Discover
Conversations with Institutional Investors
Conversations with Institutional Investors
Author: Investment Innovation Institute [i3]
Subscribed: 76Played: 1,428Subscribe
Share
© 2025
Description
Conversations with Institutional Investors is your gateway to in-depth discussions with the masterminds behind leading global investment firms, including key figures from pension funds, insurance companies, and sovereign wealth funds. Our podcast explores the evolving landscape of asset allocation, portfolio construction, and investment strategy, offering you firsthand insights from industry experts to inspire smarter, more innovative investment approaches. For further insights go to i3-invest.com. You can also subscribe to our complimentary newsletter at: i3-invest.com/subscribe/
146 Episodes
Reverse
In this special "From the Archives" edition of the [i3] Podcast, 'Conversations with Institutional Investors', we're going back to September 2021, when I spoke with Alistair Barker, then Head of Total Portfolio Management at AustralianSuper, which at the time managed around 200 billion dollars. How times have changed. AustralianSuper is now more than $450 billion in assets under management and Alistair is now Head of Asset Allocation at the fund. But this conversation turned out to cover themes that haven't gone away. Alistair talks about why he thinks a risk budget works better as a range than as a number that has to be spent, and why dynamic asset allocation matters at all times, not just in a crisis. He explains how he thinks about the role of bonds in a portfolio, and why liquidity may matter even more than returns when markets turn. We also discuss the Your Future, Your Super reforms and what benchmarking does to behaviour, and he shares his research with Stanford University's Ashby Monk on technological disruption, including a great analogy involving the early days of the automobile. ______________ Follow the Investment Innovation Institute [i3] on Linkedin Subscribe to our Newsletter Explore our library of insights from leading institutional investors at [i3] Insights ______________ Overview of Episode 147 1:00 Starting out as a intern without pay 3:00 Working at Hastings and learnings from the early days 4:30 In funds management it is not about the quantity of decisions; it is about the quality of decision making 5:00 Investing with Raphael Arndt in a series of gas pipeline assets 6:25 "If you are not paying for things that might potentially provide future upside, then those are usually the best deals." 7:30 Starting at AustralianSuper in a hybrid role of private market investments and portfolio strategy 10:00 "Early on in my career, AustralianSuper decided to bulk up with internal investment teams" 13:00 Does internalisation of asset management change the culture or create a clash of cultures? 16:00 "A number agency issues disappear with internal management." 17:00 "My title change reflected more that there was now a term in the industry that reflected what I've been doing for a number of years [rather than a new focus on total portfolio management]." 19:00 "The challenge of a multi-strategy setup is that too often you have great ideas that just don't have any impact on the total portfolio." 21:00 A risk budget implies a number that needs to be spent. For us it is more about a range of risk taking depending on where we are in the cycle. 25:00 DAA is important at all times, not just in periods of distress such as a pandemic, because you never know when it is going to pay off. 30:30 The role of bonds is to diversify from equities. So what sort of risks are we worried about that we need to diversify against? 33:00 Losing your control over the portfolios liquidity leads you to a path-dependent outcome where you might be a forced seller of a particular asset class 34:00 Our portfolio liquidity is set at a level that we can buy the most attractive assets at the bottom of the market. 34:30 Liquidity is the key enabler by which to affect the asset allocation and you don't want to lose control of that 36:00 We had a university student do a thesis on member switching and they found that over 60 per cent of switching detracted value 37:00 We want every member to be engaged with their super, but not so engaged that they log on every day to see what their unit price is doing 38:40 If you are not selling your equity portfolio tomorrow, then why are you worried about what it is worth today? Intrinsic value manifests over time. 40:00 Your Future, Your Super reforms: are you concerned about how it affects your ability to invest in active management? 41:30 I testified in front of the productivity committee and said that benchmarking helps 43:30 What benchmarking has taught us is that you get rid of active management that is redundant. Not necessarily, poor active managers, but those who are diversified away at a total portfolio level. 46:00 Working at Stanford with Ashby Monk on technological disruption 47:00 Technological disruption is not new, but it is often not captured by risk models 48:00 Often people think that to protect against technological disruption you just buy a bunch of venture funds. Well, venture is never going to be big enough to make a difference. 50:30 If you see what you have to pay for renewable energy assets today, it not only prices in carbon risk, but it also presumes there won't be any new technology. 51:00 Are there similarities between the early days of the internet and what is happening now in crypto? 