Singh runs a special-situations book — stances reflect how each name was framed in this conversation (positions, shorts, arb spreads, or evidence in the circular-AI-financing argument). SpaceX is still pre-IPO (debuts as SPCX on June 12), so it stays a private row. VCX/FDXW have no clean research pages.
| Ticker | Name | Research | View | What Singh said | At |
|---|---|---|---|---|---|
| SATS | EchoStar | QT · SA · STK · FA | Positive | His SpaceX exposure, cost basis in the 80s: EchoStar sold spectrum to SpaceX for cash + stock (struck near a ~$400B valuation), worth >$160/share at $1.75T vs ~$116 trading — already in the money; would follow an IPO rally at lower beta. Shorts are using it to hedge the IPO. | 34:09 |
| CVS | CVS Health | QT · SA · STK · FA | Positive | Long-side value work: sold off dramatically on the PBM + insurer issues but "trades extremely cheap" on a sum-of-the-parts; part of the rotation into REITs/banks/healthcare, which are very undervalued vs tech. | 53:40 |
| TE | T1 Energy | QT · SA · STK · FA | Positive | Got long for a very short-term trade — set up to be one of the leading energy providers for the AI buildout (power is the last domino in his cascade framework). | 54:00 |
| DGXX | Digi Power X | QT · SA · STK · FA | Positive | Risky small-size long from ~$3–4: won a >$1.7B AI colocation contract the market wasn't valuing (equity-raise fears); the stock more than doubled — trimmed and hedged the rest with covered calls. | 54:36 |
| CZR | Caesars Entertainment | QT · SA · STK · FA | Positive | Merger-arb long from a very low price — just got a bid from Fertitta. | 56:11 |
| MGM | MGM Resorts | QT · SA · STK · FA | Positive | Low-cost-basis arb long: $48.30 cash/share proposal from IAC/Peoples; his conservative sum-of-parts (50% BetMGM with Entain, 56% MGM China, the Strip, Japan) is ~$60 — sees potential for a topping bid. | 56:34 |
| SPCX | SpaceX (IPO — SPCX, June 12) | QT · SA · STK · FA | Neutral | "Not for the fainthearted": $18.7B revenue but a ~$17B loss run-rate (Q1-26 alone −$4.28B), ~95–100× sales / 260× fwd EBITDA, $20B of the raise repays a bridge loan, Musk holds ~82–85% votes. Yet the ~4% float + 2× oversubscription + Nasdaq-100 fast-track could squeeze it "marginally higher" — he won't play the IPO directly, holding SATS instead. | 16:42 |
| WBD | Warner Bros. Discovery | QT · SA · STK · FA | Neutral | The biggest merger spread on the board: the Ellison-funded Paramount-Skydance cash bid leaves a ~$4 (~15%) spread pending antitrust — should compress over the next year if it clears. | 57:17 |
| FDXW | FedEx Freight (spinco) | — | Neutral | Spin-off he was watching — "actually traded much better than we thought"; spin-offs/split-offs are a core special-sits category (the GE breakup doubled and tripled). | 57:00 |
| TSM | Taiwan Semiconductor | QT · SA · STK · FA | Neutral | "The most conservative people in this ecosystem" — TSMC (with the Koreans and ASML) constrains capacity, keeps Jensen from flooding the market with Vera Rubin chips, and has prolonged the AI cycle by a couple of years. | 38:55 |
| ASML | ASML Holding | QT · SA · STK · FA | Neutral | Grouped with TSMC and the Korean chipmakers as the disciplined capacity constrainers preventing the AI bubble from bursting too soon. | 38:55 |
| MU | Micron | QT · SA · STK · FA | Neutral | Memory costs up ~10× in two years — AI hardware inflation from COVID-style overordering; memory is hyper-cyclical and prices typically fall a year before demand does. Memory names sell off first in his cascade sequence; its CEO among the table-pounders. | 40:13 |
| MRVL | Marvell Technology | QT · SA · STK · FA | Neutral | Exhibit A of the pump dynamic: Nvidia put $2B in, Jensen went on live TV calling it "a trillion-dollar company," and the stock jumped ~25% — days before another big AI selloff. | 37:54 |
| GOOGL | Alphabet | QT · SA · STK · FA | Neutral | Just issued $80B of equity ahead of the IPO wave and is paying SpaceX $920M/month for compute at xAI data centers — both planks of his liquidity-drain and circular-financing concerns. | 35:35 |
| MSFT | Microsoft | QT · SA · STK · FA | Neutral | One of the hyperscalers whose capex-cut announcement could trigger the unwind — "at some point next year you could have Microsoft, Meta or one of the neocloud guys say… we're going to need to cut back." | 42:13 |
| META | Meta Platforms | QT · SA · STK · FA | Neutral | The easiest unwind trigger: "if tomorrow Meta says the EU is not letting us develop AI in Europe, or AI ROI isn't happening" — nearly all its FCF now goes to GPUs/data centers/memory. | 41:50 |
| AMZN | Amazon | QT · SA · STK · FA | Neutral | In the hyperscaler-capex tally now headed past $1T by 2028 — the number whose downward revision (via slowing cloud-revenue growth) would unwind the whole trade. | 43:46 |
| ORCL | Oracle | QT · SA · STK · FA | Neutral | Centerpiece of the circular-funding graphic (Bloomberg/Burry): OpenAI invests in Oracle while Oracle funds OpenAI's data centers. | 39:46 |
