Using data from over 100 Rule 10b-5 securities class action settlements from 2015 to 2018, this study compares total approved damages claims calculated by settlement claims administrators (Approved Claims) to estimates of aggregate damages put forth by plaintiffs in their motions for settlement approval (Plaintiff Estimates). The results show that, on average, Approved Claims are approximately 66 percent of Plaintiff Estimates and, for the median case, Approved Claims are approximately 58 percent of Plaintiff Estimates.

Estimates of aggregate damages play a key role in settlement negotiations. However, such estimates are typically based on probabilistic models of investor trading behavior and aggregate trading data that may not fully and accurately capture investor trading patterns.^[1] The results in this study indicate that such estimates, in particular Plaintiff Estimates, tend to overstate actual damages claimed by investors in settlements.

The finding that Plaintiff Estimates tend to overstate Approved Claims in settlements is consistent with a prior Cornerstone Research study that identified and analyzed two securities class actions that reached trial and where damages were ultimately awarded: In re Vivendi Universal, S.A. Securities Litigation (Vivendi) and In re Household International, Inc. (Household). That study found that the approved claims rate, defined there as the total dollar amount of approved claims as a percentage of aggregate damages estimates calculated following typical plaintiff methodologies, was 20 percent in Vivendi and 38 percent in Household, revealing that claims filed by class members after judgment were well below the estimated aggregate damages in those cases.^[2]

Calculation of Approved Claims Rates

This study uses publicly available information to calculate an “Approved Claims Rate” for each of the more than 100 cases in the sample. The Approved Claims Rate is calculated in this study as the ratio of Approved Claims to the estimate of aggregate damages provided by plaintiffs, namely, Plaintiff Estimates.

Approved Claims

Approved Claims for each case equal the total dollar value of recognized losses across all investors who submitted damages claims and whose claims were approved by the settlement claims administrator. Claims administrators calculate recognized losses for each class member based on that class member’s transactions in the security at issue. Per-share damages are calculated using a formula based on the timing of those transactions, which is described in the settlement notice, under a “plan of allocation.” The recognized loss formula in the plan of allocation is typically based on plaintiff estimates of inflation assuming plaintiffs would have succeeded in proving their allegations (with respect to per‑share damages and class period) that survived through the time of settlement.

Total recognized losses across all investors’ Approved Claims are often provided in a declaration submitted to the court by the claims administrator. The declaration may be filed as an attachment to the plaintiffs’ motion for payment of litigation expenses out of the settlement fund, including payment of the claims administrator’s fees. This motion is typically filed near or in conjunction with the motion to disburse the settlement funds to claimants.^[3]

Plaintiff Estimates

Plaintiff Estimates are obtained from the motion that plaintiffs file with courts in support of final approval of the settlement. The Plaintiff Estimates are often described as the “maximum potential recovery,” or plaintiffs’ “best-case scenario” as calculated by plaintiffs’ experts assuming plaintiffs prevail on their allegations. Occasionally, plaintiffs provide additional details regarding their estimates, such as the trading model used.

Plaintiffs’ counsel typically include the Plaintiff Estimates in a section discussing the reasonableness of the settlement and comparing the ratio of the proposed settlement amount to estimates of plaintiffs’ potential recovery against similar metrics for sets of securities cases they deem to be comparable.

While there are no publicly available data that allow a comparison of Plaintiff Estimates and the damages estimates put forth by plaintiffs during settlement negotiations, plaintiffs would not be expected to put forth higher estimates in their motion for settlement approval than those put forth during settlement negotiations. Since plaintiffs’ counsel may wish to demonstrate to the court a high settlement amount relative to potential recovery, plaintiffs’ counsel might be incentivized to report in their motion the lower end of the range of estimated total aggregate damages. All else being equal, such a selection bias, if present, would cause the Approved Claims Rate calculated in this study to exceed the rates calculated based on higher estimates of aggregate damages put forth in settlement negotiations.^[4]

Approved Claims Rates in Rule 10b-5 Settlements

The sample in this study comprises all U.S. federal securities class action settlements with settlement hearing dates between 2015 and 2018 for which there are Rule 10b-5 allegations only (e.g., no cases with additional Section 11 allegations), where the only security at issue is the company’s common stock, and where data were available to determine Approved Claims and Plaintiff Estimates. The data collection process resulted in 102 settlements between 2015 and 2018 where both Approved Claims and Plaintiff Estimates were available.^[5] The table below summarizes the number of settlements by the year of settlement hearing date.^[6]

Over the sample of 102 cases, the median and average Approved Claims Rates are 58.2 percent and 65.8 percent, respectively.^[7] In other words, the total value of recognized losses for valid claims submitted by investors seeking to recover in the settlement was on average less than two-thirds of Plaintiff Estimates.

Thus, the analysis in this study reveals that aggregate damages estimates from trading models, on average, overstate claimed damages for cases that settle.^[8]

As shown, there is wide variation in Approved Claims Rates in Rule 10b-5 settlements, which further indicates that aggregate trading models are unreliable predictors of actual claimed damages. The median Approved Claims Rate of 58.2 percent exceeds earlier findings of claims rates between 20 percent and 38 percent, discussed above.

