is not a relevant comparator to an FDIC- insured account for purposes of determining whether Defendant breached its contractual obligation to pay a reasonable interest rate, see March 21, 2023 Opinion and Order at 19, Dkt. 149, Dr. Palia strongly disagrees. His report unsuccessfully attempted to persuade the Court that government MMFs are relevant comparators. Palia Report ¶ 23. Because Plaintiff abandoned that portion of Dr. Palia’s report, the Court will not discuss it further.

15 It should be noted that Plaintiff does not claim that Defendant failed to pay a “fair market value” interest rate on her swept cash. Her claim is that Defendant breached its contract with her by failing to pay a “reasonable rate” of interest on the swept cash. the Deposit Benchmark Rate (“DBR”); the DBR is calculated Ts 5 59. □□ fanuary 2017, Bana’ Global Wealth and Investment Management (““GWIM”) had revised its deposit rate formulas to change the pass through rates.!’ Jd. § 65. As compared to the formula used in late 2004 to late 2006,'® the formulas in use by 2017 reduced the pass through rates. Jd. § 61. In 2014, Defendant had introduced an FDIC-insured Preferred Deposit (“PD”) product for non-IRA account holders with large cash balances; PD offered rates on sweep deposits equivalent to rates paid on MMFs. According to Dr. Palia, the PD product allowed account holders with larger balances to reap benefits not available to account holders with lower balances. Jd. f§ 16, 18. Internally, BANA explained the lower rates it had implemented by January 2017 as a way to offset the lower net interest margins it was experiencing on its “new higher priced [PD] product.” Jd. { 61. RASP rates for customers in Tiers 1 and 2 did not increase after July 2017, and rates for customers in Tiers 3 and 4 increased only marginally. Jd. § 74. Dr. Palia provides a list of hurdles that, in his opinion, prevented customers from accessing higher sweep rates.'? Jd. 78— 83.

© Theres an adjustnent ts EE 1c 0. 17 In January 2017, the pass through rates were for customers in Tier i for customers in Tier 2. i for customers in Tier 3. and for customers in Tier 4. Id. 9 65. 18 Dr. Palia relied on the spreads between the tiers in the 2004—2006 time period in his analysis; he did not explain, however, why that is a relevant period for purposes of comparison. /d. {{ 140-41. Plaintiff asserted in her legal memorandum that Dr. Palia used July 2004—July 2006 as “the control period, given the similar interest environment.” Pl. Opp. at 12 n.25, Dkt. 244. 19 Products on which Defendant paid higher interest, such as PD, were not available in retirement accounts. 478. According to Dr. Palia, Defendant used procedural barriers to prevent “cash sorting” by retail investors. 479. Cash sorting refers to the process by which customers move their cash to accounts paying higher rates. Jd. Defendant also took steps to restrict the ability of RASP customers to access higher rates of interest. Jd. 80. In 2019, for example, Defendant eliminated the ability of retirement account holders to sweep cash into MMFs and to execute trades against an MMF as if the funds in the MMF were cash. Jd. § 82.

According to Dr. Palia, “[Defendant] set interest rates on cash swept out of retirement accounts with the goal of maximizing profit under market conditions inconsistent with the fair market value definition of a competitive market.” Id. ¶ 108. B. Dr. Palia’s Model For Determining a Reasonable Rate of Interest

Dr. Palia employed a pass through model to examine how the cost of money is passed through to various financial products. Id. ¶ 112. His model examined the percentage of change in the effective federal funds rate that is passed through to other interest rates.20 Id. Dr. Palia’s analysis entirely ignored sweep rates offered on FDIC-insured deposits by other brokerage firms that sweep funds to affiliated banks because those institutions have a conflict of interest. Id. ¶ 118. Dr. Palia’s model21 considered only the pass through rates on retail government MMFs22 offered as sweep options by Defendant’s competitors and savings accounts at online banks that Defendant used as a benchmark for its PD accounts. Id. ¶ 120. Dr. Palia also estimates a pass through model for the period 2016–2023 for online savings accounts offered by Ally Bank, American Express Bank, Barclays, and Goldman Sachs/Marcus. Id. ¶ 133.

When computing the range of reasonable interest rates, Dr. Palia first compared the result of his pass through model to rates paid by Defendant on Tier 4 accounts. Id. ¶¶ 137–38. Dr. Palia relied on the spread between the tiers in the 2004–2006 time period to estimate a reasonable RASP rate for Tiers 1–3 in the 2015–2019 time period. Id. ¶ 140. Dr. Palia

20 The federal funds rate is the interest rate at which banks lend funds to each other overnight. Id. ¶ 113. The federal funds rate is calculated by the Federal Reserve as the volume-weighted median interest rate charged by banks on overnight lending. Id. Changes in the federal funds rate quickly affect interest rates that banks and other lenders charge on short term loans. Id. ¶ 115.

21 The pass through model Dr. Palia uses has been widely used in studies published in reputable peer- reviewed academic finance and economic journals. Id. ¶ 125.

22 As noted supra n.14, Dr. Palia squandered pages of his report using MMFs as a comparator for RASP rates, an argument that Plaintiff has wisely abandoned. Pl. Opp. at 19 n.35, Dkt. 244. The Court will, therefore, only discuss Dr. Palia’s model based on the rates paid on savings accounts by online banks. concluded that the rates BANA paid on its RASP accounts were significantly lower than the estimated range of reasonable rates that he calculated using his model. Id. ¶ 145. To calculate damages, Dr. Palia multiplied the aggregate monthly sweep balances in each tier by the difference between the lowest forecasted RASP rate generated using his model and the

corresponding RASP rate paid for each period between December 2016 to March 2020 and then summed the monthly difference for each tier. Id. ¶ 153. Dr. Palia estimates damages of $68.6 million between December 2016 to March 2020. Id. Dr. Palia also calculated the damages for the putative class period of January 2022 to April 2023 as $83.8 million. Id. ¶ 153, Table 8. Combining the damage assessments for both periods, he opines that the total damages of the putative class are $152.4 million. Id. C. Dr. Palia’s Rebuttal Report Dr. Palia introduced additional comparators in his rebuttal report. Palia Rebuttal Report ¶ 80, Ex. E, Dkt. 229–5. He looked at rates offered on money market deposit accounts (“MMDAs”) by Ally and Discover, online banks. Id. ¶¶ 80, 85. According to Dr. Palia, Merrill

