Innovating in Venture Capital
The venture capital industry is 60 years old and set in its ways. That is about to change.
VC terms are frozen in time. GPs offer self-liquidating LP funds, with a 2 and 20 list price, and a 10 year fund lifespan. The first fund with this structure was offered in 1958.
Since then the deal documents have become optimized around use of cash, incentives, and taxes. It works! Assets in LP funds have ballooned to over $5T, with over $1B in venture capital. Recently Pitchbook reported that “North American venture capital has beat out every competing asset class in three-year IRR performance, at 18.51%.”
Funds that stray from the formula have been punished. For example, securitized VC funds are not optimized in the same way. We can see a typically awful result from one of the longest-lived public VC funds, Safeguard Scientific. This chart (starting at a stable point after the Internet bubble and crash) shows Safeguard stock price changes in black, compared to the NASDAQ index of other public tech stocks.
What is new?
Recent years have seen a lot of attempts to update the VC model with tokenization, crowdfunding, DAOs, SPACS, etc. Most of them have not worked. The best of these efforts is AngelList rolling funds. The success of rolling funds is an indicator that the model is starting to change.
Rolling funds reduce the pain of starting a VC fund. Startup is painful with the old model. The partners of a new VC fund spend 1–2 years lining up the $50M in commitments that they will need to support even a small team of full-time VCs. They pitch the same set of limited partners that are being pitched by funds with longer and better track records. They spend their own money on $150K in legal and admin costs. Then, if things go well, they start to make money five or ten years later.
The rolling fund package attacks these pain points. VCs can start part-time with small commitments from angels they know. A cheap SaaS package from AngelList covers their legal and administrative needs. AngelList gives them some fancy accounting so they can get started right away, and incrementally bring new investors into new cohorts.
Rolling funds don’t work well for large deals, because money is only available incrementally. We will propose a different structure that helps emerging VCs draw larger amounts of funds to lead a deal.
The rolling fund structure from AngelList is hostile to securitization. It fragments every rolling fund into a series of very small quarterly funds with different portfolios and different values. They can’t be reassembled into a larger-cap fund because the documents prohibit secondary sales.
DAOs or “decentralized autonomous organizations” have also had some recent success. Metacartel and The LAO have grown assets and profits in line with the outstanding growth of DeFi markets. These organizations are not mechanized and “autonomous”. In fact, they are old-fashioned coops, with human members who put in both money and work, and democratic voting on major decisions. They are notable because they include a much larger number of partners than a VC firm. The technology and tactics for working together in “community” sized groups is improving rapidly.
Crowdfunding is growing with direct token offerings, and to a lesser extent, Jobs Act portals that sell Reg CF deals.
Both of these channels have had problems. The government-sanctioned Jobs Act portals have a low volume of sales and have not had much impact. The mid-sized Reg A+ public offerings have mostly been bad. They suffer from adverse selection — the tendency to try to sell securities to retail investors when professionals won’t buy them.
Crypto offerings have raised more money, have meshed better with a global market, and had a bigger impact. However, they also suffered from massive adverse selection during the ICO boom. The new DeFi deals are more sophisticated and much better at delivering value. They are also more likely to be funded by traditional VCs.
In my opinion, these crowdfunding channels are supplementing VC money, and not replacing it. A deal that skips VCs and goes directly to crowdfunding is often a bad deal. VCs do an important job when they structure a deal that will be good for investors. The good deals seem to be raising VC money, and then offering slices to their customers in one of the crowdfunding channels.
Tokenization and securitization
I have looked at dozens of attempts to “tokenize” LP funds, which really means to securitize them in digital packaging. None of these deals have been successful.
The securitized packaging picks up new costs. They have sales costs that are higher than the 2% allocated for selling a private LP fund. This is a pure loss for investors. Buyers pay in money which is not immediately invested, creating a “cash drag” that lowers IRR. Buyers pay more in taxes, or have a difficult time complying with tax laws. There are other compliance problems.
Tokenized funds often promise trading and liquidity, but they cannot deliver it. The size of the float is too small to attract marketmakers. So, if anyone wants to sell, it just drives down the price in an embarrassing way. The disclosure is not good enough. VCs have confidentiality agreements with their portfolio companies that prohibit them from releasing information that would allow traders to value their assets and bid on them.
