ORDER ON DEFENDANT'S OBJECTIONS TO EXPERT EVIDENCE AND DEPOSITION TRANSCRIPTS

NANCY TORRESEN, District Judge.

Before the Court are Defendant's Motion to Exclude Expert Evidence (Damages Model 1, offered by Plaintiffs' expert Thomas A. McAvity, Jr.) (ECF No. 97), Defendant's oral motion to exclude Plaintiffs' damages Model 2, made during the damages trial on June 25, 2013, and Defendant's Outstanding Objections to Deposition Transcript Testimony (ECF No. 125). For the reasons set forth below, the Defendant's Motions to Exclude Expert Evidence are GRANTED, and Defendant's Motion Regarding Defendant's Outstanding Objections to Deposition Transcript Testimony is DENIED as moot.

I. The Motions to Exclude Expert Evidence

Plaintiff's expert on damages, Thomas A. McAvity, Jr., prepared three models for determining what interest rates Unum, as an ERISA fiduciary, should have set for its ERISA-governed retained asset accounts (RAAs). See Trial Transcript, June 25, 2013 (Tr. 2), pgs. 248-317 (ECF No. 147). Prior to trial, the Defendant filed a written motion arguing that McAvity's first model, which purports to set the applicable interest rate on Unum's RAAs as competitive with interest rates credited by insurance companies on "spread products" such as guaranteed investment contracts ("Model 1"), should be excluded because it is inconsistent with the Court's order and with ERISA, and because it does not "fit" the question of damages (ECF No. 97). At trial, the Defendant also objected to McAvity's qualifications as an expert1 and to the Court's consideration of McAvity's second model under Federal Rule of Evidence 702. Tr. 2 at 247-48. The Court deferred ruling on Unum's motions until after the conclusion of evidence.

A. Legal Standard

Under Rule 702, "[a] witness who is qualified as an expert . . . may testify in the form of an opinion or otherwise if: (a) the expert's . . . knowledge will help the trier of fact to understand the evidence or to determine a fact in issue. . . ." Fed. R. Evid. 702. The Supreme Court noted in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 597 (1993), that the Court must make sure that "an expert's testimony both rests on a reliable foundation and is relevant to the task at hand."

B. Model 1

There are three reasons why Model 1 fails to meet Rule 702's relevance and reliability standards. First and foremost, Model 1 fails to address Unum's breach. Model 1 takes as its measure of damages a portion of Unum's spread between the amount it earned on investments backing the RAAs and the 1% interest rate Unum credited to RAA holders. Tr 2 at 250-51. It begins with the income Unum earned annually on all investments and works backward, subtracting Unum's expenses and a profit margin of .7%, or 70 basis points,2 annually. Under Model 1, everything that Unum earned above and beyond its expenses and profit margin should be awarded as damages to the Plaintiffs.3 Id.

It is difficult to overlook the spread that Unum earned as the source of damages in this case. Year after year, Unum earned in excess of six percent on the funds backing the RAAs while paying out only 1% in interest. See Tr. 2 at 290 (reciting Unum's annual portfolio yields during the class period). Every point of interest Unum decided not to credit to its RAAs was a point of interest it kept for itself. But Unum's breach in this case is not its retention and investment of the funds backing the RAAs, but its improper exercise of discretion in setting interest rates on RAAs. The measure of damages, given the violation, must focus on the propriety of the rate credited by Unum on its RAAs, regardless of Unum's earnings.4

In this case, Unum was permitted under the terms of the plans to settle claims by means of RAAs and, given the definition of RAAs under the policies, to retain the assets backing its RAA liabilities. Tying the Plaintiffs' remedy to Unum's investment earnings is not appropriate where there is no improper retention of funds. Because Model 1 is based on Unum's interest spread rather than on the difference between the actual interest rate Unum credited to its RAAs and the rate appropriate under ERISA, it is not a relevant measure of damages.

Model 1 also suffers because its calculation of damages is backward-looking. In setting damages, the Court engages in a forward-looking analysis. See, e.g., DiFelice v. U.S. Airways, Inc., 497 F.3d 410, 424 (4th Cir. 2007) ("whether a fiduciary's actions are prudent cannot be measured in hindsight . . ."); Roth v. Sawyer-Cleator Lumber Co., 16 F.3d 915, 917-18 (8th Cir. 1994) ("[T]he prudent person standard is not concerned with results; rather it is a test of how the fiduciary acted viewed from the perspective of the time of the challenged decision rather than from the vantage point of hindsight." (internal quotation marks omitted)).

