The Exchanges

Q Can I ask you final one? When you think about your biggest mistake, a fund that you regret doing, when you review, don't say it, obviously, um, I know you won't, but, uh, when you review that decision, what did you not see that was the determinant of that fate?

A I think one of the challenges with investing in venture is that there are so many unknown unknowns, um, and, and the degree of randomness that leads to That, that is involved as to whether something is successful or not is, is high. And the earlier you get that, that the greater that degree is. So I think there are certainly things that, that, you know, we couldn't have been expected to predict at the time we were doing the deal. And one of the things we do with, with, with our investments is, is sort of four or five years post investment, we'll do a decision review analysis of, of those, um, to see what can we learn from that and how can we improve our decision? One of the big things that we, Learned was we used to only take references on managers, um, from VCs that they had invested alongside. So we wanted to know what were they like as a partner? What were they like on the board? What we didn't do was if somebody was operating in a, in a particular space and we knew one of our managers was, you know, was one of the top investors there and they hadn't done any deals with that manager. We just didn't follow up on that. But now we'll actually say, we'll phone that manager up and say, like, why haven't you done any deals with this group? Like, what, is it just that you're in different parts of the market, or are there specific reasons for not having done those deals? Um, and I t…

AI assessment note: “not just to reference people who we know work with each other”

Q Can I ask you final one? When you think about your biggest mistake, a fund that you regret doing, when you review, don't say it, obviously, um, I know you won't, but, uh, when you review that decision, what did you not see that was the determinant of that fate?

A I think one of the challenges with investing in venture is that there are so many unknown unknowns, um, and, and the degree of randomness that leads to That, that is involved as to whether something is successful or not is, is high. And the earlier you get that, that the greater that degree is. So I think there are certainly things that, that, you know, we couldn't have been expected to predict at the time we were doing the deal. And one of the things we do with, with, with our investments is, is sort of four or five years post investment, we'll do a decision review analysis of, of those, um, to see what can we learn from that and how can we improve our decision? One of the big things that we, Learned was we used to only take references on managers, um, from VCs that they had invested alongside. So we wanted to know what were they like as a partner? What were they like on the board? What we didn't do was if somebody was operating in a, in a particular space and we knew one of our managers was, you know, was one of the top investors there and they hadn't done any deals with that manager. We just didn't follow up on that. But now we'll actually say, we'll phone that manager up and say, like, why haven't you done any deals with this group? Like, what, is it just that you're in different parts of the market, or are there specific reasons for not having done those deals? Um, and I t…

AI assessment note: “we used to only take references on managers, um, from VCs that they had invested alongside.”

Q compare fund sizes today to exit sizes today, and actually We've had this realization that we were wrong in COVID and that company shouldn't be for the X revenues. They should be. And we're back to the normal now. And so it's really interesting to say, Hey, project yourself for 10 years, the exercise of 10 years time. What do you think the exercise of 10 years time is then?

A Yeah, I think so in the famous words of Yogi Berra, predictions are hard, especially those about the future. Um, so I think what you have to look at, we would sort of take a step back and say, is technology becoming more or less important? Is it capturing a smaller or larger share of the economic pie? And are the market sizes for the winners in technology getting smaller or bigger? We think all of those arrows are pointing upwards. And so, yes, we know the multiples that you will see on those individual companies. Earnings or revenues were going to fluctuate, but ultimately the markets that they're playing in, and the share of the economic pie that technology is going to capture, in our view, is only going to increase. And so that gives us confidence that whatever the, the multiples are at the time of exit, the, the directionally we're going to see exits get larger.

AI assessment note: “directionally we're going to see exits get larger.”

Q compare fund sizes today to exit sizes today, and actually We've had this realization that we were wrong in COVID and that company shouldn't be for the X revenues. They should be. And we're back to the normal now. And so it's really interesting to say, Hey, project yourself for 10 years, the exercise of 10 years time. What do you think the exercise of 10 years time is then?

