Share articles to

Academy Industry Analysis Article
Blockchain DeFi Ethereum Liquidity mining OKX Insights

Ethereum fees explained: How rising gas prices impact UX and adoption

2021.03.01 Rick Delaney

An in-depth look at Ethereum’s transaction fee structure and how rising transaction costs impact the decentralized application ecosystem 

Ethereum gas prices have been a bone of contention for many in the cryptocurrency industry for years now. During times of high network demand, the cost to transact on Ethereum prices out all but the wealthiest users. Meanwhile, the network supports an increasingly exotic collection of decentralized applications, extending Ethereum’s appeal to a wider user base. 

In the following article, OKX Insights puts Ethereum’s gas problem in the spotlight. We begin by introducing the role ETH, the network’s native coin, serves on Ethereum and how the network prices transactions. We then consider the cause of gas price spikes before turning our attention to efforts addressing the degradation of user experiences resulting from high transaction fees. Finally, we conclude with a discussion about the impact rising gas prices have on the adoption of Ethereum and competing projects. 

How ETH fuels the Ethereum network

To understand Ethereum’s gas problem, it’s first important to appreciate how the network functions and the purpose ETH serves. Like Bitcoin, Ethereum is a distributed network upon which value can be transacted. However, Ethereum introduces additional complexity by also facilitating the execution of arbitrary code. Such code can be used to write smart contracts — i.e., deterministic, self-executing digital agreements between parties. Smart contracts form the basis of unstoppable and permissionless decentralized applications, commonly called DApps.

At the heart of Ethereum is the Ethereum Virtual Machine. You can think of the EVM as being like a single, real-world computer — only it has no physical presence, and no central authority oversees or controls it. A distributed network of nodes stores a constantly updating copy of the EVM. These nodes are responsible for maintaining the state of this emulated, shared “world computer.”

Users may initiate changes to the EVM’s state by making transaction requests, which vary from simply transferring ETH between accounts to deploying or interacting with smart contracts. Nodes broadcast transaction requests to the network — which, providing certain parameters are met, are grouped into blocks by miners and validators. These network participants verify transactions, execute any necessary computations and share the new state of the network for the rest of the nodes to store.

Performing operations on the EVM requires the expenditure of computational resources. Users include fees, paid in ETH, with their transaction requests to incentivize validators to commit the resources required to process them. The miner or validator that processes a specific transaction receives the fees associated with it, as well as the fees from every other transaction included in that block and a reward of new ETH created by the network. The term used to measure the computational effort required to perform a function on Ethereum is gas

A closer look at Ethereum’s gas

Ethereum’s transaction fees are paid using the network’s native currency, ETH. Gas is denominated in gwei, which is 1 billionth of 1 ETH. Gwei is actually short for gigawei, meaning 1 billion wei, the name given to the smallest unit of ETH, or 0.000000000000000001 ETH — like a satoshi in relation to BTC.

The Ethereum Yellow Paper outlines fixed gas costs for different computations. The more complex a process, the more gas it costs. For example, a simple ETH transaction between two users requires at least 21,000 gas. Meanwhile, the creation of a smart contract requires a minimum of 32,000 gas. It’s important to note that transactions — particularly those interacting with DApps — can require multiple computations. This significantly increases the total gas required.

Transaction requests also require a gas limit, which is the maximum amount of gas that the user is willing to spend on a transaction. If the limit specified is below the gas required to complete the request, the transaction will fail due to insufficient gas. For example, if the transaction was a simple ETH transfer with a gas limit of less than 21,000 gas, the ETH would remain at the original address, but the miner that attempted to process it would still receive the gas included.

Most users choose to transact with a wallet that automatically sets the gas limit based on the complexity of the transaction created. Examples of such wallets include MyEtherWallet and MetaMask.

Ethereum’s fee market

To ensure that there is sufficient time to propagate new blocks to the entire network, there is a limit on the amount of computation that can be performed within each block. This is known as the block gas limit. The current block gas limit is 12.5 million — meaning each block can process computations requiring a combined expenditure of up to 12.5 million gas. The limit has been increased several times over Ethereum’s history, most recently in June 2020.

While the amount of gas required per computation is constant, the cost of each unit of gas varies according to the demand for block space at the time. Users must set a maximum gas price that they’re willing to pay. Websites like ETHGasStation.info and CoinGecko provide current recommended gas prices, quoted in gwei, for different transaction speeds. A standard transaction speed, usually settled in five minutes or less, will cost less than a faster one. Again, many wallets determine the maximum gas price on behalf of their users.  

