Micheal Burry’s Massive Bets Against Palantir and Oracle Highlight Two Prominent AI Failure Modalities
This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.
Palantir (NASDAQ:PLTR | PLTR Price Prediction) and Oracle (NYSE:ORCL) both beat estimates in their latest reports, and Michael Burry holds put positions in both. Palantir is the valuation risk: almost no debt, but priced as if it can’t slip up. Oracle is the balance-sheet risk, with a datacenter expansion that burns cash. Their market caps sit close together, at ~$387.5B and ~$390.9B.
Palantir Sells Workflows While Oracle Sells Megawatts
Palantir’s Q3 revenue rose 95.9% to $2.31B[1]. U.S. commercial revenue reached $984M, up 148%[1], and passed U.S. government sales for the first time. That shift matters because companies buy faster than the government does. Pilots turn into contracts quickly: one industrial materials customer went from a nearly $3 million[2] trial to a $29 million ACV deal. CEO Alex Karp put it this way: “Once you go Palantir, you don’t go back.”[2]
Oracle’s quarter ran on hardware. Cloud infrastructure revenue rose 121% to $7.39B[3], while software license revenue fell 15%. GPU rentals are taking over from the old licensing business. GPU use hit 97.9%[4], so demand looks real.
| Business Driver | Palantir | Oracle |
|---|---|---|
| Growth Engine | AIP and Ontology deployments | AI training and inference capacity |
| Adjusted EPS | $0.44[1] | $1.92[3] |
| Quarterly Capex | ~$8M | $28.50B[3] |
| Free Cash Flow | $1.216B adjusted[2] | Negative $5.40B[3] |
One Risk Lives in the Multiple, the Other in the Balance Sheet
Palantir’s weak spot is its price. At 238 times earnings with a 0.54% free cash flow yield, the stock assumes hypergrowth keeps going. There are cracks. Total contract value grew just 6%[1]. Stock-based pay rose to $318.9M from $172.3M[1]. Receivables reached $1.88B from $1.04B[1]. Management says its shorter Sovereign AI contracts will get larger later. Maybe.
Oracle’s weak spot is funding. Capex jumped 235.2%[3], total liabilities rose to $236.06B, and interest expense increased 55% to $1.4B. Co-CEO Clay Magouyrk said new contracts require no extra cash: “We said it does not require additional cash from Oracle.”[4] Even so, Oracle gave no timeline for getting back to positive free cash flow.
Contract Conversion Will Decide Who Breaks First
I will be watching whether Palantir’s contracts get longer as it works toward Q4 revenue guidance of $2.649B to $2.653B[1]. For Oracle, watch the $664B[3] backlog of signed contracts. Management expects around half[4] of it to turn into revenue within 36 months.
Why I Prefer Palantir’s Problem Over Oracle’s
Over six months, Palantir rose 25.38% while Oracle fell 34.57%. I find Palantir’s risk easier to live with. Its valuation can shrink, but a company with $1.20B[1] in quarterly operating cash flow and almost no debt can absorb that. Oracle’s risk builds on itself: if AI customers slip up, the debt is still there.
I’d want Palantir’s contract bookings to speed up before I trust its multiple. Oracle’s $0.50 quarterly dividend may draw income investors, but I’d wait for a free cash flow timeline. If that timeline comes, my view changes.
Contact [email protected] for any questions or corrections.