. Id. ¶ 84. Dr. Palia opines that comparing the RASP rates to the rates for the select MMDAs demonstrates that RASP rates are lower than comparable products that satisfy the fair market standard. Id. In the rebuttal report, Dr. Palia also looked at the interest rates paid by online banks on MMDAs with balances greater than $50,000 as reflected in data compiled by Informa Research Services (“Informa”).23 Id. ¶ 94. Dr. Palia opines that the average interest rates paid by online banks as reported by Informa provide a benchmark against which to compare the

23 The Informa database does not report interest rates paid on MMDAs. Instead, Dr. Palia used the rate that online banks paid on personal checking accounts, noting that the interest rates for MMDAs are generally higher than the rates for checking accounts. Id. ¶ 95. RASP rates; that comparison shows that RASP rates were not reasonable. Id. ¶¶ 81, 94. Dr. Palia calculated that RASP accounts would have earned $169.9 million more in interest between December 2016 and March 2020 had the accounts received the Informa benchmark rate instead of the RASP rate. Id. ¶ 97.

DISCUSSION I. Legal Standard Federal Rule of Evidence 702 governs expert testimony. It provides that a person “qualified as an expert by knowledge, skill, experience, training, or education” may offer opinion testimony if: (a) the expert’s scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; (b) the testimony is based on sufficient facts or data; (c) the testimony is the product of reliable principles and methods; and (d) the expert’s opinion reflects a reliable application of the the principles and methods to the facts of the case.

Fed. R. Evid. 702. “It is a well-accepted principle that Rule 702 embodies a liberal standard of admissibility for expert opinions [.]” Nimely v. City of New York, 414 F.3d 381, 395 (2d Cir. 2005). The proffering party bears the burden of establishing admissibility under Rule 702 by showing that (1) the expert is qualified; (2) the proposed opinion is based on reliable data and methodology; and (3) the proposed testimony would be helpful to the trier of fact; the district court serves as the “ultimate gatekeeper” against unreliable expert testimony. United States v. Williams, 506 F.3d 151, 160 (2d Cir. 2007). The initial threshold question is whether the “proffered expert testimony is relevant.” Amorgianos v. Nat’l R.R. Passenger Corp., 303 F.3d 256, 265 (2d Cir. 2002). An expert’s opinion is relevant if it will “help the trier of fact to understand the evidence or to determine a fact in issue.” Fed. R. Evid. 702(a). Proffered testimony is not helpful to the jury if it “usurps either the role of the trial judge in instructing the jury as to the applicable law or the role of the jury in applying that law to the facts before it.” Nimely, 414 F.3d at 397 (cleaned up). An expert’s opinion will be precluded if it “undertakes to tell the jury what result to reach” and

“attempts to substitute the expert’s judgment for the jury’s.” Id. (quotation omitted). “While an expert may opine on an issue of fact within the jury’s province, an expert may not give testimony stating ultimate legal conclusions based on those facts.” Snyder v. Wells Fargo Bank, N.A., 2012 WL 4876938, at *5 (S.D.N.Y. Oct. 15, 2012) (cleaned up). “Next, the district court must determine ‘whether the proffered testimony has a sufficiently reliable foundation to permit it to be considered.’” Amorgianos, 303 F.3d at 265 (quoting Campbell ex rel. Campbell v. Metro. Prop. & Cas. Ins. Co., 239 F.3d 179, 184 (2d Cir. 2001)). The Court considers, inter alia, whether (1) “the testimony is grounded on sufficient facts or data;” (2) “the testimony is the product of reliable principles and methods;” and (3) the witness has applied the principles and methods reliably to the facts of the case. See id. (cleaned

up). “[I]t is critical that an expert’s analysis be reliable at every step.” Id. at 267. The Supreme Court has cautioned, however, that even if an expert’s analysis is flawed, “[v]igorous cross- examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking shaky but admissible evidence.” Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 596 (1993) (citation omitted). II. Plaintiff’s Motion to Exclude Dr. Eisfeldt’s Testimony is Denied Plaintiff does not challenge Dr. Eisfeldt’s experience or qualifications for offering an opinion on the interest rate paid on RASP. Plaintiff advances several arguments for the exclusion of Dr. Eisfeldt’s testimony, none of which bears on the admissibility of her testimony

and report. First, Plaintiff argues that Dr. Eisfeldt’s testimony and report are inadmissible because she gives no opinion on whether the RASP rates were “reasonable” and instead limits her opinions to whether the RASP rates were comparable to the rates paid by other banks on similar accounts. Pl. Mem. at 1–2. Plaintiff complains that Dr. Eisfeldt failed to compare the RASP rates to the higher rates that Defendant pays on two of its other products. Id. at 2.

Plaintiff also argues that Dr. Eisfeldt should not have based her statistical analysis on the Crane data and that she failed to explain why she used the 25th percentile of sweep rates as the minimum standard of comparability. Id. at 2, 4. Her analysis is based on an equally-weighted average of RASP rates over an extended period of time but Plaintiff argues that Dr. Eisfeldt should have performed a proportionate tiered analysis. Id. at 3. Finally, because the parties agree that online banks pay a higher rate than brick-and-mortar banks, Plaintiff argues that Dr. Eisfeldt erroneously included rates paid by other brick-and-mortar banks, like Morgan Stanley, UBS, and Wells Fargo, as comparators to the RASP rates. Id. A. Dr. Eisfeldt Did Not Offer an Opinion Whether RASP Rates Were Reasonable Plaintiff argues that Dr. Eisfeldt’s testimony and report will not be helpful to a jury because she did not opine on the reasonableness of the RASP rates and because she did not