The order in which the network prioritizes transactions is determined by this fee market. A user who wants the network to process their transaction ahead of others must create an economic incentive for validators to do so. Miners or stakers attempt to form a block using the combination of transactions that provides themselves the greatest financial reward. This means that transactions with lower gas prices will take longer to complete than those offering more ETH to the network.

It’s up to the user to determine the urgency of their transaction. For example, a user wanting to catch a specific price for an ERC-20 token at a decentralized exchange like Uniswap would be advised to choose a high gas price relative to the network average. Meanwhile, a user moving funds to a hardware wallet for long-term storage might be less concerned about immediacy and may opt for a lower gas price.

Fee market in history

During periods of heightened Ethereum usage, the network’s average cost to transact increases. If validators have a constant pool of high gas price transactions waiting, they rarely need to fill up block space with lower-gas-price requests. This encourages all users to pay higher gas prices or simply not use the network. Such bottlenecks hurt Ethereum’s adoption because certain use cases become less economically viable as the cost to transact rises.

High network activity has created spikes in transaction costs on multiple occasions throughout Ethereum’s history. One of the first notable warnings of the network’s inability to handle high demand came at the end of 2017.

Already seeing heavy usage thanks to the initial coin offering craze, the launch of a simple collectibles game, CryptoKitties, brought the Ethereum network to a temporary standstill. The digital cat breeding application, released on Nov. 28, 2017, created a massive increase in the number of Ethereum transactions immediately following its launch. 

The average cost to transact quickly rose to $1.45. Although tame in fiat terms — particularly when compared to subsequent gas price spikes — the average gas price peaked above 61 gwei in the first week of December 2017 before rising to 95.78 gwei in early January. The incident served to illustrate a transaction fee issue that would grow in significance alongside Ethereum’s own rising adoption and the increasing sophistication of DApps built on the network.

CryptoKitties and ICO speculation pushed gas prices up significantly in 2017. Source: EtherScan

Current state of Ethereum

In the years since CryptoKitties, the Ethereum network has found only greater adoption. While ICO fundraising on the platform has died down, myriad new uses now compete for block space. Among them are stablecoins, nonfungible tokens and, most notably, decentralized finance protocols.

Experiencing dramatic growth during 2020 and into 2021, DeFi has fast become the largest niche creating demand for Ethereum’s block space. At the beginning of last year, there was less than $700 million locked in DeFi applications. By February 2021, the sector was worth more than $40 billion. 

ETHGasStation shows that many of DeFi’s most popular applications are among the network’s biggest consumers of gas. Taking the top spot with a 30-day gas consumption of 1,500 ETH ($2 million) is Uniswap — the most popular Ethereum-based DEX. In spots three and six are fellow decentralized exchanges 1inch Exchange and SushiSwap. Other high gas consumers from DeFi include Aave, Chainlink and dYdX.

Vying for Ethereum block space with the DeFi sector are stablecoins. From around $2 billion in early 2019, the market capitalization of the dominant stablecoin, USDT, has grown to more than $35 billion today. Similarly, USDC, another popular stablecoin implementation, has expanded from around $250 million to almost $9.1 billion over the same period.

USDT has seen dramatic market cap growth over recent years. Source: CoinGecko

With most USDT and all USDC issued on Ethereum, the two stablecoins are consistently among the top gas users on the network. As of Feb. 16, 2021, USDT transactions generated around 10,000 ETH ($12.6 million) in gas over the previous 30-day period. Meanwhile, USDC accounted for 1,480 ETH ($1.9 million) in gas.

Finally, despite CryptoKitties’ central role during the late-2017 gas price spike, NFTs are also putting pressure on Ethereum gas prices. Data from Nonfungible.com, cited by Decrypt, shows a 2,800% increase in NFT art sales in 2020. Previously, the site provided analysis on how spikes in gas prices impact the NFT space. Its report noted that rising gas prices coincided with a decline in user activity on some NFT apps — particularly with regard to lower value digital items.

While Ethereum gas issues might discourage some users, NFTs are attracting much wider attention. From the NBA and other sports franchises to Doctor Who, numerous mainstream digital collectible games have launched in recent months. 

Additionally, Christie’s auction house is helping to legitimize NFTs as a medium for artwork. The 255-year-old institution will be using the Ethereum-based MakersPlace to conduct its first-ever strictly digital art auction. Such multidirectional expansion of the NFT space suggests demand for Ethereum block space from the sector is likely to continue growing.   

