A dive into Ethereum's network stats
Taking from Chris Burniske's recent post on crypto fundamentals (especially ETH's) being less depressed in protportion to ETH's USD denominated value, I did some digging on the fundamentals of ETH beyond Gas and hashrate. In today's version of the Daily, I'll present you with some key metrics on Ethereum and attempt to unpack where the state of the network is at present.
Key Takeaways:
The Ethereum network seems to have reached a capacity ceiling in its current form.
Smart contract usage is correlated to relative transaction cost.
ERC-20 is by far the dominant standard.
Poor quality ERC-20's are starting to go offline. That should allow for some network bandwidth to free up.
High quality smart contract useage is accelerating. That is likely to increase the demand for bandwidth - as organic (non-speculative) demand kicks in.
13% of all available ETH is currently locked in Maker DAO. That's an impressive feat, a sign of progress, and a HUGE central point of failure.
Without a scalability solution, the problems that became evident in 2018, will only keep persisting, as adoption of the platform increases.
Exhibit 1: The network has reached a capacity ceiling
These two plots are directly borrowed from Burniske's post and show the relationships between Market cap (Network Value) and daily transactions and hashrate respectively. Safe to make the case that the network holds its relative fundamental strength, despite the price of ETH collapsing. There are a few cases we can make here. The most interesting ones to me are (i) the fact that mania pushed the valuation much higher than the fundamentals, but more interestingly, (ii) the fact that the market is pricing in Ethereum's scalability woes. Plasma is starting to look like a Catch 22, Sharding is being continuously delayed and the Ethereum community has grown bigger than its current organizational structures can accommodate. Be that as it may, the fact that both hashrate and daily transactions have plateaued in 2018, remains. The next couple of graphs show the effects of the network bottleneck, in action.
The two graphs above are representations of smart contract count and the relative transaction cost (measured as Gas value/ETH value). Just a quick glance in enough to reveal that there is an negative correlation between the two constructs. Low relative transaction costs lead to proliferation of smart contracts deployed on Ethereum. Fair. What is positive here is that despite the windfall in ETH's value, we recently hit a local maximum in smart contract count. The flow of developers on Ethereum is increasing - yet, they are likely to find themselves competing for finite resources. Let's take a look on the demand side of those smart contracts.
Exhibit 2: Smart contract usage is correlated to relative transaction cost
The plots above are different permutations of smart contract calls over time. A smart contract call is similar to an API query. As evidenced by the number of unique callers over time, the Gas crisis this past summer (largely attributatble to Fcoin's introduction of trans fee mining), capped the growth of unique callers. Even as transaction costs normalised, the number of unique callers did not return to its previous trajectory. A look at the number of calls might reveal why;
The capacity ceiling in Ethereum's current form becomes evident once more. Call count has plateaued at 300k for pretty much all of 2018. When calls started reaching 400k, transaction fees blew up and the call count returned to the 300k level. The fact that many useless ERC-20s are starting to die out, some capacity should be freed up. Even yet, were this source of demand to be substituted with real users, we can expect that the same mechanics would come into play.
Exhibit 3: ERC-20 is by far the dominant standard
Recently, the number of daily unique ERC-20's in operation took a turn south. The recent drop in the number of ERC-20's coincides with the recent actions from the SEC. Even so, their overall share of the Ethereum network is so large, that makes the traction of other ERC standards (like the NFT - 721) pale in comparison. Still; in relative terms the signs of life in non-fungibles and others are encouraging.
Interestingly, the number of daily transactions of ERC-20's has been on the decrease since July, coinciding with the aforementioned Gas crisis that took place around that time. Part of this change might be attributable to some of the demand sources moving to other platforms and some dying out due to continuously depreciating prices. But there's potentially one more reason...
Exhibit 4: Open Finance Dapps like Maker are starting to get serious traction
The trend change in daily transactions of ERC-20's, as well as the number of smart contract callers, coincides with an acceleration in the trend of increase of the amount of ETH that is stored in the Maker DAO smart contract, held as collateral to issue the DAI stablecoin. At the moment, approximately 13% of the available supply of ETH is locked into that contract (no wonder a16z piled in ~$25M in Maker), while others smart contracts, such as Compound, have started gaining traction (on display in the second graph above).
Concluding remarks
That final point is perhaps what's more encouraging about the ecosystem's development. More is not always better. The evidence that the number of ERC-20s is starting to drop, while high quality smart contracts are accelerating in useage is a highly positive sign, and a trend that we can only hope persists in 2019. However, without a robust scalability solution, the woes of the past are almost certain to keep coming back as demand picks up - and demand from builders for the Ethereum platform shows no signs of slowing down.
Monthly downloads of the Truffle development framework for Smart Contracts
To top that up, the Maker smart contract is increasingly becoming a bigger and bigger bounty for the one to hack it, and a central point of failure for the whole Ethereum ecosystem. Useful to remember that Solidity (Ethereum's programming language) comes without formal verification, and thus higher probability of bugs appearing down the line. So far, so good - but then again it doesn't matter how you fall - what matters is how you land.