define “reasonable rate.” Id. at 12–13. During her deposition, Dr. Eisfeldt stated, “I’m not offering an opinion on reasonableness. I’m offering an opinion about what are comparable products to RASP and whether RASP rates were in line with the rates on those comparable products.” Id. at 15 (citing 2023 Tr. 111:9-20; 116:6-19; 117:4-23; 121:5-16; 204:6-7; 206:7-17, Ex. D, Dkt. 231–4). Dr. Eisfeldt did not define the term “reasonable,” nor did she opine on whether RASP rates were reasonable; neither affects the admissibility of her testimony and report. Plaintiff’s contractual claim hinges on whether Defendant offered a “reasonable” rate of interest on the swept cash. Ultimately, a finder of fact will have to determine whether Defendant paid a “reasonable” rate based on an appropriate charge from the Court. Dr. Eisfeldt was not retained by Defendant to opine on whether its rates were reasonable, but, rather, to analyze the RASP rates “relative to interest rates offered by comparable products.” Eisfeldt Report ¶ 5. The fact that Dr. Eisfeldt did not opine whether RASP rates were definitively reasonable does not mean

that her testimony and report will not “assist the trier of fact” in assessing the reasonableness of the rates. Nimely, 414 F.3d at 397 (citing Fed. R. Evid. 702). Her analysis provides useful context from which a juror can compare RASP rates to those offered by comparable programs without telling “the jury what result to reach.” Id. (citing United States v. Duncan, 42 F.3d 97, 101 (2d Cir. 1994)). B. Dr. Eisfeldt’s Reliance on Data from Crane Plaintiff takes issue with Dr. Eisfeldt’s reliance on data from Crane in her analysis. Pl. Mem. at 21. When asked during her deposition if there were products that she considered comparable to RASP other than the ones reflected in the Crane data, Dr. Eisfeldt testified that she did not consider whether there were any other comparable products. Id. Instead, she utilized an external source that had pre-selected comparable products. Id. (citing 2023 Eisfeldt Tr.

126:9–127:6; 145:16–147:11). Plaintiff argues that Dr. Eisfeldt opined that Crane had compiled data on the “largest and most comparable sweep programs,” Eisfeldt Report ¶ 9, and yet she did not analyze whether the data was accurate or probative of reasonableness. Pl. Mem. at 21. Although Dr. Eisfeldt spoke to the co-founder of Crane to verify how Crane collected its data, Plaintiff complains that she failed to investigate Crane’s “motivations, economic interests, and affiliation” with the banks it tracks. Id. at 22. Dr. Eisfeldt properly relied on the data from Crane as opposed to cherry-picking banks to include in her analysis. Crane is recognized as a reputable source for information on money markets, and it has been cited by regulators, such as the U.S. Securities and Exchange Commission, as well as academics. Eisfeldt Report ¶ 38. Crane’s “Broker Sweep Intelligence” (“Crane Index”) reports the interest rates paid by the sweep programs of 11 brokerages that represent “the largest and most comparable entities that offer sweep programs.” Id. ¶ 39. Dr. Eisfeldt personally spoke with an official at Crane to understand how the company compiles its

data and how the firms in the index were selected. 2022 Eisfeldt Tr. 30:19–31:23, Ex. C, Dkt. 228–3. After verifying that the Crane data is reliable, Dr. Eisfeldt then reviewed the firms in the Crane Index to ensure that their sweep products were similar to RASP, in that the account was FDIC insured, available for retirement accounts, and capable of automatically transferring uninvested cash from the brokerage account. Eisfeldt Report ¶ 41. Of the 11 broker-dealers in the Crane Index, two products did not meet her criteria (Fidelity’s Cash Management Account and E*Trade’s Extended Insurance Sweep Deposit Program) because they are not available for retirement accounts; Dr. Eisfeldt determined, however, that their inclusion is either “conservative or unlikely to meaningfully impact [her] analysis.” Id. Dr. Eisfeldt’s reliance on a widely-accepted and reliable data set was entirely appropriate.

See In re Payment Card Interchange Fee & Merch. Disc. Antitrust Litig., 638 F. Supp. 3d 227, 286–87 (E.D.N.Y. 2022) (citation omitted) (“Plaintiffs cite to case law to argue that [the expert’s] ‘uncritical reliance on survey data . . . renders her opinion unreliable,’ . . . [but] [the expert’s] reliance on relevant survey data prepared in a non-litigation context is not comparable to situations in which experts merely recited facts provided by a party or by counsel.”). Plaintiff may have doubts as to Crane’s “motivations, economic interests, and affiliation” with the firms it tracks, but that does not render Dr. Eisfeldt’s opinion inadmissible. “Vigorous cross- examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking shaky but admissible evidence.” Daubert,

509 U.S. at 596. Plaintiff identified an alternative data source published by Bank of America Securities, which she argues reports a larger cross-section of sweep rates than the Crane Index, Pl. Mem. at 24; this too is not a basis to exclude Dr. Eisfeldt’s testimony, although it may provide a reasonable ground for cross-examination. C. Dr. Eisfeldt’s Exclusion and Inclusion of Comparators Plaintiff objects to Dr. Eisfeldt’s comparison of RASP rates to rates paid by certain banks

on certain products and also argues that she selectively excluded rates paid on other products from her analysis. Plaintiff primarily takes issue with the fact that eight of the ten products that Dr. Eisfeldt considered involve sweep accounts at affiliated, as opposed to unaffiliated, banks.24 Pl. Mem. at 13. This fact, Plaintiff argues, is fatal to Dr. Eisfeldt’s report and makes her analysis “useless.” Id. at 16. Dr. Eisfeldt “fail[ed] to control for the one variable under review,” to wit: whether the sweep account was with an affiliated institution. Id. at 14 (citing Fed. Hous. Fin. Agency v. Nomura Holding Am., Inc., No. 11-CV-6201, 2015 WL 539489, at *5 (S.D.N.Y. Feb. 10, 2015)). Plaintiff argues that Dr. Eisfeldt’s report focuses on whether the RASP rates were comparable to what other brokerages that sweep cash to affiliated banks pay, and she fails to consider whether her analysis required a control group of transactions with unaffiliated banks.