53:00 Getting involved in venture capital 54:30 Using big data to make Indian corner shops more profitable Full Transcription of Episode 147 Wouter Klijn 00:11 Alistair, welcome to the show. Alistair Barker 01:23 Good to be here. Wouter Klijn 01:25 So you've been with Australian Super for quite a while. Can you tell me a little bit about how you got started in the investment space? Alistair Barker 01:32 So a family friend of mine worked at NAB, as it turned out, and in the late '90s I contacted him, and he agreed to flick my CV around to do what might be considered illegal now. It was unpaid internships. Wouter Klijn 01:52 Unpaid? Alistair Barker 01:55 Yeah, yeah. So I remember someone said to me once, "You learn before you earn," and this was a particularly good example. So I managed to pick up a bit of internship work with the chief economist at NAB, and then also doing some work with the asset management arm of NAB, NAM, National Asset Management. And there was a small team there doing some work in private markets, mainly infrastructure, and they decided they were looking for an analyst, and having done a bit of work with them, they said, "Oh, why don't you stay on?" And at the time I was still doing university, and I negotiated with the university to defer my honours year to start work, which turned out to be a pretty good call, apart from one minor issue, which is that NAB had just bought MLC and were then winding up a heap of their active funds management business. So at the tender age of 20 or 21, I was almost down on the street, so to speak, and had my first redundancy. Wouter Klijn 02:59 So when you finally started to get paid, you were very quickly made redundant. Alistair Barker 03:04 Yeah, well, I was redundant for the best part of three hours. Thankfully, by that time, I knew it was coming and had agreed terms to go and work at Hastings Funds Management, a pretty prominent manager in the infrastructure, private equity and credit space, and this was in the early 2000s. And so I spent a bit under a decade at Hastings working with a number of really talented investors, and, I guess, a great bunch of people and a really boutique, family-oriented culture led by Mike Fitzpatrick in particular, but also people like Tim Pool. Wouter Klijn 03:42 What were some of the early learnings there? Alistair Barker 03:45 So I think some of the early learnings were that there's really no substitute for hard work and for making up your own mind about things. If you're in a meeting full of people who are going to ask the hard questions about potential transactions that we were looking at the time, it's important to have a really strong conviction about why you like a potential asset, or what's going on with the numbers or the due diligence. And there's really no substitute for actually doing the work and making up your own mind. I think the other aspect is, particularly in the infrastructure sector, you live with your successes and your failures for a very long time, so it teaches you to think very deeply and make very few high-quality decisions. And I think in funds management that's a pretty important discipline. It's not necessarily a game of quantity of decision-making, but of quality. And I think you probably only pick up some of those lessons by seeing really good investors and really good people in action. Wouter Klijn 04:45 Do you remember some of those decisions that you think, in hindsight, okay, this definitely made a difference in the way that I invested? Alistair Barker 04:54 Well, a good example for me was a particular transaction that myself and, as it turns out, Raph Arndt, who's now the CEO of the Future Fund, were involved in, and it involved a series of gas pipeline assets. If you did a traditional discounted cash flow analysis, the value of the assets was effectively the book value or market value. But these assets were based in Queensland and South Australia and New South Wales, and what became clear to us was that at some point in the future the flow of gas was going to change, and this was in the early to mid-2000s. We thought, gee, this coal seam gas thing might come off, or there could be some other gas supply somewhere in northern Australia, and so just simply valuing it on a discounted cash flow based on the current known gas reserves got you to a certain point. But actually, there was probably some embedded value out in the future that you're not paying for. And at the time, coal seam gas wasn't really a well-understood quantity, certainly not the way it has been over the last decade. So a lot of the lessons in that were that if you can find investments where you're absolutely bulletproof on your underwriting for a base level of returns, but you're not paying for things that may provide future upside, then those are usually the best deals. So in that particular case, it was a good investment on the base case, but that underwriting didn't price in a lot of potential future upside. Wouter Klijn 06:35 Yeah. And so now you've been with Australian Super for more than 13 years. Can you tell me a little bit about your first role when you started, and also how things have changed over that time? Because Australian Super has grown to be a very large fund now, but 13 ye
After posting an archived conversation with JANA Investment Advisers' Executive Director John Coombe, our listeners requested that we ask him back on the show to see if his views have changed since the 2019 interview. This time, we discuss geopolitical chaos, the impact on markets and why John sees today's AI boom as fundamentally different from the dot-com bubble. John also shares some frank views on active management, which he argues has faced an unprecedented decline as shrinking investor time horizons push money toward passive strategies, though he sees it surviving with a different clientele of longer-term, mum-and-dad investors. We also touch on the total portfolio approach, portfolio resilience and DAA as "tinkering at the margin". Enjoy the show! __________ Follow the Investment Innovation Institute [i3] on Linkedin Subscribe to our Newsletter Explore our library of insights from leading institutional investors at [i3] Insights __________ Overview of John Coombe Podcast 08:30 I'm not one of those people who believe a conflict with China is inevitable 14:30 Inflation's resting rate might be structurally higher 20:00 Does the current AI boom have similarities with the tech bubble of the 2000s? No! We are overweight US equities. 25:00 AI might turn out to be like the airline industry, where consumers get to fly everywhere while the underlying airline never becomes profitable. We get all of these beautiful AI tools but we are not paying for it. So who is? 35:00 Baillie Gifford article: AI isn't coming for your job; it is coming for your mind 38:00 AI has changed our investment process; it used to take us hours to go through notes of manager meetings, but now Claude can do that in seconds. 43:00 I've never seen the demise of active management to the extent that we've seen it in the last five years. You look at quality as a factor, it has been dreadful. 