Access is everything
Successful investing in the public markets requires the right asset allocation; successful private market investing requires the right manager selection. Or so the adage goes.
Indexing has been very successful in the public markets in allowing low cost exposure to the targeted underlying assets. Indices outperform almost all active public managers in the long term. In 2024, indexed assets surpassed actively managed assets for the first time. The debate in public markets has moved to doom mongering about achieving the singularity where indices represent 100% of the market: what happens to prices if no-one is actually looking at anything? Has indexing already made the market more inefficient?
Indexing the private markets is in its first innings. We’re still working out the mechanics for exactly how to index assets with low / minimal liquidity (see Apollo’s recent private credit ETF where structural details are in short supply).
In other words, access is still everything in the private markets. This is particularly true for venture where the power law skews returns to a few managers. We have stated before that the power law works at every level: power law LPs access power law GPs which access power law assets.
While VC indices will have an increasing role to play in democratizing access to the asset class, we are confident that the center of gravity for venture capital, especially at the earliest stages, remains artisanal and bespoke.
The core skillset of investors (and LPs) is reading people. VC is still a relationship game. Information asymmetry abounds.
Given the concentration of power law returns in a small number of funds, how can an allocator new to venture capital secure access to attractive risk-adjusted returns?
Data is cool, but humans rule (for now)
Before we look at the LP layer of the ecosystem, let’s first consider how VCs access the best entrepreneurs (and therefore the best companies which deliver the power law returns).
The VC job is simple but not easy (to paraphrase Buffett). Raise LP capital -> source startups -> pick start-ups - > win deals -> add value -> exit startups - > return LP capital. Rinse. Repeat.
Winning deals is the core differentiator. You can boil the ocean to market map everything that moves. Data models can help you scan the market at scale. You can then successfully identify the best teams in a particular sector. Again, data tools can meaningfully transform the probability of success (by filtering founder data, market data, company data, investor data, etc).
But all your work is entirely academic if you can’t secure a place on the cap table. Haystack’s Semil Shah nicely summed this up: “For any good investment, from Series A on, there is at least one firm to compete with. Competition is fierce. VCs will spend over a year networking just to position around one founder or one deal, and if they lose it, it’s gone.”
How do VCs access the best founders?
This question of access to quality founders is a particularly important point for all the data-driven VCs out there. There are numerous academic & practitioner papers about how data can enable VCs to outperform the market (including this aptly named paper, Predictably Bad Investments). The data set behind venture is very manageable (read small) relative to most machine learning environments; we track 50 million data points for our model. Our heuristic model is transparent & has a simple set of rules. The challenge for data-driven VCs is not to build models which work. It is to access the competitive rounds of power law companies which ensure that their models will deliver the promised returns (given the power law dynamics of the market).
Creandum created this table (Figure 1) which compares what factors are important to founders in selecting a VC with what factors a VC think is important to founders in selecting a VC. This is also a helpful reminder that great founders pick VCs, and not the other way round. I’d be curious to see a version of this chart for LPs considering GPs (and vice versa).
Figure 1. Top reasons by founders & VCs partner with each other
Personal relationship & chemistry is #1 on both sides. This can’t be automated (yet). This can’t be scaled. Tinder’s co-founder has created an AI agent to improve customers’ rizz on the dating scene. Maybe this will come to VC too to help woo founders. But not yet. Not yet.
Accessing the best funds
Allocators have to access the best funds if they want to deliver outsized risk-adjusted returns. The top decile of VC funds deliver 25%+ IRRs. But most funds return less than the S&P 500 (and with less liquidity). Vencap’s David Clark talks about how “the best [VC] managers can consistently deliver outperformance at scale, with greater predictability and lower volatility.”
Access is particularly important because data demonstrates that persistence is evident in venture capital (unlike in the public markets). The best returns continue to come from the best investors over the long term (Figure 2). Altimeter’s Brad Gerstner has this phrase that “we had 500 runners and now there are 1,000… but it’s the same 5-10 runners that compete for the podium every week.”
Figure 2. Persistence for Venture Capital Funds
Allocating to “Tier 1” funds also has the additional benefit of minimal career risk. You don’t get fired for allocating to Sequoia Capital.
Everyone wants these power law returns, but few can access them:
Large eight figure minimum checks exclude all but the largest institutions
Longstanding LP relationships mean that these funds are many times oversubscribed. Minimal LP turnover means there is not much room for newcomers. Similar to the company level, great GPs can pick their LPs, not the other way around.
