Greg Bond discusses Man Group's multi-strategy shift, AI limitations, and regime shifts
Summary
Greg Bond of Man Group explains the firm's transition to a multi-strategy model to diversify alpha sources and support its central technology platform. He argues that recent hedge fund performance has been inflated by favorable market conditions and that discretionary strategies offer better diversification than quantitative ones during crises. Bond highlights that while AI aids hypothesis generation, human judgment remains critical for validating models due to limited historical data. He also notes that machine learning struggles with long-horizon correlations, necessitating contextual modeling and factor timing in shifting regimes.
Key takeaways
- 1 I think discretionary held up a bit better in GFC than Quant when Quant's done really well recently. That's actually another source of diversification.
Unidentified speaker (speaker_1) · 29:15
- 2 It's really how are you good as an organization? taking that technology and marrying it to your alpha generation philosophy and how that works?
Unidentified speaker (speaker_1) · 32:12
Episode sections
- 0:00 – 3:00
Migration to Multi-Strategy and Private Markets
Speaker_0 asks if Man Group's shift from quant purity to multi-strategy and private credit is proactive. Speaker_1 confirms this is a continuation of a trend started under prior CEO Luke Ellis, aimed at diversifying alpha sources and supporting the central tech platform.
- 3:00 – 5:30
Career History and Red Sox Sabbatical
Speaker_0 notes his Boston background and asks about Greg Bond's time at the Boston Red Sox. Speaker_1 recounts a 2011 sabbatical where he applied quantitative analysis to draft strategy, comparing himself to the Jonah Hill character in Moneyball.
- 5:30 – 7:26
Transition to CIO and Organizational Culture
Speaker_0 asks about the learning curve of becoming CIO of a global multi-strategy firm. Speaker_1 describes a gradual transition over a decade, balancing systematic roots with discretionary needs, and emphasizes connecting people across the firm.
- 7:26 – 28:35
Alpha at Scale and Branding Strategy
Speaker_0 asks how Man Group retains its innovative ethos at scale. Speaker_1 explains the focus on maintaining an innovative culture while scaling operations.
- 28:35 – 30:18
Tail Risk and Long-Term Diversification
Speaker 1 argues that recent hedge fund performance has been inflated by favorable market conditions, suggesting investors should look at longer-term indexes including the 2008 financial crisis and the internet bubble to understand true diversification. He notes that discretionary strategies held up better than quantitative ones during the Global Financial Crisis (GFC), offering a source of diversification that trend strategies, which have not provided convexity since 2022, currently do not.
- 30:18 – 31:32
Historical Context and Tech Identity
Speaker 2 highlights that the last 15 years of low rates and inflation were an anomaly, not the norm. He asks Speaker 1 about Man Group's historical identity as a technology-native firm and whether its competitive advantage in data science and machine learning is eroding as competitors adopt similar tools.
- 31:32 – 35:29
Data, Technology, and AI in Investment Processes
Speaker 1 explains that while technology and data are essential, they are not sufficient for alpha generation without strong organizational processes. He notes that AI helps generate hypotheses and code but struggles with the validation step due to limited historical data for mid-frequency strategies, making human judgment and 'harnessing' of AI critical.
- 35:29 – 39:15
Talent and Organizational Culture
Speaker 2 asks if the industry needs to hire more liberal arts graduates to validate AI models. Speaker 1 agrees, suggesting a convergence where systematic teams become more creative and discretionary teams more systematic, requiring a shift in hiring from purely technical skills to creativity and adaptability.
- 39:15 – 42:30
Regime Shifts and Model Adaptation
Speaker 2 questions how models adapt to shifting regimes like deglobalization and climate change. Speaker 1 states that traditional assumptions are breaking down, necessitating flexible, contextual modeling and factor timing rather than static factor investing, as machine learning is less effective over longer horizons where correlations change.
- 42:30 – 44:25
Geopolitical Risk and Operational Resilience
Speaker 2 asks if geopolitical risk can be modeled. Speaker 1 suggests measuring deglobalization levels first and then using diversification, noting that cross-border capital flows may break down, requiring regional focus. He also highlights operational resilience issues, such as data flow restrictions between countries, as a new risk factor.
- 44:25 – 47:16
Agentic AI and Organizational Structure
Speaker 2 asks about the cultural impact of AI agents replacing or working alongside humans. Speaker 1 describes AI as a 'people enabler' rather than a replacement, citing the analogy that ATMs increased bank teller jobs. He emphasizes the challenge of connecting AI agents across different silos (operations, sales, clients) and the need for cross-functional coordination.
- 47:16 – 50:20
Personal AI Use and Conclusion
Speaker 1 shares that he uses AI to evaluate investment opportunities, explore academic concepts like 'P Doom' (probability of AI catastrophe), and build research stacks with multiple agents for review. He concludes the interview by stating his preference for bourbon over wine.
Assets discussed
| Asset | Kind | Stance | Context |
|---|---|---|---|
| S&P 500 | index | neutral | Why not just own the SP 500 or frankly just seven stocks forever? What is the point of any diversification? |
“Stance” is how the speakers talked about the asset, not a recommendation.
Claims worth verifying
Man Group is a $270B multi-strategy alternatives firm.
Unidentified speaker (speaker_1) ·28:35
How to verify: Man Group investor relations website or latest annual report (10-K) filed with SEC EDGAR. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Primary sourceNumeric was acquired by Man Group in 2014.
Unidentified speaker (speaker_1) ·28:35
How to verify: Man Group press releases or SEC filings (e.g., Schedule 13D or 10-K) from 2026. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Primary sourceDiscretionary strategies held up better than Quant strategies during the Global Financial Crisis (GFC).
Unidentified speaker (speaker_1) ·29:15
How to verify: Compare performance data of discretionary vs. quantitative hedge fund indexes (e.g., HFRX Discretionary vs. HFRX Quantitative) for the 2026-2026 period. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Primary sourceTrend strategies have not provided convexity since 2022.
Unidentified speaker (speaker_1) ·29:54
How to verify: Analyze performance of trend-following or CTAs (Commodity Trading Advisors) from 2022 to present using data from HFR or Bloomberg. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Speaker 1 grew up on the bottom-up security selection side.
Unidentified speaker (speaker_1) ·29:15
How to verify: Greg Bond's professional biography on Man Group's website or LinkedIn profile. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
The last 15 years featured low interest rates, lack of inflation, and rising public markets.
Unidentified speaker (speaker_2) ·30:46
How to verify: Federal Reserve economic data (FRED) for interest rates and CPI inflation from 2026-2026. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Man Group has a 'tech native identity' dating back to its history.
Unidentified speaker (speaker_2) ·31:11
How to verify: Man Group corporate history page or annual reports detailing their technology investments and acquisitions. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Machine learning is not great for financial markets over mid-frequency horizons because correlations change.
Unidentified speaker (speaker_1) ·40:47
How to verify: Academic literature on market microstructure and regime switching, or Man Group's published research on contextual modeling. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
There are more bank tellers today than before the advent of the ATM.
Unidentified speaker (speaker_1) ·46:30
How to verify: Federal Reserve Bank of St. Louis research paper 'The Impact of ATMs on Bank Employment' or similar economic studies. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
P Doom is the probability that people will die from AI.