A Yeah, I think so in the famous words of Yogi Berra, predictions are hard, especially those about the future. Um, so I think what you have to look at, we would sort of take a step back and say, is technology becoming more or less important? Is it capturing a smaller or larger share of the economic pie? And are the market sizes for the winners in technology getting smaller or bigger? We think all of those arrows are pointing upwards. And so, yes, we know the multiples that you will see on those individual companies. Earnings or revenues were going to fluctuate, but ultimately the markets that they're playing in, and the share of the economic pie that technology is going to capture, in our view, is only going to increase. And so that gives us confidence that whatever the, the multiples are at the time of exit, the, the directionally we're going to see exits get larger.

AI assessment note: “directionally we're going to see exits get larger.”

Q There are new strategies which haven't been done before. Uh, to the extent that they have like, uh, Sam Altman, who is obviously a fantastic CEO and also invests heavily on the side. There haven't been cases like him before. And so we don't have data sets to predict forward on or to judge against. How do we think about entirely new models, which may be better?

A Yep. And I think as an investor, it's a little bit, you know, for us, we don't feel we need to be Out there trying to test these new theories. What we need to try and do is to find when something is working to jump on it as quickly as we can, and then trying to, trying to get access in there where we have more signal and, and, and more conviction that, that it's not just a good story, a good narrative, but actually it's going to result in strong performance. And so when we do look, um, so, you know, we are looking at adding new managers in the market Today, because we think that signal to noise ratio looks a lot better, because we're now starting to see a lot of those companies that, that raise money in the height of the Zurb era, you know, begin to come back down to, to, to, to, to ground and, and, and get a sense of what they're, what they're really worth. And so I do think over the next couple of years, you will start to see managers differentiate themselves. But I think for us, the earliest we would intercept a manager would probably be fund three. And that's where we've had the most success historically. So if I look at our 12 managers, probably half of them we ended up doing it from three.

AI assessment note: “we don't feel we need to be Out there trying to test these new theories.”

Q had data network effects like we do today. We've never had, you know, Microsoft doing whatever it is, three hundred million in free cash flow. No. And so, yeah, very good point. And so I worry intensely that yes, the market size increases, the proportion of, like, you know, spend to tech increase, but it concentrates. Intensely, which is not good for us. Do you share my worry on that?

A I think, I think you're right. Yeah, that was, I think you, it, it's clear that the, the, the, the incumbents today have managed to have multiple, um, iterations of their products. Um, and you know, the question, the question I would have though is, is how long is that likely to continue? Um, and I, I remember back in the, in the mid nineties reading the innovators dilemma Clayton Christensen, and it was a real eye opener at the time just around how it's very difficult for, um, those incumbents to, to, to, to really innovate and to disrupt their own business and cannibalize their own business models. And it's interesting, you know, look at what's happening with someone like Google today. Look at the, Um, the reaction to, to their AI product that we've just seen over the last, the last couple of weeks, and all those people that are now saying, you know, Google needs to really address where they're going as a business. Um, you know, look at what Elon's done with Twitter in terms of, you know, turning that company on its head. I think there are a lot of challenges for the incumbents, and, and there's no guarantees that they're going to be able to continue to hold that dominant position, particularly as We enter different technology paradigms. So one of the, one of the things, I don't know if you've, if you've had a chance to read Chris Dixon's book yet.

AI assessment note: “I think, I think you're right. Yeah, that was, I think you, it, it's clear”

Q There are new strategies which haven't been done before. Uh, to the extent that they have like, uh, Sam Altman, who is obviously a fantastic CEO and also invests heavily on the side. There haven't been cases like him before. And so we don't have data sets to predict forward on or to judge against. How do we think about entirely new models, which may be better?

