There is another part of this discussion that makes the whole “Chinese auto industry vs American auto industry” framing much less clean than it sounds. Tesla itself is a good example. Tesla is American, but its battery supply chain has never been purely American. Panasonic was crucial to Tesla’s early scale, and Tesla later diversified its battery sourcing to companies including LG Energy Solution and CATL. So even the most successful American pure-EV manufacturer was built around a global, heavily Asian battery supply chain. Brazil makes this even more interesting because we are watching several versions of this industrial integration happen at the same time. BYD took over Ford’s former industrial complex in Camaçari and is progressively localizing production. GWM took over Mercedes-Benz’s former factory in Iracemápolis. Renault and Geely are expanding their industrial relationship in Brazil. GM itself is now assembling Chinese-developed Chevrolet EVs in Ceará. The Spark EUV and Captiva EV come from GM’s Chinese industrial ecosystem with SAIC and Wuling and are being assembled locally. Toyota and BYD have a 50/50 EV R&D joint venture. Nissan has had a major industrial relationship with Dongfeng in China for decades. At some point, asking whether a technology is simply “Chinese” or “Western” stops having an easy answer. And Brazil gives us another interesting experiment: range. If you look only at Brazilian homologation figures, some EVs can look surprisingly short-legged. Cars with batteries in the 50-60 kWh range can receive official Inmetro/PBEV range figures around 300 km. But the Brazilian number is deliberately conservative. Inmetro does not simply publish the raw laboratory result. Adjustment factors are applied to produce a more conservative real-world reference. The problem is that a lot of automotive discussion here then treats that number almost as the maximum distance the EV can realistically travel. Owner experience often looks very different. We now have EV owners driving these cars across Brazil, including mountainous areas and routes that climb from the coast onto the plateau. There are owners and independent tests substantially exceeding the official Inmetro range, and under favorable conditions some results get much closer to WLTP. Larger-battery EVs can exceed 400 km in real use even when their Brazilian homologated number looks much less impressive. Obviously that does not mean WLTP is guaranteed real-world range. Drive at 120 km/h, climb continuously, add headwind, temperature changes or heavy HVAC use and consumption changes dramatically. But this exposes an interesting asymmetry in how cars are discussed here. When a small 1.0-liter ICE car achieves an exceptionally good km/l result, automotive enthusiasts and media are perfectly happy to show what the car can achieve under favorable real-world conditions. With EVs, I often see the opposite. The conservative Inmetro figure gets repeated as the defining range of the vehicle, while owner consumption and independent road results receive much less attention. A much better way to discuss EV range is to show the homologation numbers and then show actual energy consumption. If an EV does 12, 15, 18 or 22 kWh/100 km, anyone can understand what a 40, 60 or 90 kWh battery means under different conditions. That's much more informative than saying “this is a 300 km car” because one homologation system printed 300 km on the label. And all of this comes back to the original Detroit question. Brazil is becoming a useful real-world laboratory because Chinese manufacturers are no longer simply shipping Chinese-built cars here. They are buying former Western factories, hiring Brazilian workers and progressively localizing production. At the same time, established American, European and Japanese manufacturers are increasingly using Chinese partners, platforms, batteries, engineering or complete vehicles. Europe is moving in a similar direction as Chinese manufacturers localize more production there. That gives us an opportunity to separate two things that are constantly mixed together in this discussion. If Chinese manufacturers lose most of their price advantage when they manufacture in Brazil, Europe or eventually the US with local workers, then Chinese wages, subsidies and domestic production conditions were clearly doing much of the work. But if a substantial advantage survives localization, then wages cannot be the entire explanation. Battery costs, vertical integration, EV-specific platforms, supplier organization, procurement, manufacturing scale, automation, development cycles and margins all have to enter the discussion. That's why I would actually find an American-built Chinese EV much more interesting than another imported Chinese EV. Require American production. American wages. American safety and environmental standards. Apply the same rules to everyone. Then compare the products. If an American-built BYD ends up costing roughly the same as an equivalent American-built GM or Ford, we learn something important about where the original Chinese cost advantage came from. But if it can still compete aggressively on price, equipment and efficiency while paying American production costs, then tariffs didn't answer the underlying competitiveness question. They only delayed the experiment.

---

How did h notice that

---

are we squeezing or what the H?

