作者：Odaily 星球日报

This article is from Pine Analytics.

Compiled by Odaily Planet Daily@OdailyChina); Translator | Ethan (@ethanzhang_web 3)

Editor's Note: In the past few years, the crypto market once believed that transaction fee revenue from L1 public chains was the core cash flow supporting token valuations. However, this study uses on-chain data to reveal a different fact: whether it's Bitcoin's congestion cycles, Ethereum's DeFi and NFT booms, or Solana's memecoin frenzy,All the boom in transaction fees will eventually be compressed by innovation.A surge in demand leads to a revenue peak, which in turn stimulates the emergence of alternative solutions.Profits were systematically squeezed outThe compression of L1 value capture is not a periodic phenomenon, but a structural result of open networks.

The market in 2026 has long since moved beyond simply using "fee capture" to price Level 1 tokens. The price drivers for ETH and SOL are shifting from the Level 1 fee logic to staking rewards, ETF fund flows, RWA narratives, protocol upgrade expectations, and the macro liquidity environment. The compression trend remains, but the pricing anchor has moved. What's truly worth considering isn't just whether fees will continue to decline, but rather:When the market stops pricing L1 based on "on-chain profits" and instead uses "asset narrative" and "structured cash flow" for pricing, is this new logic equally fragile?And, when the narrative fades, what fundamental support will prices return to?

In the stage of large-scale development, L1 blockchains struggle to generate consistent and stable revenue from transaction fees. Every major revenue stream they've found—from transaction fees to MEV (Mean Exchange Value)—is eventually eroded by their users through various arbitrage opportunities. This isn't due to any particular chain's shortcomings, but rather an inherent characteristic of open, permissionless networks: once L1 blockchains reach a certain scale of fee revenue, transaction stakeholders will devise new methods to reduce or even eliminate this income.

Bitcoin, Ethereum, and Solana are among the most successful networks in the crypto space. Interestingly, despite processing billions of dollars in value daily, they've followed almost identical paths: transaction fee revenue surges dramatically in the short term, attracting everyone's attention, only to be quickly overtaken by L2 (Layer 2) networks, private order flows, routing tools aware of MEV (Mean Energy Expectation), or new application-layer technologies, stealing the business and revenue. This pattern repeats itself in every transaction fee model, every MEV fluctuation, and every scaling solution in the crypto industry, with no signs of slowing down.

This article argues that the reduction of L1 transaction fees is a long-standing and accelerating trend. It outlines specific innovative strategies that have reduced profits at different stages and discusses their implications for those still struggling to make ends meet.L1 tokens that include the ability to consistently generate revenue from transaction fees in their valuation.What does that really mean?

Bitcoin

Bitcoin transaction fees are almost entirely generated by the congestion during on-chain BTC transfers—everyone is rushing to transfer, naturally driving up fees. Furthermore, because Bitcoin lacks smart contracts, there are virtually no MEVs (Multi-Entity Transaction Vehicles) on the network. The key issue is that each time the price of BTC surges, leading to a spike in fees, the increase in fees is significantly weaker than in the previous cycle, relative to the scale of economic activity at the time.

In 2017, Bitcoin (BTC) surged from $4,000 to $20,000. Average transaction fees also skyrocketed from less than $0.40 to over $50. At its peak on December 22nd, transaction fees accounted for 78% of miners' block rewards: approximately 7,268 BTC in fees alone, nearly four times the block reward. However, in just three months, transaction fees plummeted by 97%, returning to their original levels.

The market reacted exceptionally quickly, and solutions were soon found. At the beginning of 2018, SegWit transactions accounted for only 9%, but by mid-year, this had risen to 36%; although these transactions comprised more than a third of the total transaction volume, they contributed only 16% of transaction fees. Exchanges also began using batch processing, merging hundreds of withdrawals into a single transaction, significantly reducing fees. These factors combined resulted in a 98% reduction in fees within six months. In addition, the Lightning Network officially launched in early 2018, specifically addressing the fee issue for small transactions; Wrapped BTC on other chains also allowed users to hold BTC exposure without having to operate on the Bitcoin mainnet.

