Matthew Sigel

explains the weakness last 2 days. fundamentals looking good per blockchain data. $FIGR https://t.co/srvux0HIut

AI investors are only just now discovering regulatory uncertainty. Bitcoiners were born in it. Molded by it.

Curious structure for this new $DCAP ETF, an actively managed fund investing in preferred securities issued by Bitcoin treasury companies. At launch, it expects to be roughly 50/50 $STRC and Strive’s own $SATA preferred. Strive is the sub-adviser but can’t provide research on SATA and has no authority to make investment decisions for the fund. So Strive is sub-advising a fund that invests heavily in Strive, while being walled off from advising on Strive 🤔

$BLSH board member Andrew Bliss sold ~$10M of stock in the open market this week, roughly 5% of his position. Disappointing optics coming right after Bullish's tokenized stock push. Worth noting: the sale was made under a 10b5-1 plan adopted back in Feb before $BLSH was public https://t.co/ayNrkGJZgS

We don’t really assume a fixed conversion rate. It's clear that energized power is the scarce upstream asset as only ~2% of the U.S. data center power pipeline actually reached commercial operation last year, and median interconnection time is now ~5 years. Once you have the power, though, it still takes a lot of capital and 12-18 months to turn it into finished DC capacity (maybe faster for IREN given the existing infra). IREN have a huge amount of energized power relative to what’s actually built, so their bottleneck is capital + execution, now power. So I’m not dismissing the value of their secured power. If the demand and the financing keeps showing up, they’re in a very good spot. As for the value, our work suggests the market is paying almost nothing for uncontracted DC capacity today. For the powered-shell names, current leases suggest roughly $15-20M of value creation per MW once capacity is leased. For neoclouds like IREN it’s harder to put a clean $/MW on the uncontracted pipeline because they have much larger financing needs and generally shorter-duration customer contracts.

Matt, great stuff, thanks for sharing. I agree with your takes. Performance hurdles, in and of themselves, are not necessarily evidence of a well-designed comp structure. The details matter to understand alignment, and there are other factors such as transparency, related party transactions, board composition, shareholder approval etc, which determine overall governance, aside from comp. With that said, please do note our line in the piece: "Strive has the best-designed incentive in the group." So please keep up the good work and communication, it's greatly appreciated. As skeptical as I've been about DATs in my own fund, I am rooting for you, Saylor and others to succeed and your clear communication to shareholders is a major positive for which you have rightfully earned a premium valuation.

VanEck proud to back @withAUSD and thrilled with this massive development, nice job @Nick_van_Eck 👏 https://t.co/WI9sD9oSNF

We see more risk that NVDA misses estimates because its customers cannot get power than that the miners fail to lease their megawatts. Our re-underwritten base case implies ~83% average upside across our powered-land holdings. https://t.co/NAkO3la6SJ

3443 TT, Global Unichip limit up again, ATH. We originally found GUC through Bitcoin. It was a major ASIC design partner for Bitcoin miners, which led us to research its unusually close relationship with TSMC and deep engineering bench. Today that same capability is being monetized by AI. GUC is working on custom chips for Google, Tesla and other hyperscalers. August sales +111% YoY, driving today's action. The contrast with Bitcoin mining machine makers is striking. CAN was in NY this week on an NDR. Despite having a competitive next-gen A16 miner ready, it has decided not to mass produce it. AI is soaking up advanced wafer capacity and pushing the economics against Bitcoin ASICs. AI crowding out Bitcoin mining has reached the chip level. The implications for Bitcoin are less obvious. Canaan’s retreat could mean still more dominance for Bitmain, with Bitdeer emerging as a credible challenger. But slower ASIC efficiency gains also mean existing machines become obsolete more slowly, reducing the reinvestment required to stay competitive. And to the extent hashrate growth slows or reverses, difficulty adjusts lower, providing yet another tailwind to margins for the miners who remain. Still, hard to frame hashrate decline a positive net net.

$BTDR entering KOL v2 era https://t.co/G77BrtehXp https://t.co/dn2YP2lFHa