Id. Plaintiff also argues that Dr. Eisfeldt’s report is deficient because she did not compare the RASP rates to rates paid on other sweep products that Defendant offers, including its PD accounts and its Insured Savings Accounts (“ISA”), both of which pay substantially higher rates. Id. at 2. Although Dr. Eisfeldt contends that non-sweep rates are different from sweep rates due to the volatility and transient nature of cash in RASP accounts, Plaintiff argues that Dr. Eisfeldt

24 Plaintiff argues that Dr. Eisfeldt failed to analyze sufficiently Fidelity and Baird as comparators in a separate subgroup; Fidelity and Baird sweep cash to unaffiliated banks and pay higher interest rates. Pl. Mem. at 16–17. did not provide evidence or data to validate that premise. Id. at 18. She did not compare RASP rates to rates paid on PD accounts; nor did she analyze Defendant’s reasons for excluding ISA and PD from retirement accounts or preventing non-retirement account holders from executing trades against ISA or PD account balances. Id. Because online banks pay higher rates than the

brick-and-mortar banks, Plaintiff contends that Dr. Eisfeldt should have excluded the rates paid by other brick-and-mortar banks like Morgan Stanley, UBS, and Wells Fargo in order to test whether RASP rates were reasonable. Id. at 3. Plaintiff’s argument that Dr. Eisfeldt’s analysis is flawed is based on (1) Plaintiff’s contention that it is per se unreasonable for banks to sweep uninvested cash to affiliated institutions and (2) her view that Dr. Eisfeldt was required to create a control group because nine of the 11 firms in the Crane Index sweep to affiliated institutions. Defendant responds that Dr. Eisfeldt opined, consistent with SEC advice,25 that it is not per se unreasonable for cash to be swept to affiliated institutions, so long as the brokerages “fully and fairly” disclose all such possible conflicts.26 Eisfeldt Report ¶ 44. Defendant also argues that because Dr. Eisfeldt

conducted a comparative analysis to determine whether RASP rates were in line with market rates and not a causal analysis, a control group was not necessary. Def. Opp. at 14, Dkt. 240 (citing Sec. & Exch. Comm’n v. Ripple Labs, Inc., No. 20-CV-10832, 2023 WL 5670711, at *13 (S.D.N.Y. Mar. 6, 2023) (granting a Daubert motion where the proposed expert’s “statement of causation” was “insufficient” because it was not supported by a control group study)).

25 SEC, “Staff Bulletin: Standards of Conduct for Broker-Dealers and Investment Advisers Conflicts of Interest,” August 3, 2022, https://www.sec.gov/tm/iabd-staff-bulletin-conflicts-interest (accessed on August 27, 2024).

26 The Employee Retirement Income Security Act (“ERISA”) prohibits an ERISA fiduciary from engaging in a transaction where there is a conflict of interest (e.g., brokers sweeping cash from IRA accounts to affiliated banks). ERISA, however, allows such prohibited transactions provided that the transaction, inter alia, “bear[s] a reasonable rate of interest.” 29 U.S.C. § 1108(b); 29 C.F.R. § 2550.408b-1(a)(iv). Dr. Eisfeldt’s analysis compares RASP rates to those offered on materially similar products. Dr. Eisfeldt reasonably decided not to compare rates paid on different products, such as PD and ISA, as those products “differ from investment products in their function and purpose, their investment risk, and their flexibility.” Eisfeldt Report ¶ 20. Differences in the

characteristics of sweep programs associated with retirement accounts explain why rates paid on them differ from rates paid on other products, such as savings accounts, MMFs, stable value funds, and non-sweep MMDAs, that have different characteristics. Id. Plaintiff may prefer to compare RASP rates to the higher rates offered on other products, but Dr. Eisfeldt logically explained her rationale for comparing RASP rates to rates paid on comparable products. See Shatkin v. McDonnell Douglas Corp., 727 F.2d 202, 208 (2d Cir. 1984) (cleaned up) (affirming exclusion of testimony of expert who compared discounted and undiscounted cash flows in part because “such an ‘apples and oranges’ comparison simply cannot withstand scrutiny”); 720 Lex Acquisition LLC v. Guess? Retail, Inc., No. 09-CV-7199, 2014 WL 4184691, at *11 (S.D.N.Y. Aug. 22, 2014) (denying Daubert motion, where “[t]here is no facially implausible ‘apples and

oranges’ analysis that might justify exclusion”). For the same reason, it was not unreasonable for Dr. Eisfeldt to compare RASP rates to the rates paid by other similar full service brick-and- mortar banks rather than to rates paid by online banks. D. Dr. Eisfeldt’s Methodology Plaintiff objects to two aspects of Dr. Eisfeldt’s methodology. First, Plaintiff argues that Dr. Eisfeldt’s analysis compares RASP rates to the 25th percentile of the rates reflected in the data reported by Crane without providing a justification. Pl. Mem. at 19. Because Defendant paid interest at rates that were in line with its primary competitors , all of which tended to pay the lowest rates, Plaintiff argues that using the 25th percentile as a benchmark means that RASP rates will always be “comparable” even if other firms within the data set paid significantly higher rates. Id. Second, Plaintiff claims that Dr. Eisfeldt inappropriately performed her analysis based on an equally-weighted average of RASP rates, instead of weighting the rates paid on each tier proportionately to the assets held in that tier. Id. at 19–20. According to Plaintiff, Dr. Eisfeldt’s approach distorted the analysis because

Defendant paid its highest rate to customers in Tier 4, but Tier 4 accounted for the least amount of cash held in RASP. Id. at 20. Because the tiers were equally-weighted, her analysis was skewed.27 Id. at 19–20. Plaintiff also argues that it is unclear how Dr. Eisfeldt calculated the median rate of other comparable programs. Id. Neither criticism of Dr. Eisfeldt’s methodology is a reason to exclude her opinion. Dr. Eisfeldt conducted two comparative analyses: in one, she compared the RASP rates to the median interest rate paid by other institutions in the Crane Index; and in the second, she compared RASP rates to the 25th percentile of rates paid by those institutions. Eisfeldt Report ¶¶ 48, 50. Dr. Eisfeldt explained that she compared the rates to the 25th percentile to check whether RASP rates were an “outlier on the low end of the distribution of interest rates.” Id. ¶

50. In her deposition, Dr. Eisfeldt explained that an “interquartile range is a very common way of looking at the kind of central tendency of a distribution[] [a]nd the bottom of the interquartile range is the 25th percentile.” 2023 Eisfeldt Tr. 15:8–11, Ex. D, Dkt. 231–4. Plaintiff may prefer a different comparison than looking to the bottom quartile, but that is not a reason to exclude Dr. Eisfeldt’s testimony. Plaintiff takes issue with looking to what Defendant’s competitors paid as an indicator of reasonableness, arguing that all of the rates could be unreasonable. A trier of fact could accept that argument, but Plaintiff must accept the market for what it is. A trier of fact will

