Investing Billions - E124: Is Venture Capital entirely Based on Luck? University of Chicago and Oxford Study

Episode Date: December 27, 2024

In this episode of the How I Invest Podcast, I interview David Clark, CIO of Vencap, to discuss the venture capital landscape. We discuss assumptions about small vs. large venture funds, unpack survivorship bias in performance data, and explore the power law dynamics in early-stage and growth funds. David Clark shares Vencap's strategy for identifying top-performing managers, insights on fund size limits, and the implications of political and economic shifts on venture capital. A must-listen for investors seeking a data-driven perspective on navigating the venture ecosystem.

One of the interesting characteristics of the venture asset class is this persistence of return. University of Chicago had the study that over half of funds that were in the top 25% continued to be in the top 25%. Have you found that to be the case? And why is that the case? Yeah, we absolutely have. I know I've seen similar research from Tim Jenkinson at the University of Oxford. He said, I think the data he was looking at was Burgess data and he was seeing a 45% chance of having successive top quartile funds. And interestingly, there was also a similar chance of having consecutive fourth quartile funds as well. So it's not just at the top quartile where that persistence plays out, it's also at the bottom quartile.

And I think it's worth just mentioning that when compared to private equity, that persistence of performance is much greater. So David, you posted a really controversial post on LinkedIn implying that smaller venture managers may not actually outperform larger managers. Tell me a little bit about that. There definitely seems to be a common belief that smaller funds and emerging managers will outperform and it feels like that that meme is being shared and repeated

without too many people actually looking at the underlying data to see if it's backed up. So you know one of the things we always try and do at Vencap is to go to the the primary data source and see if actually the conclusions are justified from the data. So this is data that we got from Pitchbook. So they have data on about 14,000-15,000 funds that were raised between 2010 and 2019. But there's only about a thousand of those that where they have performance data. So, you know, first of all, you are working with a sample that is incredibly small. And then if you look at it, you know, the sample by fund size for that smallest fund size group, the zero to ninety nine million dollars, there are only about five percent of the funds that raised where there's performance data. We're not saying that small funds don't

you know can't perform well or large funds perform better. What we're saying is that that you can't really use the publicly available data to draw those sorts of conclusions because particularly in a power law asset class where you know you do have a small number of funds that massively outperform, you just don't know if those funds are being captured in the performance data that comes out. And certainly when we look at our portfolio, the vast majority of the funds that we back are not captured in the publicly available data, that's on Pitchbook.

If you're thinking about basing your investment strategy off the fact that, well, everybody knows small funds outperform, you've got to be really careful about that because, you know, our view is it's just not supported by the underlying data. For those listening on audio only, we have a graph on the screen based on pitch book data that shows that zero to $99 million funds, only 5.1% have actually shared the data and then a hundred to 250 million, 13.9, 250 to 523% and 500 to a billion, 29.7%. So essentially, half a billion to billion dollar funds are six times more likely to have shared their data on PitchBook.

Yeah, I think that's right. And you believe this shows survivorship bias. What is survivorship bias? So survivorship bias is something that we have to contend with in the venture industry because a lot of the data sources, so Pitchbook and certainly Cambridge, rely on the managers themselves self-reporting their performance data. find is that if a manager has a really unsuccessful fund or poor performing fund, then that data tends not to get submitted. And certainly if managers maybe raise one fund and then aren't able to raise anymore, it's unlikely that they're going to continue to report their data into a Cambridge or a pitch book.

So what that ultimately does is overstate the performance in a lot of these instances because you only have the successful managers that are contributing their performance data into the dataset. So yeah, the whole survivorship bias is a real challenge. The one thing that, I think the one dataset that doesn't suffer from that is Carter

because they are able to report all of the managers that raised funds using their platform and so there's no survivorship bias there. The downside to the Carter data at the minute is that it only goes back five or six years. But as that data matures, that could be a really interesting source of much more accurate truth about what VC performance looks like. Again, it's very much concentrated on smaller funds, so they don't have any of the really established

manager funds on there. But certainly for that emerging smaller funds segment of the market, I think that's gonna be a much better indicator of how those funds are actually performing in aggregate. To use an analogy, we all have that one friend that goes to Vegas and always talks about how much money he wins, when in reality,

