According to Thielen, traders are missing the key point that BTC’s failure to move lower despite the hawkish Fed is itself a bullish signal.

“Traders are underestimating the upside risk from a less hawkish Fed, and overlooking the possibility that the four-year cycle has already bottomed,” he noted.

Other market watchers point to erratic demand for ETFs, a preferred vehicle of institutional investors, as the reason for BTC’s underperformance. These U.S.-listed funds registered an outflow of $61.53 million, snapping an equally anemic three-week streak of inflows, according to data source SoSoValue. This week, they have pulled in $626 million in investor money, the highest tally since early May, but that trend needs to hold.

“Several consecutive days of inflows will be needed to confirm a sustained recovery in institutional demand,” Vikram Subburaj, CEO of India-based FIU-registered Giottus.com, said in an email. For now, he’s watching a tight range, with support near $63,000-$63,400 and resistance between $64,500 and $66,000.

The other problem, according to leading market maker Wintermute, is that whatever ETF flow there has been may not be directionally bullish, but could instead be arbitrage.

“That ETF bid getting absorbed without moving price says the marginal buyer in spot isn’t outright long,” Wintermute noted. “Risk appetite went single-name instead, with ZEC up 10.9% on the week on DCG’s Fortitude expanding its Zcash mining footprint and HYPE adding 5% on a dead beta day. For breadth to expand, we likely need to see BTC vol off this floor first.”

ElizaOS, a token tied to the open-source Eliza AI-agent framework, has plunged to a fresh all-time low after its creator, Shaw Walters, said the asset is effectively “dead” and that the Eliza Foundation is winding down.

CoinGecko data shows ElizaOS trading around $0.000285 at the time of writing, after briefly hitting a record low of $0.000284 on Thursday. The token’s market capitalization was about $2.1 million, following a reported 19% drop over 24 hours.

Key takeaways

• Walters says ElizaOS has no future token support, describing the token as “dead” and stating he no longer owns or backs it.
• The Eliza Foundation is winding down, removing an expected institutional backstop for the token ecosystem.
• A legal dispute is central to the founder’s explanation, with a settlement referenced and parts of the case dismissed in July.
• Eliza’s software work is expected to continue even if the token is abandoned.
• The token’s collapse reverses a major prior run-up, when the project’s earlier version peaked at about $2.5 billion market cap in January 2025.

A sharp reversal after the founder’s “token is dead” message

The latest selloff follows a direct statement from Walters that he views ElizaOS as finished. In a post on X, Walters said, “The token is dead. Completely,” adding that he neither owns nor supports the token. He also indicated that development of the open-source Eliza software would continue independently of both the token and the Eliza Foundation.

For many participants, the announcement marks a rare moment where an AI-agent category token is explicitly disavowed by its founder—rather than merely experiencing a typical liquidity or adoption slowdown. The market reaction was immediate, with ElizaOS falling to a new low on CoinGecko.

It is also a striking reversal compared with the token’s prior peak. CoinGecko shows that before rebranding as ElizaOS, the asset—then known as AI16Z—reached a peak market capitalization of about $2.5 billion in January 2025, underscoring how quickly sentiment can turn when expectations about continuity and ecosystem support collapse.

Legal pressure and “no capital to keep fighting” claims

Walters tied the project’s current state to ongoing legal problems and said the team privately settled with a group of tokenholders represented by Burwick Law. He said the settlement involved giving the represented tokenholders the project’s remaining treasury and funds.

According to Walters, the lawsuit was “ridiculous,” but he argued the project lacked resources to continue defending itself. The dispute, filed in April, named Eliza Labs, Walters, Sebastian Quinn-Watson, and the AI16Z DAO as defendants. The complaint alleged false advertising, deceptive practices, negligent misrepresentation, and unjust enrichment—claims described earlier in reporting by Cointelegraph in connection with the filing.

Court docket information referenced in the article indicates that the named plaintiff’s claims were dismissed with prejudice by stipulation on July 8, while proposed class claims were dismissed without prejudice.

Cointelegraph reported that it reached out to Walters and Burwick Law founder Max Burwick for comment but had not received a response at the time of publication. That leaves an open question for investors: while settlement and dismissals are documented, the broader dispute narrative—particularly whether any additional claims or future litigation could arise—remains less clear from the available details.