While this may suggest that approved claims rates are rising, the difference in results may also be due to differences in defining and measuring “claims rates” as well as potential selection bias in the aggregate damages observations provided in the motions for final settlement approval or even due to other aspects such as levels of institutional investment.^[9]

Endnotes

^[1]  Estimates of aggregate damages based on trading models may not reflect the actual timing of investor purchases and sales during the alleged class period and do not account for potential inflation offsets among several trades by individual investors

^[2]See C. Galley, E. McGlogan, and P. Morel, “Approved Claims Rates in Securities Class Actions,” Cornerstone Research, 2017, https://www.cornerstone.com/Publications/Research/Approved-Claims-Rates-in-Securities-Class-Actions.pdf

^[3] The claims administrator declaration usually includes recognized loss amounts for valid claims submitted by the court-approved deadline for claims submission (timely claims) and recognized loss amounts for valid claims submitted after the court-approved deadline for claims submission (late claims). The analysis in this study includes the total recognized losses for timely claims and late claims, regardless of whether the court elected to include late claims in the settlement distribution

^[4]  For several cases in which some of the authors of this article were involved in the mediation, the aggregate damages estimates provided by plaintiffs and their expert(s) at mediation matched the estimates provided in the motion for final approval of the settlement. In other cases, the values provided in the motion for final approval of the settlement were lower than the highest aggregate damages estimate provided by plaintiffs at mediation, including situations in which the value included in the motion for final approval of the settlement was substantially lower than the highest value plaintiffs and their expert(s) provided at mediation

^[5] The sample in this study was constructed as follows. First, 321 cases with settlement hearing dates between January 1, 2015, and December 31, 2018, were identified based on data obtained from Securities Class Action Services, LLC, a wholly-owned subsidiary of ISS (the data are available on a subscription basis—for further details see https://www.issgovernance.com/solutions/securities-class-action-services). Second, 31 cases that did not include Rule 10b-5 claims were excluded to arrive at 290 cases. Third, 59 cases with Section 11 or Section 12 allegations were excluded to arrive at 231 cases. Fourth, 46 cases where securities other than common stock were at issue (e.g., preferred stock, corporate bonds, options) were excluded to arrive at 185 cases. Fifth, 67 cases with no reported total recognized losses were excluded to arrive at 118 cases. Sixth, 16 cases where plaintiffs did not report an estimate of their potential maximum recovery were excluded to arrive at 102 cases for the sample that is analyzed. Whenever plaintiff counsel presents multiple estimates, the highest estimate is used as the Plaintiff Estimate to reflect (or be closer to) plaintiffs’ assessed “best-case scenario.

^[6] An October 2019 article, using a similar approach to this study, cites two anecdotal cases to support a claim that trading models underestimate securities class damages. See N. Walia and A. Werner, “Trading Models Underestimate Securities Class Damages,” Law360, October 25, 2019 (Walia and Werner). Those authors identified two anecdotal settlements to support their claim that trading models underestimate securities class damages: (1) Hosey et al. v. Twitter Inc. et al. (Twitter), and (2) Menaldi et al. v. Och-Ziff Capital Management et al. (Och-Ziff Capital). The Twitter settlement involved Section 11 allegations and thus is distinct from the scope of this article. However, Approved Claims Rates have been calculated here for Och-Ziff Capital, where the claims administrator found that total recognized losses for the valid claims were $67.2 million and the estimates of aggregate damages reported by plaintiffs in their motions for settlement approval ranged from $101.6 million to $166 million. Plaintiffs’ estimate of aggregate damages overstated (not understated, as Walia and Werner suggest) total approved claims, and the Approved Claims Rates range from 40 percent to 66 percent. Rather than compare the claimed damages to plaintiffs’ estimated damages (as they did for Twitter), Walia and Werner’s claim that aggregate damages estimates understate damages for Och-Ziff was based on their claim that the number of shares estimated to have been damaged (69.2 million) is lower than the number of shares that the claims administrators concluded had valid claims (84 million). See Declaration of Josephine Bravata Concerning the Results of the Claims Administration Process, Menaldi et al. v. Och-Ziff Capital Management et al., July 24, 2019, p. 3; Memorandum of Law in Support of Class Representatives’ Motion for Final Approval of Proposed Class Action Settlement, Menaldi et al. v. Och-Ziff Capital Management et al., December 17, 2018, p. 18

^[7]    Detailed data and analyses are not provided to ascertain the assumptions and models used by plaintiffs and their experts to estimate aggregate damages in each case. Methods, models, and assumptions likely vary by case and by the expert(s) selected by plaintiffs’ counsel to conduct the analysis. Estimated aggregate damages for a subset of the cases included in this study were computed using a consistent trading model methodology across each case, and the Approved Claims Rate results were consistent with the full sample results presented in the study

^[8]    The recognized loss formula in the plan of allocation, including the length of the class period, may differ from the assumptions used in plaintiffs’ experts’ aggregate damages analyses. To evaluate the effect of such potential differences, Approved Claims Rates were also calculated for the 56 observations for which the alleged class period in the plan of allocation was similar or identical to the alleged class period in the complaint. In such situations, there is a lower likelihood that plaintiffs’ aggregate damages model is based on assumptions that differed from those that ultimately resulted in the plan of allocation. The Approved Claims Rate for this subset of 56 observations was consistent with the full sample, with a median Approved Claims Rate of 55.6 percent

^[9]    Plaintiffs do not always include their aggregate damages estimates in the motion for final settlement approval. Since plaintiffs provide estimates of aggregate damages in the context of showing the court that the settlement value obtained is reasonable in light of the maximum possible recovery through a trial victory, plaintiffs may elect to not report these figures when the aggregate damages estimates are very high relative to the settlement obtained. Plaintiffs also appear to have some discretion as to what value to report as their maximum “likely” potential recovery. In both situations, the approved claims rate would likely be lower, all else equal.

The views expressed herein are solely those of the authors, who are responsible for the content, and do not necessarily represent the views of Cornerstone Research.

© 2020 by Cornerstone Research, Inc. All rights reserved.