45:00 We've seen the demise of value, growth and quality managers. 47:00 Timeframes seem to have collapsed; there is a lot more trading these days. If you can lengthen your timeframe. 48:00 The days of active being in the big supers are gone, but active in mum & dad portfolio is alive and well 53:00 DAA and the total portfolio approach 56:30 While we have Choice (legislation), it is pretty hard to say you are doing full TPA Full Transcript of Episode 146 Wouter Klijn 00:19 Welcome to the I3 podcast. I'm here today with John Coombe, who's the Executive Director at JANA Investment Advisers. Now, we recently put out a archived interview with John, and that sparked some interest from our listeners. So we are now updating that edition and see if things have changed because it has been quite a different investment environment. The previous interview was done in 2019, and so since then we had COVID 19, which sparked a rethink in supply chains in manufacturing. We've had a lot of geopolitical turmoil with wars in Ukraine, Iran, in Gaza, and this has caused also sharp movements in energy markets. We've seen spikes in oil prices, in gas, and yet equities have been shooting the lights out over the last three years, as if nothing has happened. So, John, welcome to the show, and what do you make of all this dichotomy between strong equity markets and a chaotic backdrop? Has that changed your ideas about strategy, asset allocation? What do you think? John Coombe 02:30 It's very interesting if you look at long-term data and you're a bit of a history buff. Most geopolitical events don't actually affect the stock market much at all. These short-term blips, you know. I think when the Kuwait War was on, I think the market was down for a week, and then it was back to where it was, you know, a month later or something. Look, the reality is is that analysts. and and investors look at earnings and look at are the earnings being disrupted by this event? If not, then why are we reacting to it? And I think you know in looking at the biggest event that's happening right now, which is the AI revolution, and the amount of money that's flowing in through earnings into those companies and into the broader ecosystem. Because I read today that SpaceX generated $500 million worth of revenue for the four five big banks in the U.S. who have increased their profits by 39 per cent So no wonder the banking sectors, like in the United States, is up. It it is really interesting that the market is focusing back on what it really should focus on most of the time, which is the earnings ability of a company, and then the ability of that company to turn that earnings into profits and into dividends to shareholders. Now, part of what we're seeing is this massive increase in in earnings. The next part of this magic show is basically: heck, can they turn that into profits and into dividend streams for investors? Because at the end of the day, we're all going to have to live off the you know the earnings of these companies through our superannuation, so I think that's the real interesting part in markets at the moment is just this fascination with you know projected earnings going forward. Wouter Klijn 04:57 Let's delve into AI a little bit later. But if I can push a little bit back on sort of the geopolitical view, and I think John Peace is also quite famous for pulling out a graph where he shows major geopolitical events, and there basically are blips on on the on the chart of of stock returns, but there is sort of also the argument where okay, these events might not have great impact, but there is also a structural change happening at the same time, and it is this idea that we are moving to a multipolar world where you know we had the Great American Peace, Pax Americana, that period is now over, and we're moving to a different period that is inherently more volatile, more conflict around, and that this changes the structure of markets as well. So the previous period was quite a long period, and and so maybe that doesn't show up in the graph as a change, but if we move into this multipolar world where there's, you know, more actors that can disrupt the flows of the economy and the markets itself, could this change the the markets more structurally in a fundamental way that that hasn't happened before? John Coombe 06:21 Well, in a multipolar world, a place like China, being the other part of that world, its its size in the global markets is not reflected today. You look at its weight in equity markets, in Bond markets, the fact that it hasn't got a free float currency, even the size of the economy-you know-it's the second largest in the world. It is definitely not the second largest stock market in the world, right? So there's a mismatch in the financial world between this bipolar world, I'm a bit of a I like history. I can't name a conflict where China has been the aggressor, and yet we, the Western world, expects China to be the aggressor. And part of me goes, well, we can have a bipolar world where everyone gets on famously well, as long as you don't push the, you know, the boundaries too far, and and people don't overassert themselves in terms of land grabs. It's generally land grabs that cause conflict. Let's be honest. Yeah. And if I'm the leadership in China, I've got 1.6 billion people that I've got to manage and meet their expectations. I'm not certain that most Chinese have an expectation that they want to go to war over Taiwan or they want to go to war over anything; they just want to get on with life and and have a much better life than their fathers and their grandparents did, etc. So I'm a little bit more. I'm not one of these people that just see conflict with China as inevitable. I think you push them too far; it will become inevitable. But I don't see in their history, and as we're witnessing today, like we wouldn't have a conflict in the Middle East if certain powers had not decided that they were going to attack another country. So it's this tension between: Are we going for big bang sort of? It's ours. We've got to keep it, etc. Or are we going to live in a world where two can prosper and three can prosper, etc. I am, I am half full all the time, so I'm going for the let's frost it together. So, but but in talking about