What Walters says will (and won’t) happen next

Beyond describing the token as unsupported, Walters said there would be no remaining funds for token buybacks. He also said there would be no foundation and no future supply intervention to support the asset.

Walters further indicated he would not allow another token to be associated with Eliza while he continues building the underlying operating system. In the X post, he wrote that he is “starting over” because he owns the intellectual property, and that he would “never” let a token come close to Eliza again.

For holders, those remarks matter because they point to a fundamental shift in the project’s economic model: instead of token-driven incentives or treasury-backed market measures, the software may proceed as a standalone open-source effort.

ElizaOS itself is described as an open-source framework for building and managing AI agents. The project launched in October 2024 as ai16z with an initial goal of raising $75,000 for what was described as an autonomous investor.

In January 2025, the project rebranded to ElizaOS, according to the article’s account of an earlier development. The change followed concerns raised by Andreessen Horowitz about potential confusion with its “a16z” brand. The token later underwent a migration as well.

All of these steps show how the ecosystem evolved quickly—but the current announcement suggests the latest stage will be different: the token component may be intentionally severed from the broader building effort.

Why this matters for AI-agent token investors

AI-agent tokens have often been marketed around long-term narratives: an ecosystem matures, a product ships, and token utilities follow. In this case, the story is less about technical progress and more about governance, funding, and the legal/branding realities that can decide whether a token survives as an ongoing mechanism.

Walters’ comments also highlight a tension common in crypto projects: even when the underlying software remains open-source and continues, the token can still lose its perceived support structure. With ElizaOS described as having no future foundation intervention, investors are left to reassess what determines value—development alone, market liquidity, and any remaining community coordination—especially when the figure most directly tied to the token’s origin says he will not support it.

Finally, the token’s trajectory—from a reported January 2025 peak market capitalization to a new all-time low—illustrates how quickly the market can reprice perceived credibility and continuity in token-managed ecosystems. The current move may become a reference point for how founders handle legal disputes and whether open-source continuity can compensate for the loss of an active token mandate.

Investors watching ElizaOS next should focus on whether any additional on-chain activity or ecosystem announcements emerge after the winding-down claim—particularly around the settlement outcome, any remaining treasury controls, and how the Eliza software roadmap proceeds without token-linked incentives.

Ether’s holder base is splitting, however. Wallets holding 10,000 to 100,000 ETH have risen from about 14 million ETH in mid-2025 to record highs near 19.6 million now.

The 100,000-plus cohort fell to roughly 2.6 million ETH in mid-2025 before climbing to about 4.6 million by May 2026, an addition CryptoQuant put at roughly 1.8 million. The 1,000 to 10,000 cohort has gone the other way, peaking near 15.6 million ETH in January 2026 and falling to about 12.9 million since.

Bitcoin whales, excluding exchange and mining-pool addresses, bottomed near 2.87 million BTC in December 2025 and hold about 3.06 million now, buying hardest as price fell below $60,000 in June. That remains roughly 170,000 BTC below the 2025 bull-cycle peak near 3.23 million.

CryptoQuant calls this the last stage of the decline while stating plainly that valuation leaves room for one more leg lower before a floor is confirmed.

Ether below cost basis is the thing to watch. It is the only one of the three where the market has already capitulated on paper, and CryptoQuant noted ETH bottomed in early 2025 at a similar level and a similar distance from its lower band.

ZIGChain has secured a strategic investment from Laser Digital to expand institutional-grade investment products onchain.

ZIGChain announced that Laser Digital has made a strategic investment in ZIG and partnered with ZIG Markets. Laser Digital, the digital assets arm of Nomura Group, will help structure products and design risk controls. The ZIGChain Laser Digital partnership aims to bring private credit, PayFi, invoice financing, small-business funding, and stablecoin-based services onchain for institutions and everyday users.

ZIGChain Laser Digital partnership sets governance role

Under the agreement, Laser Digital will provide product structuring support, risk framework design, and governance for a pipeline of ZIG Markets vault products. ZIG Markets serves as the product and access layer within the wider ZIGChain ecosystem. Its role includes finding and developing financial opportunities in regional markets. The partnership focuses on emerging-market origination rather than adding another source of onchain yield, according to the companies’ joint public announcement.

