Ethereum first went live in July 2015, and its ecosystem of decentralized applications (dApps) has advanced wonderfully over the years. 0x’s first OTC markets were launched in 2017. Uniswap was first deployed in November 2018. Yearn Finance orchestrated its spectacular airdrop in 2020. Decentralized Finance (DeFi) was an ever-evolving industry whose growth, in retrospect, seemed almost inevitable.
As it is with any economy with competing products, protocols rose and fell in dominance, relentlessly innovating to form the DeFi landscape we see today. The history of DeFi might be subjective, but the Ethereum blockchain is not. In this research piece, we tell the story of Ethereum, and the growth of such protocols, through the data. Here, we plot total gas fees spent on Ethereum, aggregated by 4-week periods, from 2018 till today.
Ethereum activity was rather quiet from 2018-2019. Total gas spent barely went above 40,000 ETH per month. Activity began to pick up in 2020, and gas spending saw a parabolic run that peaked at 650,000 ETH in 4 weeks nearing September of 2020. That was when Ethereum was still $400 apiece. Time really does fly.
We give the data a closer look by analyzing the percentage of total gas fees spent by various protocol entities. In this article, we want to examine the historical proportion of gas spent by the top 30 gas spender entities through time.
Gas fees spent over time are a function of many things: how many users are interacting with a smart contract, the computational intensity of the function executed, and overall gas prices at that time.
Before we dive in: we must note that the landscape of Ethereum in from 2018-2019 was very different. A large number of contracts active in 2018-2019 are no longer used today. We can sort the top 30 gas spenders for the period of 2018-2019 here to take a closer look.
Other than entities, I also aggregated the total gas spent by token contracts in this dataset, a group which consistently took up around 10% of Ethereum’s activity. Many of the contracts in 2018 were ponzi schemes and gambling games, with a notable proportion of them originating from China. LastWinner, for example, was based on this simple mechanic: Users deposit ETH into the contract until a certain ETH cap is reached. The last person who deposits ETH then wins the entire pot of ETH. You can view it's spike in activity in the middle of 2018
Here’s what’s even more interesting: The relative proportion of gas spent for our top 30 entities today, since 2018.
The data tells an intriguing story from year to year. Let's begin with 2018.
0x takes out the middleman. Many people think Uniswap was the first swap DEX that was born into existence, but the history of trustless exchange using smart contracts went much further back. Initial DEX models in 2017 attempted to mimic the order book models of centralized exchanges, but they were so computationally intensive and slow that they were rendered unusable. 0x deployed a method for on-chain order settlement of trades in July 2017.
It was called 0x OTC, because the model was quite literally that. When people trade on OTC markets, 1. Price discovery occurs between only two parties 2. Negotiation occurs without guaranteed offers (I could rescind anytime, but on an open market, an order I place can be immediately taken up). Users would post their orders off-chain on social media platforms like twitter, but have them settled on-chain through 0x. In 2018, 0x was doing 4 million USD in transactions daily.
Bancor was the go-to DEX. The idea of having Automated Market Makers was mentioned by Bancor in a blog post in August 2018, and flipped the idea of order book style markets completely on its head. All tokens are paired against BNT, a practice that continues until today. Uniswap, by contrast, allows all sorts of pairs to be construed.
Kyber makes its mark. Kyber essentially facilitates parties to access and contribute liquidity in a decentralized manner. It’s not a DEX per se, but rather focused on integrating pools of capital across various sources, including both DEXes and centralized market makers. Think of it like a generalized Uniswap router! This liquidity can then be utilized by entities such as payment networks, which was truly groundbreaking at the time.
Chainlink grows to be the trunk of DeFi. First introduced in 2017, Chainlink’s oracles connected external, off-chain data to Ethereum smart contracts securely. People don’t realize how essential oracles were to synthetic and margin products. BZx used Chainlink oracles for its margin trading platform, while Synthetix immediately integrated with Chainlink to provide price feeds for real world assets. Decrypt even published a canonical list of Chainlink integrations in May.