Each new use case and new user it attracts to Ethereum puts pressure on gas prices. Data from EtherScan shows the average gas price has been steadily increasing in recent years, with a few significant spikes along the way. 

Most notable are the spikes during mid-2020 — a period known in the industry as the “DeFi Summer.” By September, DeFi usage and the yield-farming craze had sent the average gas price as high as 538 gwei. More recently, the average price has dropped slightly. However, many users still complain about the cost to transact on Ethereum. 

Ethereum’s gas price has trended up as adoption spreads. Source: EtherScan

How gas spikes impact Ethereum-based projects

The increased cost to transact during periods of high network demand significantly impacts Ethereum’s overall utility. 

It is something of a paradox that a completely permissionless network effectively excludes participants based on the cost of using the network. Can DeFi really claim to be democratizing the financial industry if only high-net-worth users can justify the steadily increasing cost of using the applications themselves? Similarly, high transaction costs diminish some of the advantages that stablecoins have over legacy payment systems, damaging their own appeal.  

Faced with a degrading user experience, those building on Ethereum have three options:

  • Do nothing and hope that users find enough value in the project to justify paying high fees.
  • Migrate the project to a less congested blockchain with smart contract functionality.
  • Build on Ethereum’s Layer 2 scaling solutions.

Unsurprisingly, many projects are exploring the second and third options.

Migrating from Ethereum

Ethereum’s scaling issues have forced some projects to launch on other blockchains — among them is NBA Top Shot. 

Behind the digital collectibles game is Dapper Labs, the creator of CryptoKitties. Evidently learning from its early foray into the NFT space, the company decided to launch NBA Top Shot on its own custom blockchain implementation called Flow.

Ilya Abugov, an advisor at DappRadar, commented to OKX Insights that Dapper Labs’ decision was reflective of an emerging trend within the blockchain gaming sector:

“Gaming has been one big sector that has looked elsewhere because it has been hit by the high gas costs the hardest. So, some like Dapper Labs and Sky Mavis deployed their own chains, others have opted for L2 solutions or one of the Ethereum rivals like BSC or Polkadot.”

Not having to compete with a growing DeFi sector or expanding stablecoin use, lower overall fees have quickly helped NBA Top Shot become the biggest blockchain-based collectible game in terms of total transactions and value of sales. Data from CryptoSlam shows that the game generated over 1.8 million transactions and sales in excess of $267 million since its 2020 launch.

The popularity of the world’s most famous basketball league is certainly responsible for a lot of NBA Top Shot’s use. However, a smoother user experience and reduced transaction costs are much less likely to dissuade frequent users, too. In fact, almost anything blockchain-related — including payment — has been abstracted away behind the scenes, which has helped the platform find mainstream appeal. 

In describing the motive behind creating the Flow blockchain, Roham Gharegozlou, CEO of Dapper Labs, wrote in a Medium post:

“Every other decentralized blockchain’s plans for scaling depend on some kind of sharding or side chains. Unfortunately, our team recognized early on that these approaches make interactivity between smart contracts significantly more complex and error-prone. […] Sharding is a good solution for payments, but not for an ecosystem of complex applications like games or social networks.”

Another example of a project seeing increasing adoption on non-Ethereum blockchains is USDT. Launched as Realcoin on Bitcoin’s Omni Layer Protocol in 2014, the rebranded Tether stablecoin arrived on Ethereum in September 2017. Implementations of USDT were later launched on TRON and other blockchain networks.

Recent weeks have seen a major increase in the usage of TRON-based USDT at the expense of Ethereum-based USDT. CoinMetrics data shows that there were 101,340 active USDTe addresses on Jan. 1, 2021. By Feb. 13, this number had fallen by 18.75% to 82,330. Over the same period, USDTt active addresses increased by 141.2%. At the beginning of 2021, there were 75,940 USDTt addresses, compared to the 183,210 addresses reported in mid-February.

Growth of USDTt active addresses versus declining USDTe active addresses. Source: CoinMetrics

According to Tether’s transparency report, USDT issuance on TRON is rapidly catching up to issuance on Ethereum. There are now almost 12.5 billion USDTt in circulation, compared to 19.6 billion USDTe. At just 6.6 billion in late-December, 2020, it took less than two months for the circulating supply of USDTt to almost double.  