he's only letting you know maybe the 20% of times that he wins, or he wins one night, and then he loses the next night, and he only letting you know maybe the 20% of times that he wins or he wins one night and then he loses the next night and he only tells you about the wins. What are the consequences of potential survivorship bias and small funds? I remember someone always once saying to me that the plural of anecdotes isn't data. And so you know, you need to you hear a lot of anecdotes on you know, I've heard people on podcasts like yours, David, will say that, you know, I invested in a fund that did 10 X and I invested in a fund that did, you know, 20 X and that's great. You know, it's great to have those funds in your portfolio. But a single data point doesn't give you any real conclusions around whether an investment strategy is successful.

You have to look at things over the aggregate and you have to look at things not just at a particular point in time, but we would argue you have to look at it across an entire cycle because there may be certain types of strategies that outperform when the market's going up, but they might massively underperform as the market corrects. And in a way, we're seeing that a little bit at the minute. You have data from Vencap going back to 1985, and you've seen many funds. Tell me about what you see

in terms of the returns of growth funds, and is there also extreme outcomes in growth funds like we see in emerging managers? So I posted something on X a week or so ago where we looked at the power law concentration of the growth funds that we've invested in. So in terms of how do we classify early versus growth, for us, we don't really do pre-seed. We have a small number of seed funds. And then the early, so early would be categorized

as those small number of seed funds, plus funds that are predominantly investing it around the sort of series A. And they'll do some B rounds into those, but if it's predominantly A rounds, we'll classify them as early. And then growth would be anything that's from B onwards. I would say the bulk of what we do

would be kind of early growth. So it would be sort of Bs and Cs. And we've talked about this previously about what the power low looks like for early stage funds. 60% of companies don't return capital. Around 1% of companies are ultimately fund returners. And it's those fund returners. So 1%, so you invest in 100 companies,

one of them will return the entire fund. Yeah, at least one time is the entire fund. Very often it's multiples of the fund. And so that's the typical power law dynamic for an early stage fund. We thought it would look very different for growth funds because of, again, because of the sense that, you know, growth funds are investing later, they're taking less risk.

And so it would have something that resembled much more of a kind of private equity type return distribution where you got few companies that ultimately lost money, few companies that generated, money, few companies that generated 10x or more and most companies in that kind of fat middle, let's call it a sort of 2 to 5x. What was interesting when we ran the numbers, it looked quite different to that. So even for the growth funds, they lost money on more than 40% of deals. So four in 10 deals, they wouldn't get their capital back, which was quite surprising, I think, for us. But it's also

indicative of just how much risk is still involved in these companies, even at the B, C or D round. You're not going into companies that are definitely going to make it. There's a high mortality rate even for those types of businesses. And the flip side of that was even more interesting for us, because when we looked at 10X, companies that are generated a 10X, for early stage funds, it's about five and a half percent. For growth funds, it was just under that. You're still seeing that real kind of concentration of returns in a relatively small number of companies at the growth fund. And then we looked at fund returners, and this was the thing where we really, really were shocked. So for early stage funds, we've said about one percent of their investments ultimately

in funds that we've backed. One percent of their investments ultimately turn into fund returners. For growth funds, that's currently running at 1.6%. So we're seeing a higher incidence of fund returners for growth funds than we are for early stage funds. And that was shocking. You mentioned that your highest performing fund was actually a growth fund, not an early stage fund. Explain how that could be the case.

Yeah, so our best performing funds that we've invested in, in that 2010 to 2019 sample size was an $800 million growth fund. And it was able to, and it's generating a 13 and a half X multiple. Now, again, I'm just throwing an anecdote out here. That doesn't mean that this strategy works just because we've got one fund

that's done particularly well. But I think it's interesting, because I think, again, the general view is that you can't generate those sort of multiples on anything but a small emerging manager fund. That is just absolutely not the case. This fund was able to do it because it was a relatively early investor, a B round investor in one of the largest companies that was founded in the last 15 years. The company is still private, it hasn't, it hasn't, it hasn't had an exit event. But but the manager has been able to take a

little bit of money off the table on its investment, but it's it's still holding, you know, many multiples of the fund in terms of the unrealized value there. Let's take a step back, your Chief Investment Officer of VenCap. Tell me about VenCap and what is the VenCap strategy? Yeah, so VenCap has been investing in venture funds since the mid to late 1980s, and that's all we do. So, we're 100% focused on venture. And over time, our strategy has evolved.