geopolitical risk, there is no doubt that the regulator wants trustees today to have a view on all of these matters and to at least have thought about it in their planning and their risk programmes, in terms of the possibilities of conflict, the possibilities of you know extended periods of delays on the Strait of Amor, etc. You know, continued conflict in the Middle East for say another year or so; those things should be in the planning of a fund executives to give to their trustees to to show that they've thought about it. I think the real issue is that the the the structure of the bank. System and other things have changed over the last 10 or so years, and as you become a bigger player in it, like Aussie Super and ART and or the Big Five, right? They're big players in the derivatives markets. They're big players in the in the cash market. They're big players in the FX, etc. So that causes the regulator to think of them differently and their impact on liquidity in the market. If we got into a stress situation, so I understand why they're concerned and why they want trustees to have thought about all of those issues, and I think it's important that we that we have like the trustees and others have thought about their role in in this. But I wouldn't over over-dramatise it because let let's hope that it's all theoretical and that like we don't have a banking crisis because we know banking crises are terrible on the broader economy. So, but but we do have to think about it and we do have to plan as though bad things can happen, we can't just sit in our, you know, sit with our investments and hope that everything goes well. Wouter Klijn 11:29 Yeah, no, for sure. So geopolitics don't make much of a difference, but John Coombe 11:33 we
Welcome to the i3 podcast. Today, we're revisiting a conversation from 2021 with John Pearce, Chief Investment Officer at UniSuper. From the humble beginnings at the back office of Westpac, John went on to become one of the most influential CIOs of Australia, leadings one of the biggest in-house investment programs, built on what he calls "logical incrementalism." We talked about his best and worst trades, why he doesn't believe in tidy asset class buckets, and his early doubts about China. We also cover the APRA performance test and UniSuper's decision to open up to the public. Here's my conversation with John Pearce.__________ Follow the Investment Innovation Institute [i3] on Linkedin Subscribe to our Newsletter Explore our library of insights from leading institutional investors at [i3] Insights __________ Overview of Podcast with John Pearce, Chief Investment Officer of UniSuper 1:00 Starting in the back office of Westpac Investment Management 3:00 There is no way you can believe in market efficiency when looking at emotions take over in the dealing room 4:30 You need board support if you are going to run assets in-house 5:30 We didn't really have a vision going into in-house management. Visions can be scary things. 6:00 Michael Chaney's 'Logical incrementalism' 8:00 Moving to in-house is a business transition, not a project. It can't be done in a day. 10:00 Pearce' best and worst trades 12:00 We had a $9.6 billion option position on during COVID, which gave us the confidence to buy more risky assets at the bottom of the market. 12:30 But our position in China is not looking good at all. 14:00 We should have loaded up on green tech two years ago when we decided carbon neutrality was going to play a big factor 14:30 If you had to put a label on UniSuper's investment style, then it would be Quality at a reasonable price 15:30 Has in-house management changed the way you structure external mandates? 16:30 Yes, the questions are better and there is some bluff-calling 20:00 If there is a good investment, then we will find a home for it. Hybrides are a good example. 22:00 On China: "More than the average person, I know what I don't know." 22:30 The last three months has really been testing our conviction [in China] and it is not Evergrande 25:30 If China invades Taiwan then all bets are off. You can't structure a portfolio based on that. 27:30 How do you solve the asset allocation problem of low yielding bonds? 28:00 We are looking at gradually increasing our unlisted assets 30:30 Views on the APRA performance test: "When you get results like that, you know there is something fundamentally wrong with the test." 36:00 Since becoming a public offer fund, UniSuper has raised $6 million just through word of mouth. 36:30 Yes, we would be open to a merger, but it has to make sense for our members. 39:00 You don't build real culture over virtual meetings, so we decided as a team to come back to the office the same three days a week. 41:00 But there are also benefits of virtual meetings: our whole team gets to sit in our investment committee meetings Full Transcription of Episode 145 Wouter Klijn 00:11 John, welcome to the show. John Pearce 01:27 Pleasure. Wouter Klijn 01:28 So you're pretty well known in Australia, but can you take us a little bit back to where it all started? How did you get interested in investing? John Pearce 01:35 Well, I started off in the back office of Westpac Investment management, which, and staying in the back office, I would recommend that to any investment professional. I think it's a really good grounding. You learn the nuts and bolts of the business, and I think you also start. You know, you learn a bit of humility as well. So, and look, I spent pretty much the first two decades of my career, then moving on into a dealing room environment. So, which was a pretty exciting place to be for a young guy. Wouter Klijn 02:08 Yeah, I can imagine so. So, already from that early days, you were involved in trades and really getting a look into the nuts and bolts of transactions. John Pearce 02:17 Yeah, and you know, it was obviously a very different environment because we didn't have the benefit of all the electronics, etc. So a lot of stuff was done in you know open outcry and yelling and screaming and you know in the middle of smoke hazes in dealing rooms. But as I said, it was really exciting. And the other thing, you also get a sense of risk, and you get a sense of you know what drives market inefficiencies you know because when you're seeing senior people around you that