Oral argument — Cunningham v. Cornell University

Cornell University Thank you and may it please the court. Congress enacted ERISA to impose a fiduciary duty that Judge Friendly famously described as the highest known to the law. The statute requires a fiduciary responsible for a retirement plan's assets to act with care, skill, prudence, and diligence. A duty that the Supreme Court has now twice held incorporates a duty to monitor a plan's investment options, confirming this year in Hughes v. Northwestern University that the monitoring duty applies to each investment in a plan. For years, fiduciaries of Cornell University's retirement plans did not meaningfully monitor the 300 options in which Cornell employees invested their retirement savings or their recordkeeping and administrative fees charged to their accounts. After receiving advice in 2012 that it should streamline the plan's lineup, eliminate many underperforming and redundant investments, changes that would put more money in employees' accounts by lowering fees and improving performance, Cornell still waited over five years to begin the streamlining process, continued not to monitor hundreds of funds, and never engaged in a competitive market-based process to determine if the plan's recordkeeping fees were reasonable. Despite this, the district court denied a trial because the fiduciaries periodically discussed topics such as potential disruption to participants, eventually began implementing changes due to difficulty in measuring losses from the plan's recordkeeping fees. The judgment in the defendant's favor on those issues was error because the record is more than sufficient to support a finding, both a fiduciary breach and plan losses on the investments claim and the recordkeeping claim. Can I ask you just at the start one question about the Count 5 claim? You complain about the parsing, the partial dismissal of that claim, and one of the allegations is that there were many, many investment options, and in your opening brief, you make clear that you're not saying that having a lot of investment options is per se imprudent, it's imprudent if you don't monitor all those options adequately. So I'm just having difficulty seeing what was the, if it was just one, a factual allegation in support of a general theory of imprudence, what difference did it make that the judge said I'm taking it out of the Count 5? Well, the difference it made was that at summary judgment based on having dismissed that, the court said in footnote 11 of the order that any allegation about monitoring wouldn't be considered and limited the plaintiffs to a theory that specific funds would have been removed based on performance alone. So in striking the allegation that the Tier 3 wasn't monitored, the 200 option portion of the plan, and the allegation that what has been described as the too many funds claim, which was based on overlapping funds such as 55 large cap investments, that would have been eliminated and the fees reduced if a prudent monitoring process occurred. Why isn't that harmless? I mean, does that go back to the fees, the record keeping fees? Or don't you have to show something more on that issue, whether too many funds resulted in some losses? Right. And the plaintiff's expert, Dominguez, has a calculation on that, on what the plan would have looked like under the streamlined lineup that had been recommended. And there's a calculation of the losses of what the assets would have earned in the streamlined structure, versus the structure they retained with over 300 funds. On the record keeping losses, how do you prove that without the expert testimony? Well, our argument is that the experts should not have been excluded. The finding that they relied solely on references to their experience was error, because they did more than that. They considered the size of the plan, the fact that a larger number of participants can generate lower fees, the services provided, and then back that up with other plans that paid lower fees during the time period. Aside from the experts, there was data- Just assume with me for the purpose of the question that we affirm on the district court's- Right. So there were evidence that Cornell itself submitted of the fees that were paid by other TIA clients, as well as other CAP trust clients, that's in the record. And the fact that the plan itself was able to reduce fees from over $200 per participant that was being charged by TIA to something in the range of $140. And Fidelity had a similar reduction from, I believe it was $112 to something in the 50s. Is the fact that as soon as the topic was raised that Mr. Bursick described it as TIA leaping to lower its fees, that supports an inference that the fees were high before that occurred. So then any time you would negotiate to get fees lowered, there would arguably be a claim that you could have done it sooner. Isn't that the logical implication of what you're just saying? Well, we're not relying on the fact that fees were reduced to prove that there was a breach. There's evidence that Cornell didn't understand what the fees were and didn't particularly try to find out. And that itself supports the finding that there was a breach of duty, the lack of any competitive bidding process, things of that nature. So the question then becomes, if there was a breach, did the plan suffer a harm as a result? And part of the difficulty of proving it is that there was no bidding process done that would have provided some data on what this plan actually earned. So in these circumstances, it's not unfair to look at the reductions that did occur once some negotiation took place. And I would also add that even if none of that evidence is found to be competent to prove loss, there's still the issue of the count four prohibited transactions claim. And there, the burden is on the defendant to prove that the fees were reasonable. And as the district court noted, how much the plan's paid