【帖子标题】：Meta spent $31.1B in a quarter. Price per ad is still +12% 【标题翻译】：Meta单季度烧掉311亿美元。广告单价仍保持12%增长
【帖子正文】： I don’t own Meta. （声明：本人未持有Meta股票）
Meta spent 784M of free cash flow. Stock fell about 8.6% the next day. That’s the print people marked. They also split the ad line. Impressions +14%, down from +19% in Q1. Price per ad +12%, second quarter in a row. Daily users only +3%. Ad revenue still +27% to $59.4B. The extra money is coming from charging more per ad, not from finding more people. Meta二季度支出311亿美元，仅产生7.84亿美元自由现金流。次日股价暴跌8.6%，市场用脚投票。拆分广告业务线可见：展示量增长14%（较一季度19%增速下滑），广告单价连续两季度增长12%，日活用户仅增3%。广告收入仍增长27%至594亿美元——增量完全来自单价提升而非用户增长。
Full-year capex is 72.2B last year and they didn’t guide 2027. In a business where users grow 3%, that spend only shows up if they can keep charging as impressions slow. Two quarters of +12% is not a payback. What would break it: price per ad under 8% for two quarters, impressions already in single digits. Then they bought reach they can’t charge for. Could also just be a tighter auction. They don’t publish ROAS, so I can’t split a smarter match from scarcer inventory yet. Did the $31B buy a better ad, or more of them? 全年资本支出预计1300-1450亿美元（去年722亿），且未给出2027年指引。在用户增速仅3%的生意里，天量开支要见效必须维持广告单价——即便展示量增速放缓。连续两季度12%的单价增长远未回本。崩盘信号将是：连续两季度单价增速低于8%+展示量个位数增长，届时其购买的流量将无法变现。也可能只是竞价机制收紧。由于未公布广告回报率，我尚无法区分这是智能匹配优化还是库存减少所致。这311亿美元究竟买到了更优质的广告，还是单纯的数量堆砌？
https://www.investmoat.com/research/metas-capex-has-one-receipt （原文链接保留）