The partners said the model combines ZIG Markets’ regional origination capabilities with Laser Digital’s global asset management experience. Laser Digital will apply institutional risk standards and oversight to products linked to emerging-market private credit. Nomura identifies Laser Digital as its digital asset subsidiary, while ZIGChain presents its network as infrastructure for onchain investment opportunities.

$100 million TVL target for vault products

ZIGChain is targeting at least $100 million in total value locked across the planned institutional-grade vault products. The announcement did not disclose the size of Laser Digital’s investment in ZIG. It also did not provide detailed terms for individual vaults or a timetable for reaching the TVL goal.

The pipeline will cover private credit and related financing products. The companies plan to explore PayFi, financing for small and medium-sized enterprises, invoice factoring, and stablecoin-enabled products. Their stated goal is to make categories that have often been hard to access at scale available through onchain infrastructure with formal risk controls.

Private credit forms central part of plan

The partners described onchain private credit as a leading use case within more than $30 billion of tokenized real-world assets. They said global asset managers have increased activity, but credible institutional origination from emerging markets remains limited. The collaboration seeks to address that gap through regional access and institutional governance.

Abdul Rafay Gadit, co-founder and chief commercial officer of ZIGChain, said onchain finance has attracted capital but still needs greater institutional credibility. “Our partnership with Laser Digital brings institutional governance, product expertise, and global best practices to our offering,” he said. Gadit added that the approach could make the products more accessible to banks, family offices, and other investors.

First product expected in coming months

Dr. Jez Mohideen, co-founder and chief executive of Laser Digital, said execution risk in onchain finance has often been underestimated. He said ZIG Markets brings regional depth and an origination record, while Laser Digital will apply the risk frameworks used across its broader offerings. “The shared vision remains to make the next generation of asset management products accessible to those moving serious institutional capital,” Mohideen said.

The agreement follows ZIGChain partnerships with Beehive, Taurus, and ADI Foundation, along with a growing pipeline of real-world asset products across MENAP and other markets. The first product under the Laser Digital partnership is expected to launch in the coming months. The companies said they will release more details as the product nears completion.

Circle Internet Group said on Aug. 5 that its Coinbase USDC agreement had renewed on existing terms, keeping the stablecoin at the center of the U.S. exchange’s products. 

Circle confirmed the renewal during its second quarter 2026 earnings call.The agreement’s initial term began in August 2023. Circle’s collaboration agreement filed with the Securities and Exchange Commission provides for automatic three year renewals when both companies continue meeting their contractual obligations. The latest renewal therefore extends the arrangement into 2029.

Chief Financial Officer Jeremy Fox Geen also said Circle had no plan to introduce quarterly dividends. Management intends to retain capital for products, infrastructure and other strategic opportunities rather than begin regular shareholder payouts.

Circle and Coinbase preserve their USDC economics

Circle and Coinbase established their current commercial arrangement on Aug. 18, 2023, after closing the Centre Consortium structure that had previously governed USDC. Coinbase received a minority ownership interest in Circle, while Circle assumed sole responsibility for issuing and governing USDC.

Under the filed collaboration agreement, Coinbase supports USDC across its products and receives payments linked mainly to income earned from the assets backing the stablecoin. Circle retains an issuer allocation before the remaining income is divided according to where USDC balances are held. Coinbase also receives part of the reserve income generated by USDC held outside either company’s platform.

The arrangement makes Coinbase more than an exchange listing USDC. Circle reported that 30% of USDC circulation was held on Coinbase’s platform at the end of the second quarter. By comparison, Circle held $12.4 billion, or 17% of circulation, within its own platform infrastructure.

USDC circulation reached $73.3 billion at quarter end, up 19% from a year earlier. Circle’s total revenue and reserve income rose 7% to $701 million, although a lower reserve return rate offset part of the benefit from circulation growth.

New partners can still join the USDC network

The renewed Coinbase USDC agreement does not prevent Circle from signing other distribution arrangements. Circle said it had more than 150 partners with economic incentives to integrate, distribute and support USDC across exchanges, wallets, payment applications and financial platforms.