in fees is disputed. So at this stage, it hasn't been proven as a matter of law, certainly, that the fees were reasonable as paid both to TIA and Fidelity. Well, do those rise and fall together, count three and four, whether they're reasonable or whether there's a breach or causation? Do they- I don't think so. Or can we find one way and the other and different on the other? I think you can find one way on one and a different way on the other because the burden of proof on the damage element is on the fiduciary on the prohibited transaction. And under 1108B, that's an affirmative defense that has to be proven as reasonableness. So under the exception of 408, that the burden then is on the defendant to prove the reasonableness? Yes. Thank you. Good morning, your honors. May it please the court and counsel, Mike Scodro on behalf of the Cornell defendants. The plaintiffs are striving to suggest that there's a dispute of material fact in here somewhere, but they have yet to identify one. And at the heart of this effort is to conjure a fact issue. Is plaintiff's reliance on the wrong legal standard for reviewing decisions by ERISA fiduciaries? This runs throughout the briefing and argument. Plaintiff's case presumes a legal regime in which courts are free to second-guess fiduciary decision-making at a fine-grained level. So long as plaintiffs, and they can force a trial, so long as they can muster an expert who is willing to say that they would have come up with a different fund menu or would have come up with the same menu, in this case, even faster. But that's exactly the sort of second-guessing that the Supreme Court and this court routinely forbid, recognizing that such a standard would incentivize employers not to offer retiree benefit plans in the first place. As the Supreme Court recognized in Hughes, and I'm quoting from the court, the circumstances facing an ERISA fiduciary will implicate difficult trade-offs, and courts must give due regard to the range of reasonable judgments a fiduciary may make based on her experience and expertise. And as this court held in Marino, which we also cite in our brief, fiduciary decisions are not to be judged from the vantage point of hindsight. And so long as the prudent person standard is met, ERISA does not impose a duty to take any particular course of action. If another approach seems preferable. Now, to be sure, courts will find a breach of fiduciary duty, whereas in the George B. Kraft Foods case on which plaintiffs so heavily rely, there's no evidence that planned fiduciaries were actually making conscious decisions on fund management. But here, as the district court correctly held, in a thorough opinion, there's abundant evidence of constant ongoing efforts to maintain fund performance. Indeed, that's obvious just from a review of the minutes of the Cornell Committee RPOC, which met regularly throughout the class period. And I commend those minutes to the court, all of which appear in the appendices before the court. So with that proper standard in mind, I'll turn briefly to plaintiff's claims. And I'll start with the record-keeping fees claim that was discussed a moment ago. As we show in our brief, plaintiff's record-keeping fee claim fails on all three of the required elements. But the easiest place to start is where the district court did in correctly recognizing that plaintiffs failed to show loss because they never introduced comparators here. Now, the plaintiffs recognized the need for expert testimony in the trial court, but the district court has excluded that testimony on Daubert grounds. And that is a decision that plaintiffs don't appear to meaningfully contest on appeal, though I'm more than happy to discuss it. Instead, they now contend they could make that showing without expert testimony, that required comparator showing without expert testimony. But for the reasons we detail in our brief, including the fact that one of the arguments simply misreads the data, none of those arguments carry water. And as we further explain in our brief, plaintiffs fail on the other two elements, both breach and causation as well, even if defendants bear the burden of disproving the element of causation. Before you get into causation, on the record-keeping fees, the factually similar comparator analysis that the district court used, that seems to be, I mean, the wording suggests that that may be a question for the jury. Is it not? It's not- Using TIA's own data, you can compare the numbers and make a decision. Different people, even excluding the expert testimony, might see those things differently. It's not, your honor. The summary judgment is used in these cases as a critical gatekeeping moment to ensure that in instances like this, where there is no evidence to suggest that Cornell was in any way behind the times. Indeed, they were out there, uncontested evidence, trying to get fee reductions before RPOC, before they had retained CAP trust. You're going to breach now, but isn't this, the fee, the numbers have to do with the losses? They do, your honor, but there were no relevant comparators shown. The Harvard comparison, there was an admission that it was simply chosen out of thin air to support a pre-existing theory. The Caltech- I guess my question is, isn't that itself a fact question for the jury? Whether this is an adequate comparator or not. Not under the relevant standard, your honor. As Hughes makes clear, that would be allowing the juries to make precisely the kind of fine-grained reconsideration of heavily processed, detailed decision-making made, as the district court correctly identifies, that the fiduciaries in this case made. They had a process that included, again, this goes back, I know, to some extent, to breach, but with regard to the losses, you have to have, the cases are clear. You have to have comparators that match up in terms of all the relevant, meaningful criteria. Would you need an expert to identify relevant