By 2021, when BTC prices peaked at $64,000, monthly transaction fee revenue was actually lower than in 2017. At that time, the number of on-chain transactions was lower, but the scale of dollar-denominated transfers was 2.6 times higher than in 2017—simply put,There are more online transfers now, but the transaction fees you can earn haven't kept up, and have even decreased.

The current cycle further illustrates that this trend is unstoppable. BTC rose from $25,000 to over $100,000, an increase of approximately 3 times (the original text states 4 times, slightly adjusted to reflect the actual price range, without changing the overall meaning), but standard transaction fees have not surged as dramatically as in previous cycles. By the end of 2025, daily transaction fees will be only around $300,000, less than 1% of miners' total revenue. In 2024, Bitcoin's total transaction fees reached $922 million, but most of this came from the short-term hype surrounding Ordinals and Runes, not from the stable income generated by traditional BTC transfers. By mid-2025, spot Bitcoin ETFs will hold over 1.29 million BTC, approximately 6% of the total supply, providing the market with massive BTC exposure without generating any on-chain fees. The on-chain interactions required to acquire Bitcoin assets have been largely engineered away.

In April 2024, Ordinals and Runes raised the percentage of transaction fees in miners' revenue to 50%, but as related tools matured, this percentage fell back below 1% by mid-2025. This short-term surge was more like a chance gain from MEV (Metal Equivalents), rather than a stable income generated by congestion. It stemmed more from the immature tool system surrounding the new asset than from a genuine demand for BTC settlement.

The pattern is actually quite obvious:As long as Bitcoin earns enough and conspicuous money from transaction fees, cheaper alternatives will emerge in the ecosystem. L1 can only earn a short-term peak in transaction fees from each demand, and then this profit will be gradually eaten up by continuous innovation.

Ethereum

Ethereum's transaction fee story is even more dramatic. This is because the chain once captured enormous value, only to witness its systematic dismantling.

In mid-2020, the "DeFi Summer" propelled Ethereum to the center of the new financial system. Uniswap's monthly transaction volume surged from $169 million in April to $15 billion in September. TVL grew from less than $1 billion to $15 billion by the end of the year. In September 2020, Ethereum miners' transaction fee revenue reached a record $166 million, six times that of Bitcoin miners. This marked the first time a smart contract platform had earned a consistent and substantial income from real economic activity.

In 2021, NFTs were superimposed on DeFi. Average transaction fees peaked at $53. Quarterly fee revenue grew by 1,777%, from $231 million in Q4 2020 to $4.3 billion in Q4 2021. EIP-1559, implemented in August 2021, introduced a basic fee burning mechanism, permanently removing a portion of fees from the market.At that time, it seemed that Ethereum had really solved the core problem of not being able to make money with L1.

In reality, these fees are essentially "congestion fees": users pay $20 to $50 not because the transaction itself is worth that much, but because everyone is crowding onto the blockchain, exceeding Ethereum's processing capacity of approximately 15 transactions per second (15 TPS). This inherent limitation also leaves ample room for cheaper alternatives.

Other L1 services like Solana, Avalanche, and BNB Chain offer transaction services for just a few cents; Ethereum's L2 Rollups, such as Arbitrum and Optimism, have taken a significant share of the business—they process transactions on their own networks and then send compressed batches of transactions back to the Ethereum mainnet for settlement, making it both fast and cheap.

Subsequently, Ethereum underwent a "self-weakening" process. The Dencun upgrade on March 13, 2024, introduced Blob transactions (EIP-4844), providing a cheaper data publishing path for L2. Prior to this, L2 used calldata, costing approximately $1,000 per megabyte. After the upgrade, Arbitrum's single transaction fee dropped from $0.37 to $0.012; Optimism's from $0.32 to $0.009. The median Blob fee nearly dropped to zero. Ethereum intended to retain users with this, but instead, it weakened its last important source of transaction fee revenue.

