Satisfying Fiduciary Duty Under ERISA

A fiduciary may contact an outside consultant to assist in making the necessary decisions. One such firm, NERA Economic Consulting [the author's firm], is a global firm of experts dedicated to applying economic, finance, and quantitative principles to complex business and legal challenges. As noted in NERA Economic Consulting's December 2009 report, Recent Trends in Securities Class Action Litigation: 2009 Year-End Update, the median ratio of settlement to “investor losses” has been at 4.0% or less for each of the past ten years. (See Figure 19, p. 20, of the report, available at http://www.nera.com/publication.asp?p_id=4015.) NERA's “investor losses” variable is not intended to be a measure of actual losses, nor of damages claims, but a proxy that will tend to overstate the allowable claims based on losses suffered by an investor. Not investigating likely defenses or alternative reasons for stock price declines enables the creation of a large database using a consistent methodology that allows for valid statistical analyses.

'Investor Losses' and Allowable Claims

Nevertheless, even with this difference between “investor losses” and allowable claims, the observed ratios are so small that after adjusting for the difference between “investor losses” and claimable damages, there will still be many situations in which a securities fraud case will settle for “far less than the full value of the plan's losses.” As a result, the fiduciary's burden for justifying its acceptance of, or its decision not to object to, the settlement is likely to be triggered frequently. This assertion assumes that an exemption is needed. However, the DOL did not indicate whether participating in a settlement would, in fact, be a prohibited transaction. (For a useful discussion of this issue, see Nell Hennessey, “ERISA Considerations in Litigation Settlements Involving Employer Securities and Mutual Funds,” available at http://www.erisasettlements.com/press/Settlements_073009.pdf. Also notable is an earlier version of this article stating that “securities litigators ' often overlook the requirements of a prohibited transaction class exemption that the Department of Labor ('DOL') issued at the end of 2003″ (concluding footnote omitted), available at http://files.ali-aba.org/thumbs/datastorage/skoobesruoc/pdf/Cl027-ch13_thumb.pdf.)

Settlement Discounts

It is not surprising that most securities class actions settle for far less than what investors lost. Settlement discounts occur both because plaintiffs receive their awards sooner than if they were to proceed to trial, and because they avoid the risk of losing at trial and getting nothing. The typical ratios indicate how sophisticated parties in numerous instances have judged the potential risks and costs of settlement versus proceeding to trial. Consequently, a low ratio of settlement amount to losses alone does not indicate that the settlement was improperly low and could have been expected to be improved had the ERISA plaintiffs objected or opted out.

To evaluate whether a proposed settlement is unreasonably low, data such as the types of allegations and presence of co-defendants can be used to obtain a statistical prediction of the likely settlement amount for the case. For example, NERA's statistical model based on the data described in its December 2009 Recent Trends report uses numerous variables to analyze approximately 1,000 historical securities class action settlements. This specialized analysis would seem to conform to goals behind the DOL's statement that “the independent fiduciary may wish to retain outside experts to assist the fiduciary in determining whether or not to settle litigation.” (PTE 2003-3039, p. 75635.) A proposed settlement that is similar to or exceeds the predicted settlement provides evidence in favor of the proposition that accepting the settlement is justified, while a proposed settlement that is substantially below the prediction may indicate a need to investigate whether there are factors outside of the statistical model that may explain the discrepancy.

Statistical Models

In addition to being useful for ERISA fiduciary purposes, a comparison of a proposed settlement with a statistical prediction can be of value in justifying or criticizing a proposed class settlement before a court or in the context of an insurance dispute about the size of a settlement. The use of a statistical model can be compelling in that the model will provide objective results showing that the proposed settlement is either high or low relative to the model. Because the results are derived solely from objective inputs, they will not be affected by the position of the party commissioning the study. This is particularly true if the model was developed outside the course of a particular case or dispute, though even statistical models developed to answer a particular question are constrained in a way that subjective views are not. On the other hand, because no statistical model can incorporate all possible factors, the predicted settlement will not typically be the only input in a determination of whether a proposed or past settlement was reasonable.

Conclusion

While no one measure will prove whether a settlement is reasonable, a determination of the expected amount of the settlement via statistical means provides an objective measure that can help justify or cast doubt upon a settlement amount and aid decision-makers in their task of approving or objecting to a settlement.

Dr. David Tabak is a Senior Vice President at NERA Economic Consulting. This article does not reflect the views of NERA or necessarily reflect the views of any NERA employee other than the author.

The Department of Labor (DOL) has issued guidance covering situations in which a pension plan, by virtue of its holdings of its employer's stock, is a potential claimant in a securities fraud suit: Prohibited Transaction Exemption (“PTE”) 2003-39: Class Exemption for the Release of Claims and Extensions of Credit in Connection with Litigation. 68 Fed. Reg. 75632-40 (Dec. 31, 2003). The exemption arose out of “concerns raised by the pension community regarding the impact of ERISA's prohibited transaction provisions of the settlement of litigation by employee benefit plans with parties of interest.” (PTE 2003-3039, p. 75632.)

