Hook
Ethereum’s average gas fee hit $18.72 in May 2021. The network was congested, NFTs were flipping for millions, and everyone screamed “bullish.” Two years later, those same fee-heavy chains saw daily active addresses drop 40%. The noise stopped. The ledger bleeds faster than the logic holds.
Last week, Ripple CTO David Schwartz did something rare in crypto. He stated publicly that high fees do not equal a healthier network. He said it plainly—no fluff, no agenda. Just a cold, mechanical correction to a widespread cognitive error. I have been watching this mispricing since 2017, when I manually audited ICO contracts and saw teams boast about “demand-driven fees” while their code had integer overflows. Fees are not a sign of vitality. They are a tax on friction.
Context
Schwartz’s comment cuts straight to the core of a systemic blind spot. For years, retail and even some professionals have misinterpreted high transaction costs as a proxy for value. The logic goes: if users are willing to pay $50 to move a token, the network must be valuable. But that logic is a dam with cracks. I count the cracks before the dam breaks.
To understand why, you need to look at the mechanics. A blockchain’s fee market is a function of block space competition. When demand exceeds supply, fees spike. That is not “health”—that is congestion. A healthy network has ample capacity, low latency, and low fees at scale. Bitcoin’s security model depends on fees, but even Satoshi acknowledged that the fee structure would need to scale with adoption, not with artificially created scarcity. Ethereum’s EIP-1559 burned fees, but that burn is a byproduct of activity, not a cause of value.
The irony is that many L1 projects market their high fees as proof of adoption. Solana’s fee spikes during NFT mints were celebrated. BNB Chain’s periodic congestion was called “growth.” But if you strip away the narrative and look at the order flow, the picture changes. High fees repel small users, kill composability, and create an extractive environment for arbitrage bots. I built a custom AI agent in early 2025 to trade options on Lyra. The biggest cost wasn’t the premium—it was the slippage and gas during congestion. I learned that liquidity is just borrowed time with a premium.
Core
Let me dissect the technical argument. A healthy network should be measured by three metrics: throughput, decentralization, and real user growth. Fees are a lagging indicator of stress, not a leading indicator of health.
Throughput: Ethereum processes ~15 TPS during peak. XRP Ledger handles 1,500 TPS at a fraction of a cent. Which one is healthier for global payments? The answer is obvious if you ignore the hype. Decentralization: High fees concentrate power. When it costs $10 to interact with a smart contract, only whales and institutions participate. That creates a feedback loop where developers build for the wealthy, not for the many.
Real user growth: Look at Daily Active Addresses (DAA) per dollar of fees. In 2024, XRP had a DAA-to-fee ratio over 200x higher than Ethereum. That means for every dollar of fees, XRP served 200 times more users. That is not a bug—it is the design.
I saw this firsthand during the 2020 DeFi Summer. I ran a high-frequency arbitrage bot across Uniswap and Sushiswap. Gas costs were so high during the UNI airdrop that my $45,000 profit came from a handful of trades executed with surgical precision. Most retail participants lost money to fees. The network was “busy,” but not healthy. Healthy is when I can execute a trade without paying 30% of the value in gas.
During the 2022 LUNA collapse, the death spiral was accelerated by—guess what—fees. As UST depegged, the arbitrage mechanism required moving large amounts of LUNA. The fee structure couldn’t handle the load, so the chain ground to a halt. I shorted LUNA because I saw the mechanical fragility. High fees do not protect a network; they amplify risk when black swans hit.
Now, Schwartz is not advocating for zero fees. He is simply saying that fee levels are not a reliable signal for network health. This is similar to how traditional finance does not measure a stock exchange’s health by how much it costs to execute a trade. Volume, liquidity, and stability matter. Fees are just a number.
Contrarian
Here is where most analysts get it wrong. They point to Ethereum’s fee burn as a deflationary pressure that accrues value to ETH. That narrative worked in 2021 when fees were high and supply was being destroyed. But in 2025, with L2s handling most transactions and fee revenue dropping, the burn is negligible. The network is still healthy by other metrics—activity on L2s is booming—but the fee-based story is broken.
The contrarian angle: retail traders are still buying into projects that boast “high fees = demand.” They see a chain with $100M in daily fee revenue and think “this must be valuable.” But smart money looks at the cost to serve one user. If a chain extracts $10 per user while another extracts $0.01, which one is more scalable? Which one will capture real-world adoption?
I am not shilling XRP. I don’t hold a bag. I am stating a mechanical truth: high fees are a tax on inefficiency. Ripple’s CTO is saying the quiet part out loud. He is challenging the sacred cow of “fee-driven value.” That is a contrarian position in a market that still rewards narrative over substance.
Consider this: In the 2024 ETF approval era, BlackRock and Fidelity bought Bitcoin at scale. They did not care about Bitcoin’s fees. They cared about liquidity and regulatory clarity. Fees are a noise variable for institutional allocators. They care about total addressable market and real usage.
Takeaway
What does this mean for your portfolio? Stop using fee revenue as a valuation metric. Instead, focus on user growth per fee dollar, network uptime, and developer retention. If a chain has high fees but declining DAA, it is dying. If a chain has low fees and growing DAA, it is scaling.
Watch for the next wave of L1 projects that prioritize efficiency over hype. Ripple’s CTO is not wrong—he is early. The market will eventually price this in. Until then, survival is the only alpha that compounds.
Now, back to scanning the order books. The ledger bleeds faster than the logic holds.
