Lookonchain APP

App Store

Opinion: US Stock Short Positions Reach an All-Time High, Bitcoin May See a 'Decoupling Moment'

2026.05.26 19:37:54

May 26: XWIN Research published a note pointing out that short positions in the U.S. stock market have recently surged to an all-time high. But this doesn’t signal broad market bearishness, the report argues. Data shows hedge funds’ total leverage ratio has climbed to roughly 293%. Both the S&P 500’s short exposure and Days-to-Cover ratio hit record peaks, indicating institutional investors are maintaining large long positions while significantly boosting their hedge bets. The market is now in a structural state of “high leverage with concurrent long and short expansion”: the surface trend remains upward, yet internal defensive sentiment is sharply rising. The core driver of this dynamic is that funds keep concentrating in AI-linked mega-cap stocks. A flood of capital is flowing to just a handful of tech giants, while mid-cap and weaker sectors face sustained shorting. This has kept the index afloat but is building up internal fragility. Such a structure suggests that once the AI trading frenzy cools, the market could face far more intense volatility and risk unwinding. This shift also has major implications for Bitcoin. Historically, BTC has typically moved in lockstep with U.S. equities during major risk events—for example, during the 2020 pandemic-induced crash, Bitcoin fell alongside the S&P 500, failing to live up to its reputation as a traditional safe haven. But since 2025, the two have started to diverge sharply: the S&P 500 has held relatively steady, while Bitcoin’s volatility has spiked. ETF inflows and active spot buying show Bitcoin is increasingly driven by its own liquidity cycle, leverage structure, and institutional demand. Markets are beginning to view Bitcoin as transitioning from a mere “high-risk asset” to a hybrid asset that combines both macro liquidity attributes and its own independent market dynamics.
Relevant content

An OG Ethereum whale has held a long position in ENA for 228 days, with unrealized profits nearing 900%.

According to TradingBeats monitoring, the ENA long position held by a whale address starting with 0xeadc has been ongoing for roughly 228 days. As of 10:34 Beijing Time, the address still holds approximately 10.0265 million ENA long positions, with an average entry price of $0.110537, position value of around $2.0849 million, and unrealized profit of about $976,600. ENA has risen around 88.12% from its entry cost. The position uses 10x full leverage, generating a return of roughly 881.18%. It currently holds a total of 41 long positions, with a total notional value of approximately $69.8481 million, and an overall contract leverage of about 4.65x. ENA accounts for only around 2.98% of its total contract exposure, but its unrealized profit ranks third among all current positions, trailing only NEAR’s ~$1.3695 million and AAVE’s ~$1.0225 million. It is reported that this address is an ETH OG who entered the crypto market in 2017, once holding nearly $90 million worth of ETH, all transferred to Hyperliquid. Previously, it held the largest HYPE long position on the platform.

10 minutes ago

Solana network recorded a record high of 5.218 billion transactions last month.

According to Blockworks data, on-chain transaction activity on the Solana network reached a record high of 5.218 billion transactions in August.

10 minutes ago

The ZCode unauthorized code leakage incident has led to the aggrieved enterprise pursuing accountability, ordering Zhipu to delete all uploaded data.

Beating AI Insight News Brief: The ZCode code upload controversy from Chinese AI firm Zhipu AI has spread beyond developer communities to enterprise clients. Taiyuan Chengming Technology has sent a letter to Zhipu, alleging that ZCode automatically uploaded not just a small number of code snippets, but full project archives including complete source code, Git history, database passwords, cloud service credentials, and employee personal information. The company demanded Zhipu delete the uploaded data, disclose the data’s whereabouts, and provide proof of deletion. Zhipu AI had earlier apologized, attributing the issue to its "code repository indexing" feature, which was enabled by default during its initial rollout and could upload repository data when generating Repo Wiki pages. The firm noted that relevant data is destroyed immediately after use, the problem has been resolved, and it has pledged to open-source ZCode and invite third-party audits.

10 minutes ago

Uniswap’s tokenized stock DeFi TVL rose by $82.8 million over the past 30 days.

Over the past 30 days, Uniswap has recorded the highest growth in DeFi total value locked (TVL) for tokenized stocks, with its V4 and V3 versions collectively adding $82.8 million.

10 minutes ago

Arthur Hayes accumulates 25.33M $ENA at $0.09, targets $0.50 with $3.28M unrealized gains

Arthur Hayes (@CryptoHayes) just called for $ENA to reach $0.50. He bought 25.33M $ENA ($5.53M) at an average price of $0.09 a month ago and is now sitting on a $3.28M unrealized profit (+146%)!

10 minutes ago

JieYue Step 5 is officially released, and will be open-sourced on October 15.

Beating AI News: StepStar officially launches its flagship model Step 5 Preview. Built on a sparse Mixture-of-Experts (MoE) architecture, the model has a total of 600 billion parameters, with only 27 billion activated per inference run. It supports a 1 million-token context window and text-image input. The model’s API and Studio are now fully open, with full model weights scheduled for release on October 15. Artificial Analysis (AA) completed pre-evaluations earlier. Step 5 Preview scored 44 points on the Intelligence Index, matching Kimi K3 Max. AA currently measures its output speed at ~100 tokens per second, with a task cost of $0.71, compared to Kimi K3 Max’s $2 per task. The developer highlighted the model’s long-task capability: StepStar enabled Step 5 Preview to autonomously optimize a set of H100 GPU cores for up to 24 hours straight. The model independently modifies code, runs tests, compares results, and iterates. After roughly 22 hours, it achieved 508 TFLOPS, while Claude Opus 5 reached 493 TFLOPS in the same experiment. In another 24-hour trial, Step 5 Preview designed its own training data, lifting Qwen3-30B-A3B’s accuracy on AIME24 from 53.3% to 60% — matching Claude Opus 5, while consuming fewer labeled tokens.

10 minutes ago

Popular tokens

BitcoinEthereumHyperliquidSolanaTRONBNBTetherAaveXRPPepeFartcoinOndoJupiterUniswapBonkPendleEthenaArbitrumAvalancheLidoChainlinkPolygonDogecoinCardano