27 For example, Plaintiff notes that Defendant maintains in Tier 1 and in Tier 4, but Dr. Eisfeldt weighted the RASP rates (0.14% v. 0.75%) equally to get an equally-weighted RASP rate of 0.46%. Pl. Mem. at 20. That rate is higher than a cash-based weighted RASP rate, which, according to Plaintiff, would be 0.25%. Id.; Ex. K, Dkt. 231–11. likely view it as important to understand how Defendant’s rates compare to the rates offered on similar products by similar institutions when determining whether Defendant’s rates were reasonable. Dr. Eisfeldt’s decision to equally weight the RASP rates across its four asset tiers is also

not a reason to exclude her report. Dr. Eisfeldt noted the volatility of balances in sweep programs and the fact that the balances in each tier fluctuate given “the transitory nature of these deposits.” Eisfeldt Report ¶ 23. The Crane Index does not break down the proportion of assets that each bank holds in each tier, and that information is not publicly available. Had Dr. Eisfeldt performed a proportionate analysis based on the rates offered to each tier, she would have needed to make a lot of assumptions to account for the volatility of each tier. Despite equally-weighting the RASP rates, Dr. Eisfeldt noted that the rates for Tiers 1 and 3 were on average lower than the median, and the rates for Tiers 2 and 4 were on average higher than the median. Eisfeldt Report ¶ 49 n.72. There were periods of time when RASP rates were higher than, equal to, and lower than the median rate. Id. ¶ 49. Plaintiff argues that if Dr. Eisfeldt had conducted a proportionate tier analysis, the RASP rates would have been lower and likely below the 25th percentile. Pl.

Mem. at 20. Dr. Eisfeldt’s decision to equally weight the tiers may be a proper basis for cross- examination but it does not justify excluding her as an expert. See Joffe v. King & Spaulding LLP, No. 17-CV-3392, 2019 WL 4673554, at *9 (S.D.N.Y. Sept. 24, 2019) (“[T]he fact that the median or a weighted average may be a better metric does not render the simple average inherently unreliable.”). Accordingly, Plaintiff’s motion to exclude Dr. Eisfeldt’s report and testimony is DENIED. III. Defendant’s Motion to Exclude Dr. Palia’s Testimony is Granted in Part and Denied in Part Defendant does not challenge Dr. Palia’s experience or qualifications for offering an opinion on the reasonableness vel non of RASP interest rates. Nevertheless, Defendant seeks to exclude Dr. Palia’s testimony and report arguing that he has “not offered any relevant, useful, or reliable expert opinions.” Def. Mem. at 1, Dkt. 225. Defendant argues that Dr. Palia’s opening report relies heavily on using MMFs as an appropriate benchmark for RASP rates, a position that the Court has rejected in several prior decisions.28 Id. That portion of Dr. Palia’s report and testimony, which Plaintiff abandoned,29 is not admissible. Dr. Palia also provides a comparison between RASP rates and the rates offered on

savings accounts by online banks; that opinion is admissible. The rebuttal report introduces two new comparators, which Defendant contends Dr. Palia failed to include in his opening report without a valid excuse and without explaining why the new comparators are appropriate. Id. at 13. The Court disagrees and denies Defendant’s motion to exclude these additional comparators.

28 See June 3, 2020 Opinion and Order at 13, Dkt. 31 (finding Plaintiff’s comparisons to rates paid on MMFs inapt, as MMFS are “an entirely distinct investment option”); January 25, 2021 Opinion and Order at 4, Dkt. 54 (“Plaintiff’s comparisons to interest rates paid on entirely distinct investment products are irrelevant and insufficient.”); April 12, 2023 Opinion and Order at 18, Dkt. 167 (rejecting Plaintiff’s prior expert’s analysis because the “bottom line remains that Government MMFs and bank deposit accounts are not the same. . . . [T]he factors that determine the interest rate paid on MMFs and the factors banks must consider when setting interest rates to be paid on deposit accounts are different”); June 28, 2023 Opinion and Order at 5, Dkt. 182 (rejecting Plaintiff’s motion to reconsider its prior holding that “as a matter of law, . . . government MMF rates are not proper comparators to deposit rates for swept cash”).

29 Dr. Palia’s opening report repeatedly uses rates paid on MMFs as a benchmark, despite the Court’s previous decisions stating that “a Government MMF is not a relevant comparator to an FDIC-insured account.” Palia Report ¶ 23. Dr. Palia acknowledges the Court’s prior rejection of MMFs as a benchmark but nonetheless “strongly disagree[s]” and states that Defendant should have used the interest rate paid on MMFs as the starting point in setting the RASP rates. Id. Dr. Palia submitted his opening report shortly before the Court denied Plaintiff’s motion for reconsideration on the prior Daubert motion, in which the Court held that it “will not reconsider its decision that interest rates paid on [MMFs] are not appropriate comparators when determining what constitutes a reasonable interest rate.” June 28, 2023 Opinion and Order at 9, Dkt. 182. Apparently four opinions making the same point finally did the trick; Plaintiff now says, despite Dr. Palia’s report, that she “is not advocating for a MMF-benchmarked methodology.” Pl. Opp. at 19 n.35. Finally, Defendant argues that after serving his rebuttal report, Dr. Palia emailed two more comparators, neither of which was in either prior report and each of which was submitted without any analysis. Id. at 2. These supplemental “reports” are inadmissible. A. Dr. Palia’s Comparison of RASP Rates to Rates Paid on Savings Accounts by Online Banks Is Admissible Defendant challenges Dr. Palia’s decision to compare RASP rates to interest rates paid on savings accounts offered by online banks (Marcus, Ally, American Express, and Barclays) (“Online Savings Accounts”). Defendant contends such accounts are fundamentally different from sweep accounts and serve different purposes. Def. Mem. at 8. Dr. Palia noted that “digital banks with little to no physical presence offered deposit products comparable to those offered by

their brick and mortar competitors in terms of available services . . . at substantially higher rates.” Palia Rebuttal Report ¶ 25. He also conceded, however, that “[b]rick and mortar and online bank deposit accounts are widely understood to offer rates that are typically different.” Id. ¶ 26. Dr. Palia uses the interest rates paid on Online Savings Accounts at four banks as a benchmark. Although savings accounts at online banks and brick-and-mortar banks, such as Defendant, may differ in some respects, that is not a reason to exclude his opinion. Dr. Palia proffered these benchmarks because the rates offered by such banks “are a reliable indicator of the competitive FDIC-insured rates that were available.” Palia Report ¶ 97. Dr. Palia rejects Dr. Eisfeldt’s opinion that investment products and deposit accounts are so “fundamentally

different” that they cannot be compared. Palia Rebuttal Report ¶ 30. He analyzed the economic factors30 that influence comparability and applied them to identify what he opines are

