The Exchanges

Q Neil, what do you think you can do at Duke, or you do do, that a number of your other institutional peers don't or can't?

A I don't know. There's a lot of things that there's not at least somebody that can do what we do. I don't think we're that unique. A few things I would throw out, I think Duke has always had a culture ever since I've been there of being willing to, to take some risks, a risk-taking culture, and we've done a lot of things. We were very early in venture capital. That was before my time. We are very aggressive in alternatives broadly. Privates are about half of the portfolio now, and so those are some things that I think given the, the risk culture at Duke, we're able to do. Risk Culture meaning we'll take risks, but know what the risks are, and not take them willy-nilly, I guess. We also do a lot in separate accounts. We've built up a big operational support team, so even on the hedge fund side, we do a lot of separate accounts, and that's more operationally difficult for some of our peers, and I think that's been somewhat of a differentiating edge, and then we do some direct investing also, not trying to replicate what our managers do, but on the private side, we do a lot of co-investing. On the public side, we do, I guess I won't use the term, quantum mental, We do some sort of quasi-passive portfolio construction to do tactical bets on the public side.

AI assessment note: “that's more operationally difficult for some of our peers, and I think that's been somewhat of a differentiating edge”

Q at the company level, you work your way, maybe managers are starting to use it, and then you get to the allocators, and there's, well, you know, we kind of run a few statistics here and there. So what are you doing differently with this data that you can do today because of the abundance of it that you might not have been able to a number of years ago?

A I'll give you a few examples, I guess, and I don't know that any of this is revolutionary, but we have full transparency on about 90% of our portfolio, so we have a very robust data set. We can look at daily P&Ls for all of our public managers. We have it all plugged into Bloomberg, and that in and of itself can be distracting, or it can be a very valuable tool to let you identify if something strange is going on in the portfolio. I think from an analytics standpoint, I'll just give you one example. Because we have the position level data, and we see how our managers are trading around, we do Trading analysis on our portfolios, doing things like tracking if the manager hadn't traded the portfolio this year or this quarter versus if they did, how would that have done? What are their trading costs? How much do they lose each year based upon the way that they're trading the portfolio? Things that may not be intuitive to us as we are, and have actually sometimes been very surprising to us, the outputs once we've invested with a manager, and we're trying to do better about doing it even before we invest. So that's one example of how we've used data.

AI assessment note: “tracking if the manager hadn't traded the portfolio this year or this quarter”

Q Neil, have you had any difference now that you've been in the seat for a dozen years or so of being pulled in different directions less today than in the past? Maybe an example of that.

A Being pulled less? Probably not. I mean, I think it's getting, it's been there ever since I've been at Duke, and Duke has done a really good job of really trying to focus, we've, that's why they created DUMAC as a separate, we're all Duke employees, but we're a separate kind of subsidiary of Duke, effectively. We're off campus, so they try to isolate us so that our mission is really the economic interest of Duke University. They've always done a really good job of that, but I think, obviously, with ESG and a lot of those things, we're getting more Pulled into a lot of discussions that we weren't pulled into as aggressively 10 or 15 years ago. So it's probably going the other way slightly from that standpoint for us.

AI assessment note: “Being pulled less? Probably not.”

Q Just pull that thread one more. You have a manager that you've now gone to, and they're adjusting their trading based on this input that you provided them. How does that make you feel about their sort of independence and conviction as a manager as compared to their ability to evolve?

A If they did it based solely on our analysis, that would be problematic because we're doing a very rudimentary. What we've seen is it usually pushes them to do a more thorough review of their own processes and approach and come up with their own way. We're not telling them what to do. We're just saying, hey, here's an issue that we're seeing. You're losing 600 basis points a year in trading costs. Your alpha has been flat. I mean, something's wrong here, and that's usually enough for them to go. We're not telling them, hey, you need to trade less, or whatever. They then go through and identify, we're moving our net and gross exposures too much, or in the case of a long, short equity manager, or maybe we are trading too much, and we're trying to limit the, whatever it is, we want them to come up with the answers. We're just telling them what the yellow or red flags are.

AI assessment note: “If they did it based solely on our analysis, that would be problematic”

Q How do you think about Your edge. So you go out and talk to managers, and you always want them to figure out what's their competitive advantage, what's their edge. What is your edge compared to your peers?

A I don't know about relative to our peers, but what we view as our edge internally is identifying people. I mean, our job at the end of the day, I thought when I joined UMAC, 20 years ago, it was going to be a very quantitative job. We'd put numbers into a spreadsheet and the output would be these are the managers that you should hire and learn very quickly that it's very much a people job. And I think that is our core competitive advantage. We spend a lot of time on the road visiting managers and our ability to assess the managers that we think are going to be the best performers going forward. That is, in our view, the core competency. It's challenging because data is very helpful in this process, but really returns aren't a big driver of our decision-making process. We want the best people in the portfolio working for Duke, and that's what we try to find.

AI assessment note: “what we view as our edge internally is identifying people”

Q As you have to balance managing a pool of capital and managing a team of professionals, how have you learned to manage people?

A It's funny, when I was offered the job to be CIO, I never really wanted to be CIO, because I never really wanted to manage people. I view myself as an investor, not as a manager or administrator. It's the part of the job that I was most worried about, frankly, but it's actually, it's been one of the more rewarding parts of the job to build a team and to build a group of people with a common mission. A lot of it has been trial by fire, I think. There wasn't a training program. I guess I did get an MBA, so that was somewhat of a training program, but You don't learn until you, you do, and you're in the situation, but for me, it's about developing the relationships with the people on our team, and then trying to make sure that we all have a common mission and a common direction, and that's been the challenge and the learning that I've had.

AI assessment note: “A lot of it has been trial by fire... You don't learn until you do”