Scan to download
BTC $69,937.75 +1.50%
ETH $2,089.98 +1.94%
BNB $632.13 +2.09%
XRP $1.50 +7.13%
SOL $88.52 +4.83%
TRX $0.2824 +0.28%
DOGE $0.1121 +15.77%
ADA $0.2967 +8.93%
BCH $564.22 +2.44%
LINK $9.13 +3.61%
HYPE $31.65 +0.27%
AAVE $129.31 +8.19%
SUI $1.03 +7.37%
XLM $0.1747 +6.06%
ZEC $326.95 +20.74%
BTC $69,937.75 +1.50%
ETH $2,089.98 +1.94%
BNB $632.13 +2.09%
XRP $1.50 +7.13%
SOL $88.52 +4.83%
TRX $0.2824 +0.28%
DOGE $0.1121 +15.77%
ADA $0.2967 +8.93%
BCH $564.22 +2.44%
LINK $9.13 +3.61%
HYPE $31.65 +0.27%
AAVE $129.31 +8.19%
SUI $1.03 +7.37%
XLM $0.1747 +6.06%
ZEC $326.95 +20.74%

enso

Benson Sun: Bitcoin's decline reached a rare -5.65σ, occurring only 4 times in history

Cryptocurrency KOL and former FTX community partner Benson Sun posted that Bitcoin experienced an extreme drop this morning. Calculating with a 200-day lookback period, BTC's decline reached -5.65 standard deviations (σ). The Six Sigma standard in manufacturing allows for only 3.4 defects per million occurrences, which defines "almost impossible" in human industrial civilization. Yesterday's BTC volatility was just 0.35 standard deviations away from this "industrial-grade impossibility."A -5.65σ occurrence has a theoretical probability of about one in ten million under normal distribution. Despite the fat tail effect in financial markets, this level of volatility has only occurred 4 times since BTC began trading in July 2010, accounting for about 0.07% of all trading days. Even during the deep bear phases of 2018 and 2022, such a rapid decline had not occurred within a rolling 200-day period. This poses a severe challenge to quantitative strategies.Currently, most quantitative models are built on data after 2015, and historical samples exceeding 5.65σ, apart from the anomalous "312" flash crash in 2020, occurred before 2015, leaving almost no reference precedent.CoinKarma's quantitative strategy has shown a paper loss in this round of market conditions, but due to maintaining low leverage (about 1.4 times) over the long term, it remains manageable, with a maximum drawdown of about 30%. While extreme market conditions are an expensive "tuition," contracts and on-chain data will become important nutrients for future risk control models.

Bittensor, in collaboration with HackQuest, launches the Build on Bittensor developer program

The decentralized machine learning network Bittensor has officially partnered with the Web3 education and developer platform HackQuest to jointly launch the Bittensor × HackQuest: Build on Bittensor global developer program. This program focuses on Subnet development and aims to encourage global developers, researchers, and builders to systematically participate in the Bittensor decentralized AI ecosystem.Build on Bittensor is Bittensor's first developer program. The program will combine Bittensor's decentralized machine learning and Subnet architecture design with HackQuest's experience in developer education, local execution, and community organization, providing a complete participation path from learning to practice for developers from various backgrounds.This program continues the joint tour activities previously conducted in multiple cities and officially upgrades to a global developer tour (Bittensor × HackQuest Tour). Past events have attracted over 500 developers, university students, and AI researchers, facilitating in-depth discussions on decentralized AI, Web3 infrastructure, and the Bittensor Subnet mechanism, demonstrating the developer community's strong interest in real Subnet construction and mechanism design.As an important part of the Build on Bittensor program, both parties will launch the Bittensor Subnet Ideathon for the first time, open to global developers. Participants will design and build their own Subnet solutions around Bittensor's Subnet architecture, gaining practical understanding of decentralized incentives, miner and validator collaboration mechanisms, and have the opportunity to receive cash rewards, incubation, and subsequent ecosystem support.Additionally, for developers who are new to Bittensor, HackQuest will provide a structured learning path (Learning Track) covering the basic concepts of Bittensor, understanding of the Subnet mechanism, and practical development preparation, helping new developers smoothly transition from entry-level to participating in Subnet construction.In the future, Bittensor and HackQuest will continue to promote the Build on Bittensor series of activities globally, further expanding the developer base for decentralized AI and driving more Subnets from concept to real operation.
app_icon
ChainCatcher Building the Web3 world with innovations.