Article: The Changing Landscape of Fees in Securities Class Actions
December 11, 2023
A recent Law 360 article by Edward Flores, “The Shifting Landscape of Securities Class Action Fees”, reports on the changing landscape of attorney fees in securities class actions. This article was posted with permission. The article reads:
A recent Law360 article notes that there appears to be a growing shift in how class action fees are awarded, with the benefits secured for class members taking an increasing role in determining fees rather than fees simply being a function of the settlement amount. An examination of data on the percentage of the recovery awarded to attorneys in securities class actions shows that while the percentage declines as recoveries get larger, that percentage also increases as recoveries grow relative to a statistically predicted recovery.
As part of the settlement stage of a securities class action, plaintiffs file motions for approval of the settlement, as well as attorney fees and expenses. In these motions, the attorney fees sought are typically based on a percentage of the proposed settlement, with the proposed figures often based on a sample of recent cases, the case's lodestar multiplier or relevant benchmarks. For example, a June case in the U.S. District Court for the District of Delaware, In re: Electric Last Mile Solutions Inc., noted that "Lead Counsel intends to seek an award of attorneys' fees of no more than one third of the Settlement Amount. This fee request is in line with other settlements approved in recent cases."
In another case in the U.S. District Court for the Middle District of Tennessee in July, Bond v. Clover Health Investments Corp., the plaintiffs' counsel noted that the "request for 25% of the Settlement Fund ... amounts to a lodestar multiplier of 3.53, which is within the range of multipliers commonly awarded in securities class actions and other complex litigation."
Finally, the plaintiffs' counsel in a U.S. District Court for the Northern District of California case in August, Purple Mountain Trust v. Wells Fargo & Co., noted that "Lead Counsel seeks an award of attorneys' fees of 25% of the Settlement Amount. This fee request ... is consistent with the benchmark for attorneys' fee awards in this Circuit."
Plaintiffs counsel also typically highlights the recovery they have obtained for the class and their work to obtain that recovery. For example, in the aforementioned case involving Clover Health, counsel for the plaintiffs noted that the "recovery represents approximately 13.2% of the likely recoverable damages in this case, well above the median recovery of 1.8% of estimated damages for all securities class actions settled in 2022."
However, those recoveries are, at best, benchmarked as a percentage of some estimate of losses suffered by the class. An important question is whether plaintiffs counsel receives a higher fee when they obtain an unexpectedly large settlement, potentially indicating productive work on their part. Moreover, it may be of interest to see what other factors drive attorney fees awarded in securities class actions, both in terms of examining the direction and magnitude that certain factors have on attorney fees, as well as for purposes of estimating or cross-checking proposed fees given certain characteristics of a case.
Data and Analysis
To examine what factors can help explain attorney fees awarded, we analyzed securities class actions with alleged violations of Rule 10b-5 involving common stock that settled during the period from January 2012 through June. In addition to the logarithm of the actual settlement amount, some of the variables analyzed include a potential proxy for attorney performance and the strength of a case, the time to resolution of a case, and the status of certain motions as of the time of settlement.
As a potential proxy for attorney performance and the strength of a case, we calculated the ratio of the actual settlement value to its mean predicted settlement value to examine if better-than-expected settlements are reflected in attorney fees received. For each case, the mean predicted settlement value was calculated using an econometric model that uses as inputs variables that are historically known to drive settlement values.[8] For the middle 95% of cases in the dataset, the value of this ratio ranges from 0.2 to 3.8.
We also examined the effect that the time to resolution has on attorney fees, based on the filing of the first complaint of a case. Finally, we examined the effects that the filing and outcome of the motion to dismiss and the motion for class certification have on attorney fees. In measuring attorney fees, we used two different variables: attorney fees as a percentage of the settlement, and total attorney fees and expenses as a percentage of the settlement. As these variables range from 0 to 1, a fractional logistic regression model is used to analyze the relationship between attorney fees and the independent variables.
