The next Ethereum upgrade still doesn’t have a mainnet date. The date that’s been circulating in Chinese-language crypto communities turns out to be for the testnet. After developers raised it during a meeting, the plan was pushed again to the next meeting. Countdown posters and posts about “the last chance to get on before the upgrade” ran much faster than the schedule itself.
Ethereum’s core developers maintain a fork registry. In the row for Glamsterdam, the fields for activated block, timestamp, and epoch are all blank; the status is still marked as planned; and it doesn’t list which EIPs will ultimately be included. In the same repository, the mainnet upgrade and emergency response plan also has empty date tables, and the client team’s contact list is blank too. A fork time for the Sepolia testnet was proposed for September 28. No one objected at the meeting, but it stopped at the “proposed” stage. Treating it as the upgrade date is equivalent to taking developers’ draft notes and mistaking them for the actual schedule.
This matters because, in August, $ETH delivered Ethereum’s best month of the year. If you calculate using Binance’s spot monthly chart, August rose by a little over 30%, and the ETH/BTC price relationship has been climbing for a second straight month. The market attributes this rally to three things: ETF inflows, upgrade expectations, and a macro shift. Of the three “legs,” only one holds up under scrutiny.
The ETF leg is real—but the size isn’t as thick as rumor suggested, and it only just reversed. On September 3, the U.S. spot Ethereum ETF saw net outflows of $48.07 million, ending more than two weeks of net outflow. On the same day, the spot Bitcoin ETF recorded net inflows of $101 million; the two split for the first time. Looking one layer deeper, BlackRock’s Ethereum product with staking had net inflows of $52.91 million that day—the single biggest source sucking in the money. The money is still in Ethereum; it just moved from a non-yield “shell” to one that can capture staking rewards. This kind of reshuffling provides far less marginal upward pressure on price than genuinely new capital would.
The upgrade leg is the one that can’t stand up to close examination. Glamsterdam’s current main thrust is actually the act of changing how blocks are produced. ePBS writes the relationship between proposers and builders into the protocol, and in doing so reduces the MEV fee-cut space. Block-level access lists let execution run in parallel, and there’s also a round of gas repricing. The beneficiaries are block producers, stakers, and L2s that consume throughput. But between them and what ETH as an asset should be worth, there are multiple layers of transmission.
And those middle layers are precisely what’s leaking. In Q2, Ethereum mainnet captured only 4.9% of the economic value created by the applications running on top of it; the rest remained with L2s and the applications themselves. Standard Chartered estimated that just the fee diversion from Base alone is equivalent to extracting several tens of billions of dollars from Ethereum’s market value. Matthew Sigel of VanEck has long held a negative view on the long-term value prospects of L2 tokens. That objection flips the other way just as well—once activity moves to rollups, how much can the base asset still receive? ePBS and parallel execution don’t solve this problem: they make the network run faster, but they don’t make ETH collect more money. #Ethereum
The macro leg, if anything, has been underestimated. The direct trigger for the September 3 green candle was a Waller comment: inflation can take a bit more time, and this meeting doesn’t need to move interest rates. The odds of a September rate hike dropped sharply that day; Treasury yields eased, and both gold and U.S. stocks strengthened together. ETH rose a bit more than BTC that day, but the direction was the same. Once risk appetite returned, assets with higher elasticity jumped first—that’s beta, and it has little to do with what Ethereum itself did to “create” that move.
The most credible bullish voice on the long side is Tom Lee. He set an end-of-year target above $5,000, and he gave specific reasons: stablecoin supply scale, tokenized assets, corporates putting ETH on their balance sheets, and the completion of regulatory groundwork. In that logic, there’s not a single reliance on Glamsterdam. Anyone who is genuinely bullish on Ethereum isn’t betting on that upgrade date.
In August’s rally, macro beta did most of the work. A real but not large institutional buy order did some of the remaining part. Upgrade expectations mostly only contributed sentiment—and that sentiment was built on a date that doesn’t even exist. There’s one hard piece of evidence on the long side that I need to make clear: during the period when prices rose, the open interest in perpetual futures didn’t keep stacking up; the funding rate stayed near the benchmark line. Price was bought up gradually in spot rather than via much leverage. This kind of rally is sturdier than a short squeeze, and it’s less likely to trigger a chain of liquidations on a pullback. The August gains themselves hold up; the market just gave credit to the wrong place.
There are two scenarios that would make my explanation fail. In the rest of September, if the ETF keeps seeing net outflows while the ETH/BTC ratio still climbs, that would mean there’s buying demand in pricing that I haven’t accounted for, and I would need to reduce the weight I assigned to the institutional capital leg. The other scenario is if once the Glamsterdam mainnet date is set, the ETH/BTC ratio breaks out into an independent trend—that would mean I underestimated the upgrade.
Compared with the testnet time on September 28, September 15 is the one that should stay on the calendar. The procedural voting schedule in the Senate for the CLARITY Act is set for that afternoon. You need to gather 60 votes to move it forward, and the current number of seats held by Republicans isn’t enough—you have to pull votes from the other side. The outcome of that vote directly determines the rules under which the stablecoins and tokenized assets on Ethereum will be run next year, and it’s much closer to how block production is scheduled in the protocol than how it relates to ETH pricing. If you want a time point that can help verify who’s right—bulls or bears—you can start by looking at that day.