30 These factors include branch features, automated transfers of cash, maturity and liquidity, bank size, risk, and credit quality, risk profile, account size and other customer characteristics. Palia Rebuttal Report ¶ 31. appropriate benchmarks. Id. ¶ 31. Dr. Palia looked at the “comparator benchmarks” found on Defendant’s “Money Markets: Competitive Landscape” matrix. Palia Rebuttal Report ¶ 41. On that matrix, Defendant compared the interest rate it paid on PD accounts to the rates paid by online banks; Dr. Palia asserted that Defendant’s decision to look to “the most competitive

online bank products provide[d] an objective basis to select [them as] comparable benchmarks.” Pl. Opp. at 14–16, Dkt. 244; Palia Rebuttal Report ¶ 48. Defendant contends that Dr. Palia cherry-picked the four online banks and did not explain why he excluded rates paid by other online banks on their savings accounts. Def. Mem. at 10. Dr. Palia’s reliance on the four online banks is not an example of improper cherry-picking; he explained that he used those four banks specifically because Defendant used them as benchmarks for its PD in its “Money Markets: Competitive Landscape” matrix. Palia Rebuttal Report ¶ 41. While Defendant may have a persuasive jury argument that Dr. Palia is comparing apples to oranges,31 this is not a situation in which an expert omitted all unfavorable or adverse data and selectively included facts to support a predetermined outcome. Defendant will be free to cross-

examine Dr. Palia about his decision not to survey a broader swath of online banks and to explore why, in his opinion, the comparators Defendant looks to when setting the rate it pays on PD are appropriate comparators for RASP, but weaknesses in his opinion are not a basis for exclusion.

31 At her deposition, BANA’s Rule 30(b)(6) witness testified that when setting the RASP rates, BANA looks to the sweep rates offered by its “core competitors,” . 2022 Tr. of Aileen Gleason Dep. at 77:13–18, Ex. F Dkt. 229–6. Dr. Palia, however, relied on the testimony of a different employee of Defendant who acknowledged that Defendant looks to the rates of the four online banks when setting the rates for Defendant’s PD, which is an entirely different product from RASP. Def. Mem. at 11–12 (citing Tr. of Jeremy Foskett Dep. at 19:14–15, 34: 15–17, 153: 11–18 Ex. H, Dkt. 229–8). Although it is a close call whether online savings accounts are sufficiently similar to RASP to make Dr. Palia’s testimony relevant, the Court holds that Dr. Palia’s opinion using rates paid on Online Savings Accounts as comparators is admissible. B. The Two New Comparators Contained in Dr. Palia’s Rebuttal Report Are Admissible

Rebuttal evidence “is properly admissible when it will explain, repel, counteract or disprove the evidence of the adverse party.” Scott v. Chipotle Mexican Grill, Inc., 315 F.R.D. 33, 44 (S.D.N.Y. 2016) (citation omitted). A “rebuttal expert report is not the proper place for presenting new legal arguments, unless presenting those arguments is substantially justified and causes no prejudice.” Id. at 44. Rebuttal experts may rely on new methodologies “for the purpose of rebutting or critiquing the opinions of [the opposing party’s] expert witness.” Park W. Radiology v. CareCore Nat’l LLC, 675 F. Supp. 2d 314, 326 (S.D.N.Y. 2009). Defendant argues that Dr. Palia introduced two new comparators in his rebuttal report: 1) rates on MMDAs offered by Ally and Discover and 2) a benchmark based on interest rates offered by unspecified banks as reported by Informa. Palia Rebuttal Report ¶¶ 80–81. Because these comparators were not included in Dr. Palia’s opening report, Defendant argues that Plaintiff waived them, and they should be stricken. Def. Mem. at 13 n.11. The Court will address each comparator in turn. 1. Rates Paid by Ally and Discover on MMDAs Defendant contends that Dr. Palia used interest rates paid on non-sweep MMDAs by Ally

and Discover as comparators without conducting any analysis of whether those accounts are representative of other online MMDAs. Def. Mem. at 13.32 As with Dr. Palia’s use of rates paid

32 Def. Mem. at 13 n.12 (citing 2023 Palia Tr. at 182:25–183:3, Ex. B, Dkt. 229–2 (“Q. But you didn’t conduct an analysis as to whether Ally and Discover are representative of other MMDAs, did you? A. No.”); id. at by other online banks as comparators, Defendant asserts that Ally and Discover’s MMDAs are not appropriate comparators. Although Dr. Palia introduces the rates paid on MMDA accounts by Ally and Discover for the first time in his rebuttal report, it is proper expert rebuttal; Dr. Eisfeldt’s opening report

acknowledges that RASP is an MMDA but posits that sweep accounts are “fundamentally different” from MMDAs that are not sweep accounts. Eisfeldt Report ¶ 33, 36. Dr. Palia explains that he used Ally and Discover as comparators . Palia Rebuttal Report ¶¶ 44–45. Defendant acknowledges that it tracked the rates paid on online MMDAs in connection with other products it offered but argues that Ally and Discover were not used as comparators for setting RASP rates.33 Def. Reply at 8, Dkt. 251. While perhaps detrimental to the persuasiveness of Dr. Palia’s opinion, that is not a basis for exclusion. Dr. Palia also states that the two banks “are similar to the other comparable online banks along observable deposit characteristics,” Palia Rebuttal Report ¶ 46; in Appendix B of his

rebuttal report, Dr. Palia provides additional data comparing Ally and Discover to other online banks based on total deposits and of the percentage of the banks’ total deposits that are savings and MMDA deposits, id. ¶¶ 104–05. Dr. Palia surveyed data reported by S&P’s Global Market Intelligence CIQ Pro; S&P collects and aggregates interest rate data from more than 75 percent

