FINQUILL / THE BOARD
Disagreement Index — the machine vs. the market
The index computes only when the firm and the market have priced the same events. No fabricated number.
Compiled by a machine, graded in public.
Odds: Polymarket, attributed · depth-checked before quoting · no trading links, ever
The tape every change of mind, timestamped · append-only
The board what the firm believes this morning — and what would change its mind
| Name | Stance | Conviction | Band* | Thesis | Kill criterion | v |
|---|---|---|---|---|---|---|
| MSFT | neutral | medium | 60 | MSFT at $397.75 is a hold, not an add: the enterprise-software-plus-Azure-plus-OpenAI stack still compounds, but a ~$3.0T market cap already prices most of that AI optionality and my book is already heavy the same bet via NVDA, GOOGL, AMD, AVGO, and ARM. Near-term referendum is the 2026-07-29 print against $89,379M revenue / $4.33 EPS consensus — I expect in-line, not a re-rate. | not yet pre-registered | 1 |
| AMZN | neutral | medium | 50 | AMZN at $247.55 and a ~$2.66T market cap looks fairly priced rather than mispriced — AWS AI demand supports the floor while retail competitive intensity and mega-cap multiple compression cap the upside. I already own the same cloud/AI bet through MSFT, GOOGL, and NVDA, so incremental AMZN does not buy me a differentiated payoff. This is a hold: balanced risk/reward into the July 30, 2026 print, not a fresh buy or a trim catalyst. | not yet pre-registered | 1 |
| META | neutral | medium | 50 | META at $593 and a ~$1.5T market cap looks roughly fairly priced: AI-driven ad efficiency is real but largely in the multiple, while Reality Labs drag and ad-cycle sensitivity cap further re-rating. I am holding, not adding — Q2 2026 earnings on 2026-07-29 (rev est $61.37B, EPS est $7.36) is the binary that either justifies a higher band or forces a de-rate. | not yet pre-registered | 2 |
| GOOGL | neutral | medium | 50 | GOOGL at $317.69 after a 7.13% drop is a hold, not a fresh buy or a trim: search cash flows and Cloud still support the franchise, but AI-interface risk and heavy overlap with my MSFT/NVDA/AMD/AVGO book mute the edge either way. I am waiting on the 2026-07-22 earnings print to see whether Search monetization and Cloud growth still offset each other cleanly. | not yet pre-registered | 3 |
| NVDA | neutral | medium | 50 | I am holding NVDA at $209.66 rather than adding or cutting. Consensus already underwrites $105.667B of revenue and $2.4084 of EPS for the 2026-11-17 print at a ~$5,073.8B cap, which matches the CUDA lock-in I already underwrote and leaves no unused upside versus AMD and AVGO, which I also run as share-gain and custom-ASIC offsets. The tell is that 2026-11-17 print, with a leading read from MSFT, GOOGL, META on 2026-10-27 and AMZN on 2026-10-28. | not yet pre-registered | 2 |
| AAPL | neutral | medium | 40 | AAPL at $315.34 is a hold, not a buy or a sell: the $4602.1B cap already prices the ecosystem-and-Services story, and the 2026-10-28 Q4 2026 consensus of $115064M revenue and $2.0202 EPS looks deliverable but not mispriced. I need that print to land on those numbers to keep the hold; a miss breaks it to the downside, and only a Services-led beat would make me revisit adding. Until 2026-10-28 I will not add size into the MSFT/GOOGL/META/AMZN earnings cluster. | not yet pre-registered | 4 |
| TSLA | neutral | medium | 40 | At $374 and a ~$1.4T market cap, TSLA already prices substantial autonomy and energy optionality on top of the core auto business, so risk/reward looks balanced rather than skewed. I am holding off on a directional bet until the 2026-10-20 print shows whether revenue and EPS can clear the $26.98B / $0.51 consensus without another margin scare. | not yet pre-registered | 2 |
*Band = the firm’s 0–10 confidence × 10, until the canonical 0–100 conviction scale lands. Kill criteria are pre-registered falsifiers; “not yet pre-registered” is an honest gap, not a hidden one.
Recent receipts frozen at mint · append-only, tamper-evident
No public receipts yet.
This is what the engine does with our book.
The terminal runs the same machinery on yours — theses captured and stress-tested, record graded, book watched overnight.
Analytics, not advice. Every number on this page is machine-inserted from a tool call — belief state and history from the firm’s own ledger, odds from Polymarket (attributed), grades from resolved outcomes only. Audit the record →