Circle and Coinbase have also cooperated on third party agreements. Circle moved about $4.4 billion in USDC to a Coinbase linked address through HyperEVM after Coinbase became Hyperliquid’s USDC treasury deployer. Arkham described it as the largest recorded USDC transfer at the time.

Circle later said approximately 90% of Hyperliquid’s USDC was held within Coinbase’s platform at the end of the quarter, while around 10% remained on Circle’s platform. However, management declined to disclose how the three companies divide the associated reserve income.

In related coverage, JPMorgan warned that the Hyperliquid agreement could pressure Circle and Coinbase’s margins. The bank’s estimates are external forecasts and not figures confirmed by either company.

Circle rejects quarterly dividends to fund growth

Fox Geen said Circle wanted to maintain a strong balance sheet that could support investment during different market conditions. He also said retained capital would allow the company to pursue strategic opportunities when they arise.

Circle believes shareholder returns from reinvesting in its platform will be “far greater” than returns from quarterly dividends, according to Fox Geen.

The claim is forward looking and is not a guaranteed return. Product execution, interest rates, regulation, competition and distribution expenses will determine whether the strategy produces the outcome management expects.

The position is consistent with Circle’s latest annual report. The company said it had not declared or paid cash dividends and did not expect to do so in the foreseeable future. Its board may still reconsider the policy based on Circle’s financial condition, capital requirements and other business factors.

For U.S. investors, the decision confirms that Circle continues to present itself as a growth company rather than an income stock. Circle began trading on the New York Stock Exchange in June 2025 after pricing its initial public offering at $31 per share.

U.S. regulation strengthens Circle’s investment case

Circle’s capital allocation decision comes as it invests in regulated U.S. infrastructure. In July, the company received final approval from the Office of the Comptroller of the Currency to establish Circle National Trust.

As previously reported, Circle National Trust will initially provide custody services to Circle and its affiliates. Its approved plan may later support institutional custody and management of assets backing USDC, although Circle has not confirmed when those services will begin.

The OCC is also developing rules to implement the GENIUS Act. Its proposed stablecoin framework covers reserves, redemption, custody, capital, audits, supervision and issuer applications. Circle therefore faces continued compliance spending as the federal system moves from legislation toward operating rules.

Meanwhile, distribution remains one of Circle’s largest expenses. The company recorded $330.6 million in Coinbase related distribution costs during the first quarter of 2026, compared with $303.2 million a year earlier. Circle said these expenses could increase as reserve income grows and it adds more distribution partners.

Circle reported $460.6 million in distribution and transaction costs during the fourth quarter against $733.4 million in reserve income. Those figures show why investors monitor not only USDC circulation, but also how much reserve income Circle retains after paying major partners.

What happens next

The renewal removes an immediate source of uncertainty surrounding Circle’s most important commercial relationship. Investors will now watch future SEC filings for any additional disclosure about the renewed term, distribution costs or changes to the agreement.

Circle’s dividend policy could also change later, but only through a future board decision. For now, management plans to direct available capital toward USDC distribution, regulated infrastructure, payments and new products rather than quarterly payouts.

South Korean crypto exchange Upbit opened trading for Cap’s CAP token on Aug. 6 across its Korean won, Bitcoin and Tether markets. 

According to the exchange’s official CAP listing notice, trading was scheduled to begin at 14:00 Korea Standard Time. Deposits and withdrawals were expected to open within two hours of the announcement.

The listing gives CAP direct access to Upbit’s won market about five weeks after Bithumb introduced its own CAP/KRW pair. At the time of research, CoinGecko data showed CAP near $0.02735, up about 15.3% over 24 hours, with a market value around $42.5 million. The data showed stronger activity around the announcement but did not establish how much of the move came from Upbit’s decision.

GSR increased Bitcoin’s allocation in its Core3 model portfolio on Aug. 5 as trading activity slowed and volatility eased across Bitcoin, Ether and Solana.

The latest Core3 model portfolio assigned 44.1% to Ether, 36.5% to Solana and 19.3% to Bitcoin. The weights total 99.9% because GSR rounds each allocation.

