The Case for Generalist Early-Stage VC Funds
“Rule of thumb” - what a funny phrase.
It comes from a time when people literally used their thumbs as a form of measurement. Thumbs were generally about an inch long and you always carried them with you. Quite convenient. Over time the phrase expanded its purview and has come to apply to any concept that is generally understood or agreed upon.
Sometimes when I encounter a rule of thumb I wonder if it is rooted in science or narrative. The human mind prefers narrative. Stories are easier to digest, easier to believe. That’s deeply embedded in our wiring as humans. Our connection with narrative has helped us organize and survive.
So when I have heard LPs say they prefer venture funds that are sector-focused vs. venture funds that are generalists, I’ve wondered why. It’s clear that the narrative of a sector-focused fund is stronger. When trying to decipher which of the many venture funds in the market will win or will provide additional diversification to a portfolio, it’s easier to craft a narrative about a sector-focused fund. Since their focus is by definition a bit less abstract, we can conjure stories about why a sector-focused fund should have a unique exposure or advantage. It’s not as easy to imagine what opportunities will be packed into a generalist fund, especially when both types are blind pools. Intuitively, sector-focused funds should provide unique exposure and better performance due to the specialization of their managers, right? But is this rule of thumb based on science or narrative?
I truly didn’t know. As a multi-decade venture capitalist, I had my own theories - my own narratives about why generalist funds are strategically better positioned. I could argue, for example, that a firm is better positioned to win if they get to see more pitches before they swing. There’s also some signal in the fact that most of the dominant institutional venture firms are generalists. But at the end of the day, I didn’t have any data to point to either.
I suppose that since my instincts contradicted common wisdom, I was more eager to find the truth. I was curious. So, the team at Interplay did the heavy lifting to find some answers. We invested hundreds of hours into developing a data-driven view on the comparison between the performance of sector-focused venture funds and generalists.
Let me start by saying that this analysis is imperfect. We sought to dispel myths with data and see if the rules of thumb hold up to science. But, PhDs of all walks will surely find holes in methodologies and point to a variety of biases. We’re first and foremost investors after all; research is not our forté. But we did our best with the resources at hand to deliver a directional answer on this question. Sunlight is the best disinfectant; we’ll lay bare our methodology and let you decide if it’s sufficient.
Here’s what we found.
METHODOLOGY
TL;DR: We evaluated the performance of more than 400 now-mature US-based early stage VC funds that launched between 2000 and 2014. We manually tagged each fund’s strategy based on our evaluation of the companies within its portfolio. This data set allowed us to analyze performance across strategies. Sources: PitchBook, Crunchbase, Press Releases.
Expanded: Here’s a more detailed explanation of what we did. We attempted to create a very thoughtful methodology for this analysis. As I mentioned in the introduction, however, it’s a certainty that the analysis is imperfect. Nonetheless, we think the findings are directionally relevant, but we’ll detail our methodology here so that you can draw your own conclusions.
Population:
- We included 426 funds in the sample set. Note that a firm may have more than one fund or vintage. All of the included funds were US-based, focusing on investing in early stage US-based companies.
- Since we’re writing this report in 2024, all of the funds were founded between 2000 and 2014. This means that the youngest funds included in the dataset will be roughly 10 years old and, as a result, we hope that we’ll be comparing fairly mature performance KPIs of each fund.
Performance:
- Fund performance is assessed using Total Value to Paid-In-Capital (TVPI).
- While this metric was available for all of the funds in our population, there is some risk presented by using this KPI. TVPI relies on the fund’s reported carrying value of their positions, which can be misrepresented. Our hope is that because the sample size has a decent volume, the probability of carrying values being out of whack is roughly equal in both the sector-focused and generalist fund populations.
Categorization:
- The missing link from the available datasets was a reliable categorization of each fund’s strategy. Without knowing which funds are generalist or sector-focused, there would be no way to compare the two populations.
- To categorize the funds, we manually evaluated the underlying portfolio companies of each of the 426 funds and applied the following logic:
- First, we filtered out the funds that did not specifically focus on software. We categorized any fund with less than 85% of its portfolio represented by software companies as not being software-focused. 215 funds were deemed to be software-focused.
- Second, we segmented the software focused funds into generalist and sector-focused funds. Note that sector-focused in this context refers to a focus on a sub-sector within the software industry (e.g., marketing technology, financial technology, health technology, education technology or cybersecurity). Funds were determined to be sector-focused if more than 50% of the portfolio was invested in one or two key sub-sectors.
FINDINGS
#1: Software-focused funds outperformed hybrid funds.
To get a fairly clean comparison, we needed to evaluate companies that were all truly investing in software. To do that, we peeled out any fund that had more than 15% of its portfolio invested in non-software companies (e.g., hardware, services, etc.). We refer to funds with software and other sectors in the portfolio as hybrid funds.
Once we had those two group segments, we compared their performance. The skinny: software-focused funds outperformed hybrid funds.
#2: Generalist software funds tended to raise larger first vintages than subsector-specific funds.
Within software funds, we further segmented the population into two groups: 1) generalist funds, which invest in any type of software sub-sector, and 2) subsector-specific funds, which focus on a particular software sub-sector, such as marketing technology, financial technology, health technology, education technology or cybersecurity.
Generalist funds tend to raise more capital than subsector-focused funds in their respective first vintages. The first vintage of generalist funds averaged an AUM of ~$97M; the first vintage of sector-focused funds averaged an AUM of ~$44M.
This one surprised me. My perception was that subsector-focused funds had an easier time raising their first vintage given the strong narrative tied to their subsector focus. I was wrong.
#3: Software generalist funds averaged higher returns than subsector-specific funds.
Software generalist funds delivered better returns than subsector-specific funds. The generalists generated an average TVPI of 2.5x vs. 2.1x from the subsector-specific funds.
#4: Generalist funds tend to survive longer than subsector-specific funds.
Software generalist firms tend to operate for more vintages than sector-specific funds, averaging close to 4 vintages and 2 vintages, respectively.
There are lots of potential explanations here. Maybe the GPs are proactively shutting down a fund strategy after the innovation cycle in a particular subsector has matured a bit. Maybe the headwinds of raising capital are prohibitive. It’s not clear from the data. But, we do know that these firms tend to survive for about half as many vintages.
Another interesting anecdote (on a dreadfully small sample size): three out of four sector-specific funds that performed well in their first vintage pivoted to become a generalist fund in their subsequent vintages and performance tended to improve.
CONCLUSION
What does this mean for you? Interpret it how you will.
I suppose that I found this analysis validating for our approach as a software generalist VC investor. This data generally supports the more flexible generalist model in creating durable firms that survive across many vintages.
I do, however, think that averages are average and rules are sometimes meant to be broken. I’m sure there are sector-focused funds with a unique strategy that can outperform. I’ll personally be on the lookout to invest in the outliers, but with this framing, my bar is higher.
To connect with our VC team and learn more about our fund, please reach out to Christian Mark.