Lookonchain APP

App Store

Analysis: Bitcoin's current round of decline is "different from the past," not accompanied by a major crisis but with continuous fund outflows

2025.12.17 19:56:05

As of December 17, Bloomberg reports Bitcoin is on track for its fourth annual decline in history—and the first time a pullback has occurred without a major scandal or industry-wide meltdown. Analysis notes the downturn is unfolding even as institutional participation grows, the regulatory landscape matures, and the crypto sector gets public backing from U.S. President Trump, catching markets off guard. Since hitting an all-time high above $126,000 in early October, Bitcoin has pulled back sharply. Trading volumes are currently low as investors continue exiting related products. Data shows U.S.-listed Bitcoin spot ETFs have seen net outflows of over $5.2 billion since October 10, market depth has dropped ~30% from its yearly peak, and there’s little appetite for a rebound in the derivatives market. Unlike prior bear markets, this pullback wasn’t triggered by exchange collapses, heavy-handed regulation, or systemic risk events. The three prior annual drops stemmed from: Mt. Gox’s collapse (2014), the 2018 ICO bubble burst, and the 2022 industry crisis sparked by the FTX collapse. Analysts note Bitcoin is showing signs of decoupling from U.S. stocks. The S&P 500 has repeatedly hit new highs this year, up ~16% annually—with tech stocks particularly strong—while Bitcoin remains under pressure. Apollo Crypto says despite numerous positive factors, the price lacks sustained momentum, signaling a clear weakening of market sentiment. Overall, this Bitcoin adjustment looks more like a reallocation of funds and a pullback in risk appetite at elevated levels, rather than a panic-driven collapse from a single event.
Relevant content

Trader 0x362a's 10x long on 236 $BTC nets $726K unrealized profit on Aster DEX

Trader 0x362a opened a 10x long on 236 $BTC ($19.98M) on @Aster_DEX and is now sitting on a $726K unrealized profit!

16 minutes ago

Trader 0x362a opened a 10x long on 236 $BTC ($19.98M) on @Aster_DEX and is now sitting on a $726K unrealized profit!

Trader 0x362a opened a 10x long on 236 $BTC ($19.98M) on @Aster_DEX and is now sitting on a $726K unrealized profit!

16 minutes ago

Tencent open-sources T-Mem: AI memory starts learning "association"

Beating AI Express (Insight): Tencent’s PCG Team Open-Sources Long-Term Memory System T-Mem, Paper Accepted by EMNLP 2026 Main Conference. T-Mem not only stores information but also records potential future use scenarios when memorizing content. For example, if a user previously mentioned “a colleague is allergic to seafood” and later asks “where to eat for a team building trip”, ordinary memory systems would fail to connect these two sentences with almost no overlapping keywords. T-Mem pre-associates “seafood allergy” with “restaurant selection for gatherings”, enabling it to retrieve relevant old information in similar scenarios later. Evaluations show that when queries lack keyword overlap with old memories, ordinary memory systems’ performance drops significantly, while T-Mem is less affected. Removing Horizon Trigger—the component responsible for predicting future scenarios—results in a notable performance decline for T-Mem, confirming this part is the core of its “associative recall” function.

16 minutes ago

Privacy AI protocol Phala briefly broke through $0.06, with a 24-hour rise of 66.8%.

According to market data, privacy AI protocol Phala briefly crossed $0.06, currently trading at $0.05956, with a 24-hour gain of 66.8%. Official information shows that usage of its flagship privacy AI product Confidential AI has surged recently: the product processed around 5.5 billion tokens on September 15, rising to 16.8 billion and 17.8 billion tokens in the following days, and hitting 25.451 billion tokens on September 19. Supported mainstream models include GLM-5.3 (which often accounts for over 40–70% of usage), GLM-5.3 Flash, DeepSeek series, Kimi K3, and others.

16 minutes ago

Xiaomi has publicly halted its MiMo reinforcement learning project, with two training runs costing $3.47 million.

Beating AI News Brief: Xiaomi’s MiMo has halted two publicly live-streamed reinforcement learning training runs. Both the Pro and Flash runs completed a full 30 training steps each, with total combined costs of $3.4747 million. The Pro run cost $2.6207 million, while the Flash run cost $854,000. The investment delivered clear performance gains: On DeepSWE v1.1, Pro’s score rose from 58.41 at step 1 to 72.57, a 14.16-point increase; Flash’s score jumped from 48.67 to 65.68, a 17.01-point gain. However, both curves showed noticeable dips in the middle, meaning scores do not rise steadily with longer training. Two additional benchmark tests also improved: Pro’s internal code evaluation score rose from 57.54 to 65.43, and its AutomationBench score increased from 45.2 to 53.1. The training process was not smooth: Pro encountered GPU out-of-memory (OOM) errors due to uneven expert load, plus disconnections and restarts between its training cluster and evaluator. Flash had to restart from step 15 due to infrastructure errors. Later, Xiaomi filtered out tasks that were too simple for Pro and removed the cyber dataset with abnormal rollout patterns from Pro’s subsequent training.

16 minutes ago

The total value of the "Maji" contract account has rebounded to over $10 million, earning $7.15 million in the past 24 hours.

According to TradingBeats' monitoring, the crypto entity known as "Ma Ji" posted a $7.15 million profit over the past 24 hours, pushing its total account value back above $10 million. Its current long positions consist of 30,950 ETH (valued at $84.79 million), 279 BTC (valued at $23.72 million), and 89,000 HYPE (valued at $8.48 million).

16 minutes ago

Popular tokens

BitcoinEthereumHyperliquidSolanaTRONBNBTetherAaveXRPPepeFartcoinOndoJupiterUniswapBonkPendleEthenaArbitrumAvalancheLidoChainlinkPolygonDogecoinCardano