| AMD | Advanced Micro Devices | QT · SA · STK · FA | Neutral | Named as part of the same circular-funding web as OpenAI/Oracle. | 39:46 |
| NBIS | Nebius Group | QT · SA · STK · FA | Neutral | Listed among the circular-funding/neocloud cohort alongside Micron and CoreWeave. | 39:46 |
| DXYZ | Destiny Tech100 | SA · STK · FA | Neutral | Named as another listed pre-IPO venture fund in the VCX genre (with newcomer PWRL) — VCX was simply "the most overvalued one in the market." | 51:07 |
| Anthropic | Anthropic (private) | — | Neutral | Next up in the IPO wave — an estimated ~$60B raise at close to a $1T valuation (91% odds it announces this year per prediction markets); part of the ~$280B issuance competing for the same investor cash as SpaceX. | 52:06 |
| xAI | xAI (merged into SpaceX, Feb 2026) | — | Neutral | The reason SpaceX's losses exploded: the February merger makes IPO buyers funders of a massive AI data-center play — roughly half the ~$2.5B/quarter burn is AI infrastructure, including the 220,000-GPU Colossus. | 17:23 |
| Figure AI | Figure AI (private) | — | Neutral | Cited as the opaque end of the circular-financing web — private names "where you don't even know what the financials are." | 39:46 |
| TSLA | Tesla | QT · SA · STK · FA | Negative | "The most overvalued company in the Nasdaq and in the Mag 7" — falling FCF, falling margins, dwindling EV sales, other bets years out — held up only by the Elon premium. Prediction markets put ~50–60% odds on a SpaceX merger by mid-2027; he sees few real synergies (SolarCity redux), though "anything is possible." | 15:12 |
| VCX | VCX (Fundrise pre-IPO fund) | — · FA | Negative | Flagship short: a listed pre-IPO fund that traded at $565 — 40× its ~$20 stated NAV. Shorted from ~$300 via options (now ~$150); re-shorted at ~7–11× his $30 adjusted NAV (which already credits SpaceX at $1.75T, Anthropic near $1T) ahead of the September IPO unlock. | 50:21 |
| NVDA | NVIDIA | QT · SA · STK · FA | Negative | Hub of the round-tripping: invests billions in OpenAI/neoclouds who then buy Nvidia GPUs; Jensen "consistently goes and pumps the market" (the $2B Marvell call). In the cascade unwind, GPU names sell off after opto/memory — "this whole supply chain is just one trade." | 40:45 |
| CRWV | CoreWeave | QT · SA · STK · FA | Negative | "Effectively just a conduit that buys Nvidia GPUs and builds its own data centers" — a pass-through in the circular-financing web. | 39:46 |
| OpenAI | OpenAI (private) | — | Negative | SoftBank is "overpaying for OpenAI shares"; it sits at the center of the circular graphic (investing in Oracle while Oracle funds its data centers), and its ~$60B IPO (~Sept; 88% announce-odds) drains the same liquidity pool as SpaceX. | 8:32 |
| SFTBY | SoftBank Group | SA · STK | Negative | Masayoshi Son named among the serial AI promoters — overpaying for OpenAI shares and "continuously pounding the table on demand." | 38:22 |
A jargon-free summary of the thesis behind each name — what it actually is and why he holds that view. (Plain-language companion to the table above; renders on each ticker's consolidated page.)
EchoStar is a satellite and spectrum company. Its golden asset here: it sold wireless spectrum (the licensed airwaves Starlink needs) to SpaceX and took part of the payment in SpaceX stock — back when SpaceX was valued around $400 billion. With the IPO pricing SpaceX at $1.75 trillion, that stake alone makes EchoStar worth over $160 a share by Singh's math, versus about $116 where it trades.
This is his way of owning the SpaceX story without buying a frothy IPO: he got in with a cost basis in the 80s, so he's already sitting on profits, and if SpaceX pops on debut, EchoStar should ride along (a bit more gently). It's the "clever, cheaper way in" he always looks for instead of paying the sticker price on day one.
CVS is the drugstore chain plus a giant insurer (Aetna) and a pharmacy-benefits manager — the middleman that negotiates drug prices. Those last two businesses have been under political and earnings pressure, and the stock sold off hard.
Singh's view: if you value each piece separately and add them up (a "sum of the parts"), the stock is extremely cheap. It's part of his broader rotation into the unloved sectors — healthcare, banks, REITs — which are about as cheap versus tech as they've been in a decade.
T1 Energy aims to be a power supplier for the AI buildout. Data centers need enormous, steady electricity, and in Singh's "cascade" framework power providers are the most durable part of the AI supply chain — the last to suffer if the trade unwinds. He bought it only as a quick, short-term trade, not a core holding.
A small, risky company that hosts computing equipment ("colocation" — renting out powered, cooled data-center space). Singh bought around $3–4 a share when it won an AI-related contract worth over $1.7 billion that the market ignored, fearing the company would have to sell new shares to fund it.