So, back in the day, we were similar to a lot of other investors out there in that we had a diversified portfolio, investing in 30, 40, 50 different managers. And what we found was that the average, the aggregate returns from that strategy just didn't make sense. We weren't getting the outperformance

that we wanted to see from the VC industry. But within that cohort of managers that we backed, we found there was a relatively small number that were consistently able to outperform. And by outperform, it really means hit that top quartile benchmark on a regular basis. And so we spent a bit of time trying to dig in and understand what was causing these managers to outperform. And what we found was when we looked at funds that had generated a 3X net return for us,

at the early stage, 90% of them had at least one company that returned the entire fund. And so we really then started to focus on, okay, who are the managers that can back the very best companies and do so in a way that allows them to generate fund returning outcomes from those investments. And what we found is there's a relatively small number of managers that are able to do that consistently. So really for the last 15 years or so, our portfolio has

been very concentrated into those managers. So 90% of the capital we've invested over the last decade has gone to 12, 13 managers. And we don't add a new manager very often because the bar is exceptionally high. Interestingly, we added one earlier this year. That was the first new manager we've added to our core program in probably five or six years. One of the interesting characteristics of the venture asset class is this persistence of return. University of Chicago had the study that over half of funds that were in the top 25% continue to be in the top 25%. Have you found that to be the case? And why is that the case?

Yeah, we absolutely have. I know I've seen similar research from Tim Jenkinson at the University of Oxford, who said, I think the data he was looking at was Burgess data, and he was seeing a 45% chance of having successive top quartile funds. And interestingly, there was also a similar chance of having consecutive fourth quartile funds as well. So it's not just at the top quartile where that persistence plays out, it's also at the bottom quartile. And I think it's worth just mentioning that,

when compared to private equity, that persistence of performance is much greater. So in private equity, it's about a third, I think, of funds that have, if a manager has a top quartile fund, they'll go on to have a successor top quartile fund. So, you know, venture is very different to private equity in that respect. And I think ultimately, it comes down to the power law nature of the asset class. And, you know, it was interesting. I was at Slush, you know, a week or two ago and talking to a lot of European managers, talking to a lot of US managers

that were over. And one of the US managers said something that really resonated with me. And it was almost a kind of light bulb moment, which was the best founders can raise an infinite amount of capital from anybody they like. So why are they going to choose me? You know, first of all, to sort of see the deal and to be in the network of those founders. So, you know, I think that's a, that's a prerequisite for, for doing that. But, but why is it, you know, why is a world-class founder going to choose you as an investor?

Um, I think generally it's because, you know, they're looking at. And what are some good answers for that today's episode Is brought to you by Reed Smith the practice of law has the power to drive progress To move businesses forward and support them in achieving their goals by seizing opportunities and overcoming obstacles Reed Smith is focused on outcomes They know that your time is valuable and that your matters are important is focused on outcomes. They know that your time is valuable and that your matters are important. Their deep industry insights and local market knowledge allows them to anticipate and address your needs. Reed Smith delivers purposeful, highly engaged client service that drives progress for

your business. Do I want to work with you for the next 10 to 15 years. That's one of them. So, you know, there is a personality thing that, that, you know, if you're an absolute dick, then people aren't going to want to work with you. So, I think, you know, you have to have a personality that is supportive of founders and makes founders want to spend time with you. I also think, you know, as a firm, you have to be able to, in a way, kind of add value to those companies that you're backing. And how do you do that? You know, one of the things we hear consistently is the best firms are able to lend their brand to the companies that they back. And that brand allows those companies to raise capital, to hire employees, to bring in customers. Not that the companies couldn't do that themselves, but the addition of that brand, it's been this persistence of returns for many

