Ether.fi launched a bank this week, but the bigger story is what it quietly gave up.

Two days ago, this column published a countdown: ether.fi predicted that something big would happen on August 13, and its tokens were the only one that kept rising in the market that was suffering a sharp decline at that time. Now, the mystery has been revealed as scheduled: a bank. Tokenized stocks, metal trading, portfolio loans realized through Aave, fiat channels covering more than 30 currencies. Good news. Arriving on time is a rarity in this industry. But as everyone was reading the press release, few noticed another page-ether.fi's own documentation that showed it was moving away from the core business that made it famous.

Actual released content

According to the company's official announcement, the "Summer" version will be officially launched on August 13, 2026, covering web, iOS and Android platforms. The feature list is quite impressive: support for tokenized stock and metal trading (parallel to cryptocurrencies); integration of the Aave Marketplace on Optimism, allowing users to borrow against their entire investment portfolio; access to more than thirty currencies through channels such as Apple Pay and Cash App; and programmatic ETHFI repurchases built into the app.

CEO Mike Silagadze positions it as a bridge connecting decentralized finance with daily financial needs. In the company's own words, the app has been deliberately rebuilt to "weaken the tendency to cryptocurrency" to appeal to a wider audience. It should be noted that trading in tokenized stocks and metals is not yet open in the United States and some other markets-a detail that is often ignored, but is crucial for readers in Ohio.

In terms of underlying architecture, ether.fi Cash migrated its credit backend to an Aave V4-based deployment on the OP main network, replacing the internal system with a target capacity of hundreds of millions of dollars. This is a real infrastructure upgrade, not a marketing gimmick.

So: They promised big things and fulfilled big things. This is worthy of recognition.

A key number

is less than 1%.

According to ether.fi's own forfeiture risk documentation, this is the proportion of assets it is currently re-pledged on EigenLayer, compared with about half of the proportion at the beginning of 2026. The same document shows that the remaining share will fall to zero in the third quarter of 2026, and the company plans to remove EigenPod withdrawal vouchers from its validators by the fourth quarter, completely severing its final structural association with EigenLayer.

Do you understand what this number means? ether.fi was once the largest business established on EigenLayer's re-pledge model. The core selling point for the entire 2024 is that one token bundles both Ethereum pledge income and re-pledge exposure. It is this combination that allows the agreement to grow rapidly. Last week, it separated the two: weETH is now just an ordinary liquidity pledge token, and any user who really wants to repledge must switch to another token built based on Symbiotic-weETHs.

Before the announcement was released, the chain had already begun to shrink. As of this writing, there are no blog posts explaining the decision, and some ether.fi documentation still describes weETH as automatically re-pledging on EigenLayer. When a company changes its core philosophy when it was founded and updates the document first and then the external narrative, the document itself becomes the new core narrative.

This number is the shadow behind the release of the new bank.

Now, let's take a look at what background it is, making this consumer application release look more like a strategic transformation rather than a purely product roadmap update.

According to DefiLlama data, ether.fi's pledge business holds approximately US$3.3 billion and remains the largest liquidity re-pledge agreement, ranking third overall, behind Lido and Binance Staked ETH. The numbers look good.

But it peaked in August 2025, when it reached US$12.43 billion.

This means a decline of about 73% in one year. The core business that started the company has shrunk by nearly three-quarters. This week, the company launched a bank. These two facts should be placed in the same paragraph, and I have not seen such statements anywhere else.

A well-intentioned interpretation is a reasonable starting point: a management team, watching yields compress and re-pledge demand fade, took the initiative to build a second business based on existing card products and thirty currency channels, catching up with the first before a business becomes a problem. It's a sign of ability, and it's harder than it sounds. Many agreements choose to die out in a fading narrative rather than acknowledge reality.

But the skeptical interpretation is equally worthy of attention: consumer fintech is the most difficult market in the world, and ether.fi now has to deal with real banks and brokerages, not other DeFi protocols; and its products cannot enter the world's largest consumer market due to geographical limitations. It's easy to announce a bank, but it's extremely difficult to run.

A lingering threat

Ethereum has a draft, called EIP-8363, sometimes referred to as the "Tapered Issue Burn," that destroys some verifier rewards when pledge rates rise, which could lead to net pledge yields approaching zero at high participation rates. The original text and discussion of the draft are disclosed in the Ethereum EIPs repository.

With this background in mind, let's take a look at this strategic transformation. If the pledge yield is structurally compressed, then a company with pledge income as its core product must urgently find new products. There is no information in the public record to suggest that the release of the "Summer" version was due to EIP-8363, and this article does not intend to make such an assertion. But at the same time that Verifier Economics is facing structural draft proposals, a company is embarking on diversification, which is no coincidence that can be easily ignored.

Repurchase Program and Dependency Conditions

ether.fi DAO has authorized the use of agreement revenue to conduct ETHFI repurchases of up to US$50 million on the open market when the token price falls below US$3, and the "Summer" version integrates these repurchase functions directly into the application. Compared with its low market value of hundreds of millions of dollars, this authorization is huge in relative proportion and far more aggressive than the annual market value of approximately 1.2% of Chainlink's reserve repurchase program.

But the same caveat still applies, and now it is even more critical: authorization is just a ceiling, not a timetable. The repurchase funds come from agreement income, which previously mainly came from a pledge business that had just shrunk by 73%, and the replacement income came from consumer-grade applications that were launched only yesterday. Please pay attention to the actual execution status on the chain through Etherscan, rather than the authorization limit in the title.

Summary

ether.fi delivered on the promised date and released a truly ambitious product. At the same time, it has quietly ended the core arrangements that made the company, watching its core business fall from US$12.4 billion to US$3.3 billion. Now it is betting its future on retail banking and competing with a global market that excludes U.S. users. Both stories are true. But only one of them appeared in the official announcement. If you are trading this token, then "bank" is the front page and "exit" is the core logic. The latter, on the other hand, will continue to play a role for a long time after the candle line on release day disappears.