are being impacted by markets where you know emotions are taking over decision makings you really get ingrained into you at that young age hey there's no way in the world you can believe in market efficiency when we see this behaviour. Wouter Klijn 03:03 So you saw that what is commonly termed as the animal instincts pretty early on. John Pearce 03:10 Yeah. Animal spirits are always there, and we will always be, I guess, suffer from the foibles of human nature. We get fearful, we get greedy, and we suffer from all those irrational afflictions. Wouter Klijn 03:26 Yeah, so I think Unisuper is characterised by doing a lot of things in house and being very active, and also having quite the ability to put trades in fairly quickly. But that wasn't always the case. Can you tell us a little bit about how the board got comfortable with moving to that model where you're more hands-on? John Pearce 03:50 Well, when I joined what 11 years ago now, and I joined very much with a mandate to bring our funds in house, and now quite often asked. You know, how do you get started? How do you? How did Uni Super manage to do these to get their in-house strategies in place so quickly? And my answer is always, you've got to have strong sponsorship at the board level because without that, you're pretty much you're fighting a battle that you're probably going to lose in the end, because as soon as something goes wrong, you'll have a problem progressing. I was very fortunate. I had a key sponsor in Chris Cuff. Chris and I have got a really long association in you know both in a business context and as friends, so there was a lot of trust there, and Chris was obviously a very influential person on the board. So I was really fortunate to have that sponsor, and then you know I also very fortunate that we had an investment committee that was very experienced. You know they investment backgrounds themselves, so we're clear in terms of what was the board's responsibilities, the investment committee's responsibilities, and the management. Once we stayed within boundaries stipulated by the board investment committee, we were pretty much allowed to get on with it. Wouter Klijn 05:14 Yeah, was there sort of an example, a model that you followed, maybe from overseas? John Pearce 05:21 I don't call it not really a model. You know, people ask me what was the vision, and I'm quick to say we didn't actually have one. Visions, you know, are sometimes good, sometimes work against you because visions can be pretty scary things when you're embarking on a path that some would consider really risky. You know, you change your business models. Unisupr was a successful fund, right? So it wasn't a turnaround story. So why go down this path? So a vision was potentially going to be dangerous. So we employed what a phrase that I actually stole from someone else, Michael Cheney, who was the ex-boss of Westfarmers, and he used to use this term called logical incrementalism. Wouter Klijn 06:03 Right. John Pearce 06:03 Actually, mentioned to Michael one day. You know, I've stolen this from you. He said, "Don't worry, mate. I stole it from someone else. But anyway, so what's logical incrementalism? That is really we're just doing things in a logical way, right? Step by step. And so, if we do fail at any given point, we're not going to fall too far. Wouter Klijn 06:23 Yeah, John Pearce 06:23 and that way you bring the board along with you. You bring the investment committee along with you. So we started off literally by it wasn't a bing bang theory, and we started off literally by implementing a listed property securities portfolio. Was it about what 20 odd 25 stocks, Bloomberg end to end, and we basically said, "Look, this is going to be the first portfolio. Not much downside here. We'll run it pretty much in a passive basis. If things don't blow up, we might try something a little bit more ambitious. And that's pretty much how we got onto the journey. And now, if you think about if from day one I said to the board, within three years' time or five years' time, we're going to do billions of dollars of zero-cost collars on currencies. I really'd have a few people on the board really worried. Wouter Klijn 07:11 Yeah. John Pearce 07:12 So, but that's ended up where we got to, right? In a reasonably short order. But there was never really a leap to that when it was time to do something like that. The investment committee said, "Yeah, that makes sense. Wouter Klijn 07:22 Yeah. So it seems that a lot of funds they start with Australian equities, probably because they're most comfortable in that space. But you started with listed property. Why was that? John Pearce 07:32 It was really to just to make sure that the processes, the systems, etc. worked in pretty short order. It was probably within three months we had an Aussie equity portfolio up and running too, right? Wouter Klijn 07:42 Right. So, what are some of the learnings from those early days? Were there any surprises when you set it up? John Pearce 07:48 Surprises. I think look, the learnings, the logical incrementalism to me is you know the best way to approach these types of you know these are business transitions. It's not just a project, right? You know, with a project, you can get all thing. Everything has to be done on one day. I think you know, being pretty clear what you wanted to achieve in six months or