comparators? Is that the problem? You absolutely would, your honor, and the two experts here were excluded. We see the need for the experts, even in the errors made in the briefing, without experts, in terms of interpreting the relevant data, the top quartile point that's discussed in our brief. So you absolutely need expert evidence on this. And in the absence of it, to merely throw before jurors, well, here are a couple of instances where, notwithstanding all these efforts and so forth, here are a couple of instances where we think potentially comparable funds, maybe not, maybe a different TIA composition, maybe more or fewer participants, whatever it might be. Regardless of all those differences, here are some other places that achieved lower fees. And by the way, there's uncontested evidence that we were at the very bottom of the fee scale. But even if that weren't true, even if that weren't true, that would not be a jury question under the standard the Supreme Court reaffirmed in Hughes. Before I cut you off earlier from getting into causation, but I did want to ask you about that. Could you address the burden shifting rule in particular that NYU case seems to adopt the common law burden shifting to defendants to prove losses? And so I just want you to- Sure. Thank you, Your Honor. Yeah, we acknowledge the language in Sasserdoti that does appear to place the burden on the fiduciary defendants on that issue. And therefore, while we disagree with that for all of the reasons detailed in our brief here and our briefs below, and we preserve certainly our concern with that decision, we think it's incorrect. There's more than enough, as we point out in our brief, and there's more than enough evidence here to carry our burden on causation, even if we hadn't prevailed on loss, or excuse me, yes, on causation. And as we point out in our brief, we also, there's more than enough to prevail on preach in light of the evidence of process involved here and the ongoing efforts to bring down these fees. And indeed, as to Your Honor's earlier question, the ultimately successful efforts to bring down fees. So we contend that we prevail on all three elements. We recognize the language in Sasserdoti, but contend that that would have no impact on our capacity to prevail on that element, Your Honor. Could you address the prohibited transaction claim, the dismissal of that claim, and particularly the statutory interpretation issue? Sure. Well, Your Honor, it's as every, to my knowledge, every appellate court to address this issue has concluded, and that includes Ramos in the Tenth Circuit, Suida in the Third, I understand there's a recent Seventh Circuit opinion to the same effect that's post-briefing. The notion that the very transaction that creates the party in interest is also then the prohibited transaction would ensure that you would have litigation, as the Ramos Court puts it. You would have litigation over every routine hiring of a vendor, of a record keeper. It would be, it cannot possibly be what Congress had in mind. And indeed, while there's a defense, it would mean you're in court and the fiduciaries are having the burden of proof to defend under the statute. That cannot be the kind of, you know, again, the case is made clear. Congress was concerned about transactions that by their nature are likely to be detrimental to the plan. And this is the furthest thing. These are the routine transactions that must occur with regard to every plan. And the statute contemplates that there are necessary, elsewhere contemplates that there are necessary services that must be provided. Correct, Your Honor. Correct. But there is sort of a plain meaning argument. I mean, the argument in the Suida case seems to be persuasive as to the overall purpose of the statute. But are there other textual arguments that might be drawn from the statute? Well, I think nothing in the statute precludes, the statute leaves open whether or not that party and interest, that interested party is one that has some sort of pre-existing relationship. And that is certainly an easy reading of the plain language. And it's the only one that makes sense. If you had already hired someone as your fill in the blank on behalf of the plan and that your attorney, and then you turned around and hired them as your record keeper, one would imagine why Congress would be interested in requiring the fiduciary to justify the costs there. But the plain language leaves open that possibility. And it seems that to the Suida point, to the Ramos point, the only plausible reading. And now I'm over my time, your honors. I'm happy to entertain any other questions from the court, of course. Thank you. We'll hear from your colleague. Thank you. Thank you, and may it please the court. My name is Caroline Wong of Sidley Austin, and I represent Cap Financial Partners, also known as Cap Trust. Cap Trust was the plan's investment advisor, and so its role in this case is relatively limited. The only claims against it that are at issue on appeal are the prudence claims in count five, all of which relate to the plan's investment options. Three claims in count five made it to summary judgment, and the district court correctly entered summary judgment for Cap Trust on all three. There are a host of reasons for that, many of which are also covered in the Cornell defendant's briefing. But I'd like to focus today on the element of breach as specifically relevant to Cap Trust. So plaintiff's first theory was about the Kreft stock account and the TIA real estate account. Summary judgment on that theory should be affirmed, because Cap Trust had a prudent process for monitoring those options. To give a brief overview, Cap Trust evaluated and made recommendations to the Cornell committee with regard to those two options in November 2012, within a year after it was first retained. It evaluated those options again consistent with the committee's investment policy statement the