decades, but there's also been this discipline around venture fund size over many decades. First principles would say that as funds get much larger, those returns are going to get smaller. Why would that be the wrong way to think about it? Yeah, I think you have to look at the relationship between fund size and exit size. So if fund sizes increased disproportionately to exit sizes, then I would agree with you. I mentioned the quarterly review we did, just looking over the last few months about some of the

valuations that the best tech companies are able to raise capital at today, with a view that by the time they ultimately exit, they could be worth multiples of that. I think that's really tying into our perspective, which is that the size of the best tech companies is only going to continue to increase. And so what you need to do is to compare the size of funds today with where we think the value of the best companies is going to be in 10 to 15 years time when they ultimately exit. Is there a fund size where you basically say, OK, guys,

it's been a great relationship, but we can't back you in a $10, $20 billion fund? I think what we're seeing is from early stage, the largest early stage funds we're seeing are probably these days around a billion dollars, maybe a billion five. So what do you have to believe that for them to return a billion dollars on a single investment,

they need to own 10% of a $10 billion company. Do we think that manager is capable of doing that? Well, the first question we look at is how often have they done that historically? And if they've done that multiple times historically, then that gives us much greater conviction on their ability to do it going forward. So I think it's not necessarily about

is there a fund size that is too big. It's much more about what is the maximum fund size that we can sensibly underwrite a fund returner for this particular manager. Tell me about the state of the venture market post-2020 for election. What changes, if any, have happened in the space? I think it's really interesting, you know, that people focus on kind of short-term events and

thinking what's changing here. Like, nothing's changed in a way, you know, in the long term. You know, for us, venture's about, you know, can you get the best managers backing the best founders? Like, that's what venture is all about. And don't try and overcomplicate it. And don't try and impose what might happen over the next four years into how that affects your long-term strategy. Because I can guarantee you that Trump is not gonna be

President of the United States when companies that are founded today ultimately exit in 10 to 15 years time. At least I hope I can guarantee that. But so I think you can get as an LP, you can get really sort of fixated on what's happening in the near term and lose sight of ultimately but how do you build a portfolio and a program

that's gonna deliver over the long term? And as I said, for us, it's about like just find those managers that can consistently back the very best founders. Now, having said that, clearly we have seen some changes in the market sentiment since Trump was elected. We do invest in a number of crypto funds and blockchain focused funds. It feels as if there will be a much more sensible regulatory environment for those firms to operate in and for their portfolio companies to operate in. Mark Andreessen was very vocal on Joe Rogan recently.

Yeah, great episode. Something everybody should recently. Yeah. Great episode. Something everybody should watch. Yeah. Yeah, it was brilliant. And for those that haven't seen it yet, and haven't seen the all the conversations on X, he was highlighting the fact that the traditional banking system was effectively debanking people that were in the crypto space. And we've seen this firsthand as well with our portfolio, how difficult it's been for some of the funds, the crypto funds to open bank accounts and what they have to go through and some of the things they've been talking about with their portfolio companies.

So this is real and it does feel like in that part of the market, there is going to be a much more sensible regulatory framework. I think for tech more broadly, again, Mark Andreessen was talking about the regulatory environment that a lot of tech companies had to operate in, which was stifling innovation and stifling the ability to grow. It does feel as if that handbrake will be lifted when the Trump administration comes in. I think what we don't know is what the second and third order consequences of a Trump presidency are going to look like for the tech industry, particularly if

the US becomes more protectionist, if tariffs come into play. So I think our sense generally, as we feel in the short term, it could actually be a positive for the tech industry. But we are kind of reserving our judgment on ultimately what that will look like longer term until we see exactly which bits of policy they end up implementing. I think there's a bit of a paradox when you think about politics or trying to play macro investor from one perspective.

