表面看是个反直觉的故事：周四(10月1日)公布的9月非农仅增2.9万人，失业率升到4.2%，前两月合计还被下修6万人——这种数据按常理该让市场避险，但BTC不跌反涨，现报86,667美元，24小时涨3.5%，7天涨2.6%，仍处在7天高点附近。真正的因果链不是"坏消息=利好"这么简单。第一层机制是流动性预期：疲软就业数据让市场押注美联储会继续降息,这是直接驱动。但更关键的第二层,是买盘的真实性,10月1日BTC现货ETF净流入1.03亿美元,一举扭转了前一天1.487亿美元的单日净流出。同时衍生品市场的未平仓合约和资金费率同步上涨,说明这是新多头在实际加仓,而不是单纯的杠杆噪音。OpenClue也在10月2日当天把BTC的立场从"谨慎看多"上调为"看多",依据正是24小时内出现处于自身波动90%分位的涨幅、价格突破此前区间,并有ETF净流入和交易所余额下降这类真实买盘信号支撑——这是过去30天BTC立场三次反复切换以来最明确的一次确认。值得注意的是,ETH ETF已经连续3天净流出,和BTC的资金流向出现明显分化,说明这波买盘目前集中在BTC而非整个板块。如果ETF流入能延续,叠加恐慌贪婪指数已回升到72(贪婪),这个"降息预期→真实现货买盘→情绪走强"的循环短期内可能自我强化;但今明两天美股连续闭市,ETF数据不会更新,循环能否持续要等下一个交易日的流向数据来验证。
On the surface this looks backwards: Thursday's (Oct 1) US jobs report showed just 29,000 new payrolls, unemployment at 4.2%, and a 60,000 downward revision to prior months — textbook risk-off data. Yet BTC is up, trading at $86,667, +3.5% on the day, +2.6% over 7 days, still near its weekly high. The mechanism isn't "bad news is good news" — it's layered. First-order driver: weak jobs data pushed rate-cut bets higher, a liquidity story. Second-order and more telling: BTC spot ETFs saw $103 million in net inflows on Oct 1, reversing the prior day's $148.7 million single-day outflow. Open interest and funding rates rose in tandem, signaling real new long positioning rather than leverage noise. OpenClue upgraded its BTC stance from cautiously bullish to bullish on Oct 2 itself, citing a 24-hour move in the 90th percentile of its own volatility range, a breakout above the prior trading band, and real buying evidence — ETF inflows and falling exchange balances — the clearest confirmation after three stance flips over the past 30 days. Notably, ETH ETFs have logged three straight days of net outflows, a clear divergence from BTC's flow pattern — this buying pressure looks concentrated, not sector-wide. With the Fear & Greed Index back at 72 (Greed), the loop of rate-cut bets feeding real spot demand feeding sentiment could self-reinforce short term — but US markets are closed the next two days, so ETF data won't update until the next trading session confirms whether the loop holds.