Commit to numerous fund cycles to get into the fund that actually has the power law returns. If you don’t invest in one fund, you might lose your access forever.
However, like company investing, allocating to the best VCs is simple but not easy. Some considerations to take into account:
Fund sizing. Some VCs, notably Union Square Ventures and Benchmark, have remains disciplined with their investment strategy and fund size. The Union Square Ventures 2012 Fund with investments in Coinbase, MongoDB, and Duolingo serves the successful exemplar of this model. Others have moved to asset accumulation mode. Contrary’s Kyle Harrison writes about this bifurcation well here. Large AUM should deliver something more akin to high quality VC beta. Safe but not stellar.
Generational change. The private market industry is famously bad at passing on the baton to the next generation. In the aftermath of the ZIRP era, talented young GPs might conclude that the larger funds raised in 2020/21 are unlikely to deliver carry. So why not strike out on your own? Turnover is high at the moment. See Sunil Nagaraj (ex Bessemer), Tomasz Tunguz (ex Redpoint) or Chemistry (ex Bessemer, Index & A16Z).
Fund structure. GPs are creating ever more funky fund families and structures which may not be aligned with how LPs are seeking to allocate. See Sequoia’s holding company. Or A16Z’s family of funds; good luck to the LP who says that they are not particularly interested in allocating to the gaming fund or the crypto fund but they do still want access to growth.
Emerging managers
The alternative for allocators is to pursue emerging managers, which historically have outperformed their more established competitors (Figure 3). Emerging managers are smaller, so the multiples on the fund can be larger. For example (based on publicly available information), Lowercase Fund I (Uber and Twitter) returned 250x+, K9 Ventures Fund I (Twilio and Lyft) returned 50x+, and Initialized Capital Fund I returned 50x+ (Coinbase, Instacart). In addition, emerging manager sector specialists tend to outperform generalists.
Figure 3. New & Developing Funds Are Consistently Among Top 10 Performers (from Allocate)
But the variance of outcomes from emerging funds is also high (see Figure 4). And arguably the proportionate performance of emerging managers is poor given that 80% of all funds raised are <$100m. It is hard work for LPs to sift through fairly undifferentiated emerging fund managers to identify the winners. Chapeau to the small cadre of professional allocators who do this work. The growth of the asset class has made the task harder still. 25,000 GPs have raised at least one fund since 1990 in the private markets. Access ain’t the problem for allocators focused on emerging managers, it’s discerning quality.
Figure 4. Pitchbook’s VC TVPI simulation by manager experience
Those emerging managers might become the brands of tomorrow. In which case, an LP is well positioned to maintain their allocation in future funds for the next Thrive / Acrew / Haun Ventures. But many emerging managers do not succeed in the long term. 37% of first time funds are unable to raise a subsequent fund, and it appears to be getting harder to raise the next fund.
Fund of funds
An alternative is for an investor to consider a fund of funds model. This can be a good place for investors new to venture capital to start because it increases diversification with your dollar being spread more widely into more managers and more underlying assets. This increases the probability of hitting a power law company. Diversification is the only free lunch in finance. A single fund of fund investment is also much simpler from an administrative perspective (rather than managing multiple capital calls across multiple funds).
Jaap Vriesendorp has a helpful introduction to fund of fund investing as a means of achieving cost-effective diversification. He also covers some of the challenges with this model, including layering fees, the mediocrity (or at least lack of distinctiveness) of many fund of funds, and the challenges of liquidity.
At SignalRank, we like the idea of a single investment allowing all investors access to top tier venture capital. We believe that our model improves upon the fund of funds model:
Liquidity: our intention is to achieve a direct listing, which will allow investors to convert TVPI into DPI at will (by selling their shares in SignalRank)
Lower fees: our intention at scale is to offer SignalRank at 0.50% per annum (and there is no layering of management fees as our seed funds do not charge management fees for the SPVs). By hard coding SignalRank’s own management incentive at the point of a listing, we also ensure that the incentive is lower than in a traditional fund of funds.
Exclusively invest in break out companies: we cut off the long tail of zeroes and underperforming companies because we only invest in the break out companies of seed funds at the point of Series B. By reducing the number of zeroes, we improve returns.