Speaking with CoinDesk, Nate Maddrey, an analyst at CoinMetrics, attributed the growth of TRON-based USDT to differences in the cost to transact. “As fees go up, [Tether] transactions tend to shift from Ethereum to Tron,” he told the media outlet.

Interestingly, the total value of transactions of Ethereum-based USDT continues to outweigh those on TRON. Meanwhile, the actual number of transactions has often been greater on TRON. This supports Maddrey’s view that those using USDT to make lower value transactions are increasingly favoring TRON-based USDT for its cheaper transaction costs. 

Are “Ethereum Killers” benefiting from ETH’s gas issue?

In addition to those projects moving chains, non-Ethereum blockchains are increasingly supporting similar DApps to those found on Ethereum. The hopes of projects like Avalanche’s Pangolin DEX and Polkadot’s synthetic asset exchange Laminar are that the lower-cost transactions on their respective networks will encourage the migration of those Ethereum users put off by gas price spikes.   

There is evidence to suggest that this is the case, too. Launched as recently as Feb. 9, 2021, Pangolin has already attracted around $174 million worth of liquidity and has seen daily trading volumes touching nearly $32 million. The totals might fall well short of Uniswap’s current $4 billion in liquidity and roughly $1 billion in daily volume, but they still remain impressive — considering that the Avalanche mainnet only went live in September 2020.     

Meanwhile, Polkadot’s Moonbeam is one of several direct attempts to lure Ethereum developers to alternate networks. Moonbeam is a parachain implementation that provides those building on Polkadot familiar tools to deploy smart contracts via a full EVM implementation. 

Still not quite out of its testnet phase, the project has already been announcing plans to work with multiple Ethereum-based DApps since it was conceived in April 2020. Such efforts may help to eventually dethrone Ethereum as the dominant smart contract platform as users seek lower fees elsewhere.  

Exploring Ethereum’s scaling solutions

Meanwhile, for those building DeFi applications, the appeal of remaining on Ethereum is clear. The network offers, by far, the most liquidity. Also, the composability of smart contracts makes fragmenting the sector across multiple chains undesirable. Similarly, Ethereum boasts the largest developer and user communities, as well as the most diverse token economy. For these reasons, Layer 2 offers the only solution worth pursuing, for many.

DappRadar’s Abugov explained to OKX Insights:

“The big advantage is the perception of staying within the Ethereum ecosystem. User base, developer base, network support, and capital are big factors.”

Jesse Johnson, COO at Pixelcraft Studios, told OKX Insights about Aavegotchi’s own decision to build on the newly rebranded Polygon — formerly known as Matic:

“After lots of due diligence, we confirmed that Polygon does in fact offer a Layer-2 experience for Ethereum users that is incredibly user-friendly, fast and virtually free of fees.”

He went on to describe the Layer-2 scaling implementation as being tailored to NFTs and gaming while noting similarities between the current mainchain user experience and Polygon. The project is an effort to scale Ethereum that describes itself as “Ethereum’s Internet of Blockchains.” The project offers a range of Layer-2 scaling solutions — including the previously launched Matic PoS Chain and Plasma Chains — as well as solutions based on zkRollups, Optimistic Rollups and other emerging technologies.

According to the team, Polygon makes Ethereum “a full-fledged multi-chain system.” Comparing the project to multi-chain efforts like Polkadot and Avalanche, Polygon notes that other networks bridging blockchains are not supported by the same network effects, security or flexibility as Ethereum.  

Another major project heading to Layer 2 in search of lower fees is Synthetix. In a 2020 blog post, Kain Warick, the project’s founder, explained why Optimistic Ethereum was chosen. He described other options — such as alternate blockchains, zkRollups, state channels and Plasma — as being ill-suited for a project as complex as Synthetix. The ability to reuse already-tested code and to complete general computation, as well as having an established security model, made the Optimistic Ethereum implementation most appropriate.

The launch of Synthetix on Optimistic Ethereum is a multiphase effort. The first phase saw staking go live in January 2021. A blog post explicitly stated that, while full functionality of Synthetix on Layer 2 is still lacking, the newly launched implementation may be appealing to those holders priced out of staking smaller amounts of SNX on the mainchain by gas prices. 

ETH 2.0 and the future

While Layer-2 solutions and the migration of some projects to other blockchains look set to alleviate some of Ethereum’s scaling woes in the short-term, real improvements to the architecture of the mainchain itself will take longer. With the introduction of sharding, the Ethereum 2.0 rollout should address the network’s capacity issues and, by extension, reduce gas prices. However, the roadmap for Ethereum 2.0’s full launch is expected to extend well into 2022.

