The US Treasury Department Meets with Three Crypto Companies to Discuss Bitcoin Reserve Custody Solution

Latest data from Bitget shows that the platform’s TradFi (Traditional Finance) contract trading volume surpassed $100 billion in July, hitting a single-day peak of $10 billion, reflecting surging global user demand for trading traditional financial assets in crypto derivatives markets. As a core growth engine of the TradFi segment, Bitget’s stock contracts have exhibited strong market penetration. Currently, roughly one in three contract trades on the platform come from stock contracts. As of press time, Bitget supports 272 popular U.S. stock and ETF contract underlyings.

On-chain analyst Murphy states that all BTC positions bought in 2025 are currently underwater for their holders. Thus, aside from wallet migrations, the reduction in the size of 2025 BTC holdings most likely signals holders selling at a loss. Data shows that as of now, roughly 4.77 million BTC positions bought in 2025 remain, down 41.5% from their December peak. The downward trend of these holdings has unfolded in two phases: a sharp decline before February this year, followed by a significant slowdown in the reduction rate after February, though a steady downward slope persists. Murphy notes that 2025 BTC holdings represent the largest potential supply side in the current market. By contrast, BTC holdings accumulated in 2024, 2023, and 2022—still holding unrealized gains—have largely completed the release of high-level trapped positions, with their decline slope gradually flattening, indicating easing selling pressure from long-term holders. Historical data shows that during the bottom of the past two bear markets, high-level holdings saw notable declines: at the 2022 bear market bottom, holdings bought at the 2021 peak fell by ~51%; at the 2018 bear market bottom, 2017 peak holdings dropped by ~62%. Drawing on historical cycles, Murphy estimates that during the current bear market bottom, the decline in 2025 peak holdings could hit 50%-60%, with the current 41.5% drop leaving room for further liquidation. However, this assessment does not account for BTC purchased by institutions including spot ETFs and MicroStrategy; these holdings are mostly locked in long-term positions, which could reduce actual market supply pressure.

Ethereum Layer2 network Robinhood Chain has recorded abnormal user growth recently, with daily active addresses surging from a previous normal level of around 280,000 to 1.9 million on August 11, and further exceeding 5.2 million on August 12, drawing widespread market attention. On-chain activity indicates this growth is primarily driven by three applications: Uniswap (trading, liquidity pools, and Poolstrade launchpad), StonkPit (stock token trading), and OpenSea (NFT trading), with Uniswap accounting for a large share of the activity. Robinhood Chain integrated Uniswap as its core AMM liquidity infrastructure at launch, and also supports OpenSea trading of stock tokens, NFTs, and community tokens. Robinhood’s key advantage is its native user base: through Robinhood Wallet integration, low-barrier experience, and potential incentive mechanisms, the platform can rapidly convert traditional finance users to on-chain users. When paired with meme trends, new product launches, or ecosystem subsidies, it can easily generate short-term explosive growth. That said, on-chain daily active address data requires cautious interpretation. Such a large short-term surge typically includes a significant number of bot addresses, wash trading accounts, incentive farming participants, and low-quality interaction addresses, meaning the actual number of valid users may be far lower than the reported figure. The market will closely monitor the sustainability of Robinhood Chain’s user growth moving forward. If activity is mainly fueled by meme speculation, short-term events, and gas subsidies, user data could see a sharp decline once the hype subsides.

According to GMGN market data, the stock-themed meme token UTILITY on Binance Wallet has seen its market cap drop back to $7.3 million, with its 24-hour price surge narrowing to 18.5x and 24-hour trading volume hitting $20.3 million. UTILITY’s market cap once briefly surpassed $10 million last night. Earlier, on January 30, CZ posted that GME should issue a utility token on the blockchain, preferably on BSC. Today, bStocks’ official account reposted CZ’s old tweet and announced that GMEB, the tokenized GameStop stock on bStocks, is now tradable on the platform. BlockBeats Note: UTILITY’s trading pair is UTILITY/GMEB, not the usual liquidity pairs like BNB or USDT. This is part of the popular "stock meme" trend on BSC, using bStocks’ tokenized US stock (here GMEB) as the liquidity pool, centered on the classic narrative of GameStop’s retail investors vs. Wall Street and meme stocks. BlockBeats reminds users that most meme coins have no real use cases, experience high price volatility, and investment requires caution.

The average rate on 30-year fixed mortgages in the U.S. rose to 6.67% this week, hitting its highest level in over a year, up from 6.43% in early July and 5.98% before the outbreak of the U.S.-Iran conflict. The recent rise in the yield on 10-year U.S. Treasury bonds is one of the key factors driving the increase in mortgage rates. As a result, the average monthly mortgage payment for a typical U.S. home has risen by nearly $150 since the start of the year. In the UK, the average rate on 5-year fixed mortgages climbed to 5.66% in July, marking its first month-over-month increase since April. Germany’s 10-year fixed mortgage rate rose from 3.3% in early July to 3.7% this week, while France’s 10-year fixed mortgage rate increased from 3.02% in June to 3.15% in July. Market analysts note that the U.S.-Iran conflict has pushed up oil and energy prices, stoking concerns about a resurgence in inflation and driving up financing costs including U.S. Treasury yields and UK swap rates. High interest rates are further dampening real estate transactions in Europe and the U.S. The "low-rate lock-in effect" persists in the U.S. market, while signed home purchase agreements in the UK fell by nearly 10% year-on-year in July.

According to Reuters, citing three people familiar with the matter, Apple (AAPL.O) has trained a large language model specifically for the Chinese market, a shift from its previous strategy of relying mainly on third-party models to power AI features in China. The AI model was developed through a partnership between Apple and Alibaba, and completed training with Alibaba’s support, the sources said. The news of Apple training a China-exclusive AI model has not been reported before. Previously, Apple preferred to use models from local Chinese partners to integrate generative AI features into iPhones and other devices sold in China. Neither Apple nor Alibaba responded to requests for comment. Apple’s AI toolkit, Apple Intelligence, is expected to launch in China in the coming months following an iOS operating system update.