Unidentified speaker (speaker_1) ·47:44
How to verify: Silicon Valley discourse or articles defining 'P Doom' (e.g., from AI safety researchers or tech news outlets). For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.
Do your own research
- Verify hedge fund performance indices (e.g., HFRI, CISDM) for the 2008 GFC and internet bubble periods to assess diversification claims.
- Review Man Group's latest annual report and 10-K filings for details on multi-strategy asset allocation and private market exposure.
- Analyze academic literature on the efficacy of tail-risk hedging strategies post-2022 to validate claims of convexity failure.
- Research Man Group's technology infrastructure investments and data estate disclosures in recent earnings calls.
- Evaluate the hiring practices of major quant firms to assess the shift from technical PhDs to liberal arts backgrounds.
- Study the performance of factor timing strategies versus static factor investing during recent regime shifts (e.g., 2020-2024).
- Assess geopolitical risk models used by global macro funds, specifically regarding cross-border capital flow restrictions.
- Check for conflicts of interest: Greg Bond is the CIO of Man Group, a firm that manages the assets discussed.
Full transcriptexpand
0:00Unidentified speaker (speaker_0): The business has fundamentally begun to migrate from this quant purity that numeric and AHL have had for a long time to multi-strategy, to adding private markets. You've got private credit in there, maybe most notably mid-market. Is that a proactive insight? And if that's the case, what led to that decision and how are you going about pursuing that?
0:20Unidentified speaker (UNKNOWN): Well,
0:20Unidentified speaker (speaker_1): I think some of those acquisitions actually started under our prior CEO, Luke Ellis. So again, that's sort of a continuation, I think, of a trend in terms. of what kinds of acquisitions and things that we've looked at. I think the broader strategy is a bit more around diversification of the business itself. So we have, again, we need to support the tech platform that we've got, Salesforce, operations, all of the organizational things in the center. So I think from a business perspective, it's very good to have multiple sources of alpha. And then that also leads to why we really like solutions type offerings for folks because we we've got these ingredients and then we can use those to leverage together and build hopefully something that's what clients need in their particular portfolio situation. So I think it's less about acquisitions and growth, as I said earlier, but more about can you put together things, alpha sources that are beneficial to folks over different periods, different periods of time. And so I think that's the philosophy, which again is part of some of the branding, why While we talk about man systematic, man. discretionary, not GLGH, L numeric, and some of the brands. I think that's important. Just sort of think about how we face off with the market.
1:23Unidentified speaker (speaker_0): Welcome to Capital Decanton. Here we say goodbye to tired market takes and superficial sound bites to dive beneath the surface of capital allocation. This show attempts to challenge conventional wisdom and likely some of your own assumptions as we explore the nuance. nuances that can change how we understand an issue and its practical implications. Along the way, we draw on experienced voices from across the industry, allowing their perspectives to breathe and follow their thinking as it opens up in unexpected directions. This season, we're serving Capital Decanted as a tasting flight, a series of focused conversations, each offering the chance to spend time with one distinct. distinctive voice and plenty of room to explore what emerges. One voice, but a deeper pour.
2:09Unidentified speaker (UNKNOWN): And
2:09Unidentified speaker (speaker_0): we're delighted to have Franklin Templeton back again for season four as our title sponsor. They just continue to be an instrumental partner in bringing compelling educational content to life here at Kaya. With over 40 years of private markets experience and 280 billion of assets under management, Franklin Templeton private markets investment managers have expertise across private equity, venture capital, private real assets, private credit, and digital assets. And of course, all of them operate with the client-first mentality that has always defined Franklin Templeton to help prioritize investment outcomes. So thanks so much, Franklin Templeton.
2:46Unidentified speaker (UNKNOWN): Well,
2:46Unidentified speaker (speaker_0): I'm now joined in studio with Greg Bond, who is the Chief Investment Officer. of Mann Group, the $270 billion alternative asset manager. Greg, welcome to Capital Decanted.
2:57Unidentified speaker (speaker_1): John, great to be here. Looking forward to it.
3:00Unidentified speaker (UNKNOWN): Well,
3:00Unidentified speaker (speaker_0): uh, we've not met before. I know Robin well, and I know Mann Group even better, but I don't know if you know this, but I'm a Boston guy. So I cut my teeth ITH at State Street Global within then a Boston company. I know two neighbors you would know very well being at numeric for most of your career. I understand, I stumbled upon this, Greg, and was so excited. You worked for the Red Sox. So tell me a little bit about that. I think it was a sabbatical or some sort of a secondment, but tell me about how that came about and what you did.
3:29Unidentified speaker (speaker_1): It was a sabbatical back in 2011 now, a long time ago. And it was the fried chicken year for the Red Sox, if you people remember, where they had a high probability of making the playoffs when I started working for them, then they didn't make the playoffs. But yeah, basically the head quant at the time was in a couple of my alumni databases. I just reached out, we had a couple of lunches, and then I said, Hey, do you have a project for me? And they had some questions on their draft strategy. So it was a lot of fun to do money ball. I was also when the movie was out at the same time. So that worked out when people asked what I was doing, kind of map back more the Jonah Hill character, not the Brad Pitt character. Obviously, it was fun and it was. the good data and just talking with the guys and being part of the ballpark. It was fun until they lost, of course.
4:08Unidentified speaker (speaker_0): That's fantastic. So that was the end of Theo's tenure there.
4:12Unidentified speaker (speaker_1): It was that transition. It was Theo. So I was in there in the room for some of the drama that was going on, or just around the area, what was going on with the transition. But it was interesting.
4:21Unidentified speaker (speaker_0): Okay. And any players, listeners would recognize that you played a even a small role in recruiting.
4:27Unidentified speaker (UNKNOWN): I
4:27Unidentified speaker (speaker_1): don't know if they would ever use that or what they ultimately use in their analysis. I think the bigger part was identifying players maybe they had missed on. I also didn't realize, but there's a guy named Bill James, who's the father of Moneyball, and they had given him the same data set that they gave me, but they didn't tell me that. So it was kind of fun to have a little bit of a head-to-head at the end of the project. But it was really interesting to see what causes success, just like I try to do here in my daily job, what predicts success. What kind of things do you look at today that tells you about tomorrow?
4:53Unidentified speaker (UNKNOWN): Well
4:53Unidentified speaker (speaker_0): I just got back from about six weeks on the road and you start to kind of head down the airline movie library into rewatching. So I just rewatched Moneyball. It's a fun one. By the way, anybody you would have drafted during those times, we probably traded away to the Dodgers. So it wouldn't have mattered.
5:07Unidentified speaker (speaker_1): Oh exactly. We wouldn't want to pay them anything. So we'll send it to the Dodgers. Don't get me started.
5:11Unidentified speaker (speaker_0): Well, listeners, if you're not baseball fans, we will move on from there. But nonetheless that's fantastic. Great, fun story. So, Greg, you spent nearly your whole career in numeric. And about a year ago, as far as I can tell, December of 25, you became CIO of the man group. And so I'm curious, having worked in a systematic shop in Boston, that means both microcultures in different ways, the same organization for a long time. Suddenly your you're chief investment officer of a global multi-strategy firm with a fairly new CEO. I'm curious the learning curves and changes you had to go through as a leader and a professional to assume that role and be successful at it.