Looking at the data makes it even clearer. In 2024, L2 generated $277 million in revenue, but only paid $113 million back to Ethereum. By 2025, L2 revenue had plummeted to $129 million, while only about $10 million flowed back to Ethereum, less than 10% of L2 revenue, a year-on-year decrease of over 90%. L1 transaction fee revenue, which once averaged over $100 million per month, had fallen below $15 million by the fourth quarter of 2025. This chain, which generated $4.3 billion in revenue in a single quarter, had shrunk by approximately 95% in just four years.

Bitcoin revenue was compressed because people could obtain BTC without going on-chain; Ethereum revenue was compressed in two waves.The first wave came from other alternative networks, which lured away users who didn't want to pay high congestion fees; the second wave came from Ethereum's own scaling plan, which reduced the cost of L2 data transmission to almost zero, making it impossible for Ethereum to make money from settlements anymore. Either way, it means that L1 either built or allowed tools that were taking away its revenue to emerge.

Solana

Solana's profit-making logic is completely different from that of Bitcoin and Ethereum.—It earns almost no fees from congestion. The base fee is a fixed 0.000005 SOL per signature, so cheap it's practically negligible. Approximately 95% of its fee revenue comes from priority fees and MEV tips paid through the Jito Block Engine. In the first quarter of 2025, Solana's Real Economic Value (REV) reached $816 million, with 55% coming from MEV tips. In 2024, validators earned approximately $1.2 billion, while operating costs were only about $70 million, indicating a substantial profit margin.

The key to Solana's fee surge was memecoin trading. Pump.fun, launched in January 2024, earned over $600 million in protocol revenue in less than 18 months, at its peak contributing a staggering 99% of memecoin issuance. Its daily DEX trading volume once reached $38 billion. The launch of the Trump token in January 2025 caused daily priority fees to soar to 122,000 SOL, with MEV tips reaching 98,120 SOL. In 2024, the top 1% of memecoin traders contributed $1.358 billion in fees, accounting for nearly 80% of total memecoin fees, almost entirely driven by MEV.

Currently, two types of innovation are squeezing this revenue.

The first category is proprietary AMMs. Protocols like HumidiFi, SolFi, Tessera, ZeroFi, and GoonFi use private vaults managed by professional market makers, quoting prices internally and updating them multiple times per second. Because liquidity is not publicly disclosed, MEV bots cannot engage in squeeze trading. More importantly, proprietary AMMs route orders through aggregators like Jupiter, actively selecting trading counterparties, rather than passively exposing themselves to any participant willing to pay MEV tips, as is the case with public pools. By maintaining pricing privacy and continuous updates, they eliminate the problem of "stale quotes"—the very source of Solana's massive MEV revenue. HumidiFi processed nearly $100 billion in trading volume in its first five months after launch. Today, proprietary AMMs account for over 50% of Solana DEX trading volume, with an even higher share in high-liquidity trading pairs like SOL/USDC.

The second type involves Hyperliquid directly migrating its most profitable spot trading activities off Solana. Leveraging its self-developed HyperCore technology, it created a native bridging tool that allows tokens on Solana to be deposited into Hyperliquid, withdrawn back, and traded on its spot order book. When Pump.fun launched its PUMP token in July 2025, pricing was done on Hyperliquid, not Solana's DEX, using HyperCore for cross-chain bridging. Prior to this, Hyperliquid had already tested this model on SOL and tokens like FARTCOIN—gradually moving these transactions off Solana during the initial price surge, when price spreads were largest, volatility was highest, and MEV gains were most easily earned.

These two approaches have compressed Solana's revenue in two ways: the proprietary AMM has reduced the number of MEV transactions remaining on Solana, while Hyperliquid has moved the most profitable spot trading off-chain. By the second quarter of 2025, Solana's REV had decreased by 54% compared to the previous quarter, leaving only $272 million; daily MEV tips had dropped by more than 90% from the January peak, to less than 10,000 SOL per day.

