Analysis: Active buying by long-term holders has helped provide short-term support for BTC, but a potential cyclical low should still be considered.
Related
Glassnode: Long-term holders are starting to accumulate Bitcoin again, but the market may still need to undergo one last round of consolidation.
According to Mars Finance, Glassnode's latest weekly report on July 3rd stated that despite Bitcoin falling below $60,000, continued outflows from spot ETFs, and increased risk aversion in the options market, on-chain data shows that long-term holders (LTH) have re-entered a buying phase, with multiple wallet addresses simultaneously shifting towards accumulation, indicating that patient funds are gradually absorbing market selling pressure. Currently, approximately 10.83 million Bitcoins are in a loss-making state, higher than the approximately 9.22 million Bitcoins in a profitable state, reflecting that the market is still under significant pressure, and holdings are gradually shifting towards investors with high conviction. Meanwhile, US spot Bitcoin ETFs continue to experience net outflows, indicating that institutional risk appetite has not yet recovered; Coinbase's order book shows a significant increase in buying, showing that institutions are gradually providing liquidity and rebuilding market support. In the derivatives market, Hyperliquid leveraged traders continue to increase long positions, which could drive a rapid rise if the market rebounds, but a break below key support could trigger a new round of long liquidation. The Deribit options gamma structure is beginning to stabilize the market, helping to reduce volatility, while demand for put options remains high, indicating that investors are still generally defensive. Bitcoin is gradually transitioning from a distribution phase to an accumulation phase, and the foundation for a long-term recovery is forming. However, before the trend truly reverses, the market may still experience one last round of shakeouts triggered by macroeconomic factors or leverage liquidation.
Multiple Wall Street firms are collectively advocating for "buying on dips" in the semiconductor sector: the long-term logic of AI remains unchanged, but investment is entering an era of selective buying.
According to Mars Finance, on July 7th, amidst the recent continuous correction in the semiconductor sector, several Wall Street institutions have voiced their opinions, generally believing that the current adjustment presents an opportunity for investors to "buy on dips." However, unlike past recommendations to allocate across the entire semiconductor sector, institutions generally believe that AI investment has entered a phase of selective stock picking. Goldman Sachs stated that AI chip trading has entered a more selective phase, and does not recommend continuing to "buy a basket" of semiconductor stocks. They remain optimistic about specific sub-sectors such as CPUs, ASICs, memory, and semiconductor equipment, specifically highlighting AMD and Applied Materials. JPMorgan Chase believes that the recent correction in semiconductor stocks presents a good entry window, as AI chip demand remains in a long-term upward cycle, with new capacity not expected to be significantly released until around 2028, and the industry's supply and demand structure remains healthy. Bank of America maintains its optimistic outlook on the long-term boom cycle of AI semiconductors, believing the industry is still in the middle of an 8- to 10-year growth cycle, and the global semiconductor market size is expected to continue to expand. They recommend focusing on industry leaders such as Nvidia, Broadcom, Lam Research, and KLA. UBS stated that the long-term investment logic for AI remains unchanged, and the short-term fluctuations in the semiconductor sector actually provide long-term investors with opportunities to gradually build positions. They recommend taking advantage of market corrections to buy on dips. Morgan Stanley believes that the long-term prospects for AI chips remain positive, but with the sector's significant rise, the market will focus more on earnings realization capabilities. Funds may gradually rotate from some chip stocks to AI infrastructure beneficiaries such as cloud computing, and investors should pay more attention to individual stock selection. Overall, several Wall Street institutions, including Goldman Sachs, JPMorgan Chase, Bank of America, and UBS, have recently released similar signals: the semiconductor correction is not the end of the AI rally, but rather provides a new window for investment. However, the market has moved from a phase of "broad-based sector gains" to one of "selecting leading companies," and future performance will depend more on companies' earnings realization capabilities and the sustainability of demand for AI infrastructure.
Analysis: Strategy's capital restructuring alleviated short-term liquidity pressures, but structural risks remain.
