Arthur Hayes: I bet that Bitcoin will surpass $110,000 before retracing to $76,500

Renowned trader Killa stated in a post tonight that data from Bitcoin’s multiple bull and bear cycles shows the depth of bear markets is gradually shrinking, and the bottom of this cycle may have already formed. The expectation of some who missed out on gains that Bitcoin will fall to $50,000 in October is highly unlikely. However, Killa also pointed out that if Bitcoin drops to $61,000, the expected value of long positions being liquidated could reach as high as $20 billion, noting that market makers have the motivation and possibility to liquidate long positions before re-establishing new ones. Killa, a quantitative trader focused on BTC, previously predicted the peak of this bull market in May 2025 and has over 200,000 followers on X. In mid-April, he shorted Bitcoin at $74,688 and switched to long positions during the broad market sell-off on June 5.

According to HTX market data, Bitcoin has rebounded, briefly breaking through the $78,000 mark, and is currently trading at $78,007.56, with its 24-hour decline narrowing to 0.28%.

According to GMGN market data, some meme coins on the BSC ecosystem saw notable rallies today. Among them, "4" surged over 52% in a single day, with a current market capitalization of $17.78 million; "My Horse Is Here" rose 21.35% daily, standing at $13.64 million in market cap; "Lobster" gained 18.11% in a day, with a current market cap of $71.27 million. BlockBeats reminds users that related tokens are highly volatile, so investors should exercise caution.

Crypto KOL @0xkioto today summarized the common trajectory of tokens that have reached or are near the 100 million yuan market cap since the launch of the Robinhood Chain. He pointed out that tokens like CASHCAT, AI, and PONS have strong market appeal from the start, attracting massive capital inflows. However, once the hype cools, all short-term buyers get washed out. All these tokens have seen crashes ranging from 60% to 95% after their initial rally. @0xkioto argues that these crashes transfer supply to loyal holders who will not sell, as well as a "team" accumulating tokens, laying the groundwork for the next rally. When new demand emerges, it will face thin sell orders, possibly paired with a catalyst to accelerate gains—this is when the token starts surging sharply. "The Robinhood Chain belongs to holders, not disruptors," he said. BlockBeats notes that the above summary may hold some representativeness and logic behind token consolidation. However, the performance of these tokens is also heavily influenced by overall market sentiment fluctuations since the Robinhood Chain’s launch. The Robinhood Chain launched in early July with impressive performance, but later fell into a liquidity crunch and faced challenges with its narrative due to factors like capital outflows, which is one of the key reasons behind the sharp corrections seen in the aforementioned tokens.

Beating AI Insight News: Terminal-Bench has launched version 4.0, recalibrating time, CPU, and memory parameters for agent task execution, fixing 19 tasks, and removing 8 tasks plagued by issues including saturation, refusal to answer, public solutions, or quality problems. All tasks now have a uniform maximum execution time of 8 hours, aimed at reducing the impact of timeouts and environment-related issues on performance scores. In the latest ranking, Opus 5 + Claude Code tops the list with 51.8%, Fable 5 takes second at 44.5%, while GLM-5.3 + Claude Code claims third place with 41.8%, surpassing GPT-5.6 Sol + Codex’s 37.3%. Among the top three, GLM-5.3 is the only model not developed by Anthropic. In Terminal-Bench 3.0, GLM-5.3 ranked fourth with 32.4%, trailing GPT-5.6 Sol’s 34.6%; in version 4.0, GLM-5.3 jumped to third, leading Sol by 4.5 percentage points.

Serenity has published an analysis note questioning Sivers Photonics (SIVE)’s development strategy, arguing the company remains overly focused on the Swedish market, while US investors likely prioritize its future growth potential and actual economic value derived from orders. Serenity noted US analysts are more inclined to focus on the implications of SIVE’s recent two wafer fab capacity allocations, and how much revenue and operating leverage these capacities will unlock amid supply bottlenecks and rising Average Selling Prices (ASP). Additionally, factors such as the 2028 volume ramp-up of CPO players including NPO/CPO, pluggable optical modules, and Ayar, ELS products’ collaboration with O-Net, and the scale and potential Total Addressable Market (TAM) of six new pluggable customers also warrant in-depth discussion. By contrast, local market concerns center on issues like “how to stem the bleeding”, why private clients cannot be disclosed, why the focus on transceivers, and what the “business opportunity pipeline” actually entails, forcing management to spend substantial time addressing these queries instead of discussing future growth. Serenity believes the more time SIVE devotes to the Swedish market, the more its valuation will be constrained by Swedish market investment logic, hence the need to actively showcase future growth opportunities and their underlying economic value to US investors.