The underlying principle is the same as the previous two chains; only the way to make money is different.Solana's transaction fees were essentially short-term profits earned through MEV (Mean Equity) during the early, chaotic stages of new trading methods. Once proprietary AMMs optimized trading efficiency and Hyperliquid absorbed high-value orders, these profits quickly dwindled. L1 (Level 1) can generate substantial profits during market euphoria, but the market always finds new ways to prevent such short-term gains from lasting.

Impact on token price

The patterns presented by the three chains mentioned above are not merely hindsight descriptions; they also possess a degree of foresight. Each L1 fee mechanism follows the same trajectory: new demand leads to revenue peaks, these peaks attract innovation, innovation compresses profits, and once this compression occurs, it is difficult to reverse. Following this line of thought, we can make a general assessment of the future of the four tokens.

Ethereum: Continuous Fee Collapse 

Ethereum transaction fees have yet to see a clear bottom. In 2024, L2 fees contributed $113 million to the Ethereum mainnet; by 2025, this had plummeted to approximately $10 million, a drop of over 90%. With each new L2 instance, the demand for Ethereum mainnet block space decreases, and Ethereum's own scaling plans continue to reduce data transmission costs. EIP-4844 is not a one-time repricing, but rather the starting point of a structural shift—Ethereum actively subsidizes infrastructure tools outside its fee market. Currently, monthly L1 fee revenue has fallen below $15 million, and the downward pressure is intensifying. If Ethereum cannot create new sources of native L1 demand, the token price will continue to reflect this compression trend.ETH is increasingly resembling a low-yield infrastructure asset rather than the high-growth smart contract platform it once was. 

Solana: Activity hits record high, price may not be... 

Solana will almost certainly set a new record for on-chain activity in the next cycle—its ecosystem is deep enough, its developers are numerous enough, and its infrastructure is mature enough—but transaction fee revenue may not necessarily increase accordingly. The memecoin craze of late 2024 to early 2025 is equivalent to Bitcoin's "SegWit moment" for Solana: a peak in transaction fees supported by new demand, which was then quickly compressed by innovation.

Currently, proprietary AMMs handle over 50% of DEX trading volume, significantly eroding MEV. Hyperliquid's HyperCore technology is moving the most profitable pricing process off-chain. Even if on-chain activity is 2 to 3 times higher than in January 2025, its fee system is already so mature that it's difficult to convert this increased activity into corresponding validator revenue. Daily MEV tips are currently down over 90% from their peak, but on-chain activity remains healthy.Without sufficient transaction fee revenue to support its valuation, even if Solana's usage rate reaches a new high, it is unlikely that SOL will break through its historical high in the next cycle.

Hyperliquid: The Before and After of Boom and Compression 

Hyperliquid is the most noteworthy case becauseIt represents the next stage of this "making money - being squeezed" cycle.However, the market has not yet realized how the second half of this cycle will unfold.

Hyperliquid is now a leading decentralized exchange for trading perpetual contracts (perps) of traditional financial assets. During the recent peak in silver volatility, the HIP-3-deployed market captured approximately 2% of global silver trading volume, with median spreads for retail-scale trading even outperforming COMEX. At certain times, traditional financial instruments accounted for approximately 30% of the platform's trading volume, with daily notional trading volume exceeding $5 billion. The platform's revenue in 2025 is estimated at approximately $600 million, with 97% used for HYPE buybacks and burns.

We expect Hyperliquid to continue to dominate perpetual contract trading for TradFi assets.Its products do have advantages: commodities and stocks can be traded 24/7, even when traditional markets are closed; new trading markets can be added without approval through the HIP-3 proposal; and it offers leverage of up to 20x on assets for which the CME requires an 18% initial margin. In the next bull market, if trading volume and fees continue to rise, the HYPE token may be repriced, similar to Solana's rebound from its bear market lows. If trading volume in traditional financial assets continues to expand, HYPE will likely follow a similar path. Investors may very well use a single quarter's high revenue to predict its future earning potential.