---

Alright fuckers let's do a blowoff top, go divide your stocks forward estimated EPS by current price and make sure it's more than the 5% bond yield (memory companies h

---

From my friend Claude - Worth saying upfront: I'm made by Anthropic, and the entire thesis hinges on what Anthropic is worth. Take my read with that in mind. I've tried to check the claims rather than form a view. **What checks out** Most of the facts are right, which puts this above the WSB average. CEF Connect shows a share price of $34.24 against NAV of $34.30, and a 52-week range where price hit $72.87 and NAV hit $34.30. The May comparison is accurate: the stock traded at $61.66 against a NAV of $24.56, a 151% premium, and at its April 2024 peak buyers paid over 400% above the value of the underlying assets. The insider buying is real: Sohail Prasad made 3 purchases buying 56,000 shares for an estimated $1,767,343 with no sales. And the New York Times reported in late-August 2026 that an eventual IPO could value Anthropic at $2 trillion, against a Series H in May 2026 at $965 billion post-money. **Where it falls apart** **The "10% discount to NAV" is a discount to a stale number the manager produced.** CEF Connect's 8 September page lists the NAV as of 31 March 2026. Destiny's own documentation says the fund strikes a quarterly NAV, and fair value determinations are made by the adviser's valuation committee with assistance from an independent third-party valuation firm. You are not buying a discount to observable assets. You are buying a discount to the manager's own estimate of illiquid private stakes, months after the fact. **The Anthropic re-rate is mostly already in the NAV.** The $965B round closed in May. NAV went from $19.97 at the end of Q4 2025 to $34.30, and total return on NAV over 12 months is 395.66%. The markup has happened. Run the actual numbers on the post's own inputs: Anthropic at 14.4% of a $34.30 NAV is about $4.94 per share. Going from $965B to $2T adds roughly $5.30, which is where the post's "$40 NAV" comes from. But that requires the full $2T. At the more sober base case, secondary markets price Anthropic at $1.05 to $1.15 trillion and FutureSearch projects a median first-day market cap of $1.10 trillion, a 14% premium over the last private round, which adds about 70 cents per share. Not several dollars. **The $2T figure is a forward-forward projection.** It rests on projected 2028 revenue of $190 billion to $200 billion. That is two years of flawless execution priced in today, leaked to journalists by parties who benefit from it being believed. **The real thesis is "the premium comes back," and the post's own data argues against it.** Every price target listed is NAV multiplied by an assumed premium. That isn't valuation, it's a bet that other people will pay more. But look at what actually happened: the 52-week average premium was 74% and the 52-week high premium was 290%, while the 6-month average is a 1.92% discount. The premium collapsed from triple digits to zero **through** the SpaceX IPO, the single biggest catalyst this fund has ever had. SpaceX listed in 2026 as one of the two largest IPOs ever. The post frames the May buyers as lunatics, but they were buying the same "premium inflates into the IPO" thesis it's now recommending. It didn't work. That's the precedent, not the exception. **The fee structure is worse than "a giant pile of T-bills" sounds.** The 2.50% annual management fee is charged on gross assets, not NAV, and the total expense ratio for 2024 came in at approximately 6.28% of average net assets, making DXYZ one of the most expensive publicly traded funds available. You're paying 2.5% a year to hold treasuries. The T-bill pile is a cost centre, not a feature. **Post-IPO, the fund's reason to exist shrinks.** Once Anthropic lists you can just buy Anthropic. And 2026 US IPO proceeds have hit a record while new listings have lagged the market. **My read** The one genuinely good point is buybacks below NAV, which are accretive, and insider buying alongside them. That's a real floor mechanism and I'd credit it. But the structure of the argument is: buy at fair value, sell at an irrational premium to a stale self-mark. That works until it doesn't, and the chart says it already stopped working four months ago. Context also matters for timing. The Fed decision lands in hours with a hike priced at 91%, and AI-related stocks sold off sharply on 14 September as concerns over AI risk intensified. There's a certain irony in Anthropic trying to convince Wall Street it deserves one of the biggest debuts in market history while simultaneously saying the pace of AI development needs to slow down. A high-beta, illiquid, 6%-fee closed-end fund is about the worst vehicle to hold into that combination. Separately: you're already long this theme through NVDA, MU and CRWD, which is where most of your gains sit. Adding a leveraged-sentiment wrapper on the same trade concentrates rather than diversifies. If you want the Anthropic exposure, waiting for the S-1 and buying the actual listing is a cleaner expression with no fee drag and no premium risk. I'd skip DXYZ.

---

Trump is fucking fuming right now lol. Hike t h at shit

---