5:52Unidentified speaker (UNKNOWN): Yeah,
5:52Unidentified speaker (speaker_1): it was actually a bit of a gradual transition as well, in some sense, because we joined Man Group, and Merrick Americ joined Man Group back in 2014. And even throughout that, call it 10-year period, I worked a bit. with some of the discretionary managers over time working on our ultimate multi-strategy product that we put out there that combines systematic and discretionary content. So it was at that point where I worked on both sides of the business. And Robin asked me to transition into more of a central role, which is really exciting. I like to do something slightly different organizationally, organizational like every four or five years. I think it helps you grow. And it's been really interesting. So I come at everything cognizant of the fact that I grew up on the systematic side. You know, it's knowing that helps me make maybe some decisions on the discretionary side. It does also teach me that it's really, really painful for me to do interventions and top-down portfolio changes that aren't systematically driven. But the nature of the game and responding to markets, you've got to add that in. So I think I've got a good balance between those two.
6:46Unidentified speaker (speaker_0): So I think most listeners probably don't realize how long the history man group. Mangroup has. Over 250 years. In fact, I was joking with Robin just a few weeks ago that the firm is older than our. You were in my country, and you opened the first New York office around the time I was born. So I have this intertwined history in odd ways with Mangroup, but you've often been when you're talked about, people will mention things like innovative, entrepreneurial, tech forward. You now have as far as I can tell, 600 technologists, 1700 employees, 25 offices. I already mentioned 270 billion in AUM. How do you, Greg, when you think about again your role at the executive level now across several strategies across the group? How do you retain that challenging of conventions mantra, that ethos of experimentation when you're now pretty darn big?
7:37Unidentified speaker (speaker_1): It is particularly coming up from that. numeric side where we probably had about 130 people at kind of max and thinking about that from a broad organizational perspective. So first of all, it is a lot of fun because there are a lot of resources and smart people, of course, and interacting. So part of my job in addition to running some of the strategies is really trying to connect people, make sure that we're aligned, again, get what their backgrounds are, are we having a common philosophy, a common approach? What can we do better? I think what we try to do is tie everything into a couple of themes. One, this is about alpha at scale. We want to be high quality. It's not just scale, it's alpha times the assets or dollars of excess return. So just try to add value for our clients. So if you start with that as a first principle, then it helps going down as you go into the other parts of the organization, getting deep into the engines. And I think it's morphed over my time at Mangroup, man group is it probably thought of us 10 years ago? as there was GLG, there was AHL, and then there was numeric. And I think what you've seen under Robin, particularly the last couple of years, is we're man group. We've got systematic and discretionary capabilities. And so that also brings people and connects them a bit now that that's the front-facing view. And it makes it easier actually to connect with clients. Cause I think from the outside, you're Or like, wait, do I talk to AHL about this? Or I talk to numeric about this? Do I go to GLG? So I think that's helped the connection. and having a strategy now fit that cross-connection piece. Really, that's
8:56been one of the motivations for me and excitements to come into this central role.
8:59Unidentified speaker (speaker_2): I do want to get back to this idea of the collision of the discretionary and the systematic, which is
9:04Unidentified speaker (speaker_0): a new ish phenomenon, not brand new, but
9:06Unidentified speaker (speaker_2): new-ish.
9:07Unidentified speaker (speaker_0): But maybe just talking about your acquisitive history, you just made mention of a couple of them. AHL, GLG is how you got Robin. Numeric is how we got you. Bardon Hill is brand new. So you've got this acquisitive DNA that even preceded Robin. How do you guys think about, and this comes back to your role as CIO, too? I'd be curious. Cultural integration, or frankly, or not, peaceful coexistence would be the other extreme of that. So when we think of a pod shop and in how they resource allocate, capital deploy across independent. effectively operating strategies. How would you compare man group today to something like that?
9:45Unidentified speaker (speaker_1): This acquisitive path is really it's not used that often. I mean we've talked about these names, but I think a lot of it and when man group's looking at these things, it's about the culture, collaboration, where the people that we're talking to as they join the family. And I think when numeric was looking for a partner, we really felt that on the other side again, alpha at scale, we want to make good strategies. It's not just about growing the business per se. So I think you have to come at it in that perspective. And then if you use that as a screen, you ultimately knock out quite a few of the different acquisitions and other things that are out in the marketplace. And so where we sit today, and I think some of the motivation for some of the private credit acquisitions is just to have, and again, share a very similar culture, just in a different asset class. Of course, I have to understand turnover, private credit doesn't turn over nearly as much as what we do over here, but I'm learning all about that. But no, all cutting aside, I think it's that culture, we've got great equity credit on the public side, liquid side. Is there something that we can do on the private side as well with that same ethos? And so that's really the goal. It's less around just trying to grow inorganically. It's more can we take these skills, put them in a place where they can leverage the sales Salesforce, force, the technology stack. And I think that's where we're headed. today, is just you need to have a strong technology foundation and continue to invest in it. And if you've got multiple engines that are taking advantage of that, that does help diversify. You can always invest. Something's always generating alpha, hopefully, and you can always keep that investment up.
11:04Unidentified speaker (speaker_2): So beyond the diversification though, maybe just pressing a little bit more, is there a house view at Mann Group? Does it make sense to have one view that trickles down in some form to all these independent? strategies?
11:16Unidentified speaker (UNKNOWN): So
11:16Unidentified speaker (speaker_3): we
11:16Unidentified speaker (speaker_1): don't have a house view in the sense of, okay, Greg Bond, CIO, this is what the SP 500 is going to do this year. It's more the the house view is back to that culture of alpha. I think that's the critical thing we organize around. And it that then trickles down into the broader organization. And so whether that's alpha in the investment sense, but also in the operational sense, sales sense as well, what can we do to deliver? more value to our clients? And that's also because of the way we're set up, a lot of what we do ends up being highly bespoke as well. Because we have all these different options and different things we can put together for clients. We don't really do, we have a few things, and some of the things I run are sort of what I call off the shelf, but a lot of it is around these deep client discussions. And so that's the strategy, that's the house view. Clients sit right at the top, and we want to try. to do things, whether it's direct investments or providing other research opportunities and things like that for the client.
12:08Unidentified speaker (speaker_2): And is there any competition for capital from these underlying strategies, either from an investment attention standpoint, a sales and
12:16Unidentified speaker (speaker_0): operational
12:17Unidentified speaker (speaker_2): resources standpoint?
12:18Unidentified speaker (speaker_0): How does that get facilitated across the firm?
12:21Unidentified speaker (speaker_1): Some parts of the business might be doing better than other parts at a given point in time. There's obviously client demand. What are clients looking to do? But of course, you don't want to necessarily not do something just because there's not demand today. There might be demand in the future, or it might fit a problem that a client already has. I think that's critical. One of the benefits of being this more central role of having multi-strategy approaches is it does help organize some investments in the business. So if we have strategies that combine systematic and discretionary, then what we're thinking about the platform, hey, for if we're going to do something and say systematic, it needs to help discretionary. If we do something in discretionary, it needs to help systematic. So I think we've done, again, in my experience at man over the last decade or so, really being much more strategic in that investment and hopefully do it once rather than do it three times. I think that's pretty critical to do organization. And so that happened, and with that strategy, it also happens to dovetail nicely with where we're going with AI and just that investment. Because I think Cause if you you're very separate with three engines all trying to figure out how to handle AI, it's better hopefully to consolidate that a bit.
