Smart Contract Platforms, Developer Network Effects and Pareto Distributions. Oh My!
We just passed the three year anniversary of Joel Monegro’s Fat Protocol Thesis. In crypto, three years is the equivalent of a decade or two in some other industry, so it’s remarkable that the it remains the driver for the broad majority of of crypto-investor theses today. It’s not just that investors continue to pursue base layer protocols at insane valuations based on testnets or less, it’s that the innovation happening on Ethereum, including layers 2 and above, continue to fly well under the radar. And the disparity is growing.
Money Still Chasing Base Layer Protocols
In June, Algorand raised $60M in a dutch auction token sale that coincided with the launch of its mainnet. The sales pitch for Algorand’s technically superior blockchain is that it can handle 1000 transactions per second with latency under 5 seconds.
According to the Algorand block explorer, daily transactions have hovered around 3000 per day with consistency over the past week. This suggests that signs of product market fit or any kind of meaningful adoption are no where in sight. $126M raised to get to a mainnet launch.
Another “scalable” blockchain, the NEAR Protocol just released their beta testnet with a mainnet launch expected in late 2019. The NEAR Protocol raised a relatively modest $12M to get their beta launch. Their own description of why developers will want to build on their protocol is as follows:
Towards the end of 2019 the NEAR Foundation will deploy the NEAR mainnet. Around this point a ton of people who hold NEAR tokens will likely want to use the apps on the network they now own access rights to, and people who hear about the launch of this new decentralized protocol will want to come and see what the fuss is all about.
There’s a super competent team building the base-layer protocol and developer tooling, and they’ll be available, responsive, and extremely open to suggestions for when things inevitably don’t go as planned. If you need 100% certainty, this program isn’t for you.
Will these value propositions attract developers? I talk to startups and developers in the Ethereum ecosystem every day and I don’t hear much talk about migrating off Ethereum. When Binance announced its own chain, there were a number of rumored attempts to strong arm Ethereum developers and projects into switching over and that effort didn’t get anywhere either.
Apologies to the impressive teams and projects at Algorand and NEAR, who just happened to be the most recent examples in the news. There are more egregious examples and most any base layer smart contract blockchain fits some version of this critique.
Developer Network Effects and The Blockchain Trilemma
Much of the interest in these base layer blockchains stems from justifiable concerns about Ethereum’s scalability and uncertainty about the transition to ETH2.0. The competitor base layer smart contract protocols are promising some combination of faster finality or higher throughput.
But protocols are subject to network effects, and protocol network effects are directly correlated to concentrations of developers. As Electric Capital’s blockchain developer data shows, the dominant majority of blockchain developers are working on Ethereum.
Beyond the vibrant developer community, Ethereum has emerging network effects around standards and financial primitives. This combination of network effects creates an opportunity for astronomical value creation on all layers of Ethereum, not just the base layer.
Base Layer Value Relative To Higher Layers
It’s a popular and intuitive hypothesis that the value of the base layer must be larger than the value of the assets secured by the smart contract platform, if only to ensure the security of the assets.
Placeholder VC just proposed a new metric to track this called “Network Value-to-Token Value (NVTVRatio)”.
If this intuitive hypothesis is true, the largest networks of value must be stored on the highest value blockchains, which would deepen existing network effects and power laws.
Power Law Curves and Pareto Distributions
The Fat Protocol thesis was correct in that massive value would accrue to the base layer assets of the winning protocols in a way that was never possible for HTTP and SMTP. But power laws and emerging value capture mechanisms ensure that the fat protocol thesis cannot be applied linearly.
There won’t be hundreds of vibrant developer communities, and in fact they tend to consolidate around the top few platforms or protocols. PalmOS, Blackberry and Windows Mobile found out the hard way what it meant to finish outside of the top 2 in the mobile OS wars. To use a protocol analogy, Superhuman is the most successful of the many and varied attempts to “reinvent email”, but no one tried to replace SMTP.
We’re still learning about true value capture in blockchain networks, but we’re also starting to see real signs of Ethereum being used as financial collateral in the higher layers. Perhaps we should consider that the upper layers on Ethereum will accrue more value than the base layers of even some (most?) of the larger eventually successful protocols. While this chart is an overgeneralization, this is what a Pareto distribution for protocol layer value capture might look like.
If the above assumptions are directionally accurate, then all layers of Ethereum are underrated right now. Per Placeholder’s NVTVRatio, EOS is trading at 234X while ETH is trading around 2X.
While it’s true that Ethereum will never scale to handle the kind of massive throughput common in legacy high frequency trading applications, that may not be necessary in a crypto-native financial system.
Legitimate scaling technology (Starkware) and legitimate scaling hacks via product and UX (Burner Wallet) are emerging such that even ETH 1.0 looks like it will scale sufficiently to address global financial use cases. With multiple network effects compounding every day, Ethereum “just has to work” to absorb a massive chunk of financial value, and “it won’t scale” has historically been a bad bet.
(Note: There are some other wild card factors, like Cosmos and Polkadot, that I’m leaving out of the discussion for now)