Some VCs have continued the experiment. For example, Cosimo X is a securitized VC fund that is solving the size problem with an evergreen structure. New investors buy into the same fund as older investors, and the fund grows in size over time. Investors will find it easy to get into this fund, because they can buy it incrementally. They may also have opportunities to get liquidity out of the fund by selling the tokenized security.
This brings up some questions:
- How will investors get their money back? Cosimo proposes to distribute half of every exit in the form of a buyback, and reinvest the other half.
- What is the price for new investors? We can think of the offering of new VC shares to fund a deal as a sort of PIPE — a private investment in new shares that goes around the trading market. Statistically, a PIPE needs to be priced at about a 6% discount to trading shares in order to entice new investors to take a large block. In fact, Cosimo is holding down the price of new shares by offering them at the price of the last NAV assessment. This creates conflicts of interest between new investors, old investors and the fund.
- How will this security comply with (archaic) US rules for exchanging private securities? The most common answer, and probably the best answer, is to offer it mostly in non-US markets.
A SPAC can buy a late-stage VC deal. And, it’s a securitization channel that takes the deal public. This is actually a useful way to securitize because it focuses on getting to a tradable size and disclosure. We can expand on this concept.
What is next?
Here are some ideas.
In many cases, we are looking for solutions to the problem of erratic returns. On average, venture capital is a good investment. VCs make most of their money on unicorns, deals that pay back more than $1B. If you have a piece of one of those deals, your fund will do well. If you hold a piece of Google or Facebook or Coinbase that goes to a $100B IPO, you can pay for all of your less successful investments and buy a private island. However, only about 1.28% of funded early stage deals go on to become unicorns. Many VC funds do not contain one of the winners, and so they lose money. They need a large portfolio to have a good chance of making a profit. With early stage deals, they need 100 or more. If they buy 1% of the deals that they look at … they have to look at a lot of deals. This situation cries out for industrial-scale organization.
A direct solution is to roll up early stage deals into a much bigger fund that is a sort of ETF for a whole VC sector like AI or electrification. The fund managers would trade out of VC deals that are exiting their target phase, into rollup fund shares. Eventually the rollup would grow big enough and predictable enough to trade. Then it would get a liquidity premium, and be able to provide lower cost capital for early stage funds.
There are a couple of reasons that this doesn’t happen. The trivial reason is that secondary sales are discouraged by US regulations. A rollup concept would need to be proven in a non-US market. It could come to the US market when it got big enough for a full public filing.
A more substantial reason is that GPs are doing active management of their deals. They participate in management, and they put in new money. If they sell the deal to a fund, they stop doing that job, and investors lose a source of value. Active management is very important in PE deals.
The rollup concept works better in a market structure where more investors buy a smaller share of each deal, and take a less active role. In fact, that’s where the VC market is headed, with more shareholders in each deal.
For decades, VCs have been telling entrepreneurs to limit the number of investors and “keep the cap table clean”. They said that a smaller number of investors would result in more committed investors, and faster response. Often, the VCs use their “fast response” to fire founders. The crypto boom revealed this recommendation to be self-serving and incorrect. Deals that include more investors have more supporters, and often, stronger supporters.
The strongest supporters are customers, and the strongest structure in crypto is mutual ownership, where customers get rewarded with project shares and meaningful ownership.
Coop funds and bonding curves
It might be a good idea for LPs to get some mutual ownership in their GPs. It would also be better for the GPs because they could raise money faster. In fact, early LP investors often negotiate for shares in a GP. In DAO structures, the GP is eliminated entirely, and the DAO coop is controlled and operated by the investors. We can imagine a number of points in the middle where LPs get GP shares.
Ideally, they would get shares on a bonding curve that rewards early commitments with more shares. The earliest commitments provide the most important momentum. If this accelerates fundraising, it will also improve returns by reducing fundraising costs, and reducing delays in getting to hot opportunities.