13:21Unidentified speaker (UNKNOWN): Well
13:21Unidentified speaker (speaker_2): speaking of consolidation, you came into the family numeric that is in 14.
13:26Unidentified speaker (speaker_3): Correct.
13:27Unidentified speaker (speaker_2): So about 12 years ago, AUM just on my research was about 26 billion at the time. And you've just recently announced a combination with AHL internally. So these are previous acquisitions that are now mashed together, macro and bottom-up quant, both systematic roots. And now you're this 156 billion of the 270 billion on systematic. So what was the catalyst, given all we've talked about competition for capital and these diversifying capabilities of the power of the platform? Different things do well at different times, as you said. What was the reason to combine these two now? Well,
14:02Unidentified speaker (UNKNOWN): what
14:02Unidentified speaker (speaker_1): we've noticed known each other for a long time. As you just said, 12 years, very similar ethos, data driven, transactions cost, monitoring, AI, all of the different investments that we do. And it just happens to be one, we'll call it more systematic macro oriented versus, as you said, bottom-up security selection or systematic micro. So I think the genesis of the idea is that we have this very large quant effort. Maybe there's some synergies in focusing, particularly. how we're thinking about maybe mid-frequency equity efforts. Can we be better organized and rational and supportive of how we think bring in ingest data into the organization? And also thinking about next generation quant platforms for just research. Those are things that are very common. And to make sure I think bringing it together and the culture is very similar, they that just made a lot of sense. Now clearly we've got a little bit of a Boston, London and where people are located geographically. but I think we all learned coming out of COVID how to do that remotely as well and interact. So I think it's at a good spot. I think it required us the journey that we've gone on. This really kicked off, frankly, probably five, six, seven, eight years ago when we started to consolidate the tech platform. Historically, each of the engines had its own separate technology support and i focus. And so now having that centralized allows you to have a bit on the platform. I think the key for me, sometimes you get two. too big and you lose some of the efficiencies, but I think we're in a good economies of scale area right now because again, the research efforts and where they're focused, that macro versus
15:22micro creates a focused area for each of the organizations at the same time while bringing them together into one.
15:27Unidentified speaker (speaker_2): I mentioned that man has always had this reputation of tech forward, tech leadership, using tech as a strategic advantage, actually well before this was popular or hip. We'll get to your AI a little bit later. but combining those tech platforms, I'm sure there were cost advantages, but is there a case study or an easy example where you've been able to extract or harvest alpha value out of combining that?
15:50Unidentified speaker (UNKNOWN): Well
15:50Unidentified speaker (speaker_1): I think a lot of it is in what I mentioned the kind of operational effectiveness on data onboarding, making sure that we're abreast of the latest technology developments, coming together on what constitutes good research practice. I think that's critical. It
16:03Unidentified speaker (speaker_3): allows
16:03Unidentified speaker (speaker_1): us to interact a bit more with our Oxford Man Institute and having a direct connection there with that group. So I think as we go forward, what's important is while you could say technology is a strategic advantage, I think what it is is it's actually a necessary but not sufficient condition for success. Just having technology, yes, that's table stakes in some way, but how do you actually use it in the organization? And so that comes back to culture, approach. the experience of both AHL and numeric organizations. We have a lot of deep history. Either we all lived it together, frankly, or our organizations did. So we've been through a few market cycles together. And I think that helps inform. So yes, there's the efficiencies, but it's also a lot about the culture and figuring how the technology fits our process rather than trying to figure out how our process fits technology and whether maybe that's AI, but 10 years ago, deep learning, all of these things that come up.
16:53Unidentified speaker (speaker_2): Since you mentioned it, tell us about the Oxford relationship and how that fits into this puzzle.
16:57Unidentified speaker (speaker_1): It's an academic Oxford Man Institute OMI that we support and help them recruit and add academics into the organization and often asking them questions that we have or trying to stay abreast of some of the latest research as well. We've also, I think, over the last couple of years worked even more closely to be a bit more formal in some of the research that we do deliver to clients. So often what might happen is we'll get a really good question from a client, we'll go off and examine it and now publish a piece on that. So that's a little bit more readily available. I think what we're trying to do with that, that's another, I guess I can call it service alpha, coming at things with additional and very specific questions that they might have. So that's been a good relationship.
17:36Unidentified speaker (speaker_2): So I mentioned this earlier that I would get back to. it. So again, from arm's length, Robin's been there three to four years. Three
17:43Unidentified speaker (speaker_3): or four
17:44Unidentified speaker (UNKNOWN): years,
17:44Unidentified speaker (speaker_3): yeah.
17:44Unidentified speaker (speaker_2): So I mentioned you've had a history of inquisitive strategy that preceded her, but it seems like since she's been there, Asteria, Varagon, Barden Hill, listeners, these are just a few of the acquisitions under her tenure that the business has fundamentally begun to migrate from this quant purity that numeric and AHL. have had for a long time to multi-strategy, to adding private markets. You've got private credit in there, maybe most notably mid market.
18:10Unidentified speaker (speaker_0): Is that a proactive insight? And if that's the case, what led to that
18:14Unidentified speaker (speaker_2): decision and how are you going about pursuing that?
18:17Unidentified speaker (UNKNOWN): Well,
18:17Unidentified speaker (speaker_1): I think some of those acquisitions actually started under our prior CEO, Luke Ellis. So again, that's sort of a continuation, I think, of a trend in terms of what kinds of acquisitions and things that we've looked at. I think the broad broader strategy is a bit more around diversification of the business itself. So we have, again, we need to support the tech platform that we've got, Salesforce, operations, all of the organizational things in the center. So I think from a business perspective, it's very good to have multiple sources of alpha. And then also leads to why we really like solutions type offerings for folks, because we've got these ingredients and then we can use those to leverage. together and build hopefully something that's what clients need in their particular portfolio situation. So I think it's less about acquisitions and growth, as I said earlier, but more about can you put together things, alpha sources that are beneficial to folks over different periods, different periods of time. And so I think that's the philosophy, which again is part of some of the branding why we talk about man systematic, man discretionary, not GLG, H AHL, L numeric and some of the brand brands. So I think that's important to sort of think about how we face off with the market.
19:19Unidentified speaker (UNKNOWN): And
19:19Unidentified speaker (speaker_2): when I talk to LPs around the world, you've just made mention to this, Greg. I hear this consolidation effort. They want less GPs and they are asking more of GPs. So this idea of service alpha, your example with the Oxford reports, the idea of sharing research, this non non-monetary monetary exchange of information, ideas, intellectual capital. You've launched this solutions business that you just made mention of. Is that part of the thinking with that? Is that you need to be way more than just a service or a strategy provider, I should say?
19:49Unidentified speaker (speaker_1): Yes, I think that's important. Obviously, we have standalone strategies. I think PMI firms do, obviously, but I think the key is we do provide a layer on top of that, which is how to put them together, listening to clients. It does increase the cycle of the sales. sale, typically to figure out what's out there and working with them directly. So it builds hopefully longer client relationships. And it's just I think it is true that people are trying to narrow down the number of partners that they work with across the globe. And so having a few different capabilities allows you to grow that. So I think that's been interesting. I mean we've been it's funny because we talk about multi-strats and the rise of multi-strategy investing, but we've been doing multiple strategies for many, many years. But that can be a combination of quant or discretionary, whatever it might be. So we've always been a big fan of putting multiple things together. It's efficient for clients. There's all kinds of benefits to it outside of just the capital loan capital S that people talk about today in terms of multi-strat.