Despite Bitcoin receiving the smallest weight, it remained the strongest of the three assets in 2026. BTC had lost 24.82% year to date. Ether was down 35.49%, while Solana recorded the deepest decline at 40.21%.

Over one year, Bitcoin fell 47.08%, compared with losses of 44.73% for Ether and 54.89% for Solana. The figures show broad weakness across the three assets rather than a decline isolated to one blockchain.

GSR Core3 model shifts toward Bitcoin

GSR said market conditions remained subdued during the latest week. Price changes were modest, while trading activity and volatility declined. The firm interpreted those conditions as a quieter market without a strong directional trend.

As a result, the model raised its Bitcoin allocation and cut its Ether position. GSR attributed the change to proprietary signals rather than recent price performance alone.

The shift is clearer when compared with GSR’s July 15 allocation. At that point, the model held 53.1% in Ether, 37.6% in Solana and only 9.2% in Bitcoin. Bitcoin’s weight therefore rose 10.1 percentage points by Aug. 5, while Ether’s fell nine points.

Ether still delivered the strongest return over the most recent 30 days, gaining 5.16%. Bitcoin rose 1.26%, while Solana lost 9.64%. However, GSR said the lower Ether weighting indicated “relatively stronger forward looking opportunities elsewhere.”

That assessment is a model based view, not a confirmed prediction of future returns. GSR said the portfolio uses quantitative signals that can change as prices, volume and volatility move.

Solana leads losses despite falling volatility

Solana’s 40.21% year to date decline was the largest among the Core3 assets. It also lost 54.89% over one year, compared with the portfolio’s 57.78% decline.

However, SOL produced the best weekly result, rising 1.86%. Bitcoin gained 1.19%, while Ether fell 0.54%. The mixed figures support GSR’s description of a market with limited short term direction.

Solana’s measured volatility also eased sharply. Its 30 day volatility stood at 37.39%, below Ether’s 41.69%, although still above Bitcoin’s 29.89%. Solana’s 60 day reading remained higher at 54.92%, showing that its calmer recent trading followed a more unstable period.

Solana previously approached support near $60 after whale selling, weaker decentralized finance activity and market liquidations weighed on the asset. The network’s planned upgrades continued to support its longer term development case, but they had not prevented sharp token losses.

The broader weakness also extended beyond SOL. As previously reported, the crypto market excluding Bitcoin and Ether lost almost 23% during the first half of 2026. The decline occurred even as some blockchain networks continued recording strong usage.

Core3 trails a basket weighted equally

The Core3 portfolio lost 37.86% year to date and 57.78% over one year. An alternative portfolio allocating equal amounts to Bitcoin, Ether and Solana declined 33.99% and 49.84% over the same periods.

Core3 therefore trailed the equally weighted basket by 3.87 percentage points in 2026 and 7.94 points over one year. Its larger exposure to Ether and Solana increased losses when those assets fell more sharply than Bitcoin.

The result also shows the difference between a changing allocation model and a passive basket. The model attempts to adjust exposure based on GSR’s signals. An equally weighted portfolio simply maintains broadly similar exposure to all three assets.

Earlier allocations show that those signals can change rapidly. On July 8, GSR held 46.7% in Ether, 40.1% in Solana and 13.1% in Bitcoin after Ether led weekly performance and volatility declined.

Ether has struggled against Bitcoin during 2026, with the ETH to BTC ratio reaching multiyear lows. Institutional demand concentrated in Bitcoin, competition from Solana and questions over Ether’s value capture have weighed on its relative performance.

What traders will watch next

GSR’s next weekly allocation will show whether the model continues moving toward Bitcoin or reverses the shift. Trading volume, relative momentum and changes in volatility will remain central to that decision.

The model’s 30 day volatility stood at 38%, compared with 35.87% for the equally weighted basket. Its 60 day volatility was slightly lower than the comparison portfolio, at 42.64% versus 43.69%.

Investors should also distinguish the model results from returns available through a live investment strategy. GSR said its figures are hypothetical, exclude transaction and management fees and do not include staking rewards.

The firm also said the material is intended for professional investors and does not constitute investment advice. GSR may trade the assets for its own account, take positions that differ from its published commentary and sponsor products using related methods.