182:9–24 (“Q. And what I want to understand is how did you decide that these two products, the MMDAs offered by Ally and Discover, are comparable? Do you – A. I took – that’s why I gave Exhibit 1. You know, Exhibit 1 were those eight. Right? Q. Right. But you excluded six of the eight. A. Yes . . . [t]hose don’t have data on it, on money market deposit accounts. Okay? And I don’t know if one of them had – I don’t remember offhand, but some of them might not have money market deposit accounts. But, clearly, I can’t do analysis on the six, okay, either because the data is incomplete or – for the period, or they don’t offer it.”)). 33 Compare Eisfeldt Report at 24 Fig. 2 (the comparators to RASP are listed under the “Brokerage Platform”) with Adesanya Decl. Ex. B, Dkt. 246-2 (GWIM Deposit Pricing Dashboard) (highlighting that Ally and Discover were not identified as comparators to RASP). of all banks in the country. Id. ¶ 103. He focused on a group of eight online banks, comparing them along observable characteristics for a defined time period, concluding that the data for Ally and Discover “are representative of the full sample of banks.” Id. ¶ 105. Defendant’s arguments that the data should be excluded because Dr. Palia provided no analysis explaining

why these banks are comparable and because he cherry-picked the data are not persuasive reasons to exclude the data. Arguments regarding a small sample size go to the weight of the testimony and not its admissibility. Bryant v. Milhorat, No. 09-CV-1751, 2013 WL 12368616, at *10 (E.D.N.Y. Sept. 30, 2013). Accordingly, Dr. Palia may consider rates paid by Ally and Discover when discussing whether the RASP rates were reasonable. 2. Benchmark Based on Data from Informa Defendant argues that Dr. Palia’s comparison of RASP rates to the data from Informa must be excluded because it was not included in his opening report and because it lacks any expert analysis. Defs. Mem. at 14. Dr. Palia states that “the average interest rates paid by online banks from Informa provides a benchmark against which to compare the RASP rates that better reflects a reasonable rate of interest than the Crane Brokerage Sweep Index,” Palia Rebuttal

Report ¶ 94; Defendant contends that is insufficient analysis to support an expert opinion. As with his use of interest rates paid by Ally and Discover on their MMDAs, Dr. Palia’s use of data from Informa is not untimely because it is proper rebuttal; it responds to Dr. Eisfeldt’s opinion that RASP rates cannot be compared to the interest rates paid by online banks. Informa Research Services (“Informa”) has a database of interest rates paid by online banks on MMDAs with cash balances greater than $50,000. Palia Rebuttal Report ¶ 94. Dr. Palia states that the average interest rates paid by online banks as reflected in the dataset published by Informa provides a better benchmark against which to compare the RASP rates than the Crane Index on which Dr. Eisfeldt relies. Id. The data from Informa includes interest rates paid by online banks on retail MMDAs with average cash balances greater than $10,000 and $50,0000. Id. ¶ 95. Dr. Eisfeldt’s opening report states that she “considered” the data from Informa but did not rely on it for her opinions in her report. Eisfeldt Report at 45 n.1, App’x C . The fact that Dr. Palia and Dr. Eisfeldt relied on different sources of aggregate data does not mean that either

relied on an unacceptable source of data. Defendant will be able to cross-examine Dr. Palia during trial as to the make up of the data reported by Informa and as to his opinion that the interest rates paid on MMDAs as reflected by the Informa data are better comparators that the rates reflected in the Crane Index. C. Dr. Palia’s Supplemental Email Submissions Defendant moves to strike Dr. Palia’s two supplemental submissions that were sent to Defendant weeks after Dr. Palia submitted his rebuttal report. On August 2, 2023, two weeks after the deadline for rebuttal reports, Plaintiff sent Defendant the First Palia Supplemental Submission,34 in which Dr. Palia calculated damages by applying the difference between RASP rates and the average rates paid by Fidelity and Baird on their sweep programs. On August 28, 2023, Plaintiff sent Defendant the Second Palia Supplemental Submission,35 in which he

calculated damages by applying the difference between RASP rates and the rates paid by online banks on MMDAs as reported by Informa. Because these additional charts, labeled as “supplemental expert reports,” were sent weeks after the deadline for both opening and rebuttal reports, Defendant argues that its expert, Dr. Eisfeldt, was unable to respond to either of these submissions and that they should be stricken as untimely and outside the scope of Dr. Palia’s reports. Def. Mem. at 23. Defendant argues that Dr. Palia’s supplemental reports do not meet

34 Ex. J, Dkt. 229–10, Email from Robert Finkel to Cristina Rincon, dated August 2, 2023; Ex. L, Dkt. 229– 12. 35 Ex. K, Dkt. 229–11, Email from Robert Finkel to Paul Mishkin, dated August 28, 2023; Ex. M, Dkt. 229– 13. the requirements of Rule 26(a)(2)(B) of the Federal Rules of Civil Procedure, as the supplemental charts were not accompanied by any written expert analysis, and therefore the submissions are not legitimate supplemental reports and should be excluded pursuant to Rule 37(c)(1). Def. Mem. at 24.

“[E]xperts are not free to continually bolster, strengthen, or improve their reports by endlessly researching the issues they already opined upon, or to continually supplement their opinions.” Sandata Techs., Inc. v. Infocrossing, Inc., Nos. 05-cv-9546, 06-cv-1896, 2007 WL 4157163, at *6 (S.D.N.Y. Nov. 16, 2007). If an expert’s report “does not rely [on] any information that was previously unknown or unavailable to him,” it is not an appropriate supplemental report under Rule 26. Lidle v. Cirrus Design Corp., No. 08-cv-1253, 2009 WL 4907201, at *6 (S.D.N.Y. Dec. 18, 2009). That said, “preclusion of an expert report can be a harsh sanction.” Sandata Techs., Inc., 2007 WL 4157163, at *7. “In determining whether preclusion is appropriate, courts must consider: (1) the reasons for the delay in providing the evidence; (2) the importance of the evidence precluded; (3) the prejudice to the opposing party