However, Hyperliquid's pricing model has already sown the seeds of compression.The platform charges takers a notional fee of 4.5 basis points, with discounts of up to 40% based on trading volume and collateral. This is drastically different from the pricing logic of traditional financial derivatives. On CME, the exchange fee for an E-mini S&P 500 contract is approximately $1.33 per side, which is irrelevant to the notional value of the contract (over $275,000) and amounts to less than 0.001 basis points. For a notional position of $10 million: CME fees are approximately $2.50, while Hyperliquid fees are $4,500, a difference of approximately 1,800 times.

This price difference exists because Hyperliquid's current user base is primarily comprised of retail and crypto-native users. However, TradFi perpetual products will bring TradFi's future potential. As trading volume increases and institutional participants enter the market, the pressure to move towards a CME-style economic model will significantly intensify. Hyperliquid's own fee structure has already revealed this direction: the HIP-3 growth model will reduce taker fees in new markets by over 90%, down to as low as 0.0045%; top traders will see rates even below 0.0015%. The protocol is actively pushing for fee reduction. Competitive perpetual DEXs, and traditional exchanges offering on-chain products in the future, will further accelerate this process.Ultimately, there are only two possible outcomes: either Hyperliquid loses trading volume due to its high fees, or it changes its fee structure to a fixed fee model similar to CME. Regardless of the outcome, the high long-term returns currently anticipated by investors are unlikely to materialize, and the HYPE token price may experience a rapid decline.

Bitcoin: Price must rise before transaction fees. 

Of these four types of assets,Bitcoin is unique because the logical relationship between its transaction fees and token price is reversed.For Ethereum, Solana, and Hyperliquid, the logic is: transaction fees generate revenue, revenue supports valuation, and when transaction fees are compressed, the token price will fall. But Bitcoin is different; the logic is the opposite. Miners must rely on a continuous rise in the token price to survive after each block reward halving—because transaction fee revenue has proven insufficient to fill the gap caused by the reduction in block subsidies.

The 2024 halving reduced the block reward from 6.25 BTC to 3.125 BTC, and the daily issuance decreased from 900 BTC to 450 BTC. By the end of 2025, the average daily transaction fee was approximately $300,000, accounting for less than 1% of miners' total revenue. Although Bitcoin's total transaction fee revenue reached $922 million in 2024, most of this came from temporary peaks in Ordinals and Runes, rather than sustainable natural demand for fees. Currently, the contribution of transaction fees is almost negligible, and miners' revenue relies almost entirely on block subsidies, which halve every four years and are denominated in BTC. The only way for miners to remain profitable during a halving cycle is for the USD price of Bitcoin to roughly double within a similar timeframe, thus offsetting the 50% reduction in BTC-denominated revenue. Historically, this condition has held true. However, this foundation is extremely fragile. The chain's security budget is not funded by usage but by a sustained increase in asset prices. If the price of cryptocurrency stops rising after a halving, mining will result in losses, the hashrate will decrease, and network security will be affected. It may even lead to a vicious cycle of "price drop → hashrate decrease → security deterioration → price drop again".

This also makes Bitcoin's "sustainability" more fragile than it appears.The price of Bitcoin can support network security with almost no transaction fees, a mechanism that is difficult for other chains to replicate because Bitcoin is first and foremost a monetary asset, not a smart contract platform.

People buy BTC to hold it, not to use its block space. This gives Bitcoin an advantage that the other three blockchains don't have:The price of a cryptocurrency is driven up by demand from its users, which allows for network security even with low transaction fees. 

However, this also means that Bitcoin's long-term security relies entirely on one assumption—that the price will keep rising, something no one can guarantee. Whether this chain can consistently serve as a secure settlement layer depends not on creating applications that generate transaction fees, but on maintaining a narrative and market environment that encourages people to buy BTC. So far, this model is functioning normally, but whether the block subsidy can be further reduced from 3.125 BTC to 1.5625 BTC and then to 0.78125 BTC, and whether the price can still fill the gap after the next three to four halvings, will be the most critical unknown in the crypto space.