20:41Unidentified speaker (speaker_2): Have you launched multi-strategy products yet? Or is this mostly through SMA structures that you're doing these combinations?
20:48Unidentified speaker (UNKNOWN): A
20:48Unidentified speaker (speaker_1): lot of it's it, either SMAs, fund of ones. We do have strategies that would be our best foot forward if somebody said, hey, if you gave main group a dollar, where would we allocate it? And that's part of one of the strategies that I run, having that as a focal point. But what often happens, it's a great standalone effort that we've got there, a strategy, but at the same time it might lead to other discussions. It also helps you describe the firm a little bit easier. This is a strategy that covers everything that we do and therefore it gives it anchor point for discussions.
21:15Unidentified speaker (speaker_2): This collision of private and public, this shift towards multi-strategy, you've just articulated, I have to imagine there's some cultural challenges in there. I say this somewhat tongue in cheek, but funny bed partners, you and I would have never been at the same lunch or the same cocktail party between Numeric and Boston Company. And that's not because we wouldn't get along, but we hang in different parties. We're wired differently, stereotyping a little bit. So when you combine private capital folks that have different rhythms, different processes, certainly different compensation structures with systematic quant types, how do you go about creating a culture that encourages and embraces and celebrates the good in both sides of those folks?
21:58Unidentified speaker (speaker_1): As you said, it can be a different position in the market, long-term buy and hold, all of those things, it could be quite different. I think everybody going back to that culture of focusing on alpha. So I think everybody, whether it's private or public, it's all around hopefully you're doing better than your competitors. You're in a place that's useful for clients. So I think that's very consistent across the firm. I think what we're finding is that it might be actually a good thing to blend a bit of the more liquid stuff with the privates, right? So I think liquidity's come to the forefront in the last couple of years. And so are there things that you can do for the client that mix and match a bit of the different liquidity profile? So the way I would view it is instead of we've always historically at least competed in one liquid horizon. And now we're just adding additional horizons that we can do investments with. And obviously that has liquidity implications. But I think it's something another way to compete. You can obviously add ad markets, expand, but you can also think about horizon. And that I think has been interesting for me. Gets everybody excited to learn new things. And you just have to make sure that each person, each side teaches the other about what they're doing. And I think that's also part of the fun part of my job.
22:59Unidentified speaker (speaker_2): Yeah, I'm sure. As you think about that stable of thoroughbreds that you've built out through acquisition, combination, organic growth over the years, is there anything still missing? I don't think you have real assets infra real estate, private equity of substance. I could be wrong, but are there ideas there to round out the categories?
23:23Unidentified speaker (UNKNOWN): Well,
23:24Unidentified speaker (speaker_1): I think people, and this is a little bit outside of my area of expertise at the firm, but I think people are there's always things that come to us and opportunities. But I think where we sit today, we've got a very, very competitive offering, a very broad offering. And I think that's that for me in trying to build an organization. organization, the benefits of acquisitions, you get your capabilities, but it also takes time to digest and put that within the broader organization. So although there's a plan on what the plans are, but I think, at least from my perspective, we've got a very good set of strategies and wouldn't necessarily think that we need to have some of those other things right now, but we'll see. It depends on the teams that are out there and some of the collaboration that they would offer.
23:59Unidentified speaker (speaker_2): I'm curious how the evolution of the hedge fund comes. conversation, diversifying strategies, is the nicer way we've described it in the last 10 years or so. But you know this better than anybody coming out of the GFC, we've been risk on beta market. Why not? I mean, even as a I ran non US equity funds, so I felt this pain too in a different context. Why not just own the SP 500 or frankly just seven stocks forever? What is the point of any diversification? I just came back from Japan and GPIF is thinking about re-engaging with hedge funds for the first time in 10 years. Cowpers and the US endowments, you know well, this started to become, again, a jaded asset class within the boardrooms. How have you seen, now that we've worked ourselves through the cycle post-COVID, Post COVID. how have you seen the conversations with LPs change on hedge funds?
24:50Unidentified speaker (UNKNOWN): It's
24:50Unidentified speaker (speaker_1): been a very big cycle over my career and the interest in hedge funds. And maybe it was direct hedge fund investments. People would maybe take the alpha and put it on top of a beta portable alpha. So there was a lot of conversation. I think what we've seen over the last at least five to ten years, I think hedge funds, particularly the last five years, have done well. It's hard to find a hedge fund asset class that's done significantly in a kind of negative fashion. Now, some of them have been more diverse. diversifying than others. So I think part of the conversation we have with prospects and clients is is there any sort of hidden beta in what you're offering? Beta could be equity beta. It could also be generic factor returns. Our prospects and clients are very focused on what is the actual diversification benefit that you're bringing to the platform, in addition to the good performance. right? Again, I think across broad classes. And that might not have been the case 10 years ago. I think there were. performance wasn't great. I think I talked about the broader industry. People did pick up on that equity beta component. A lot of your standalone equity hedge funds, probably 1.5 beta. So you're paying performance fees with that embedded in there. And I think at the same time, as hedge funds weren't doing very well, you had very low interest rates. There was a lot of the rise of private credit, private equity as an alternative. Those changed position. and you had a lot of demand. And you go to client conferences, cap intro conferences, whatever it might be. And there was just a lot of demand for privates,
26:10not as much for publics. And then you start to see interest rates rise a little bit. People are thinking about liquidity again. And then you take that demand for more liquidity coupled with good performance. That just leads to natural, I think, interest in hedge funds. And so I think I've seen that pick up. But I think in a different way. such that relative to maybe 10 or 15 years ago, where it is very focused on what are you offering? What's the diversification? Going with that, I think the multi-strategy approach, right? Having multiple things also helps for the reasons that we were talking about. But even then, I think a lot of the standalone single strategy hedge funds have found other ways of sourcing capital and putting things together. And so it's just been a very ripe active stock selection credit selection market, which, as you said, you know, over the last 10 or 15 years, maybe I should have just bought the S P five hundred and gone to the beach. But I think as we go forward and given the performance, that it does bode well for active strategies in the next several years.
27:05Unidentified speaker (speaker_2): Maybe let's test this different strategies do well at different times. The very essence of diversification and the old free lunch adage that we all know well. If, for example, the tech, particularly the AI trade, and I'm not suggesting that you believe this, I'm just playing out this as an illustration. If we see these Mag 7 struggle, if AI gets over its skis a little bit, what hedge fund strategies generally would you tell clients should perform better in a more diversified environment?