If successful, Ethereum 2.0 will be capable of processing many more transactions per second than is currently possible. By “sharding” the network into smaller, semi-independent blockchains, transaction processing will not require every node to verify and store all data. Instead, nodes only need to concern themselves with data from the shard that they validate. Meanwhile, the Beacon Chain will coordinate the activity of shards and serve as a single source of authority for all data saved to the Ethereum 2.0 network.

A multi-chain future?

Although Ethereum has both an increasingly adopted Layer 2 and a plan to overhaul the entire network to optimize for greater throughput, there is no shortage of “Ethereum Killers” waiting to draw users and developers from an overcrowded network. Projects like Polkadot, Cosmos and Avalanche each attempt to provide a multi-chain architecture, alleviating competition for finite block space by linking independent blockchains. Meanwhile, efforts like Polkadot’s Moonbeam will only make it easier for Ethereum developers to apply their craft elsewhere. 

The industry is fairly divided on whether such projects will successfully lure a mass of developers and users from Ethereum. Tushar Jain, a managing partner at Multicoin Capital, is among those who believe eventual DApp users will prioritize good user experience and low fees over any allegiance to a particular platform.

On the contrary, many still believe Ethereum’s first-mover advantage best positions it to remain at the center of the future DApp ecosystem. In conversation with Bankless, Polygon co-founders Mihailo Bjelic and Sandeep Nailwal spoke about the importance of Ethereum’s established and growing network effects. They reason that many developers will find transitioning to Layer-2 solutions like Polygon easier than migrating to an entirely new chain.

Bjelic and Nailwal also spoke about Ethereum’s proven security record. Newer smart contract platforms have not had their security assumptions tested with the same rigor as Ethereum. Indeed, efforts like Polkadot are forced to encourage the formation of their own distributed network with which to provide security. For now, at least, Ethereum is much more resilient to vulnerability or attack based on its longer uptime and its more widely distributed network.    

Equally important to the Polygon team is user perception. Ethereum’s longevity has encouraged greater trust in the network — as evidenced by the ever-growing total value locked in the network’s DeFi sector. Supporting this view in a CoinDesk article is Alex Treece, co-founder of fintech firm Zabo. 

Treece explained the importance of brand recognition when it comes to consumer choice. Comparing Ethereum and Bitcoin to Coca-Cola and Google, he argued that Ethereum’s interconnectedness, brand awareness and “fanatically loyal community” of developers have widened the moat between Ethereum and other blockchain networks.

Similarly doubtful that another chain will usurp Ethereum as the dominant chain for DApps is DappRadar’s Abugov. However, he does believe that we’ll continue to see projects launch on alternate blockchains — particularly as familiarity builds with bridging solutions: 

“Ethereum remains Number One, as the majority of capital, developers and volume are there. So, a lot of innovation, ecosystems, and network effects are on Ethereum. It’s not likely that Ethereum will simply lose its status as the Number One, but rather that the ecosystem will migrate to a multi-chain paradigm where certain chains have specialties or better conditions for certain projects. As interoperability improves, being on the same chain will matter less for projects.” 

Whether Ethereum’s network effects will be enough to stave off competition from the growing list of “Ethereum Killers” remains to be seen. However, increasing interest in Layer-2 technologies certainly demonstrates an allegiance to the network that may prove difficult to break. 

Either way, it seems likely that gas prices will continue to force many established DApps to migrate from the Ethereum mainchain. Whether that migration will be largely toward Layer-2 solutions or alternate blockchains may depend on the nature of the applications themselves. Many of those DeFi applications benefiting from composability and large liquidity pools will likely prefer the familiarity of Ethereum or a Moonbeam-style EVM implementation on another chain. Whereas those use cases that require low-cost transactions but less interconnectivity may well find one of the ever-increasing number of alternate smart contract platforms more suitable. 

OKX Insights presents market analyses, in-depth features, original research & curated news from crypto professionals. 

Follow OKX Insights on Twitter and Telegram.

Disclaimer: This material should not be taken as the basis for making investment decisions, nor be construed as a recommendation to engage in investment transactions. Trading digital assets involve significant risk and can result in the loss of your invested capital. You should ensure that you fully understand the risk involved and take into consideration your level of experience, investment objectives and seek independent financial advice if necessary.