Independent Fiduciary

PTE 2003-39 provides an exemption under certain conditions, one of which is that an independent fiduciary approve the plan's actions in the settlement process. The DOL noted in its discussion that, “If the fiduciary takes no action, and the case is settled for far less than the full value of the plan's losses, the burden will be on the fiduciary to justify its inaction. The fiduciary responsible for authorizing settlement of class action claims must decide, not only whether or not to opt out of the class action, but also whether to protest the proposed settlement during the fairness hearing.” (PTE 2003-3039, p. 75636, concluding footnote omitted.)

A fiduciary may contact an outside consultant to assist in making the necessary decisions. One such firm, NERA Economic Consulting [the author's firm], is a global firm of experts dedicated to applying economic, finance, and quantitative principles to complex business and legal challenges. As noted in NERA Economic Consulting's December 2009 report, Recent Trends in Securities Class Action Litigation: 2009 Year-End Update, the median ratio of settlement to “investor losses” has been at 4.0% or less for each of the past ten years. (See Figure 19, p. 20, of the report, available at http://www.nera.com/publication.asp?p_id=4015.) NERA's “investor losses” variable is not intended to be a measure of actual losses, nor of damages claims, but a proxy that will tend to overstate the allowable claims based on losses suffered by an investor. Not investigating likely defenses or alternative reasons for stock price declines enables the creation of a large database using a consistent methodology that allows for valid statistical analyses.

'Investor Losses' and Allowable Claims

Nevertheless, even with this difference between “investor losses” and allowable claims, the observed ratios are so small that after adjusting for the difference between “investor losses” and claimable damages, there will still be many situations in which a securities fraud case will settle for “far less than the full value of the plan's losses.” As a result, the fiduciary's burden for justifying its acceptance of, or its decision not to object to, the settlement is likely to be triggered frequently. This assertion assumes that an exemption is needed. However, the DOL did not indicate whether participating in a settlement would, in fact, be a prohibited transaction. (For a useful discussion of this issue, see Nell Hennessey, “ERISA Considerations in Litigation Settlements Involving Employer Securities and Mutual Funds,” available at http://www.erisasettlements.com/press/Settlements_073009.pdf. Also notable is an earlier version of this article stating that “securities litigators ' often overlook the requirements of a prohibited transaction class exemption that the Department of Labor ('DOL') issued at the end of 2003″ (concluding footnote omitted), available at http://files.ali-aba.org/thumbs/datastorage/skoobesruoc/pdf/Cl027-ch13_thumb.pdf.)

Settlement Discounts

It is not surprising that most securities class actions settle for far less than what investors lost. Settlement discounts occur both because plaintiffs receive their awards sooner than if they were to proceed to trial, and because they avoid the risk of losing at trial and getting nothing. The typical ratios indicate how sophisticated parties in numerous instances have judged the potential risks and costs of settlement versus proceeding to trial. Consequently, a low ratio of settlement amount to losses alone does not indicate that the settlement was improperly low and could have been expected to be improved had the ERISA plaintiffs objected or opted out.

To evaluate whether a proposed settlement is unreasonably low, data such as the types of allegations and presence of co-defendants can be used to obtain a statistical prediction of the likely settlement amount for the case. For example, NERA's statistical model based on the data described in its December 2009 Recent Trends report uses numerous variables to analyze approximately 1,000 historical securities class action settlements. This specialized analysis would seem to conform to goals behind the DOL's statement that “the independent fiduciary may wish to retain outside experts to assist the fiduciary in determining whether or not to settle litigation.” (PTE 2003-3039, p. 75635.) A proposed settlement that is similar to or exceeds the predicted settlement provides evidence in favor of the proposition that accepting the settlement is justified, while a proposed settlement that is substantially below the prediction may indicate a need to investigate whether there are factors outside of the statistical model that may explain the discrepancy.

Statistical Models

In addition to being useful for ERISA fiduciary purposes, a comparison of a proposed settlement with a statistical prediction can be of value in justifying or criticizing a proposed class settlement before a court or in the context of an insurance dispute about the size of a settlement. The use of a statistical model can be compelling in that the model will provide objective results showing that the proposed settlement is either high or low relative to the model. Because the results are derived solely from objective inputs, they will not be affected by the position of the party commissioning the study. This is particularly true if the model was developed outside the course of a particular case or dispute, though even statistical models developed to answer a particular question are constrained in a way that subjective views are not. On the other hand, because no statistical model can incorporate all possible factors, the predicted settlement will not typically be the only input in a determination of whether a proposed or past settlement was reasonable.

Conclusion

While no one measure will prove whether a settlement is reasonable, a determination of the expected amount of the settlement via statistical means provides an objective measure that can help justify or cast doubt upon a settlement amount and aid decision-makers in their task of approving or objecting to a settlement.

Dr. David Tabak is a Senior Vice President at NERA Economic Consulting. This article does not reflect the views of NERA or necessarily reflect the views of any NERA employee other than the author.