According to Odaily Odaily, Alex Thorn, head of Galaxy Research, said that the capital management reforms recently launched by Michael Saylor's Strategy (MSTR) have effectively alleviated market concerns about its liquidity and preferred stock system in the short term, but they are more about "buying time" than fundamentally solving structural problems. Strategy has faced pressure on its preferred stock "digital credit" system in recent weeks, with its STRC ("Stretch" preferred stock) briefly falling below par value, hitting a low of approximately $71.25. This has raised concerns about the company's ability to pay preferred stock dividends, given its declining BTC price and shrinking dollar reserves. The market subsequently focused on three stress scenarios: selling Bitcoin, issuing new MSTR shares to dilute shareholders, or reducing/suspending preferred stock dividends. In response, Strategy announced a comprehensive capital management restructuring on Monday, launching a "Digital Credit Capital Framework" that includes five tools: a board-approved dollar reserve policy, adjustments to the STRC dividend mechanism, a $1 billion preferred stock buyback mandate, a $1 billion MSTR common stock buyback mandate, and a Bitcoin monetization mechanism. Simultaneously, the company increased its annualized STRC dividend yield from 11.5% to 12%. The market reacted positively, with both MSTR and STRC rising sharply that day, and Bitcoin also rebounding in tandem. Alex Thorn points out that this adjustment improved market sentiment in the short term, allowing Strategy to extend its cash coverage period to approximately 17 months and enhance its funding buffer through new financing. However, the company still faces approximately $6.7 billion in convertible bonds maturing in 2027-2028, and long-term structural risks remain. The core issue is not whether Strategy has enough BTC (approximately 847,000 coins), but rather that its dollar liquidity is insufficient to cover preferred stock and capital structure obligations without harming the interests of any party, leading to a squeeze on the interests of various shareholder groups. However, the key significance of this adjustment lies in enhancing the "optionality" of the company's capital instruments, shifting it from a one-way BTC accumulation strategy to a more proactive asset and liability management model, thereby preventing short-term liquidity problems from escalating into a systemic crisis. Although the current Bitcoin market environment is weak and may not have bottomed out yet, Strategy's new framework has, to some extent, bought the company a window of opportunity, allowing it to wait for more favorable market conditions.
Bitcoin short-term holders hit 30-day profit streak as bull-market odds improve: CryptoQuant
Bitcoin short-term holders approached a full month in partial profit as analysis saw improving odds of an enduring bullish BTC price reversal.
MIRA, an open-source 5B multiplayer world model, uses DINOv3 representation to mitigate long-term drift and can simulate 2v2 battles in Rocket League in real time.
According to Beating, AI research firm General Intuition, in collaboration with French AI lab Kyutai and Epic Games, has launched MIRA, a multiplayer interactive world model. As a generative game simulator supporting real-time multiplayer interaction, MIRA can simulate 2v2 battles in Rocket League in real time, based solely on historical footage and player button presses, without requiring a physics engine, rendering engine, or explicit 3D representation for inference. Unlike the "decoupling of logic computation and image rendering" approach adopted by companies like Odyssey, MIRA takes a generative simulation approach based on video latent space. MIRA boasts 5 billion parameters, and its core design builds the latent prediction space on a frozen general-purpose visual encoder, DINOv3-L. Leveraging pre-trained visual features, the generated latent states can more stably fall within the effective representation space, significantly mitigating image drift and divergence during long-term prediction. For multi-screen alignment, MIRA stitches the latent images from four player perspectives into a unified grid, enabling spatial attention mechanisms to operate naturally across viewpoints and improving the spatial consistency of vehicles, the soccer ball, and key events across multiple perspectives. The Action Dropout introduced during training also helps the system complete the game behavior of vehicles not controlled by commands when parts of the motion flow are missing. Currently, MIRA can run in real time at 20 frames per second on a single NVIDIA B200 graphics card. The team has open-sourced the training and inference code and released the Rocket Science dataset, which contains 1,000 hours of matches, approximately 4,000 hours of video, motion flow, and physics data from four perspectives; the complete training of the model used approximately 10,000 hours of clean match data.
TRON oracle WINkLink has officially launched a long-term WIN buyback and burn program, with 100% of actual business revenue being repurchased quarterly.
Mars Finance reports that WINkLink, the core oracle project of the TRON ecosystem, has officially announced the launch of a long-term WIN token buyback program. According to the official announcement, starting from the third quarter of 2026, 100% of the actual business revenue generated by WINkLink's oracle service will be used quarterly to buy back WIN tokens. The repurchased portion will be burned quarterly as planned, and the on-chain burning data will be published in the next quarter to ensure full transparency. This move establishes a value support mechanism for the WIN token driven by real business revenue, enhancing liquidity and market confidence while providing a more solid value anchor for long-term holders.