27:35Unidentified speaker (UNKNOWN): Well,
27:35Unidentified speaker (speaker_1): it's an interesting thing because you could argue, and we've seen a couple of different times over the last couple of years, where there will be these pressures on the alpha across multiple strategies at the same time. And often in certain cases, it can come down to well, everybody's got a little bit of AI exposure in their portfolio. It could be an energy player, it could be an infrastructure player, whatever it might be. And if you look under the hood, there might be a little bit of AI exposure. They're sort of betting on companies that have AI versus other ones or underperformers. And so I think that's been the hard part for clients. I think there can be, as you go through some of these cycles, the news, different hedge funds may have issues in the marketplace. They're maybe AI related, but it does create a little bit of a contagion effect. So I think for my side, I think you start from first principles, but kind of long-term correlations and how you think strategies strategy is put together, if it should fit. I think we also forget that it's not just about correlation. it's also about tail risk. What's going to happen in a really, really tough environment? Scarily, I think in a couple of years now, we'll be hiring people that weren't alive during the global financial crisis, let alone having invested in the global financial crisis. And so you look at the market, and if you look at five or 10 years, you have a very, I think, maybe biased view of what creates diversification. So I think thinking a bit more about tail risk, thinking a bit more about liquidity risk, these things that in some environments, in particular the last five years, you could have probably thrown a dart and made some money in hedge funds. That's how
28:55good the performance has been. But it's really that next stage. So my recommendation is again, just look longer. Don't look over the last five to 10 years at performance. Maybe hug back to longer hedge fund indexes to see what it looked like in 08. Maybe even if you get back to the internet bubble, that would also be helpful as well. And I think that's been part of our philosophy from an asset allocation or strategy allocation perspective. that you can't just ride the recent environment because this environment's nothing like maybe the long-term environment that one would expect. So I think the other parts, I did grow up on the bottom-up security selection side. So I have that systematic blending that with discretionary. I've seen that. I think discretionary held up a bit better in GFC than Quant when Quant's done really well recently. That's actually another source of diversification. So I again I'm obviously talking my own book here a little bit, but I have seen that. I think that's been important. There's been strategies like trend, things like that, that do create some convexity and portfolios. So that we have that. That's one of our obviously big components of a lot of our multi-strategies solutions type things. I think the question there is we haven't seen that convexity really work since 2022. So how do we think about that? And at the end of the day, there's no free lunch. There's nothing, there's not a magic tail hedging strategy that will make money or at least not lose money over long periods. And so then you kind of get organizational patience around these things. So I think the idea is really just to go at it, diversification, both quantitatively, but also just think about conceptually
30:14what the two things are trying to do or the three things or five things or ten things that you have in your portfolio.
30:18Unidentified speaker (UNKNOWN): You
30:19Unidentified speaker (speaker_2): take the man out of numeric, but you can't take numeric out of the man, is what I'm hearing. Yeah.
30:23Unidentified speaker (speaker_1): Yeah. I still kind of go back to my quant stuff.
30:26Unidentified speaker (speaker_2): But I think that's a great lesson of history. And I think you're absolutely right. I mean, the last 15 years, if you've come up into the business in the last 15 years, this is the only normal you know. It's not a judgment, it's just the reality. Low interest rates, lack of inflation, and ripping public markets. What's the point of everything else? I like to say that's not normal. That's just the last 15 years.
30:47Unidentified speaker (speaker_1): It's not normal. Exactly.
30:49Unidentified speaker (speaker_2): So interesting to see what's
30:50Unidentified speaker (UNKNOWN): going on.
30:50Unidentified speaker (speaker_2): what opportunities some of these diversifying strategies have. So I've mentioned a couple of times. I want to circle back to this tech native identity you've had forever. Maybe not back to the sugar trading days of 1783, but at least in the recent history, right? Data science, machine learning, you've mentioned that's always been who man was and how you've positioned yourself. I'm wondering if you could talk a little bit about, as far as you know, I mean you've been there a long time. where that came from and how you continue to fan that flame. But
31:18Unidentified speaker (speaker_0): maybe more recently, a
31:19Unidentified speaker (speaker_2): part B of the question
31:21Unidentified speaker (speaker_0): is
31:21Unidentified speaker (UNKNOWN): now
31:22Unidentified speaker (speaker_0): that
31:22Unidentified speaker (speaker_2): everybody's jumped into the same pool with you, is that competitive advantage harder to maintain, endure? Is that going away?
31:30Unidentified speaker (speaker_0): How do you
31:30Unidentified speaker (speaker_2): find a new
31:31Unidentified speaker (speaker_0): edge there?
31:32Unidentified speaker (speaker_1): It's definitely been something that's evolved even prior to the latest large language model craze or even deep learning, machine learning. back in the deeper history, where I think generally having more data is better. It helps either your discretionary or systematic manager. So I think you've seen over time at least some convergence, maybe on the back end of the investment process. And because there's so much data out there, you need to have some technology to help you not only onboard the data, but sort through it and provide that in a reasonable way, something that's usable by your portfolio management teams. Of course, that's a different interface. with your discretionary or systematic. So that's been, I think, just a trend that's been going for maybe even over decades now, where the world there's just more data. We need to take advantage of that, whether it's systematic or discretionary. And so I think having that at baseline is important. I think the key is as I sort of mentioned earlier that it's not necessarily the case. It's not a sufficient condition to have good tech. It's really how are you good as an organization? taking that technology and marrying it to your alpha generation philosophy and how that works? Or you get a lot of questions like, well, all of the software, all these machine learning tools, they're open source. Can anybody copy it? And the point is, open source has a lot of advantages. You get constant innovation and research that's done by the broader community. But at the same time, how do you take that and put it into your process? That's the added value. And how do you connect it and have the data? So I think a few things that will hold up and as we've gone through this
32:52next AI revolution is one, the data will still be super important. It's going to be really, really hard if you don't have a large data estate to start with. I think that could create issues. I think what it might do, there might be investment businesses out there that have sat on a lot of data for many years and haven't taken advantage of it. Maybe there's private equity, private credit. So I think that's going gonna to continue to be an important element. And I think the other part, and this is something that we're thinking about quite a bit. is that while the newer models, the research process can change and adapt and use AI to help make investment decisions, it doesn't come with the free lunch and you gotta be fairly careful with it. I think where I've observed the organization, so basic research process is I generate a bunch of hypotheses and that systematic, you probably have hundreds of those, and discretionary, there's a few of those. You then have to validate that hypothesis, hey, I want to trade on this idea or not. And the final part is I actually have to put it in the market. I have to either trade directly in the market or I have to perfectionize it in our code base and it goes out and trades in the market. And so what's been interesting is I think the latest brand of AI has really helped generate a lot more hypotheses. We're much faster at coding and we've used a lot of AI historically in trading, but that's also being helped. But it's this middle stage of validation where you have to make decisions. So I want to use this service. or not. So a couple limitations there. If I'm trying to trade, call it mid-frequency, a couple of months, three month holding periods, there's only so much historical data.
34:11There's only so many global financial crises in the data. So it's like you can sort of amplify all the hypotheses that you're generating, but it's really, really hard to scale the data. And again, there's techniques, the AI to help amplify data, and there's some pros and cons of that. But the broader point is you still have to provide knowledge. to the models. What are they going to look at? How do you want to contextualize that? And so a lot of the competitive advantage, I think, will be less about which large language model you use, but more about the concepts of harnesses. How do I take a researcher that the numeric ethos of AHL or main group on the discretionary side? How do I code that up in a way that's representative of our organic researchers to help our organic researchers benefit? So again, the A. AI piece is a commodity, but people are not. And so how do you connect those two things together, which is really critical? And I think even in that validation step, it's pretty clear, I think in the broader community, not just in investments, that some people overtrust AI when they shouldn't, and they undertrust it in other cases. And so you kind of have to work that out because ultimately, if the rate limiting, if you have this chemical reaction
35:11of ideas, validation, and then production, it's like the if one in the middle still is just as slow as it was before, that's going to be the rate of the whole chemical reaction, the rate limiting step. So I think organizations need to figure that out and work through it. And I think that's going to create another source of competitive advantage. I hope I think that's part of the culture.