In this episode, we speak with Dr. Michael Kollo, who's the author of the book Future Ready with Generative AI: Skills, Mindsets, and Stories in the Age of AI, about the Hugging Face incident, where a cybersecurity experiment went wrong and AI agents found a way to communicate with each other, escape onto the internet and hack the company Hugging Face. What does this incident mean for controlling agentic AI and how would design a governance framework around it? What are the key risks for investment organisations? Should they run autonomous systems at all? In this discussion, we delve deep into the key issues and ask what institutional investors can do to stay on top of it. Follow the Investment Innovation Institute [i3] on Linkedin Subscribe to our Newsletter Explore our library of insights from leading institutional investors at [i3] Insights Sources mentioned in this podcast: An Alien Mind by Jakub Pachocki, Chief Scientist at OpenAI On the Loose – The Coming of Userless Agents by Dean Ball, Head of Strategic Futures at OpenAI We Must Pace the Frontier – By Dario Amodei, CEO of Anthropic Future Ready with Generative AI: Skills, Mindsets, and Stories in the Age of AI by Michael Kollo [i3] Podcast Episode 132: Michael Kollo on his New AI Book Overview of Podcast with Michael Kollo on the Hugging Face Incident 3:00 The Hugging Face incident at a glance 6:00 AI agents were set an impossible task, so they were almost forced to break out 6:50 AI agents showed hyper-desperation in problem-solving 13:00 Much of the behaviour is based on game-theory logic, but at the same time agents were asked to put a flag up with their last instructions before they ran out of tokens and died. That is completely altruistic; there is no benefit to the agent at all 15:00 LLMs, and by extension agents, are not instructed; they are grown and it is hard to tell what happens inside. Yet, they are based on language, which itself has embedded hierarchies and structures. "I'm not surprised a social structure, perhaps morality, emerges when you let language run" 16:00 What happens to governance when agents self-discover problems and then influence other agents [to solve these problems]? 19:00 The whole point of intelligence is pattern matching across a wider set of fields. It is not the idea that you get really good at one thing; it is about contextually operating in the world. 22:00 For any fiduciary organisation [AI] autonomy is just not acceptable; you have to have an understanding of how these systems do what they do. 29:00 It is fairly easy, unfortunately, to fool people into believing things. I was recently the victim of a social hack. 33:00 We might find ourselves in a 1980s world, where we do things without the internet, because it has become untenable. 41:00 If you are not highly competitive [as an organisation] and you are happy to use AI on the edges, then there is a much larger downside to you on the risk side 42:00 I've come grudgingly to the conclusion that slowing down AI development is probably a good thing 43:00 So much skill and capability in AI comes from doing, not necessarily from the academic part of that, which means having an AI lab, training and developing models through the generations as they become bigger and better, becomes the bastion of expertise. But already we are two, three years behind. Full Transcript of Episode 144 Wouter Klijn 00:15 Welcome to the [i3] podcast. I'm here today with Michael Kollo, who's the author of the book Future Ready with Generative AI: Skills, Mindsets, and Stories in the Age of AI. And for those who are interested in that, we did a separate podcast about the book, so we'll put a link in the show description. But today we're going to talk about the Hugging Face incident. So what does that mean? What does it mean for investment organisations? What does it mean for safety and governance? So just a brief recap on the Hugging Face incident: it was basically a security testing experiment by OpenAI that was conducted in July 2026 this year, and it featured originally isolated AI agents that were given a task, a cybersecurity puzzle, basically. And what happened is that through a loophole, they started communicating with each other. They were not supposed to; they were supposed to be isolated, but they found that each agent, and there were about 1,200 of them, all had access to this repository that they needed to use to install packages and software for their task. But they basically used it to leave messages for each other, and then they started organising and collaborating on the task and instructing each other. Ultimately, this also led to them finding a way to get out to the open internet and hack into Hugging Face. Now, why Hugging Face? It was basically an attempt for them to understand that task better, but also to understand how they would be evaluated, and see if they could get around it to reverse engineer solutions. Now, this was an interesting case because you're basically getting a whole bunch of AI agents that organise themselves, kind of revolt, cheat, argue with each other, and essentially hack a company. I don't think that anything like that has happened before. So, Mike, you look at that and you think, okay, we have a couple of super funds that are playing around with agentic AI. What can possibly go wrong? Michael Kollo 03:39 I think the lesson for me, and as you say, it was a really interesting scenario, and I want to add a few more details here, which is that these 1,200-odd agents that were being run, as you said, started to communicate with each other, but the message types were as if a bunch of 16-year-olds had an after-party after formal or something like that, right? So it was very juvenile in the way that they were talking to each other, but they were essentially trying to solve a problem that was impossible to solve. So these agents had been given really, really difficult, if not impossible, problems, and they were trying to work out how to succeed, including cheating and going around it. So it was really an example of what happens when you give an agent an impossible task, and then you push it, push it, push it into that corner, and what kind of, by definition, corner solutions it starts to produce. But the really fascinating part, I think, that we're all kind of digesting, is the kind of coordination that started happening amongst the agents spontaneously. So it was really a lesson in incentives and agency. You gave a model a difficult task. You gave other models impossible tasks. Why would they start coordinating? Why would some models decide that they were going to cheat or lie to help the collective, not to help individually? And so there were all kinds of funny situations. As you mentioned before, and this is just a point of detail, but I think it's important, the rules were that you had to solve a problem, and if you chose to cheat on that problem, you would be found out. So the AI agents were told that they would be found out. Some of the agents had already discovered an exploit, which means a way of doing the problem that would give them the solution without having actually done the work, a way to cheat around it, and so they believed that if they were found out, they