As the coefficients of this model can be hard to interpret, we examined the marginal effects of the independent variables in order to get an idea of the effect that a change in one variable has on attorney fee percentage, holding all others constant.
Results
From Table 1, we see that the ratio of the actual settlement to the mean predicted settlement has a positive and statistically significant effect, possibly highlighting how a better-than-expected outcome is rewarded in attorney fees. In terms of magnitude, we find that a 0.1 increase in this ratio results in an additional 0.1% in attorney fee percentage.
One interesting question that our data does not let us answer is whether higher-than-predicted settlements reflect a particularly strong prosecution of the case by plaintiffs counsel or case factors favorable to the class that are not captured by our model. Overall, we find that the variables in the model using attorney fee percentage as the dependent variable can explain nearly 20% of the total variation in attorney fee percentage.
Looking at the individual variables beyond the ratio of the actual to predicted recovery, the logarithm of the settlement value is unsurprisingly the strongest predictor of attorney fees, with higher settlement values being associated with a lower attorney fee percentage, consistent with prior research.
On the other hand, the time-to-resolution variable had a small and statistically insignificant effect, with each year contributing an additional 0.22% in attorney fee percentage. Thus, to the extent that courts are rewarding plaintiffs counsel — or plaintiffs counsel are willing to ask for more — through the fee percentage, that would appear to be based more on results than time spent.
When analyzing the motion variables, we find that the filing of a motion to dismiss has a statistically significant negative effect on attorney fees. Meanwhile, the effect of a motion to dismiss being denied or partially granted is positive and statistically significant, contributing an additional 1.5% in attorney fee percentage and again potentially showing that plaintiffs counsel is rewarded for achieving outcomes favorable to the class.
As over 90% of settled cases have a motion to dismiss filed, this means that cases where the motion to dismiss is denied or partially granted tend to have a 1.5% higher attorney fee percentage relative to cases that settle before a court decision was reached in the motion to dismiss.
The filing of a motion for class certification has a small and not statistically significant effect on attorney fees. However, the effect of a motion for class certification being granted is positive and statistically significant, contributing an additional 1.6% in attorney fee percentage, again consistent with the idea that plaintiffs counsel is being rewarded for achieving outcomes favorable to the class.
We also performed this analysis using total attorney fees and expenses as a percentage of the settlement size as the dependent variable. Under this model, the explanatory variables explain roughly 40% of the total variation in total attorney fees and expenses.
While the results look similar for several of the variables analyzed, there are a few notable differences. For example, the time-to-resolution variable was statistically significant in this model, with each year contributing an additional 0.62% in attorney fee percentage and expenses. In other words, plaintiffs counsel appear to be compensated for actual expenses, but are not being compensated through the fee percentage for time spent.
The filing of a motion for class certification was also positive and statistically significant, contributing to an additional 2.8% in attorney fee percentage and expenses.
Conclusion
Our analysis of settled cases shows that in addition to the settlement size, plaintiffs counsel in securities class actions appear to be rewarded for good settlement outcomes relative to a statistical prediction, with certain outcomes for the motion to dismiss and motion for class certification also affecting attorney fees awarded.
As a starting point, this means that in securities class actions, attorney fees are not just based on "the amount of money involved in a settlement," but also to at least some degree on the difference between that amount of money and the expected amount of money based on characteristics of the case.
That said, it is possible that analyses such as those described here may also be useful for purposes of estimating or cross-checking proposed fees given certain characteristics or outcomes in a case, particularly if one would like to reward successful advocacy by plaintiffs counsel.
As noted above, however, while our proxy variable for attorney performance and strength of case incorporates several quantitative factors that drive settlement values, it may not account for other qualitative strengths and weaknesses of a case.
Nevertheless, given that there are some areas that we can measure, it may make sense to see how a proposed settlement compares to a statistical prediction as part of the consideration of the relevant amount of attorney fees.
Edward Flores is a senior consultant at NERA Economic Consulting.