DeFi Weekly #55

Optimism's $3.5m round (Investors hedging across the stack), Uniswap vs Uniswap (Protocol Defensibility), Binance and DeFi Tokens (Cashflows vs Moneyness)

Latest this week

  • Just last week the infamously known Plasma Group managed to raise $3.5m and transition to a for-profit startup under the name Optimism. Paradigm and IDEO led the round mostly on a bet on the team and their talent around layer 2 scalability. Now what I find interesting with this announcement is that Paradigm has also invested in Starkware for scalability. Metastable is another case where they've invested in Starkware and Nervos (not exactly a pure layer 2 but demonstrates the point). As the crypto stack becomes more fragmented investors are going to increasingly hedge on all layers of the stack until it's clear where the value will accrue over a long enough time horizon. One thing that seems increasingly obvious is that full stack approaches are a losing game based on the evidence we have this far. Any team which has raised on the promise of being a full stack blockchain + application hasn't remotely gained any traction. My personal thesis around this is the fact that making a base layer is ridiculously hard from a talent and hiring perspective.

    Take for example Celo, Silicon Valley darling with all the right people behind it and $30m from their last round - they still haven't launched. The number of consensus layer engineers is easily less than 1000 worldwide. Furthermore there's two kinds of developers - missionaries and mercenaries. Missionaries work in the Ethereum/Bitcoin ecosystem and mercenaries go to the highest bidder. Layer 1 mercenary bounties are high as the largest players have warchests in the hundreds of millions or billions (EOS). It's even harder with the fact that not many of these layer 1s in San Francisco don't hire remotely when they really should given the tiny talent pool they're competing for. What's even harder is that you also need plenty of good blockchain understanding engineers to build the stack above node clients. That talent pool is even more fierce to compete with as developers have many more options and more interesting than the "new chain, lots of money" pitch. Once layer 2 tokens become all the rage we're going to see a similar war for layer 2 scalability talent which is almost like consensus layer stuff but a specialized skillset nevertheless.

    Talent wars are an underappreciated topic in the industry. Any crypto company that doesn't allow remote workers has to accept the fact that hiring will be a bottle neck in their execution plans. Moving away from the full stack approaches described above, a more interesting pitch to me is the semi-full stack approach. This is when businesses decide that building their own chain isn't a worthwhile pursuit but build large ecosystems around it instead. Examples of components in this stack are layer 2 ecosystems, developer platforms, custom wallet etc. I've got more thinking around this that I'll share in another edition.

  • As time progresses, on-chain liquidity is going to be increasingly important. Off-chain order books are good for certain use cases, however more primitives will rely on deeper on chain liquidity as time goes on. The sad reality is that no one has nailed a solid value capture model in this space. Kybers new tokenomics are mildly amusing to the market but still a loser overall. Uniswap is a fascinating case - and not for the reasons everyone raves about it for.

    Point 1 - Uniswap has solid defensibility as it has moneyful state in it. The more liquidity in the protocol, the more liquidity it attracts. Although Uniswap for better or worse, has no premine and no transaction fee cuts. Good for users, not so good for investors. Eventually when VC pressure kicks in, the team will need to add fees in v2 or v3 in the further out future. However the biggest threat to Uniswap will be the non rent seeking behavior of v1. Just like 0x v2 is a non rent seeking protocol (or a poorly attempted rent seeking protocol), 0x v3's biggest problem is its predecessor. If you think about it, the best time to launch and fund a new Uniswap competitor will be during the launch of a new version. I'd imagine that v2 could offer enough improvements that migrate liquidity over, however if v2 is similar to v1 (in terms of monetisation) then v3 is going to be harder to push as majority of improvements will be taken care of in v2.

    When developing a DeFi protocol having an auto migrate function after 1 year would be good trigger to forcefully move over liquidity in old protocol versions to new ones. Sounds simpler than done but could be a massive pain saver for teams in the future. I'm not sure how the community would react to such mechanisms but doesn't really mean much as long as liquidity isn't impacted.

    In a similar vein, DeFi Zap has been doing great in terms of ETH gone through the system but unlike Uniswap fails to build meaningful liquidity moats as value passes straight through their system. Their contracts are fully open source and can be redeployed without any cuts (should they be implemented). Front end defensibility only goes as far as the value differentiator in the product. One potential way to mitigate this would be to have money go through their smart contract wallet contracts and then move into the DeFi debt refinancing game which Instadapp is a lone wolf in. Zaps in their current form of a business aren't very strong as any revenue earned will result in DCF valuations (which isn't lucrative in crypto yet unless you're in the margin/exchange business).

  • My final point today is a follow-up on the above about cash flows but also BNB. In case you missed the memo, Binance low key changed their white paper to change the fact that BNB burn is based on traded volume and not profits. This could be a regulatory necessity or just Binance laughing on plebs who are hoping for that juicy 20%. Regardless, I don't think it really matters. A token which relies purely on DCF and provides no means of meaningful speculation will have a hard time in public markets. Binance's winning move was changing the primary use case as a pseudo loyalty rewards token to moving towards being money via it's IEO platform. Binance IEOs are some of the only in the industry that consistently produce good returns. By holding BNB you're hoping to accumulate more so you get first pick in the lottery and can differentiate yourself from other retail users.

    Access to profitable IEOs is worth far more than the hopeful bump in token price when Binance "burns" BNB (used quotations as the burn is super weak as it doesn't exert any buy pressure on markets). Every token should eventually be competing to be a SoV/moneyful - including Ethereum. It's always surprising when I hear pitches saying that a token just wants to be a MoE as it's like saying you'd never want your token to become something people deem as scarce. I think MKR faces this issue as well, the game theoretic component of it is a good but practically doesn't play out. In order for the MKR price to increase, ETH has to increase as well. In that case, the alpha from ETH is far higher than anything MKR would produced as MKR appreciating is a second order effect from ETH appreciating. The exception to this is SNX, I think they've really nailed this aspect and clearly demonstrated it by providing far higher returns than MKR or ETH in a one year time frame. I still don't hold any SNX but am potentially coming around to why it might be a good idea. Many irk away from it due to certain mechanics but as mentioned before, could be a good buy given the valuation it can grow into. In fact, SNX is the first token in DeFi that aggressively goes to funnel value back to the token.

    The ETH/DeFi purist view I think is unprofitable and will lead to the gradual demise of many DeFi projects and tokens if they don't change their thinking around value capture and tokens. Selling equity is an easy path but you're going to become beholden to monetisation/regulatory issues soon as the lack of revenue will cap you off from further rounds and probably lead to a token generation event eventually.