Ethereum just slipped under 10% market dominance for the first time since 2020. The exact number this morning: 9.95%. On that same morning, Ethereum still hosts 52% of all DeFi value locked across crypto. Both numbers are correct. They are simply pricing two different things.
The thesis: in 2026, Ethereum is no longer one asset. It is one ticker covering two markets. The asset market is mediocre. The infrastructure market is dominant. Treating them as the same thing is the most expensive misreading currently happening on crypto Twitter. The panda has been staring at this for an hour and has receipts.
What does "Ethereum's two markets" even mean?
According to CoinGecko's global market data, Ethereum's share of total crypto market capitalisation stood at 9.95% on May 13, 2026, with ETH at $2,300 and a market cap of $277.69 billion. Total crypto market cap was $2.79 trillion. Bitcoin dominance held at 58.29%.
Now look at the same chain through a different lens. According to DefiLlama's chain ranking, total DeFi TVL across all chains stood at $86.65 billion on May 13, 2026, and Ethereum alone hosted $45.51 billion of it. That is roughly 52.5%, a majority share, larger than Solana ($6.19B), BSC ($5.70B), and the rest of the long tail combined.
Two metrics, two stories. Asset dominance says ETH is losing the race for the unit of account. Infrastructure dominance says ETH is winning the race for the layer where capital actually does things. The asset side is voting "underwhelming". The infrastructure side is voting "central". Same chain, two markets.
Mais voilà, this bifurcation is not a new idea. What is new is the magnitude. ETH dominance below 10% with infrastructure share above 50% had not been observed simultaneously since 2020.
The numbers that don't agree with each other
If asset dominance and infrastructure dominance were tightly coupled, they would track each other. They used to. The "ETH is digital oil" framing assumed that more activity on Ethereum meant more demand for ETH, meaning higher price, meaning higher market cap share. Utility, then fee burn, then scarcity, then price, then dominance. A clean chain of logic.
Sauf que the L2 era cut that pipe. According to DefiLlama's Ethereum page, the Ethereum mainnet still holds the lion's share of locked value, but a large portion of "Ethereum-aligned" activity now happens on Arbitrum, Base, Optimism and similar rollups, where fees are paid in ETH at one or two orders of magnitude below mainnet. Fee burn dropped. Issuance held. The deflationary thesis loosened.
The ultrasound money mechanism has been quietly de-rated. ETH the infrastructure still captures value from being the settlement layer for $45 billion in locked capital and a serious chunk of stablecoin flow. But the value capture is no longer a price function. It is a settlement function. Different market, different valuation logic.
The cleanest tape: BTC is up 0.24% in 24 hours, per CoinGecko's Bitcoin page, to $81.20K. ETH is down 0.31% on the same window. Same regime, opposite tape. On asset terms, ETH is a beta that is failing to play.
Counterarguments: maybe asset and infrastructure are still the same thing
Here are the fair objections. None of them are dismissable.
The first is that the bifurcation is a temporary artefact. Maybe ETH's DeFi infrastructure share will eventually drift down toward its asset share. Layer 1 competitors like Solana have grown fast. BSC is quietly accumulating TVL too, +2.18% over seven days. If TVL leaks out of Ethereum at the same rate that dominance has, the "infrastructure dominance" thesis evaporates within 18 months.
The second is that the asset metric is being read wrong. ETH at 9.95% dominance still implies a $277 billion asset. That is not "losing" in any historical sense. It is being measured against a Bitcoin that absorbed a generational ETF flow. Strip the ETF distortion out and ETH's relative position is closer to its 2023 baseline. There is no real decline, only a relative one against a uniquely large Bitcoin event.
The third is existential. If validators do not get paid enough, because L2s siphon fees and burn slows, the security budget of Ethereum mainnet eventually fails. No security budget, no settlement layer. No settlement layer, no DeFi infrastructure. The two-markets frame assumes the infrastructure side is robust forever. It is not. The Ethereum Foundation tracks this risk openly in its ongoing roadmap research.
Objection three is the serious one, and it has no clean rebuttal yet.
What to watch next
Three dated checkpoints turn this from a clever frame into an actual call.
By September 30, 2026, Ethereum's DeFi TVL share should remain above 50% on DefiLlama's chains dashboard. If it drops below 48% for two consecutive weeks, the infrastructure thesis weakens fast. By December 31, 2026, the ETH/BTC ratio, currently near 0.028, should reclaim 0.035 or the asset-side weakness becomes structural rather than cyclical. By March 31, 2027, Ethereum's annualised fee burn should stabilise above $1 billion or the ultrasound mechanism is effectively dead.
Each of these is publicly verifiable. None requires insider information.
If two of three hold, the two-markets thesis survives. ETH becomes investable in two distinct ways: as a modest asset and as an infrastructure proxy. If two of three fail, the bifurcation was slow asset decline dressed up in a smarter narrative. Spoiler: on l'avait vu venir.
Where Dadacoin and BSC sit in this story
The two-markets frame is not just an Ethereum story. It changes how every other chain gets graded. BSC has never had a credible "asset" thesis for BNB beyond exchange utility. But it has a real infrastructure thesis: $5.70 billion in DeFi TVL, +2.18% week over week, deep memecoin liquidity, predictable fees. For tokens that live as infrastructure plays on a credible chain rather than as beta on an L1 narrative, the bifurcation is good news. The bar shifts from "be the next ETH" to "be useful on a chain that settles".
This connects directly to the agentic wallet thesis around ERC-8004: if AI agents end up the dominant settlement users by 2027-2028, both ETH and BSC win as infrastructure regardless of where their dominance metric sits at the end of any given quarter. The parallel argument applies on chain comparison too, as covered in our boring-chain thesis on BSC versus Solana. For more on this lens, the ai-agents topic cluster tracks adjacent developments.
The market still trades Ethereum as one thing. It hasn't been one thing for a while. Pas une opinion, juste l'arithmétique.
Disclaimer: This article is not financial advice. Always do your own research (DYOR) before investing.
Researched and edited by the Dadacoin team. AI-assisted writing, reviewed for accuracy.
Cover photo by Jonathan Borba on Pexels.