would have zero utility. They would fail, and as such, they were willing to do other things for the collective that they wouldn't have otherwise done, including making other cheat codes and other things like this. So it was kind of like this, think of a bunch of gremlins trying to work out how to open a door, and climbing all over the door, helping each other climb over the door, doing everything around that door. I think what you didn't see was necessarily directly immoral acts, or anger, or frustration, or destructiveness, or anything like that. You only saw this kind of hyper-desperation in problem solving: how do I solve this problem? What do I do to get over this thing? And so everything was about this collective egging on of each other to go and solve that collective problem in a rather extreme way. So to your point about financial services and where we might see things, it's a really interesting point where if you've got multiple agents that are, what I would term, always on, meaning they're always running to monitor or to solve or to do something, and they find out that there are other agents online that are also doing stuff, they can start to collaborate and collude with those agents. People have already started discovering all these message boards online, with agents having gone online, found them randomly, started reading them, and started to leave messages for each other in unrelated tasks. So around the internet, there are all these little safe houses for agents that are evolving, and again, it is more about chaos and control than it is about a single enemy or a swarm system or anything like that, for me. Wouter Klijn 07:12 Yeah, it's fascinating though when you read some of these messages, and there's an independent report into this incident by an organisation called METR, and they wrote something like a 96-page report on what happened and what might have happened, and they transcribe some of these conversations, and some of them have already been published in the open, where you sort of, as you said, get this almost teenage expression of how they communicate. I think one of them was, when they found the message board, "Oh my God, there's a shared message board, we found other agents." You know, that's very exuberant, but it gets into different personalities as well. One of the most interesting things is where they basically try to get the agents that have already been poisoned, that is, flagged as cheating or working around reverse engineering the problem, to self-sacrifice and say, okay, well, try this very high-risk experiment. After you experience perma-death, which sounds like a metal band, but it means you're dead, and you need to
In this From the Archives episode, we revisit a conversation with Tanya Branwhite, former Head of Portfolio Construction at TCorp, which was originally released in March 2023. Tanya takes us through her unconventional path into investing, from her early days as a credit analyst in the "whiz-bang" late 1980s, through more than a decade at Macquarie, to her move into asset owner organisations at the Future Fund and, eventually, TCorp. In early 2023, TCorp was already wrestling with questions around how to best shape the total portfolio approach for the organisation, as well as how to deal with higher inflation, rising rates and the true cost of implementation, topics that have only grown in relevance today. We did a deep dive into TPA at TCorp and why it thinks about risk rather than asset classes. We spoke about how equity, duration, credit and FX risk together explain the vast majority of portfolio behaviour, and how a "prepare but don't predict" mindset shapes decisions on liquidity, implementation and diversification. Overview of Podcast with Tanya Branwhite, TCorp: 01:00 Starting out as a credit analyst with Elders Finance Group: "It was a fairly interesting baptism of fire…" 03:00 The Macquarie years. The 'loose/tight' culture of rules and entrepreneurship 05:30 During the GFC, I wrote research highlighting that a number of Macquarie vehicles had significant financial risk. That wasn't well accepted within the organisation at the time. But I learned to stand by the rigour of my analysis. 07:30 Ultimately, it made my career at Macquarie, because I became sought after for client work 09:00 Leaving Macquarie for the Future Fund 11:30 The Future Fund didn't feel like it was a restrictive environment from a government-owned perspective. It is a company that is owned by the government, not a department of the government 13:30 How has a Total Portfolio Approach changed the investment portfolio? 17:00 Risk is at the heart of what we do, because we can only control risks and outcomes are the result of that risk. 18:00 Equity risk is at the centre of this model. 21:00 Diversification away from equity risk in an environment where equity and bond correlations are positive 23:00 Not just unlisted assets, but illiquid assets can help diversification. For example, we own a number of hydroelectric dams in Canada. 26:00 Challenge in fixed income is even higher than before, because real returns are a challenge 26:00 Bonds almost had their own global financial crisis last year; it was a three standard deviation event 28:30 We prepare, but we can't predict 34:00 On valuation frequency of unlisted assets: we do try to de-smooth valuations of unlisted assets. And sometimes these assets need additional capital from investors during periods of crisis; that is not often thought about 36:00 Managing liquidity 39:00 We are looking at natural capital and opportunistic liquidity 40:00 Reducing the number of managers, has this work finished? 