A Yep. And I think as an investor, it's a little bit, you know, for us, we don't feel we need to be Out there trying to test these new theories. What we need to try and do is to find when something is working to jump on it as quickly as we can, and then trying to, trying to get access in there where we have more signal and, and, and more conviction that, that it's not just a good story, a good narrative, but actually it's going to result in strong performance. And so when we do look, um, so, you know, we are looking at adding new managers in the market Today, because we think that signal to noise ratio looks a lot better, because we're now starting to see a lot of those companies that, that raise money in the height of the Zurb era, you know, begin to come back down to, to, to, to, to ground and, and, and get a sense of what they're, what they're really worth. And so I do think over the next couple of years, you will start to see managers differentiate themselves. But I think for us, the earliest we would intercept a manager would probably be fund three. And that's where we've had the most success historically. So if I look at our 12 managers, probably half of them we ended up doing it from three.

AI assessment note: “we don't feel we need to be Out there trying to test these new theories”

Q Do you care if your VCs add value? I know it sounds strange, but isn't?

A I think, I think, I think we care if our VCs, we want our VCs to understand when they need to get involved and when they need to get out of the way. Because there will be certain points in a company's life where they do need help. Nothing, no success happens in a straight line. There are, you know, you look at most of the, the companies out there that have, that have been successful. At some stage, they had a near-death experience. And I think the role of a founder is, It can be at times incredibly lonely, so I think there are times when a VC needs to be there for that founder, um, and needs to give them a hard, have a hard conversation with them, and needs to deliver a bit of tough love, but also needs to be supportive, um, and, and, and, and, and also, in a way, to be a psychological support for, for that founder. So, I think the best VCs are able to do that and can pick and choose their times. I think there are clearly VCs out there who just need to back off a whole heap.

AI assessment note: “we want our VCs to understand when they need to get involved”

Q How, sorry, I didn't mean to age you there, but how did you first become an LP, and when was that, I want to do this as a career?

A Yeah, funnily enough, I didn't grow up thinking, actually, my life's ambition is to become a, an LP in VC funds. Um, so I, I, I grew up in a small village in Northumberland. Um, I, I'm pretty sure that nobody in that village had ever heard of venture capital, and, and, and I hadn't heard of venture capital. Um, Um, but I, I sort of finished university, um, I actually was keen to kind of go on and, and, and do a PhD, um, but, but kind of life intervened at that time, and, and I had to end up getting some real work, and I was just looking for pretty much anything. Um, I saw an advert, so my, my girlfriend at the time, my wife now, was, was living in Oxford. I was still living with my parents up in Northumberland, and so I, when I came down to see her, um, Um, we were looking for, I was looking for a job. I saw an advert in the Oxford times, numerate graduates required for global finance firm. And I thought, well, it doesn't sound very interesting, but if I don't apply for it, she's going to see it and she'll kill me. That was, that was, um, and I was fortunate enough that, that, um, you know, my, my first boss was willing to take a punt on a, you know, a spotty, fresh graduate with no experience. You'd never heard of VC before.

AI assessment note: “I saw an advert in the Oxford times, numerate graduates required for global finance firm.”

Q know when things change in the firms that you're in? Like, I don't know the firms you're in and respect that, but I'm sure I could tell you some horror stories right now of, um, things that are happening in your firms because my friends are in them and I, Like, it's like founders know founders. Do you know what I mean? Do you think you're close enough to know?

A It's probably a fund later than the issues would start to emerge, but at the same time, I think you also, you probably hear a lot of, a lot of just general scuttlebutt about what's going on in there, and, and, you know, how much of that actually plays out. So I think, you know, you can, you can almost hear too much and not understand, or find it a challenge to really appreciate, you know, what's, what's gonna, what's material for the firm, and, and everyone likes to moan about where they work. That's never changed. Um, I think, I think for us, what we, what we tend to look for is, is to make sure that firms have that process where, you know, we're seeing a continual flow of new people coming in, and they're, Being valued for the work that they're doing and the senior partners stepping aside.

AI assessment note: “It's probably a fund later than the issues would start to emerge”

Q new strategy. I see, I hear a lot of endowment funds say, oh, we, we, we've got the Yale David Swenson model. And I'm like, he did that in the eighties when there were much fewer managers to select and venture had much better returns. That is not a comparable strategy today. Do you not worry that that is lagging data that you are now acting on in today's environment?

