Scorecard
Theses we're tracking
Every piece states in advance what would prove it wrong. This is the running list, and we mark each one when the data comes in, whichever way it goes.
16
Open thresholds
0
Tripped so far
8
Pieces tracked
Open thresholds
| Piece | What would prove us wrong | Next check | Status |
|---|---|---|---|
| $NVDA is locking up suppliers with commitments, not cash Oct 21, 2026 | the Q3 FY27 10-Q shows supply and capacity commitments back below ~$119B (the pre-jump level), meaning the lock-up is unwinding. Or, if NVIDIA breaks out prepaid supply agreements again, a balance above ~$10B (about 2x the FY25 peak), meaning the lock-up is moving into cash. | Q3 FY27 10-Q | Open |
| Not-yet-started leases vs cash capex Oct 21, 2026 | The Q3 10-Qs (late October) show the not-started total flat or down, meaning commitments are converting, not stacking. | Q3 10-Qs (late Oct; $ORCL in Dec) | Open |
| Not-yet-started leases vs cash capex Oct 21, 2026 | Leases commencing show up as a step-up in lease assets and liabilities with no new signings. | Q3 10-Qs (late Oct; $ORCL in Dec) | Open |
| Frontier labs are buying into industries Oct 21, 2026 | industry-workflow deals fall to 25% of lab deals or less over the next year (27% now), or Anthropic's next filing still shows usage at 80%+ of revenue with no industry product line broken out. | Next lab filing / deal count | Open |
| Power meter: the grid is the limit Oct 20, 2026 | ERCOT's eligibility audit (report due Dec 10) cuts the conditionally included load by more than half. | ERCOT audit report (Dec 10) | Open |
| Power meter: the grid is the limit Oct 20, 2026 | PJM's next auction clears below the cap with no shortfall. | PJM 2029/30 auction (Dec 2026) | Open |
| Power meter: the grid is the limit Oct 20, 2026 | Indiana's regulator, once the contract is filed, rejects or materially changes the customer-funded structure. | NIPSCO contract filing at the IURC | Open |
| The $AMZN supplier-warrant ledger Oct 20, 2026 | the pattern breaks if new warrants move away from optics, power and server parts, or if vesting stalls in the filings (spend not showing up). | Next $AMZN 10-Q | Open |
| Chip output vs Korea chip exports Oct 20, 2026 | the volume-value gap closing below 1.5x on our same calc (our pre-set line), or MOTIE crediting the gains mainly to shipment volume rather than prices. MOTIE currently cites both. | Next MOTIE monthly release | Open |
| $BRK.B: defensive by cash, cyclical by business Oct 19, 2026 | The Q3 10-Q (early Nov) shows BNSF and the manufacturing, service and retail group holding or growing year over year while $SPY earnings estimates fall. | Q3 10-Q (early Nov) | Open |
| $BRK.B: defensive by cash, cyclical by business Oct 19, 2026 | The 1-year beta climbs back above ~0.6 by Q1 2027 without a change in the cash pile. | Q3 10-Q (early Nov) | Open |
| $BRK.B: defensive by cash, cyclical by business Oct 19, 2026 | Cash and T-bills drop below ~20% of total assets (from 28.9%) through deals or share repurchases. | Q3 10-Q (early Nov) | Open |
| $BRK.B: defensive by cash, cyclical by business Oct 19, 2026 | Interest income stops falling year over year, even if the Fed cuts. | Q3 10-Q (early Nov) | Open |
| $BRK.B: defensive by cash, cyclical by business Oct 19, 2026 | $SPY falls more than 10% within 3 months of Oct 16. One case wouldn't overturn 79, but we'd note it. | Q3 10-Q (early Nov) | Open |
| Oil spikes vs stocks: an event study Oct 19, 2026 | Future spikes push the 12-month hit rate below 70%, or the median more than 5 points under the all-days rate, with 40+ events. | Next qualifying spike | Open |
| Oil spikes vs stocks: an event study Oct 19, 2026 | A spike outside a recession or bear market is followed by a 12-month loss of 15% or worse. | Next qualifying spike | Open |
Statuses describe our own published thresholds, not securities. Not investment advice.
Get each new piece by email
New research, free by email on Substack.