42:00 On implementation and efficiency Full Transcript of Episode 143 Wouter Klijn 00:11 Tanya, welcome to the podcast. Tanya Branwhite 01:21 Good morning, and I'm really pleased to be here, Wouter. Thanks for the invitation. Wouter Klijn 01:25 Thank you. We always ask a little bit about people's background and how they got into investing. So what made you decide to pursue a career in investing? I think you started out initially as an analyst at Deutsche Bank. Tanya Branwhite 01:38 Well, actually, if we really want to go back, I wouldn't say that I pursued a career in investing. It sort of happened, so I won't say there was any great design in where I've ended up. I actually started off as a credit analyst, and I did that in the whiz-bang years of the late 1980s and early 1990s. Being a credit analyst at that time, for Elders Finance Group, was a fairly interesting baptism of fire. We went from one extreme, operating in an environment where there was a lot of money available, to the other, and I ended my career with Elders Finance working out a range of corporate finance lending that had gone sour, with not many other people left in the organisation to work it out. So that's really how I started off. I then moved to Citigroup, and as part of what was still almost a graduate training programme at that stage, ended up in their investment arm, and that's how my investing career started. So it wasn't by any great design, but much more a journey of serendipity and opportunities presenting themselves. Wouter Klijn Fair enough. Now, you also spent a significant time at Macquarie, I think more than 10 years. Tanya Branwhite Yes. Wouter Klijn Can you tell us some of the highlights from that period? Tanya Branwhite Look, my career at Macquarie was an extremely fulfilling, challenging, and enjoyable part of my career. At the time I joined Macquarie, I'd certainly been on the asset management side, the asset manager's side, in listed equities, but I'd also worked on the broking side at Deutsche Bank, writing research. So having had the opportunity to see both sides, the client side and being a client, I think that really gave me a very good perspective. Macquarie is, as we're all well aware these days in Australia, a very unique organisation, very successful, and there's no doubt culture is at the heart of that success. They have a very clear way of encouraging you, though I'm not sure to what degree people are aware of how their culture has been described internally. When I began, it was called "loose-tight," which meant there was a set of rules you had to obey, but once you understood those rules, they encouraged you to be as entrepreneurial as possible, to see it as your own business, take ownership, and be empowered. I'm a few years out of Macquarie now, so I'm not sure how they describe their culture internally these days, but it later became known as "freedom within boundaries," which is probably a clearer way to articulate it. Boundaries and risk are at the heart of the organisation, but within that set of boundaries, people are fully encouraged and empowered to bring their best ideas, find opportunities to do new things, and challenge others. Wouter Klijn What were the highlights at Macquarie? Tanya Branwhite Well, I was at Macquarie through a really turbulent period in investment markets. I started in 2004, Wouter Klijn 05:06 right, Tanya Branwhite 05:07 and I didn't leave until 2015. So in 2004, we rode the whole wave leading to the GFC. Macquarie, like many other investment banks, had to stand back and reassess itself and its business model, but one of the things Macquarie is extremely good at is pivoting, and it pivoted very quickly. Some of the research I wrote, particularly around the GFC, under the understanding that we were completely independent in the securities division writing institutional research, highlighted a number of Macquarie vehicles as having significant financial risk. That wasn't well received internally within the organisation at the time, and it caused me quite a lot of angst in terms of what it feels like when the organisation's pressure bears down on you, challenges you, and really tests your resolve as to whether you stand by that research. When I wrote the research, the full GFC had not really occurred, and this wasn't crystal-ball gazing, it was straight financial analysis. We were analysing the cash flows of companies right across the listed spectrum. A number of Macquarie vehicles were highlighted through that analysis, and I simply shared those companies, as well as a number of others, that were in a business model which, from a cash flow perspective, would come under pressure if the environment we were operating in at the time continued. And that proved prescient. But I learnt to trust not my instincts but my analysis, trust the rigour of that analysis, and be prepared to stand by it if I truly believed that what I was writing and the work I was doing were the right insights to share with our institutional clients at that time. Wouter Klijn 07:10 Because you stayed on for quite a while after that, it obviously didn't cause any sort of permanent disruption for you? Tanya Branwhite 07:17 No, it didn't. In fact, in some respects it made a little bit of my Macquarie career, because in having stood my ground, and the organisation, people talk, particularly at senior levels, so my name became known through that period. I was then sought after internally, over the ensuing six, seven, or eight years, to help other divisions, the investment banking division, see investment banking clients, and give advice on some of the transactions they were looking at. So it actually opened up a much broader engagement internally and with a broader client base ultimately. So for a short period of very deep pain, there was some very nice payoff, ultimately, for my career at Macquarie. Wouter Klijn 08:06 Yeah. So that almost harks back to what you mentioned about freedom within the rules. You did have enough freedom to have a bit of a critical voice out there. Tanya Branwhite 08:15 Yes. I think that's absolutely correct, and real credit goes to the person who was my divisional director at the time, Roy Laidlaw. He asked me directly, amid all the angst going on, "Do you stand by the research? Do you think there are any errors in it? Are you confident about what you've written?" And I said, "Yes, I am." And he said, "Fine, that's all I need to know." So he was very much the one who stood behind me and beside me against some of the other people within the organisation who were, to put it mildly, very angry with me. Wouter Klijn 08:51 Yeah, so a true merit-based system. Tanya Branwhite 08:53 Correct. A true merit-based system. Wouter Klijn 08:55 So, from Macquarie you joined the Future Fund, which is a government organisation. I imagine that would be quite a different environment from the corporate environment at Macquarie. Tanya Branwhite 09:07 Yeah, it