Open Access Launchpools
Venture capital runs on trust. That prevents it from scaling into global markets. It makes the VC and startup business less inclusive of newcomers from non-traditional backgrounds. VCs raise money from a small circle of LPs that trust them. They have to trust the LPs to honor their commitments to send money in response to capital calls. This reduces the number of people and entities that can even be an LP. Entrepreneurs need to build trust in a similar way when they raise money.
Launchpools can open startup deals, and LP deals, to a global group of investors that we don’t know. Then the deal sponsor can run a process to build trust and get commitments.
Investors can show their interest in a launchpool deal by “staking” — sending some assets into a blockchain escrow pool. They can withdraw the assets at any time. On the other hand, if the deal gets hot, their place in line may gain some value, and they will be rewarded for showing up early. Eventually they will get a specific offer that they can buy.
Staking uses a trick from DeFi. Stakers can continue earning interest and appreciation while they stake. Normally when investors send money to buy a SPAC or securitized fund or other securitized deal, the recipient of the money invests it in low-yield bonds or bank accounts. This is dead money that reduces returns. In the world of DeFi, investors can buy their own interest earning or appreciating assets, and then stake those in tokenized form. The opportunity cost of staking is low.
This type of escrow opens doors for:
- Startup deals that go out to a global market, instead of to a small circle of VCs that know the founders.
- A lightweight SPAC. The launchpools version is cheaper and faster for everyone, and might be applied as a sort of single-deal VC fund
- A capital call pool. Investor stakes can go toward funding a year’s worth of deals from a VC. The VC gets instant access to capital for qualifying deals. Later, the investor claims the fund shares, and any unused stakes.
Sell follow-on rights
When VC’s buy a series A deal, they usually negotiate the right to buy similar shares of future B, C, D fundings. They say that they can make a higher return by doubling down on their winners. However, research from Hatcher+ indicates that later stage rounds make a lower return than earlier stages. If this is true, then follow-on rounds benefit VCs rather than the LP investors.
- The later stage rounds are bigger, and VCs make a percentage of the invested funds. It’s a labor-saving device for VCs to buy one small deal, and get paid again for follow-on fundings.
- VCs reduce their risk by shifting into later-stage deals. Those deals have a higher likelihood of a payoff than the 1 in 100 chance of a very early stage deal. However, follow-on rounds have an ambiguous effect on LP investor risk. They reduce the number of holdings in their portfolio, compared with a system that would force VCs to find a larger number of deals.
In theory, early-stage specialists could make money by stripping off the follow-on rights and just sell the rights to later stage investors. This might be more economically efficient for LP investors.
VC’s have been getting fast returns in the form of liquid “tokens” from crypto and DeFi. Maybe they are just getting lucky in a hot category. However, there are reasons to believe that we can build systematically on the acceleration provided by DeFi. DeFi doubles up returns by allowing one investment to be used as collateral for a second investment. It increases velocity by reducing infrastructure to weightless software. It increases scale by naturally reaching a global market.
Putting it all together
We can imagine a turbocharged system where:
- Investors put stakes into launchpools to support “emerging” VCs.
- VCs can instantly pull cash out of escrow to get to the front of the line for hot new deals.
- Investors can get liquidity from deals as they mature out of the target investment stage by redeeming for tokens, or swapping shares into salable sector funds.
This would open up access. It would give investors a bigger portfolio at their target investment stage. It would be a lot of work. It would run on DeFi.
We can think of many variations. We can give early stakers participation in GP shares (coop fund). We can make commitments be automatic for a selected VC (rolling fund). Investors often have narrowly targeted interests, and they might post to a sort of private SPAC that takes proposals from VCs that want to lead a deal in that category (like a search fund). Investors could use their stakes as pooled collateral to get a place in several categories. We can engineer a liquidity system so investors can cash out more rapidly into tokens, either through a swap into an ETF rollup (opportunity fund), or by just investing in transferable crypto and DeFi tokens (tokenized fund).
We’re seeking contributors to figure out how to bring DeFi magic to venture capital. If you are interested, please contact me on Discord.
Maxos builds DeFi for the real world. We bring together experts to figure out how to rebuild real-world financial services such as venture capital. Then we create, validate, and launch.