Finally, Model 1 suffers from a reliability problem. Model 1 sets a rate that is purportedly competitive with insurance "spread" products, such as guaranteed investment contracts (GICs). See Tr. 2 at 343-44. A GIC is a contract offered by an insurance company in which the insured pays a lump sum premium and the insurer guarantees repayment of the premium plus some amount of interest over a predetermined period of time. See Trial Transcript, June 24, 2013 (Tr. 1), pgs. 46-47 payments were required. It would make sense in a Mogel-type case to craft a remedy that targets the insurer's earnings on retained funds because in such a case the retention of funds is improper. (ECF No. 146). The GIC could be structured to pay out periodically over the life of the GIC or in a lump sum at the end of the GIC. Generally, GICs are multi-year contracts involving large amounts of money. See id.; Tr. 2 at 344-46. But the funds in the RAA pool were constantly in flux, and the total amount in the pool varied during the class period from over $500 million to approximately $350 million. Joint Exh. 7. Any GIC that Unum might have purchased with the funds backing its RAAs would have been too inflexible to cover beneficiary withdrawals from the RAAs.5 See TR. 2 at 344-46. Thus, the annual interest rates Model 1 proposes to credit to the RAAs are not competitive with any actual market rates that could have been achieved for the Plaintiffs.

For the foregoing reasons, Model 1 fails Rule 702's relevance and reliability standards, and is excluded.

C. Model 2

Unum worked with a third-party administrator that managed its RAAs and that also managed the RAAs of other insurers. Tr. 1 at 66-68. This administrator regularly compiled data regarding the interest rates that each of its clients credited to their RAAs. See id. It provided this data to Unum (and presumably to its other clients) in a blind format that did not reveal the identity of the particular insurers or the policies from which the RAAs stemmed. See id. It simply reported the crediting rates. See id. Unum accepted these reports and referenced them in its internal communications regarding what interest rate to credit on its RAAs. See id.; Pls.' Exh. 6.

The Plaintiffs used the data in these blind reports to establish a "top quintile" rate, which was the average interest rate credited by the top one-fifth of insurers in the reports in each year throughout the class period. Tr. 2 at 279-282; Pls.' Exh. 75, Slide 5.6 At its best, the data in the vendor reports encompassed approximately 120 out of 400 companies offering life insurance products. Tr. 1 at 68; Trial Transcript, June 26, 2013 (Tr. 3), pg. 506 (ECF No. 148). In the last two years of the class period, the number of companies in the reports dropped to about 70 as insurers, including Unum, chose different vendors to administer their RAAs. Tr 2 at 381. MetLife, which has the largest share of the group life insurance market, services its own RAAs, so data from MetLife was never contained in the third-party administrator's report. Tr. 3 at 506. There is no other published source of the data used to compile Model 2. Tr. 2 at 385-86.

Not much is known about the individual insurers on the list. There is no way of knowing who the insurers are, whether the rates reported were on ERISAgoverned RAAs, or whether the rates were preset under the terms of the policy or were set after the death benefit became due as Unum was doing. Tr. 1 at 84-87; Tr. 2 at 382-390; Tr. 3 at 506-07. It is not clear which if any of the insurers were acting as fiduciaries under ERISA, or whether, if they were subject to fiduciary duties, they understood themselves to be fiduciaries in setting rates on ERISA-governed RAAs.7

The RAA rates offered by the different companies ranged from under 1% to over 4% and the rates changed over time. See Pls.' Exh. 37. Different companies may have had different strategies. Insurers could have been treating RAAs as shortterm products and setting rates competitive with interest-bearing bank accounts. Others may have been treating them as longer-term products and setting interest rates competitive with certificates of deposit, mutual funds, or money markets. Tr. 1 at 91-92, 126-127; Joint Exh. 4 and Pls.' Exh. 7. Unum posited that some insurance companies might have used an artificially high RAA interest rate as a "loss leader" to entice beneficiaries to conduct repeat business with the insurer. Tr. 2 at 390-91; Def.'s Exh. 65 at 45-46.

The Plaintiffs argue that because Unum relied on the vendor's reports to set the RAA's interest rate, they are reliable enough to establish the rate that Unum, as a fiduciary, should have set. The argument is unpersuasive. Rule 702 requires that expert testimony be based on sufficient facts or data. This data falls short of the mark. Model 2 cannot be relied upon to establish any single competitive market rate, and McAvity's testimony on Model 2 is therefore excluded under Rule 702 on both reliability and relevance grounds.

II. Outstanding Objections to Deposition Transcript Testimony

Prior to trial, Unum objected to portions of deposition transcript testimony the Plaintiffs anticipated offering into evidence at trial. (ECF No. 125). The Court does not rely on any of the offered deposition testimony in its forthcoming findings of fact on the damages trial. Unum's objections are therefore overruled as moot.

CONCLUSION

For the reasons stated, the Defendant's motions to exclude expert evidence offered by Plaintiffs' expert Thomas A. McAvity, Jr. are GRANTED, and its outstanding objections to deposition transcript testimony are OVERRULED.

SO ORDERED.