An alternative oracle service was Tellor, which had a momentary surge in usage in October. Up till 2021, Chainlink’s share of Ethereum gas fees has consistently remained above 4%. Talk about a winner.
Advanced trading features sprout. Dydx, which initially hosted lending and borrowing markets, began developing a margin trading feature for up to 4x leverage with a sleek UI. Nothing really came closer to a CEX experience.
Synthetix finds its own niche. Very few today know that Synthetix was originally known as Havven, and was in fact designed to be a stablecoin protocol akin to Maker. You can almost tell from the gas charts how dominant Maker was throughout the years. Synthetix pivoted to create marketplaces for synthetic crypto and inverse value assets, hitting 60 Million USD of volume by July.
Interest in compound begins to compound -- Compound V2 was launched in May with a fresh new look. Very little has changed with Compound’s money markets over the years, but its consistent gas throughput (1.5-5%) is a testament to the longevity of its design. Not surprisingly, you’ll find that lending protocols incur lower gas fees than DEXes over the years
Uniswap dominates the AMM market. Uniswap was first launched in November 2018, but officially overtook Bancor in trading volume in February 2019. While both DEXes subscribed to the model of 50/50 reserves, Uniswap’s design was much more gas efficient and user friendly. Uniswap’s design also enabled permissionless listings of assets, making it much more composable with the larger DeFi ecosystem. A quote from a blog post in 2019 is telling:
Our token was listed on Bancor for several months before we switched to Uniswap. The process for listing a token on Bancor required us to contact the Bancor team and work with them to transfer ETH and equivalent amount of our token to addresses provided by the Bancor team. The Bancor team also required us transfer no less then $60,000 USD of ETH during the setup to be used to provide liquidity. The process took a day or two of going back and forth with the Bancor team from the time we decided to list to the time to our token appeared on the Bancor site. In contrast the process for creating a Uniswap contract was as simple as filling out a short form and clicking a button. The process for add liquidity to the contract was just as easy. The process took a minute or two and we did not need to contact the Uniswap team. Nor was there any requirements for how much liquidity to add to the contract.
1inch enters with a bang. Perhaps the fastest growing DeFi protocol in 2020, 1inch barely raised its seed round in August before taking up as much as 6% of the gas market in November. It offered a DEX aggregation service that splits and routes orders through various liquidity pools to find the most cost-efficient trade. 1inch ended the year with a retroactive airdrop to users, hitting 10% gas dominance in December.
Forsage spikes in activity -- which pops quickly. This was an odd one to look at. Forsage apparently asked users to pay ETH to use their platform, and promised to pay them ETH for every person they referred. Talk about pyramid schemes. The project still seems to be running today.
DeFi users feel the Yearn. No history of Ethereum would be complete without giving a nod to Yearn Finance's airdrop of YFI. This "valueless governance token" multiplied by 35 times in price within 7 days. Since 2020, Yearn has been at the forefront of DeFi, and the structure and form of its growth is a guiding light for new protocols today.
Tether and Centre represent a surprisingly large proportion of ethereum activity -- they account for almost 12% of gas fees spent today. This is a good proxy for the volume of assets being transferred into and out of Ethereum, as gas is used to mint and burn USDT and USDC.
The Wrapped Ether contracts remain widely used -- the backbone of Ethereum DeFi -- continues to be widely used. WETH was a pretty revolutionary idea at that time. Tokenized Ethereum could be used as collateral, a means for the exchange, and a pricing benchmark for other tokens. WETH is an example of a DeFi lego that fills a need, is widely used, and mostly stable. And no, there’s no governance token for that!
A living, breathing ecosystem of products form: An array of swap protocols now own their fair share of Ethereum’s activity -- though Uniswap still dominates. Nansen alone tracks at least 94 DEX protocols today -- each with their own quirks and value proposition. More than 2.8 Million contracts have been deployed since the start of this year. And around 198 thousand ERC20 contracts have been deployed to date.
You contributed to this chart. Every swap, stake, deposit, withdrawal, mint you have made to date is recorded on the Ethereum blockchain. You are a participant in the forest of Ethereum, which continues to survive, grow, adapt, and thrive. Where do you think we're headed to next?