You want to be long the asset class from another perspective, you want to be smart in how you allocate your investments. There's also, I think it's true, obviously, we're going to have a different administration, 8, 12, 16 years. But there is this reflexive nature to startups where they become self-fulfilling. If suddenly nobody is investing in crypto or nobody is investing in new AI companies, it becomes a self-fulfilling prophecy. So sometimes the prognostication on the future can actually affect even seed companies today. I think you have to understand where you are in the in the kind of food chain those is LPs. You know, I feel like an LP strategy is a little bit like an oil tanker. You know, you can't be navigating what's happening on a day to day basis. You know, you've got to set a course and take the volatility that's in there invariably going to come as you kind of play out your

investment strategy. And I think it's actually dangerous to try and jump around too much because, you know, again, our sense is that it's it's almost impossible to time the market. It's almost impossible to pick, you know, individual verticals that are going to generate those next, you know, the next set of top one percent companies for LPs. And so, we would much rather take a view that we want to invest consistently.

We want to keep a fairly consistent annual investment pace. And we really want to focus on, as I said, those managers that are consistently able to back the top 1% founders and not try and not try and overthink it. We've seen some of the greatest companies of all time started during recessions like Microsoft, Airbnb. What needs to happen for other LPs to become excited about the venture asset class again? I'm a bit of a contrarian by nature. So I get excited about the venture asset class when everybody hates it.

And it's really hard to raise capital. That's what gets me excited. I think if you're talking what will it take for more dollars to come into the VC asset class, I think the answer there is we need to see that exit market open up. And again, the early indications we've seen over the last few weeks are that companies are starting to have those discussions again.

And so I wouldn't be surprised if we see 2025 as a much stronger year when it comes to IPOs and then hopefully we can get some clarity on the M&A market as well. You know, clearly we haven't seen much happening in that space over the last couple of years. I'm curious.

A lot of LPs are concerned that venture funds are marking their positions higher than they're actually worth. Looking back at Vencap's data going back to the 80s, have you found that TVPI or where a venture fund marks its book closely tracks DPI, which is actual capital returned back to investors? No, we haven't. So again, we looked at... So the skepticism is warranted.

Yeah. But again, I think so we looked at the data to say, how does DPI or even TVPI in the air five, how does that correlate to where a fund will ultimately end up? And the answer is, it's not really a great predictor. And so, the early performance of a fund to us is kind of irrelevant. And what we need to see is evidence of one of those, one or more hopefully of those potential top 1% companies in the portfolio. And I think that's a much stronger signal for us than early liquidity or early write-ups,

because you have to understand where you are in the cycle, because it was, we saw a lot of seed funds do exceptionally well in terms of getting those unrealized gains through 2019, 2020, 2021. And interestingly, we still see indications that they are holding a lot of those companies at last round value, not at what they would be worth today. You know, in contrast, when we look at our established managers, they were probably pretty quick to write things down and did so, you know, reasonably aggressively. You know, it varies by manager to manager, you know, not everybody is as conservative as perhaps we would like them to be or to do it as quickly. And so, you know, what but what I would say is it feels as if we've gone through that wave, certainly with our managers now, and we think that portfolios are possibly undervalued on an aggregate basis.

How do people follow you online? So I'm on X at Dave Clark 85. I'm also on LinkedIn as well. So it's a mixture of stuff. Some things I'll post on X, some things I'll post on LinkedIn. Kind of a slightly different audience there. So yeah, feel free to follow me on either of those.

Also, Vencap is also active on LinkedIn, so follow us there. We try and post things that are data-driven because that's how we think of the VC industry. That's how we approach the VC industry. I think there's a lot of people in VC that tend to have strong opinions, but those opinions aren't necessarily backed by data. So we prefer to...

Strong opinions loosely backed. As opposed to weakly held, yeah. Yeah, so I think for us, you know, we like to have a very strong foundation of why we do things. And, you know, we recognize the data is not perfect and the data is backward looking. And, you know, there are reasons when, are reasons when we'll go against the data if we think it's not relevant moving forward. But I think understanding the data and using it to really push back on some of the

kind of common myths that are out there in the venture industry, I think is really important for investors to do. One of the things we say internally is that when someone says, well, it's common knowledge that this is the case in venture, we say, well, common knowledge in venture is usually wrong. So, don't just take these assumptions at firsthand.

You need to go and really examine the underlying data and understand what's driving it. Well, David, it's been great to have you back on the podcast. I look forward to sitting down in New York or London very soon. Yeah, that sounds good. Brilliant. Thanks so much, David. That's been fun.

Thank you, David. Thank you, David.