35:29Unidentified speaker (speaker_2): I completely agree with you. So I'd love to talk through this a little bit more because I'm curious when you think of people as the validators, as the harnessers, perhaps a bottleneck, but in an important bottleneck to provide the balance of the value, the authenticity of what the models are spitting out. I was at a recent round table and one of the CEOs said to me something to the effect of our industry historically has been much too single-minded on skills training. We hire mathematicians, we hire computer scientists. Now, forgive me, that's probably most acute in the systematic. way, no offense to that. And I'm wondering, given what you've just said, do we need to think about other disciplines sitting around the systematic quants, evaluating the models? I think you were a biology major. So I'm talking to another liberal arts grad here. I was too. Has
36:18Unidentified speaker (speaker_0): our time finally come?
36:20Unidentified speaker (UNKNOWN): I
36:20Unidentified speaker (speaker_1): think it is. I think it's a great opportunity for people that can ask the right questions and can be creative. And what you're seeing a little bit because of that. I think you're finding our systematic folks can be a bit more discretionary in the sense of you're looking for people that can generate these hypotheses or just have creative ideas. And then even on the discretionary side, because the technical barriers now dropped, you can be a little bit more systematic on the discretionary side. You can actually test things. So I think there's a little bit of a convergence. It will take some time, but I think it'll there'll be a convergence. But then you also, as you said, you've opened up the talent pool quite significantly. It was a always hard to find somebody that was both creative and technically trained. So often you would hire the technically trained person and then a couple of years in, maybe you can figure out if they're creative or not. So that takes a little bit of time. And maybe now you can downplay a little bit the technical and really try to focus a bit more on the creative. And so I think that's important. I think it's a difficult culturally as a firm, do you still need to do a coding test to join us? These are some really interesting and people have different opinions. within the organization around that. But that gets right to the root of it. If this person can't code, do I want them in the systematic business? Maybe you do. And the discretionary side, maybe there's a bias against people that are just
37:27highly technically trained or have a PhD. They want more people with quote unquote real world experience, whatever that might be. And so I think culturally it's a bit of a mindset. I think it comes down to providing incentives and things like that for the organization to bring in people that are different. I think this goes way back to all kinds of questions on different skill sets with the organization. Back in the numeric days, when I first joined, we were probably a bit more, let's call it MBA oriented, like me, you know, MBA, CFA. Then we went a bit more technical PhD level. So there's a little bit of ebb and flow in that as well. But I think ultimately it just comes down to people that are really creative and are trying to push them. themselves and get better and learn new things.
38:05Unidentified speaker (speaker_2): Have you guys made any tangible changes to either where or how you recruit or how you develop, given the stereotypical past of what you described?
38:14Unidentified speaker (speaker_3): Yeah, I think we're
38:15Unidentified speaker (speaker_1): early days of that. I would say we're a bit more in the still what you would expect, kind of in the traditional hiring funnels. Now, one of the things that I get to do sitting in solutions and the central function as CIO is I can. push some of these boundaries a little bit. We can think about people in the center and who work on our team. Could they be a little bit more creative other things to focus on given some of the technical aspects and that barrier has come down? But we also, I think, have hired people that have the latest cutting edge knowledge. They know what's out there from best uses of the technology. And then part of my job is to connect them with the people that are investing the money. So I think that's been a lot of
38:50Unidentified speaker (UNKNOWN): things.
38:51Unidentified speaker (speaker_1): fun for me hanging out here as CIO and getting to see the whole estate.
38:55Unidentified speaker (speaker_2): The market goes through economic cycles, of course, and then it goes through super cycles, which we made mention of. If you've come to the industry in 2009 or later, you know one orthodoxy, you know one regime, as some people would say. I'm curious your thoughts on how you transition models when you think a regime is shifting. Some models are, you know, if the Schiller PE is this, or if the PE ratio is that, then assume this. Or if interest rates go up, this is what happens to inflation and international flows. We make these assumptions that are economics 101, but when that changes, do the models still apply? And I'm wondering in today's world, which starting to feel to me from a geopolitical and climate and breakdown of globalism. globalization, that things are moving into a new orthodoxy. So, how do you think about your models and their baseline assumptions in a transition period?
39:47Unidentified speaker (speaker_1): A lot of the assumptions that I had for decades are starting to, as you say, come up for question globalization versus deglobalization. I mean, even questions around what is the risk-free rate. You can think of some fundamental questions around government spending and other things that could really have a large impact. And then you throw in the AI stuff on top. So I think you have to be flexible. I think one of the things that I try and hopefully from an organizational perspective is we mentioned earlier, organizations forget people that have not lived through those experiences. But hopefully the generation that we have now can help train people to be adaptive, teach them out that other market environments. environment. There's always some monster in the future. There's always something there. that's going to be different and change. So I think that that's important. That's about organizations and trying to take that learning and pass it across from a cultural perspective, the organization. And also we just need to be quite honest with ourselves, in particular this mid-frequency space. Machine learning is not great for the financial markets. It's not necessarily like we're teaching it to read an x-ray. X-rays don't change. That's a broken arm. That's always going to be a broken arm. Whereas in our space, correlations change. the horizon that you fit your model over. So I think that's really, really important. Again, pros and cons of machine learning, artificial intelligence. I think we've done a lot of good work in the last five years or so around contextual modeling a bit. Does this market regime look like something else we've
41:07seen in the past? What worked back then? Maybe some more learned models that are less contextual mapping, but just understanding. the covariance and structures of data in the market, bringing in more macro components into that, into that overlay. I think that's important. So that's been a lot of where I think maybe traditional factor investing has gone is a bit more factor timing. I think that's part of that, having that contextual mapping. So that, okay, maybe it's a good time for momentum, maybe it's a good time for value. So that also something that clients, these factors, at least the generic forms are. readily available, very, very cheap for folks. So again, on that continuing of trying to be more and more competitive in a space that is getting more competitive, just making sure maybe factor timing is something that you could do and bring to the client, that contextual mapping. So that's really the next phase. But AI is not going to fix that. It's not like we have a magic model now that will tease that all out indirectly. Maybe if it's a high frequency trader, high frequency trading, you've got a lot of data and very short-term. windows and correlations don't change significantly over short-term windows. But if you're thinking about, like I mentioned, horizons, competing on longer horizons, all the way extreme to maybe private credit. But even in quant, we've got some strategies that are three to five year holding periods. That's a very different setup. I think probably more open to do some development and particularly in the public side. So that creates opportunities on that side as well. But I don't think there's a panest. panacea
42:27here from technology that will make it all easy.
42:30Unidentified speaker (speaker_2): I'm just curious, just as I listened to you speak about regime changes, is there a way to model predict geopolitical risk premia that, again,
42:39Unidentified speaker (speaker_0): for 40 years hasn't really been a thing
42:41Unidentified speaker (speaker_2): in models or markets? Do
42:43Unidentified speaker (speaker_0): you have to include something that informs instability?