Ethereum’s core developers maintain a fork registry. In the row for Glamsterdam, the fields for activated block, timestamp, and epoch are all blank; the status is still marked as planned; and it doesn’t list which EIPs will ultimately be included. In the same repository, the mainnet upgrade and emergency response plan also has empty date tables, and the client team’s contact list is blank too. A fork time for the Sepolia testnet was proposed for September 28. No one objected at the meeting, but it stopped at the “proposed” stage. Treating it as the upgrade date is equivalent to taking developers’ draft notes and mistaking them for the actual schedule.
This matters because, in August, $ETH delivered Ethereum’s best month of the year. If you calculate using Binance’s spot monthly chart, August rose by a little over 30%, and the ETH/BTC price relationship has been climbing for a second straight month. The market attributes this rally to three things: ETF inflows, upgrade expectations, and a macro shift. Of the three “legs,” only one holds up under scrutiny.
The ETF leg is real—but the size isn’t as thick as rumor suggested, and it only just reversed. On September 3, the U.S. spot Ethereum ETF saw net outflows of $48.07 million, ending more than two weeks of net outflow. On the same day, the spot Bitcoin ETF recorded net inflows of $101 million; the two split for the first time. Looking one layer deeper, BlackRock’s Ethereum product with staking had net inflows of $52.91 million that day—the single biggest source sucking in the money. The money is still in Ethereum; it just moved from a non-yield “shell” to one that can capture staking rewards. This kind of reshuffling provides far less marginal upward pressure on price than genuinely new capital would.
The upgrade leg is the one that can’t stand up to close examination. Glamsterdam’s current main thrust is actually the act of changing how blocks are produced. ePBS writes the relationship between proposers and builders into the protocol, and in doing so reduces the MEV fee-cut space. Block-level access lists let execution run in parallel, and there’s also a round of gas repricing. The beneficiaries are block producers, stakers, and L2s that consume throughput. But between them and what ETH as an asset should be worth, there are multiple layers of transmission.
And those middle layers are precisely what’s leaking. In Q2, Ethereum mainnet captured only 4.9% of the economic value created by the applications running on top of it; the rest remained with L2s and the applications themselves. Standard Chartered estimated that just the fee diversion from Base alone is equivalent to extracting several tens of billions of dollars from Ethereum’s market value. Matthew Sigel of VanEck has long held a negative view on the long-term value prospects of L2 tokens. That objection flips the other way just as well—once activity moves to rollups, how much can the base asset still receive? ePBS and parallel execution don’t solve this problem: they make the network run faster, but they don’t make ETH collect more money. #Ethereum
The macro leg, if anything, has been underestimated. The direct trigger for the September 3 green candle was a Waller comment: inflation can take a bit more time, and this meeting doesn’t need to move interest rates. The odds of a September rate hike dropped sharply that day; Treasury yields eased, and both gold and U.S. stocks strengthened together. ETH rose a bit more than BTC that day, but the direction was the same. Once risk appetite returned, assets with higher elasticity jumped first—that’s beta, and it has little to do with what Ethereum itself did to “create” that move.
The most credible bullish voice on the long side is Tom Lee. He set an end-of-year target above $5,000, and he gave specific reasons: stablecoin supply scale, tokenized assets, corporates putting ETH on their balance sheets, and the completion of regulatory groundwork. In that logic, there’s not a single reliance on Glamsterdam. Anyone who is genuinely bullish on Ethereum isn’t betting on that upgrade date.
In August’s rally, macro beta did most of the work. A real but not large institutional buy order did some of the remaining part. Upgrade expectations mostly only contributed sentiment—and that sentiment was built on a date that doesn’t even exist. There’s one hard piece of evidence on the long side that I need to make clear: during the period when prices rose, the open interest in perpetual futures didn’t keep stacking up; the funding rate stayed near the benchmark line. Price was bought up gradually in spot rather than via much leverage. This kind of rally is sturdier than a short squeeze, and it’s less likely to trigger a chain of liquidations on a pullback. The August gains themselves hold up; the market just gave credit to the wrong place.
There are two scenarios that would make my explanation fail. In the rest of September, if the ETF keeps seeing net outflows while the ETH/BTC ratio still climbs, that would mean there’s buying demand in pricing that I haven’t accounted for, and I would need to reduce the weight I assigned to the institutional capital leg. The other scenario is if once the Glamsterdam mainnet date is set, the ETH/BTC ratio breaks out into an independent trend—that would mean I underestimated the upgrade.
Compared with the testnet time on September 28, September 15 is the one that should stay on the calendar. The procedural voting schedule in the Senate for the CLARITY Act is set for that afternoon. You need to gather 60 votes to move it forward, and the current number of seats held by Republicans isn’t enough—you have to pull votes from the other side. The outcome of that vote directly determines the rules under which the stablecoins and tokenized assets on Ethereum will be run next year, and it’s much closer to how block production is scheduled in the protocol than how it relates to ETH pricing. If you want a time point that can help verify who’s right—bulls or bears—you can start by looking at that day.