A Yeah, and I think, totally accept that, that one of the challenges of venture is that the feedback loop is so long. And, and, and to be honest, most LPs probably aren't going to be in the same job when, when that, when that feedback comes. So they're, they're more worried about deploying than they are about what happens 15 years down the line when, you know, when the performance data is, is actually in. But then I would kind of flip that around and say, let's go back to the first principles. So I totally accept that. Let's go back to the first principles. Let's go back to the underlying performance data and let's see what strategies, you know, can be successful. Um, and, and for us, it it's, you know, we think venture is such a power law industry.

AI assessment note: “totally accept that, that one of the challenges of venture is that the feedback loop”

Q new strategy. I see, I hear a lot of endowment funds say, oh, we, we, we've got the Yale David Swenson model. And I'm like, he did that in the eighties when there were much fewer managers to select and venture had much better returns. That is not a comparable strategy today. Do you not worry that that is lagging data that you are now acting on in today's environment?

A Yeah, and I think, totally accept that, that one of the challenges of venture is that the feedback loop is so long. And, and, and to be honest, most LPs probably aren't going to be in the same job when, when that, when that feedback comes. So they're, they're more worried about deploying than they are about what happens 15 years down the line when, you know, when the performance data is, is actually in. But then I would kind of flip that around and say, let's go back to the first principles. So I totally accept that. Let's go back to the first principles. Let's go back to the underlying performance data and let's see what strategies, you know, can be successful. Um, and, and for us, it it's, you know, we think venture is such a power law industry.

AI assessment note: “totally accept that, that one of the challenges of venture is that the feedback loop”

Q know when things change in the firms that you're in? Like, I don't know the firms you're in and respect that, but I'm sure I could tell you some horror stories right now of, um, things that are happening in your firms because my friends are in them and I, Like, it's like founders know founders. Do you know what I mean? Do you think you're close enough to know?

A It's probably a fund later than the issues would start to emerge, but at the same time, I think you also, you probably hear a lot of, a lot of just general scuttlebutt about what's going on in there, and, and, you know, how much of that actually plays out. So I think, you know, you can, you can almost hear too much and not understand, or find it a challenge to really appreciate, you know, what's, what's gonna, what's material for the firm, and, and everyone likes to moan about where they work. That's never changed. Um, I think, I think for us, what we, what we tend to look for is, is to make sure that firms have that process where, you know, we're seeing a continual flow of new people coming in, and they're, Being valued for the work that they're doing and the senior partners stepping aside.

AI assessment note: “find it a challenge to really appreciate, you know, what's, what's gonna, what's material”

Q noise and far more signal sounds wonderful. Uh, sadly, when you have far more signal, there's far more noise, and traditionally, I agree with you in terms of supply of capital in cycles, but there is a permanence to the wall of capital that has entered venture, I think, and you're seeing a new size of that with sovereigns and pensions doing what they're doing today. Is that possible anymore?

A I think, again, it comes down to understanding What, how venture works and going back to the first principles of the industry. And that's one of the things that, that we've always tried to do is to, is to really go deep on understanding what is it that makes a great fund? What is it that really drives our performance in the venture industry? And the thing that we constantly come back to is that venture is a power law industry, and it's one percent of the exits that ultimately generate the bulk of the returns created by the entire industry globally. So we're looking at, at around 30 companies a year that generate more than half of the total exit value for the VC industry. And when we look at who are the investors in those companies, it tends to be the same names time and time and time again. And so for us, we'd much rather spend our time trying to access those very best names than trying to find that needle in a haystack, that one in a 500 new emerging managers that might potentially do that. And if it means we miss out on one of those managers, then we're fine with that because we've got enough in our roster that continue to find those key companies and drive that out performance.

AI assessment note: “the thing that we constantly come back to is that venture is a power law”