The stock more than doubled quickly. He took some profit and protected the rest by selling covered calls (collecting option income in exchange for capping further upside). The lesson he draws: tiny AI-adjacent bets can be multibaggers, but you size them small so a blow-up can't hurt you.
The casino giant. Singh owned it from much lower prices, and it has now received a takeover bid from Tilman Fertitta. This is classic merger arbitrage — owning a company that's being acquired and earning the gap between today's price and the deal price — except here he also caught the jump when the bid landed.
MGM owns Las Vegas casinos, more than half of MGM China (Macau), half of the BetMGM betting app, and a new Japan resort. A buyer (the former IAC, now "Peoples") has proposed taking it private at $48.30 a share in cash.
Singh values the pieces separately at roughly $60, so he thinks the bid underpays — which is exactly why he expects a possible higher competing offer. He owns it from a low cost basis, collecting the spread either way.
SpaceX lists Friday at $135 a share — the biggest IPO ever, valuing it at $1.75 trillion, more than every listed aerospace and defense giant combined. The catch: it brought in $18.7 billion of revenue but is on pace to lose about $17 billion a year, mostly because Musk folded his AI company xAI into it — so IPO buyers are largely funding a giant AI data-center buildout, and $20 billion of the money raised immediately repays a loan. Musk keeps ~82–85% of the votes, more control than any public-company founder ever.
So why isn't Singh short? Because only about 4% of the company will actually trade, the order book is twice oversubscribed, 30% is reserved for eager retail buyers, and Nasdaq fast-tracked it into the Nasdaq-100 index (forcing index funds to buy within weeks). A tiny float plus forced buying can squeeze the price higher regardless of fundamentals. His answer: don't play the IPO at sticker price — he owns it indirectly through EchoStar at a steep discount, and he'd watch the six-month "unlock" (when insiders can finally sell) as the real test.
Warner Bros. is being bought by Paramount-Skydance with cash backed by Larry Ellison. The stock trades about $4 (roughly 15%) below the deal price because regulators haven't approved it yet — the widest big-company merger spread on the board.
For an arbitrageur that gap is the product: if antitrust clears the deal over the next year, the stock rises to the deal price and the 15% is collected. The risk is the deal being blocked, which is why the discount exists.
TSMC manufactures essentially all the world's cutting-edge AI chips. In a market full of cheerleaders, Singh calls the Taiwanese (with Korea's chipmakers and Holland's ASML) "the most conservative people in the ecosystem": they refuse to recklessly expand capacity, which stops Nvidia from flooding the market with chips.
That discipline is, in his view, what has kept the AI boom from overheating and bursting — it has stretched the cycle out by a couple of years. It's the same role OPEC plays in oil: the supplier who controls the spigot ends up controlling the cycle.
Singh calls Tesla "the most overvalued company in the Nasdaq and the Mag 7": free cash flow falling, margins falling, EV sales dwindling, and the futuristic businesses (robotaxis, robots) years away. What holds the price up is the "Elon premium" — funds and retail investors with deep conviction in Musk because he's made them rich before.
He brings it up as a warning for SpaceX buyers in both directions: the premium means SpaceX may also stay expensive far longer than fundamentals justify — but it's faith, not cash flow, doing the work. On the rumored Tesla–SpaceX merger (prediction markets say ~50–60% by mid-2027), his banker's verdict is that there are few genuine cost savings to justify it — though Musk merged SolarCity into Tesla without synergies too, so "anything is possible."
VCX is a listed fund holding stakes in hot private companies like SpaceX and OpenAI — a way for small investors to "own" the names they can't buy directly. The problem: the fund's shares traded at $565 when the assets inside were worth about $20 a share — investors were paying 40 times what the holdings were actually worth, purely to touch the SpaceX story.
Singh shorted it (via options) from around $300; it has been cut in half. He re-shorted recently: even crediting SpaceX at its full $1.75T IPO value and Anthropic near $1T, he gets a generous value of ~$30 a share against a ~$150 price. The catalyst is the September "unlock," when the 80–90% of shares currently frozen become sellable — the artificial scarcity that propped the price up disappears. It's his template trade: short the wrapper, not the asset.
Singh's worry isn't Nvidia's chips — it's the financing loop around them. Nvidia invests billions in customers (OpenAI, cloud startups), who then use that money to buy Nvidia's GPUs. That "round-tripping" inflates apparent demand. He also points at the promotion machine: Jensen Huang put $2B into Marvell, then went on TV calling it a future trillion-dollar company — the stock jumped 25%.
His scenario if it cracks: a cascade where memory and optical-component makers sell off first, then the GPU names like Nvidia, with power providers holding up last. The trigger would be a hyperscaler (Meta, Microsoft) announcing capex cuts. With semis ~20% of the index, he expects 10–15% market-level "hiccups" within two years — not a prediction of immediate collapse, but a warning that the whole supply chain has become one crowded trade.
Summary & timestamps derived from the public YouTube video (transcript in transcript.txt) for personal study. Not investment advice. © The David Lin Report for source material.