Recycling compounds returns: we recycle exits into the next cohort which compounds returns for our shareholders. Over the same 10 year timeframe, we expect our model to outperform a fund of funds model because of this compounding effect.
Figure 5 shows how our model shifts potential returns further to the right compared to a fund of funds model. Harder to visualize in 2D is the compounding effect.
Figure 5. Comparing SignalRank to fund of funds models
Introducing SignalRank
At SignalRank, we have created a new way of accessing the highest potential companies at scale. We believe this is the model to enable access for new investors to the asset class.
We partner with seed investors to support them in investing their pro rata in qualifying Series Bs. Our partners have access but no capital. The top 200 micro VCs are investing in 2x more unicorns at an early stage than the top three branded VCs combined (see Figure 6). Yet, on average, a seed manager only invests in 1.2 rounds per unicorn. Most seed managers do not have opportunity funds and are tapped out by Series B.
Figure 6. Top 200 Micro VC vs Brands - Unicorns at Pre Seed, Seed or A since 2012
SignalRank has built a model to identify the highest potential Series Bs by ranking investors across rounds. Our model increases the probability of identifying a company which can deliver at least 5.0x MOIC from Series B to 30% (compared to 10% market average). We do this by reducing adverse selection more than by identifying winners (ie eliminating zeroes).
Why would seed partners share dealflow with SignalRank? Surely LP co-investment agreements absorb any outstanding pro rata?
Every LP under the sun requests co-investment rights. Few institutions are set up to diligence specific assets. Even fewer can move quick enough to participate in a competitive round. As such, SignalRank is very happy to complement co-investment agreements by taking any remaining pro rata after an LP has participated / passed.
We have clear product/market fit. We offer 20% deal by deal carry to our seed partners. So if we invest $1m into a company which achieves a 10x increase in value from Series, our seed partner would earn $1.8m upon a liquidity event for their 20% share from the $9m of profit.
Perhaps more importantly than the economics is the speed & certainty. Our algorithmic model enables us to make decisions very quickly (and without needing to meet management). This is a strong competitive advantage. It also empowers our seed partners to request firm allocations into competitive rounds because they know that we already have capital on balance sheet and can move with pace.
We have so far completed more than 20 Series Bs in the last 15 months, making SignalRank the #2 most active Series B investor for high quality Series Bs since May 2023 (after Andreessen Horowitz).
Does SignalRank ever have an access problem?
We exclusively invest with the best investors, including multiple investments alongside Andreessen Horowitz, Sequoia and Lightspeed Venture Partners. But does SignalRank ever have an access problem? Let’s look at our anti-portfolio.
We have analyzed every Series B since May 2023 which would have qualified for our product and asked the question of why we did not secure access to a round. There are three major reasons:
No seed partner. In about 40% of cases, the company had a micro VC / seed manager on the cap table but we simply did not have a relationship with the relevant micro VC. This suggests that SignalRank continues to fly somewhat under the radar (and/or that our partnership team needs to get out more). Once we are more public about our activities, we anticipate that this issue becomes much reduced.
No seed partners. In about another 40% of cases, there were zero seed partners on the cap table with whom we could partner. If a company is exclusively backed by say Sequoia Capital at seed and Series A, then by Andreessen Horowitz at Series B, then there are zero seed partners with whom we could work with. We have a potential product here which we believe will be very disruptive: founder pro rata. We simply finance the founders’ pro rata with the same economic model we have for our seed partners (ie 20% upside), thereby helping the founder defend their position and voting rights. It is the most founder friendly capital on the market. Watch this space.
Competitive. In about 10% of cases, the round was too competitive and our seed partner was unable to secure access because the lead investor squeezed out existing investors. This does happen every so often, but not regularly as the lead investors can attract a poor reputation among seed investors (for not allowing them to invest their pro rata) who in turn will recommend to their founders not to take money from these same lead investors.
Conclusion: access isn’t everything, it’s the only thing
To do venture capital properly is hard. It is all very well for the cocktail party investing set to brag about their SpaceX investment (best not to ask whether it’s via a two layer 2/20 forward contract SPV). But for large asset allocators seeking the potential returns offered by venture capital, access remains a significant problem. We anticipate that this access issue will worsen as the asset management industry’s interest in VC increases.
SignalRank solves this issue for our investors by offering low cost access to quality at scale. We expect more innovation to follow from new market entrants.