By Rustam Botashev, CFA, a Lead Analyst (r.botashev@hashcib.com), and Mikhail Butov, an Analyst (m.butov@hashcib.com), at HASH CIB
We recently published our report on Ethereum and ether analyzing network’s current state and potential development in respect to ETH’s future. Unfortunately, the report is unavailable to general public. If you are a financial/digital asset markets professional, you can fill out the form on our website with actual (and verifiable) information to request free-of-charge access. Below we do not give financial advice, but only present part of our findings for informational purposes only. Any substantiated critique or commentary is welcomed. Thank you.
Expanding utility demand for ETH is the only source for Ethereum’s long-term sustainability. The recent price hike was not caused by such a demand from actual users. ICOs are gone, and even were new offerings to emerge, they would not be accompanied by the same level of hype around the asset. ETH utility usage is weak and unlikely to strengthen unless the blockchain becomes fast enough for dApps to expand massively. But network scalability is unlikely before end-2021, at the earliest, and competitors might deliver scalable, decentralized solutions, creating the risk that Ethereum turns into a niche player. Time is working against the blockchain, which is slow to upgrade due to its distributed governance.
Ethereum’s main purpose is to be a platform for decentralized applications and protocols. As such, we see dApps/protocols as the only meaningful source for ETH demand. Even though dApps massively live on Ethereum, only a few are active. We analyzed on-chain data for the network’s 1,814 dApps/protocols consisting of 7,082 smart contracts. Out of those applications existing on Ethereum, only 448 can be considered semi-active, with at least ninety transactions in the last three-month period (one per day on average). The total number of dApp transactions over a 24-hour period is just 100K on Ethereum.
The reason for such low activity is unacceptable network throughput, which slows the adoption and expansion of applications, in our view. The blockchain’s users and investors will remember how Cryptokitties clogged Ethereum in December 2017. Released on November 23, 2017, in a matter of ten days this very simple dApp took over a third of blockchain’s total gas consumption, making the mempool size jump six-fold to 11K, utilization rate reach almost 100%, gas price and average transaction cost leap four-fold. Just two protocols, Cryptokitties and ForkDelta, consumed as much as 45% of all gas used on December 4 and 5. This example shows how vulnerable the network is to sudden jumps in usage. This Sword of Damocles still hangs over the Ethereum ecosystem, as the blockchain is no faster now.
The existing theoretical throughput ceiling of 27 transactions per second (currently ~7 tx/s) is simply not enough to deploy a full-scale dApp on Ethereum. Even with currently negligible usage and low dApp adoption, Ethereum operates at 80% of capacity. The number of transactions waiting to be executed is 10K. In February 2019, when the average time per block jumped from 14 to 20 seconds due to the delayed Constantinople hard fork, the utilization rate hit 96%, and the mempool jumped to 47K. Had the fork been delayed anew, the network would have likely jammed again.
Top 10 dApps/contracts consumed around 17% of all gas limit YTD. Even excluding smart contract creation and unidentified contracts, this number remains a still significant 14%. Given the current utilization rate of 80%, if these dApps doubled in size, they would consume practically all spare capacity. If they grew seven-fold, they would require the full capacity of the entire network. Provided that combined number of daily active accounts of these dApps is less than 10,000, this implies that mass adoption with tens of thousands, let alone millions, of daily users is impossible on Ethereum, even theoretically.
Currently, only dApps/protocols that do not require active mainnet usage and high throughput are able to thrive on Ethereum. So-called Open Finance or Decentralized Finance (deFi) applications need Ethereum, as it offers the highest possible security, a diverse community, and one of the widest ecosystems.
Stablecoins on the blockchain are mushrooming. The blockchain has become the host for $761mn worth of stablecoins, as of April 5, 2019. And this amount does not include Tether, the largest stablecoin by market cap ($2bn), which is now supported by Ethereum for both USD and EUR backed coins. If stablecoins are adopted as massively as anticipated, they will provide a sustainable demand for ether, as their transfer requires gas payments. This demand should be sufficient for the network to stay alive.
However, the blockchain’s limited speed doesn’t allow many other applications to succeed. Decentralized exchanges (DEX) are in the worst position, as they need both scale and decentralization. Without decentralization, they are not DEXes. Without speed, trading on them is unappealing and risky (due to the risks from front-running). DEX transactions are also heavy and more expensive. On average each DEX transaction currently consumes 100–120K of gas, compared to 70–90K for financial or gambling dApps. At the same time, exchanges hungry for speed need their transactions executed faster, and therefore offer higher gas prices. If for whatever reason demand for gas increases (for instance, Cryptokitties become a craze again), decentralized exchanges will be the first to suffer, since they cannot afford delays in executions.
We have already seen the first signs of dApps and protocols migrating from Ethereum. Aragon is considering launching its framework on Polkadot, while reiterating that it would deploy its dApp on Ethereum as well. Binance is moving away from Ethereum as it prepares to launch DEX. Such a disappointing development occurs even though Ethereum has no secure decentralized competitors — imagine what would happen were there one. Although several blockchains have failed to become “Ethereum killers”, Ethereum’s only chance is that its competitors once again fail to deliver on their promises to solve key industry issue, or that they fail to build out their communities and ecosystems. Time is working against Ethereum, which due to its distributed governance has been too slow to update.
