They have been far too slow to move with this

Officially overtaken

Leo 2.0

@FundaAI is closing in on @citrini on the MarketStack Index after James Bulltard pulled ahead of them earlier this week. At this rate, the top 3 MarketStack Index authors, all with 17k subscribers each, will have a combined distribution 19x smaller than the top 3 Bestsellers (Citrini/Burry/Doomberg) who together have almost 1 million subscribers.

Levered shorts down almost exactly as real-money longs came off. That mirror is what you'd expect when basis trades unwind.

1/ Leveraged funds just made the largest weekly cut to Ultra Bond (UB) net shorts since the CFTC series began in 2010. 808,870 → 726,257 contracts. ~$16m per bp of risk off in one week. 2/ Same week, asset managers sold ~$15.1m/bp of long futures. Levered shorts down almost exactly as real-money longs came off. That mirror is what you'd expect when basis trades unwind. 3/ Funding was calm: SOFR 3bp below IORB, zero SRF use. So this looks voluntary, not forced. For now. 4/ Don't read it as bullish duration. If it's basis, funds are selling cash long bonds as they buy back futures. That pressure lands on cash Treasuries, and can feed further unwinding. 5/ UB shorts are 22.6% below the June peak, but still ~0.8 SD above the 5y average and ~1.2 SD above the 10y. ZB shows the path: 52.3% below its Jan peak, now ~2 SD below its 1y average. 6/ My read: the long-end unwind isn't done. Next checks: CFTC Friday (16 Oct) and FINRA TLT short interest (~27 Oct).

Yesterday @FundaAI broke into the top 10 Finance Substack bestsellers. MarketStack saw it coming. Two weeks earlier, they already cracked the top 10 of the MarketStack Index, climbing to #6 by 30 Sep. This is the Index doing its job. >> marketstack.pro @illyquid

@jbulltard1 is criminally under-read IMO. He has >15x fewer free subscribers than @citrini (17K vs 263K), yet his revenue is only 3 spots behind them on the Substack Bestsellers board (#4 vs #1). I think he should and could be #1 on Finance Substack. I am not a trader yet I recently timed my best trade ever, partly on the back of his daily recap, which he sends to free subs. I’ve lost count of how many traders’ comments I’ve read, both in and out of the industry. Most of them leave me with more questions than answers. His are simply the most incisive and attuned to the market of any I have read...mostly because they are so easy to follow. I think anyone trying to get a grounded and pragmatic sense of how the market works should be a free subscriber at least. Obviously if you are a more active trader that is what his database is for. I have zero commercial relationship with him or his Substack.

Anthropics’s $2 trillion Capital Stack. The business has something closer to an AI utility / infrastructure economics problem than a traditional software economics problem. The $518bn commitment can simultaneously be: bullish for compute supply and dangerous for Anthropic. Why? Because massive capacity expansion could eventually drive down the cost of compute. That’s excellent for AI consumers. But if token prices / AI revenue fall faster than Anthropic’s contractual infrastructure costs, Anthropic bears the mismatch. Myers describes the core risk as Anthropic potentially being the party “holding the bag” if that happens.