following year. It developed and then used custom benchmarks to evaluate those two options with the committee. It also even had a team that spoke with the managers of the Kreft stock account to supplement that evaluation. We detailed Cap Trust's process with regard to those investment options in our brief. Was there a process for dealing with underperforming investments? Absolutely, your honor. So the investment policy statement that the Cornell committee finalized and approved in November 2012, set forth very specific criteria for benchmarking investment options, including those two in addition to the others in the plans, as well as for scoring those investment options based on both quantitative and qualitative factors. Cap Trust then implemented those investment policy guidelines, and at the very next committee meeting after that set of investment guidelines was approved, Cap Trust presented the committee with a comprehensive benchmarking analysis that covered the hundreds of investment options in the funds, including those two in particular. Those criteria took into account performance of the funds, but again, took into account many other factors, various aspects of performance, fees, qualitative factors relating to the managers of each investment option. So in short, yes, there were specific metrics laid out in those guidelines to address underperformance, and Cap Trust implemented those guidelines in the analysis it provided to the committee. The process with regard to the Kreft stock account and the TIA real estate account in particular was at least as robust as the process that the NYU committee used in the recent NYU case that came before this court, and the court found to be sufficient. So we submit that retaining them, Cornell's decision to retain them based on Cap Trust's and Cornell's process was very comfortably in the range of reasonable judgments a fiduciary can make, to take a phrase from Hughes. The district court also correctly entered summary judgment on the claim regarding the other investment options that plaintiffs have challenged. And of course, there's some issues about whether plaintiffs have waived that challenge to some extent, but the bottom line is Cap Trust had a prudent process for monitoring those investment options as well. And that process is actually why dismissal of the other and sort of overlapping claim in the amended complaint about the defendant's purported failure to monitor certain investment options prior to October 2014 was harmless. The parties took discovery on Cornell's and Cap Trust process with regard to monitoring, and the process showed, the discovery showed that that process was quite robust. I touched on that process a little bit, covering the November 2012 to July 2013 time period, after the investment policy statement was put into effect, but Cap Trust even had a process before that time. In January 2012, just three weeks after Cornell retained Cap Trust for the first time, Cap Trust reviewed the plan's entire lineup and recommended eliminating the challenged investment options, as plaintiffs acknowledge, both in their briefing and in their argument earlier today. Cap Trust also made similar recommendations to the committee in the April 2012 committee meeting that followed. And then again, worked with the committee to put together an investment policy statement and then provided robust benchmarking consistent with the guidelines in that policy statement after it was approved by the committee. That process is right there, all laid out in the committee meeting minutes and materials. Those core facts about the process are not disputed. We detail it in our brief. The district court also thoroughly described that process in its summary judgment opinion. And plaintiffs have nothing to say about those core facts or any basis to dispute them in their reply. The remaining theory relevant to Cap Trust that made it to summary judgment was plaintiff's theory about lower cost share classes. I will not address that today because plaintiffs have not challenged that aspect of the summary judgment ruling on appeal, at least with regard to Cap Trust. Just a few words about the theories that were discussed at the motion to dismiss stage if your honors have any questions about them, but I see I'm over my time, so otherwise I will sit down. Thank you for your argument. Thank you. Thank you. Thank you. Just to address the point that Cornell's council cited, the comment in Hughes that a fiduciary has latitude to apply its experience or expertise. The court is simply saying there that if there was a prudent process, the courts will not second guess a choice between two reasonable alternatives. But the issue here is, the evidence shows that they did not apply their experience or expertise to make decisions. There was a recommendation to make changes, and there was a five year delay in implementing that. At a minimum, that shows a lack of diligence. And these were changes that were going to benefit the participants and eliminate things that were harming the plan. There's evidence that other clients were able to accomplish the changes in six months to a year. So the five year delay that occurred, there's questions, at least genuine disputes about whether Cornell followed a prudent process. The point about whether we misread the data, there's no misreading of the data. Cornell's own expert agrees that costs are tied to the number of participants. And this plan had among the top 15% largest participant size plan in the data set. Yet, they were paying sometimes record keeping fees in the 40%. So Hughes doesn't say anything about a standard for data to support losses as Cornell suggested. The disputes about whether the data is properly interpreted, those go to weight and not whether it's sufficient to support a finding of loss. And unless there are any other questions, that's all. Thank you. Thank you all. Well done, and we'll take the matter under advisement.