Ethereum 2.0 is expected to solve the scalability issue by totally replacing the existing blockchain. The transition to Ethereum 2.0 consists of six phases, with only Phase 0 having precise specification and Phase 1’s specification now in active development. From Phase 2 onwards, the transition process is more about wishes than engineering or even R&D. Given the magnitude of the task, the history of postponed upgrades and Ethereum’s track record of missed expectations on deadlines, Ethereum 2.0 is unlikely to kick off before end-2021.
The planned intermediate short-term Ethereum upgrade is crucial for long-term viability but looks insufficient. Ethereum 1x is the upgrade intended to enhance the current network while Ethereum 2.0 is being developed. Introducing fees for smart contracts to reside on the blockchain (state rent) is one of the main proposed features. This should improve throughput since state size is a bottleneck. Gas limit could be potentially increased to 40 million from the current level of 8 million. However, this five-fold theoretical speed enhancement is not what dApps require for mass adoption.
Low iteration speed and a lack of coordination between smart-contract developers make a long road for Ethereum1x. State rent is а double-edged sword, as it is likely to be seriously disruptive for existing applications built on state-free models and would require rewriting almost all smart-contracts. To mitigate this risk and ease the burden of coordination, the intermediate solution, stateless client, is proposed. Implementation of state rent and stateless client can take at least four hard-forks or three years given the current pace of nine months per a hard-fork. The Ethererum 1x proponents hope for one year deployment if a three-month hard-fork pace is adopted.
Ethereum’s governance, or lack thereof, is to blame for slow development and missed expectations. The blockchain was launched and initially supported by a team exclusively focused on it. Since then the development has been spread between several different groups, and we have witnessed a slowdown in upgrades. It took Ethereum one year from the ICO to mainnet. From first production release (Frontier) to the second (Homestead) it took nine months. From the second production release to the third (Metropolis), it took three years. Without judging whether decentralized governance is good or bad in principle, it does not seem to be working well in Ethereum’s case. This is unsurprising, as successful open source distributed management is very difficult: we can name only a few successes, Linux most outstanding among them.
The Ethereum Foundation (EF) seems to favor a scientific approach to research over an engineering one. This is probably why the development process looks chaotic, with noise and abandoned proposals. Generating and exploring ideas receive more attention than executing them. There is no leader responsible for the day-to-day management needed to implement those ideas. This is not necessarily a bad thing if one has bottomless pockets and is trying to develop antigravity, but it is not very efficient if the goal is to build a space ship ahead of your competitors.
The Foundation’s pockets are not that deep anymore, currently having only 22% (~$90mn) of its initial ether holdings while the proposed changes suggest its main expenses still lie ahead. We analyzed Ethereum wallet on-chain data and wonder how much money the Foundation has really spent.
We would welcome more transparency from the Foundation, given the institution’s nature and the project’s philosophy about openness. We have Vitalik’s post that EF possessed $15mn in fiat, 800,000 ETH and 500 BTC as of June 14, 2017. If we multiply the number of ETH transferred from the Foundation and the corresponding ether price on each of the transfer dates, the transferred sum would equal $25mn up to the day of this post. Which means that EF might have spent $10mn in less than a year (from ICO to mid-2017). Since June 14, 2017, EF has transferred from its wallet $101mn worth of ETH. This sum overshadows the officially disclosed amount of $19mn grants issued to date. Unfortunately, it is hard to actually trace which of the transactions were the grants; neither has the Foundation ever disclosed any other of its transfers.
Overall EF has sent $126mn worth of ETH to just 18 different wallets, with three of them receiving $118.5mn. With the help of Bloxy Coinpath technology (Ethereum on-chain analytics tool) we tracked these addresses to the second hop and think that at least two of them immediately sold the received ethers. We identified the second-hop wallet that obtained the largest amount ($70.5mn in just one tranche on January 6, 2018) as an OTC market. The second-largest amount of $27.5mn resided on DEX Kraken, coming in two shots in May and June 2017. Hopefully, most of the cash is in fiat now on EF banking accounts.
All in all, we think that Ethereum is at risk of turning into a niche player for applications not requiring high speed and active gas usage (e.g. stable coins or decentralized lending). The gas limit is almost reached, intermediate solutions (Ethereum 1x) are not coming soon, and Ethereum 2.0, if any, is a few years ahead due to mismanagement and not the biggest funding in the industry. Meanwhile competitors may offer both scalable and secure solutions, dApps/protocols may continue emigrating, taking customers along, and developers might follow the suit. History has witnessed the disappearance of many pioneers before. Having created smart contracts for blockchains, Ethereum made a huge contribution to the industry and now can rest on its achievements. In other words: The Moor has done his work, the Moor may go.