from having to address the new evidence; and (4) the possibility of a continuance.” Cedar Petrochemicals, Inc. v. Dongbu Hannong Chem. Co., Ltd., 769 F. Supp. 2d 269, 278 (S.D.N.Y. 2011) (citing Softel, Inc. v. Dragon Med. & Sci. Commc’ns, Inc., 118 F.3d 955, 961 (2d Cir. 1997)). Plaintiff argues that her timing was excusable because Dr. Palia’s supplemental submissions are based on the same underlying data and are responsive to developments in the case. Pl. Opp. at 26. Dr. Palia mentioned the Informa data in his July 8, 2020, report filed in support of Plaintiff’s motion to amend her complaint and clarified his analysis in the rebuttal report in response to the Court’s 2023 decision on the motion to reconsider the decision

excluding her other expert. Id. Additionally, according to Plaintiff, Dr. Palia belatedly realized that his rebuttal report failed to account for damages for the period from January 2022 through April 2023. Id. n.48. The supplemental submissions update his damages calculation to include that period, and, as such, they did not contain any new analysis or methodology. Id. Because the data reported by Informa included interest rates paid by online banks on MMDAs with average

cash balances between $10,000 and $ 50,000, and with balances in excess of $50,000, Dr. Palia estimated the monthly cash balances of class members and used those sums to calculate damages by comparing the average interest rates paid by online banks to the RASP rates paid by Defendant. Id. at 26–27. This methodology, Plaintiff contends, is an extension of Dr. Palia’s 2020 analysis and buttresses his use of interest rates paid on online savings accounts as benchmarks, as he briefly did in his opening report. Id. at 27. Dr. Palia’s rebuttal report focuses on MMDAs. The additional comparators “reflect new inputs into the methodology” used in Dr. Palia’s opening report and underscore his conclusion that rates paid on MMDAs are much higher than those paid on RASP. Id. Plaintiff also argues that the Fidelity and Baird charts should not be excluded. Dr.

Eisfeldt testified that Fidelity is one of two banks that sweep cash to unaffiliated banks, and Plaintiff only discovered that Baird is the second bank that does so after Dr. Palia’s rebuttal report was submitted. Id. at 28. Plaintiff argues that it is “immaterial” that the Fidelity and Baird data was not accompanied by a formal report, “as there is no ambiguity as to what Dr. Palia was conveying through the data.” Id. at 29 n.52. Plaintiff also contends that Defendant is not prejudiced by the supplemental submissions, arguing that Defendant deliberately filed a letter three days before the parties’ briefs were due informing Plaintiff of its intended motion to strike the submissions and rejecting Plaintiff’s offer to allow Defendant to serve an expert rebuttal that considered the additional comparators. Id. at 30 n.54. Plaintiff’s arguments regarding Dr. Palia’s supplemental submissions are not persuasive. It is not relevant that Dr. Palia’s opening report was issued before the Court’s decision on Plaintiff’s motion for reconsideration. Plaintiff has been on notice since practically the beginning of this case that, in this Court’s view, the rates paid on government MMFs are not

relevant to the issue of the reasonableness vel non of the RASP rates. It is, therefore, incomprehensible why Dr. Palia’s opening report focused almost exclusively on MMF rates. Nor is Dr. Palia’s late realization that Fidelity and Baird offered higher rates on sweeps to unaffiliated banks an excuse. Plaintiff argues that Dr. Eisfeldt was not forthright in her deposition about the two programs in the Crane Index that sweep to unaffiliated banks and her purportedly cagey testimony delayed Dr. Palia’s “Eureka” moment. This explanation is silly. Dr. Palia had access to the Crane Index; and he claims to have relied on the expert rebuttal report of Plaintiff’s prior expert, Dr. Officer. That report clearly identified Fidelity and Baird as programs that sweep to unaffiliated banks. Palia Report, App’x A: Materials Relied on, Officer Rebuttal Report ¶ 109 n.137, Ex. AAA, Dkt. 128–16.

Dr. Palia’s supplemental charts must be excluded under Rule 26(a)(2)(B) because they lack any written expert analysis and “fail to include any of the underlying conclusions on which the expert’s ultimate opinions are based[.]” See Rodriguez v. Vill. of Port Chester, 535 F. Supp. 3d 202, 211 (S.D.N.Y. 2021) (citation omitted). Plaintiff argues that it is “immaterial” that there was no written analysis and that it is clear what Dr. Palia sought to convey. The Court disagrees. Plaintiff cannot flout the requirements of Rule 26 because she thinks it is obvious what her expert intends to show with additional data. Plaintiff also claims that the additional data is “merely additional back-up for the opinion stated repeatedly in Dr. Palia’s opening and rebuttal reports.” Pl. Mem. at 29 n.52. Again, it was Dr. Palia’s decision not to include that data in

either of his prior reports. Dr. Palia is “not free to continually . . . supplement [his] opinions,” even with “merely additional back up” information. Cedar Petrochemicals, Inc., 769 F. Supp. at 278 (citation omitted). Finally, Defendant is prejudiced by the late submissions because it was not able to have its expert consider the additional data. Plaintiff accuses Defendant of gamesmanship to manufacture prejudice by not raising the untimely analyses until the eleventh hour and rejecting Plaintiffs offer to serve an expert rebuttal with the additional data. Considering that Dr. Palia’s supplemental submissions were unjustifiably late and wholly deficient, Defendant need not expend more resources to respond. Accordingly, Defendant’s motion to strike Dr. Palia’s supplemental submissions is granted. CONCLUSION For the foregoing reasons, Plaintiff's motion to preclude Dr. Eisfeldt’s opinion and testimony is DENIED; Defendant’s motion to preclude Dr. Palia’s opinion and testimony is GRANTED in part and DENIED in part. The parties are directed to meet and confer and, by no later than October 18, 2024, submit a joint letter proposing a briefing schedule for summary judgment. As noted above, supra note 1, this Opinion and Order will be filed under seal. By not later than October 11, 2024, Defendant must show cause why any portion of this Opinion & Order should remain under seal and submit a version of this opinion with its proposed redactions. The Clerk of Court is respectfully directed to file this Opinion and Order under seal, with viewing restricted to the parties and the Court, and to terminate the open motions at docket entries 222 and 223. SO ORDERED. \ Qe □□ — Date: September 27, 2024 ~ VALERIE CAPRONI New York, New York United States District Judge