Series As from the investor perspective
The other side
Last week I wrote about Series As for founders.1 The data shows there’s a lot of dilution (though less than there was in the recent past). It also made me think about a number of other questions relating to this type of fundraising event.
In particular, I saw that “What’s the clearest indicator of success in a startup that’s raised a Series A?” has a couple interesting answers.
As an investor, my favorite is “They raised a Series B!”
This week I want to look at the same type of transaction, the Series A, from the perspective of an investor.
I worry that it’s perhaps not as clear as it ought to be to founders what success looks like for investors, and what they should be considering as they ponder who — if anybody — they want to partner with.
I see this as critical to founders who are looking to create a win-win relationship with their investors, and start things off on the right foot.
I’ll approach this first from a fund level, and then from a deal level.
WARNING: (simple) MATH AHEADWhat does fund-level success look like to a Series A Firm?
The cynical view, which is as good a point to start as any, is that the job of an investor is to stay in business by picking startups that return enough capital to their investors to raise future funds.
Limited Partners (LPs) tend to look at VC performance in terms of quartiles.
In November 2020, Harris, Jenkinson, Kaplan, and Stucke looked at persistence of VC and Buyout funds in a paper called “Has Persistence Persisted in Private Equity?”.2 Table 4 Panel B looks at data that LPs would have had available to evaluate VC fund performance (not limited to Series A focused funds) at the time those firms were looking to raise new funds. The chart below shows the data from the table — which considers only data about funds raised after 2000.
There are some trends here worth noting. Key among them:
A lot more 1st (top) quartile fund raised their next funds than 3rd or 4th quartile funds.3
There seems to be a connection between past and future performance — though there is plenty of movement between quartiles.
In terms of fundraising, emerging managers (which we’ll consider here as “VCs without enough data to place themselves in a performance quartile”) performed better than everybody except the 1st quartile in terms of both ability to raise future funds, and the results of those funds.
This begs the question: What’s it take to be a top quartile fund?
To look at this, the simplest metric is the Multiple on Invested Capital (MOIC).4
Any MOIC value above 1.0x is a profit, and anything below 1.0x is a loss.
During the period covered in the plots (vintage years 2001-2014, and measured in 2020), the top quartile had an average MOIC across the fund of 3.84x, and the second quartile had an average MOIC of 1.85x.
Using older funds from the same data set (1984-2000, measured in 2020), the top quartile had an MOIC of 5.34x, and the second quartile had an average MOIC of 2.36x.
These outcomes from older vintages might be larger in part due to macroeconomic trends. They are certainly are larger than the more recent funds in part because all these funds’ investments had either exited or gone to 0 by the time the data set was collected. Looking at the data set as a whole, the average unrealized portion of the funds remaining was 24%, and almost all of that was from funds started in or after 2001.
At the end of the day, the MOIC target for an entire VC fund at the early stages is probably going to be at least 3x. That means after the winners are sold and the losers go out of business, the GP is aiming to return at least 3x the amount of money they deployed to their LPs. They need that large a return in order to justify to LPs why they should lock up their capital for such a long time.
To put some quick numbers on it, if I were to go and raise a $50MM fund, I’d plan to invest with the goal of returning as much as possible, but at least $150MM, to my investors — that would make me confident that I’d be able to go out and raise another fund.
MOIC is my preferred way to look at this with founders because it’s much easier to connect it to VC incentives (compensation) than the alternative metrics.
The classic VC compensation structure is “two and twenty” (though the specific numbers may vary slightly). They earn a management fee each year around 2% of the fund size, no matter how well they do. This keeps the lights on and pays for staff. And then, to incentivize performance, VC firms earn something like 20% of the profits:
This means VC partners (and employees who earn part of the 20% called “carried interest”) can earn more money by pulling two different levers:
They get the startups in their portfolio to become really big companies, and get great MOICs — which is better for founders, would lead to larger paydays if done right, and won’t affect their fund’s strategy, but isn’t consistent
They grow their assets under management — which is more consistent, but may affect their fund’s strategy
What does investment-level success look like to a Series A Firm?
VC investing into startups follows a power law.
The likelihood of a particular startup’s success, based on past performance, is an exponential function:
A relatively small number of startups in a fund will ultimately generate the overwhelming majority of returns for an investor.
Scott Kupor, a Managing Partner at Andreessen Horowitz, took a look at this next plot on X a few weeks ago, and identified something really interesting.
He pointed out that great VC funds don’t necessarily pick all that many more great startups than poorly performing funds. It seems like they put more money into their startups that are going to have a disproportionate impact on the fund’s returns, and so have a higher ownership percentage when those startups have an exit event.5
Putting more money into winners is essential to outsize returns because the startups that make lots of money have to pay for the startups that lose money (which will, according to the plots above, be the plurality of the startups funded).
Given the likelihood of success at Series A, the conventional MOIC target at this point for any individual investment is 10x.
You can financially engineer your way to this number. My understanding is that if you look at a Discounted Cash Flow for a startup out to raise their Series A and discount it additionally by its likelihood of success, you typically need about a 10x return in most cases to get a positive Net Present Value. But that’s not the simple math I promised. More importantly, I don’t have a startup financial model at this stage I can share as an example.
My intuition for that is that historically, the odds of a Series A startup getting to Series B eventually were about 65% — and once a Series B is raised, it’s likely to have some sort of exit. Basically, my view is that a Series A startup has a 33% likelihood of a venture-scale exit.6 If I want a 3x MOIC on a hypothetical 3 investment fund, and I’m investing the same initial amount in each startup, I need at least a 9x MOIC on my winners.
When you look at how much capital a VC actually deploys,7 in order to return 3x on the committed capital, the actual deal MOIC required from their successful investments is closer to 10x.
The key financial questions that I would be considering if I were in the business of investing in Series As are “Will this startup be worth 10x more than it is today within 10 years? Why?”
Thinking about revenue as a driver of this, at least in the context of B2B SaaS, may make my questions a bit more clear.
The graphic above, which frames the issue in B2B SaaS terms, asserts that readiness to raise a Series A is indicated by the presence of multiple institutional customers, some of whom eventually become recurring customers. This creates a great connection between the financial view of fundraising and a product-led approach to company growth.
From a product perspective, Series A is all about validating Product-Market Fit (PMF). Much of the diligence I’d be inclined to perform as an investor looking at a B2B SaaS startup centers around whether or not their PMF hypothesis, if correct and successfully scaled, can create a 10x outcome.
I wrote about Series As because there’s data about them, they are early stage, and they are priced rounds — which means the economics are more immediately clear.
Part of the reason I’m building off this paper is because it’s the first data set I found that’s relevant and helpful. But I should also note that I’m in Kaplan’s course on Entrepreneurial Finance and Private Equity this quarter, and really enjoying it!
The data doesn’t seem to be controlling for AUM or fund size, which might affect this.
The paper looks at Internal Rate of Return (IRR), which is used more on the LP side. The paper also looks at Public Markets Equivalent — which compares fund performance to public equities. This matters because an alternative that LPs have to investing in VC (and through them, startups) as an asset class is investing their money through an index fund.
A large part of that is a function of a fund’s follow-on strategy, which is an important and separate topic.
Though it’s worth noting that even so, returns to an early-stage fund may be negatively impacted by things like liquidation preferences.
VCs have to eat and make payroll too! Since it comes out of the management fee, it cannot be invested — but since it’s capital called from LPs, VCs still have to pay it back to get into carried interest.