Those disclosures matter because the portfolio’s allocations are not neutral market forecasts. They reflect a proprietary framework whose positions and past returns may not translate into future performance.

Mysten Labs cofounder and chief technology officer Sam Blackshear said on Aug. 6 that he is leaving the Sui developer to join Anthropic for defensive security research. 

Blackshear announced the move in posts on X, saying the role would let him return to hands on technical work while studying a new problem area. He did not disclose his starting date, reporting line or specific projects at Anthropic.

Blackshear said he was motivated by a period in which the “balance of power between attackers and defenders is shifting.” The comment is his assessment, not a measured industry result. However, Anthropic’s recent research supports the broader concern. The company said Claude Opus 4.6 could identify high severity software vulnerabilities at scale, while its Frontier Red Team studies how advanced models could strengthen defense and offensive activity.

Blackshear will remain involved with Mysten and Sui

Blackshear said he would remain a close adviser to Mysten Labs and respond to Sui ecosystem builders who seek his help. Mysten’s leadership page identified him as a cofounder and CTO before the announcement, alongside cofounder and chief executive Evan Cheng. Blackshear said Cheng will now set the company’s technical vision, although Mysten had not published a separate succession statement.

The departing executive credited Cheng with recruiting many of Mysten’s strongest technical employees and noted his previous leadership experience at Apple and Meta. Blackshear also thanked Mysten Labs staff and the wider Sui community. His advisory role may provide continuity, but the company has not disclosed whether it will appoint another CTO or redistribute his duties.

Mysten was founded by former Meta engineers and became the original contributor to the Sui layer one blockchain.

Mysten Labs partnered with Google Cloud to use generative AI in code development and auditing. The collaboration included tools trained on Move and an AI based auditing system designed to identify vulnerabilities in Move, Rust, TypeScript and Solidity code. That work placed artificial intelligence and blockchain security within Mysten’s strategy before Blackshear joined Anthropic.

Move Foundation plans remain at an early stage

Blackshear created Move while working at Meta and has spent more than eight years developing the language. Sui Foundation materials describe Move as a smart contract language designed around ownership, scarcity and safer handling of digital assets. Its resource model aims to prevent certain programming errors, while Sui’s version adds an object centered structure and programmable transaction blocks.

Blackshear said a Move Foundation is “in the works” and that he hopes to remain involved. The statement is forward looking. He did not provide a launch date, governance structure, leadership team or funding plan. It is also unclear whether the proposed foundation would oversee the language across several blockchains or focus on narrower technical standards.

A separate foundation could provide a neutral forum for language development, documentation and governance. Still, no official organization or timetable accompanied Blackshear’s departure, so its eventual scope remains uncertain.

In related coverage, the $260 million Cetus Protocol exploit showed that a secure programming language cannot remove every application level risk. Researchers said the affected contracts executed permitted instructions, while the failure involved pricing and liquidity controls above the language layer. The incident showed why smart contract safety also depends on protocol design, monitoring and operational response.

Anthropic is expanding defensive security research

Blackshear joins Anthropic as the company increases research into AI enabled cyber threats. Anthropic reported in June that it studied 832 accounts linked to malicious activity and mapped their behavior across the MITRE ATT&CK framework. It said it banned those accounts and found AI assistance across reconnaissance, malware development, evasion and other attack stages.

Anthropic is also testing defensive uses of its models. Project Glasswing brought together about 50 partners to search critical software for vulnerabilities, while separate work with Pacific Northwest National Laboratory used Claude in simulated attacks on a water treatment system. These company reported results do not prove defenders will consistently stay ahead, but they show the research field Blackshear is entering.

Blackshear’s background in programming languages, static analysis and automated bug finding fits that work. Before Mysten, he worked on Move and related security tools at Meta. Sui’s official archive identifies him as Move’s creator and a former principal engineer at the company.

What happens next depends on announcements from Anthropic and Mysten Labs. Anthropic has not detailed Blackshear’s team or research agenda. Mysten has not confirmed a replacement CTO. Developers will also watch whether the proposed Move Foundation receives formal leadership, funding and a launch schedule.