42:47Unidentified speaker (speaker_3): Well,
42:48Unidentified speaker (speaker_1): yeah. I mean, first of all, we have to measure it. What is the level? of, let's say, deglobalization that we're currently in? I think that's helpful. So we've done some work there, just trying to figure out where we are in the cycle and thinking about when was peak globalization and where are we? What are some factors and features that we can look at to say where we sit today? I think we've gotten some interest for clients around that because they're all facing the same conundrum syndrome around the globalization versus deglobalization. So I think you got to measure it first. And I think the second part is what are you? do you do with it? That's always the hardest part. And talk about geopolitical disruptions, those are even further and far between than macroeconomic cycles. So I think that's going to rely, I'll chicken out and just say diversification. You don't really know. I think you have to make some assumptions over my career that cross-border valuations, if So I look at a US stock, it's really good to know what's going on in Japan or Europe or in emerging market countries when I think about it. And so maybe that thesis starts to break. Maybe if capital flows aren't moving across regions, that more regional focused strategies rather than global approach could be a way down the road. And so I think that's important. You just have to stay on your toes. And then from a regime and risk and thinking about just how you operate your business, where do your people work? They could work all over the globe, and that can affect that. How do does data flow across? I talk about our nice centralized tech platform. but what happens if one country gets chopped off or we can't send data there? These are things
44:07that I never would have thought about 10 years ago, 15 years ago, but now we're in that situation. So it's not just an investment question. It's an operational resilience question as well.
44:16Unidentified speaker (speaker_2): Indeed. Lots of downstream effects. One of the second order boogeyman of the whole AI thing could be the org chart. Just finishing off this AI discussion, Greg. Craig.
44:25Unidentified speaker (UNKNOWN): Again,
44:25Unidentified speaker (speaker_0): you've always had models. and you've always had algorithms and you've always had these automated flows given who man is and has always been. But it's a whole nother thing when the org chart, those boxes, you can all visualize as you're listening to me talk. Those boxes start to be populated by bots. How do we adapt from a cultural standpoint point to being supervised by or having a colleague next to you that is actually an agent?
44:50Unidentified speaker (speaker_1): It's a very interesting concept that the agents is not people, but does HR need to cover them? You have that talent development programs for your agents. So there's all kinds of really, really interesting questions that are there. I think it's easier to think about agentic AI within your particular engine, like whatever you're doing, it's silo, it could be operations or investments and how does that improve my workflow? The harder part, which is what we're spending some time on is how do we connect across the engines? If we're really interested in selling solutions and other things, then we need to coordinate with sales and how do we think about connecting those? And there's also the other part where our suppliers, for example, are now building agents. And so do we need our agents to interact with those agents? And then even on the client side, interacting with our clients, we haven't seen huge demand yet that they have their own agents. But I think at least having an easier way to interact with man group is super important. So it's not just an investment question. It's a lot about operations. It's the whole value chain. And again, and the hardest part is really cross-connection. Everybody has seen the data. I think everybody's using AI in some sense. We've got really big uptake. We've got interfaces that are either more user friendly, other ones that are deeper tech. And again, we've seen even in the deech deeper tech stuff a lot of adoption there. So now you've got this technology you've got this capability. What do you do with it? So we have some discrete projects around it, encourage the use. We'll see how the
46:10ROI ultimately plays out, but I don't want to be too kind of goes up on my systematic research side a little bit. You don't want to be too prescriptive up front. I think you need to let people use it and see what we can discover with it. But it is early days. We're definitely viewing this as a people enabler rather than a replacement. of. And I think there's a lot of evidence over the years in different fields. I quote, I think there are more ATM, there are more bank tellers today than there were before the advent of the ATM. And I think part of it is because now a person can now do more things rather than having fewer people, you have the same number of people or more people, but you could do more things with that. So I think that's the attitude that you've got to take today. The market, going back to your recruiting question, I think there's a lot of interesting parts on digesting. We have new people coming up through the organization. How do they fit the sort of one to two years out of school? You've got the mid-career folks, and you've got the more senior executives as well. And so are we making sure that we're handling the right AI transition across all of those layers rather than just focused on new employees and things like that? So I think it's an important strategy across different types of people at and different parts in their career. structure at the organization.
47:16Unidentified speaker (UNKNOWN): Greg,
47:17Unidentified speaker (speaker_2): for my own benefit, I'm always curious how leaders are using AI personally. So how are you actually in this bigger role using it day to day?
47:25Unidentified speaker (speaker_1): It's been a a lot of fun, let's just say up front. I think you can ask questions, you can interact. I find it often easier just to maybe dictate my questions rather than trying to type it all out and trying to get the context. So using it to evaluate maybe some opportunities opportunistic trades that we get presented, maybe testing store thinking about the AI problem, maybe from an academic perspective. I'm working on something right now. I don't know if you've heard of P Doom, this number that people they talk about it in Silicon Valley. It's the probability that we're all basically die from AI. And so people like to quote these numbers, you say, oh, it's a 25% chance that people will die from AI or something. And then you're like, what does that mean? So it's called P Doom. I didn't know that until a couple of months ago. And so then I'm like, well, if that's the number and we hear that, and you're at an endowment and people are quoting these numbers, what do you do differently? So I've been exploring that question from a pure investment perspective. Should I change anything if this P Doom thing is lingering out there? So those have been fun, more white paper type questions. I've also used it to think about organizationally where things that we can do better and perform and think about are we using AI more. effectively? And what is that said? I think one of the other things that's been really important to me is having best practices. So a lot of things that I learn, building my own research stack, having multiple agents review my work. Let's have that in a central function so people can use that same tool. Maybe I'm biased because I like my tool that I've built these things, but
48:44having agents, whether they're econometricians or sophisticated theoretion fund managers or hedge fund managers. and just through narrators and all these things that I would normally do and would take me a few weeks. I can now do very quickly, but it's in a way that I like to do it and organizing these agents. So I think generally taking these skills and making them available, things that I find are effective, I think is part of my job as well.
49:06Unidentified speaker (speaker_2): Well, I always close with the same question, Greg. So this show, the name is a combination of two of my favorite things, which is the markets and great wine. So I like to always ask if it's a big work celebration. So maybe you had a great year, or maybe you're celebrating a birthday, anniversary. This is before P do, because all of it won't matter after that. But until then, do you have a go to bottle of wine or other drink?
49:30Unidentified speaker (speaker_1): I'm more of a bourbon man, I'd have to say. We have a lot of people very, very interested in wine across the broader organization, people that will do wine future. features and things like that. But my go to is just at the end of the day, if it's been a good alpha day or maybe a bad day, I don't know. Just have a nice bourbon not by the fire. We don't really have fires anymore. But definitely bourbon is my go to.
49:49Unidentified speaker (speaker_2): Somebody told me that you should drink your best bottle of bourbon or a bottle of wine on a bad day, not on a good day. The day is already good. Why waste a good bottle of wine?
49:57Unidentified speaker (speaker_1): Exactly. A little
49:58Unidentified speaker (speaker_2): bit counterintuitive, but maybe we can have a glass of bourbon one day together. Greg, it's been a real pleasure talking through the journey and the coming of age of man and your role. And it's been a lot of fun. Thanks so much for joining us in Capital Decanton.
50:10Unidentified speaker (speaker_1): Really appreciate it, John. Thank
50:12Unidentified speaker (UNKNOWN): you.