Three counterintuitive lessons from a crypto investoooor

gm Substack! I finally started a Substack and for my first article I’ll be writing about three counterintuitive lessons I’ve learned from being a crypto VC over the last few years. Don’t worry I’ll get into more spicy topics like politics in future articles.

---

1. Portfolio construction is more important than picking the right companies.

This lesson is the most counterintuitive. But the math is actually quite simple. If you invest 0.5% of the fund in a company that returns 100x, you still don’t return the fund. And since VC returns follow a power-law distribution, 100x winners don’t come by very often, so every time you come across one you have to make that investment count. Concentration > spraying and praying.

Portfolio construction is why just because a VC collects a bunch of nice logos on their fund website’s portfolio page doesn’t mean that their returns are actually good. And also why it baffles me how $400M+ funds are still writing seed checks.

Some will argue that small initial investments are to get access to “shots on goal”, where the plan is to double down massively on winners in subsequent rounds. What actually happens in practice is larger late-stage funds come in with sharper elbows and get almost all the allocation for themselves (i.e. rounds become more zero-sum the later you invest). On top of that, if you weren’t the lead investor in the seed round you likely didn’t get pro rata rights, so your ownership stake gets diluted by a lot (I’ve seen 90% in one instance).

If you take this lesson to its logical extreme (i.e. picking the right companies doesn’t matter at all), then the optimal portfolio construction is 100% dollar-cost averaging into ETH; this is otherwise known as beta. And truth be told, the dirty secret of crypto VC is most funds started in the last cycle have not outperformed DCA into ETH. Assuming a reasonable cost basis of $200 if you dollar-cost averaged ETH over 2018 to 2020, your fund TVPI would be 15x at today’s prices.

Many point to this famous study done by AngelList that shows on average funds that make more investments have better returns, as a counterexample. I think crypto is different. Because the public market equivalent benchmark return (e.g. DCA into ETH) is already so high, you need to be concentrated in asymmetric returns in order to have a shot at outperforming. Otherwise, the average returns of crypto VC will be lower than just buying ETH over time. And it’s pretty damn hard to outperform ETH in the long run.

So with every new investment, you should think “Will this outperform my ETH bags and return the fund?” and make a sizable high conviction bet.

2. Before product-market fit, there’s very little correlation between how “hot” a round is and what the eventual outcome is.

If you look historically at what the biggest winners are from the last cycle, almost all were not hot deals at the seed round.

• DeFi: Uniswap may have been a hot deal but Aave, back when it was called ETHLend, was available for any retail investor to pick up on the public market for pennies. And really all of Ethereum DeFi were not sexy investments (while new BitMEX competitors were super hot deals) until DeFi summer happened.

• NFTs: While DeFi was becoming hot with all the degen yield farming, SuperRare cryptoart was still completely overlooked. The floor prices of XCOPY and Pak pieces were still in the single digit ETH.

• L1s: Ironically, Solana was one of the only “VC chains” that was not a hot deal at the time (unlike Dfinity, Oasis, Algorand, ThunderToken, NEAR, etc.), and turned out to be the best performing alt L1 investment.

This is why it makes sense to stay very disciplined on valuation at the seed stage. Increasingly I’m seeing VCs aping into $60-100M valuation pre-product seed deals. The only exception I can think of where this makes sense is L1s because the TAM/upside is so high, but otherwise you can even buy publicly traded coins with traction at a cheaper FDV than some pre-product seed deals.

However, after product-market fit, it’s the complete opposite: the best investments are the most obvious winners. And that’s because humans naturally have a difficult time internalizing what exponential growth feels like (i.e. look at how many people dismissed COVID in early 2020); we tend to underestimate how much winners can actually win and become monopolies. OpenSea’s Series A valuation at $100M pre-money seemed high at the time, but very quickly became a crazy bargain given how fast their volume was growing.

This is a good example of dialectics (opposing truths at extremes). The best risk-reward investments are either cheap pre-PMF companies or expensive post-PMF companies, nothing in between.

3. It’s difficult picking the winner in hot narratives and crowded spaces.

A trend I’ve seen in the past year is web2 founders gravitating towards building for the hottest web3 narratives and the most crowded spaces. Many VCs point to this as a sign of great talent entering crypto, but I see it more as good credentials entering this space not necessarily good founder-market fit. Crypto may be different from other industries when it comes to credentials; historically almost all of the most successful crypto projects were founded by those who did not have Ivy League/Silicon Valley pedigrees.

There are a few concerns I have about investing in obvious ideas.

• These ideas attract mercenary founders. These founders are good at copying what’s already working (e.g. Ethereum DeFi onto other L1 chains, existing web2 SaaS product for web3 DAOs) and aggressively marketing their product. But inevitably there will be a drawdown in the founder’s sector as crypto rotates to the next hot narrative. For example, we’re seeing this happen right now as Ethereum DeFi coins are down 70-80% from all-time highs while DeFi on other chains is the new hot narrative. Of the Ethereum DeFi projects started during 2020 DeFi summer, the mercenary founders have pivoted to angel investing while the missionary founders have the product vision to keep building and innovating.

  A good way to suss out mercenary vs. missionary founders is to walk through the idea maze with them. Can the founder talk about all the approaches that came before them and what they’re doing now that’s better? It’s always a red flag if a VC knows way more about a particular space than the founder does.

• These ideas are super competitive. When there are a dozen projects trying to build the same thing (e.g. Solana lending protocol), it’s much harder to pick the winner. Each category is still very much a winner-take-all or duopoly-take-all business. If you’re taking a concentrated investing approach (per lesson #1 above), then you can’t spray and pray into competitors due to conflict of interest.

• These ideas have high pre-product valuations. The minimum valuations I’m seeing now for <DeFi clone> onto X chain is $40-60M pre-product, sometimes reaching $100-200M. The risk/reward might be good for traders looking for a quick pre-sale to token launch flip, but not for VCs looking for outliers to return the fund.

---

I don’t have a conclusion so I’ll end this piece with one spicy hot take: The funds that showed their paper gains to LPs to raise massive mega funds are going to underperform ETH once their paper gains are materialized.