Intel Corporation [INTC]
No position is recommended. This memo issues no verdict. What follows is the implementation study, with every instrument priced from a chain actually pulled.
Spot $86.24 (2026-07-28 close). Market cap $434.9bn.
| Reading | |
|---|---|
| Quality Criteria | PASS — a genuine INFLECTION on every measure the Criteria names |
| Valuation Criteria | FAIL — required terminal margin 39.7% vs a demonstrated maximum of 24.6%, −15.1pp |
| 12-month target | UNIDENTIFIED — bound −71% to +108% |
| Downside case | −57% to −59%, named cause: Foundry wins no material external customer |
| Momentum Criteria | strongly positive — 12-1 momentum +484.9%, 1.30x the 200-day |
The long fails on the binding Criteria. Nothing here supports a long, and the honest reason is not that the business is bad — it is visibly improving — but that a 347% move has priced a terminal margin Intel has never earned.
Nor does anything here support a short, and that must be said plainly rather than left implied:
The correct output is no position and a monitoring trigger, which is the shape of the analysis rather than a hedge.
Realised volatility (252d) 77.8%. Beta 2.69.
INTC 15-Jan-2027 calls (Alpaca options snapshots, live quotes and open interest):
| Strike | Open interest | Bid | Ask | Spread | IV | Delta |
|---|---|---|---|---|---|---|
| 80 | 14,668 | 18.72 | 19.45 | 4% | 88.7% | 0.624 |
| 85 | 5,057 | 16.92 | 17.53 | 4% | 88.4% | 0.585 |
| 90 | 4,166 | 15.23 | 15.99 | 5% | 88.5% | 0.548 |
| 95 | 5,772 | 13.38 | 14.42 | 7% | 87.4% | 0.509 |
| 100 | 12,249 | 12.37 | 12.82 | 4% | 87.4% | 0.475 |
| 110 | 5,533 | 9.64 | 10.46 | 8% | 86.1% | 0.407 |
| 115 | 4,611 | 8.64 | 9.44 | 9% | 85.8% | 0.377 |
| 120 | 18,022 | 7.77 | 8.34 | 7% | 85.2% | 0.347 |
| 125 | 9,990 | 7.24 | 7.37 | 2% | 85.3% | 0.322 |
| 130 | 11,908 | 6.54 | 6.96 | 6% | 86.0% | 0.302 |
This is the deepest, tightest chain in the cluster — 12,000–18,000 open interest on round strikes with 2–7% spreads. Liquidity is not the constraint here.
IV ≈ 87% against 77.8% realised: a ratio of 1.12. A normal volatility risk premium, not punitive.
Nothing is recommended. For completeness, the only expression consistent with a Quality PASS and a Valuation FAIL is a small, defined-risk, long-dated call on the Foundry conversion — i.e. buying optionality on the single event that would make the terminal margin identifiable rather than betting on the current price.
Jan-2027 $100/$130 call spread, priced: - Buy $100 at the ask 12.82; sell $130 at the bid 6.54. Net debit $6.28. - Max value $30.00. Max profit $23.72. Risk/reward 3.78 : 1. - Breakeven $106.28, +23.2%. Max payoff above $130, +50.7%. - Both legs among the most liquid on the board: OI 12,249 and 11,908; spreads 4% and 6%. - Why defined risk and not equity: the named downside is −57% to −59%. A 3.78:1 defined-risk structure caps that at the premium while retaining the upside if external foundry revenue inflects. Maximum loss is the entire $6.28 and the probability of that is not small — the breakeven requires +23% inside six months, against a name already at the 96th percentile of its own multiple range.
Rejected, with reasons: - Long equity: rejected. Valuation Criteria is BINDING and it fails by 15.1pp. - Short equity: rejected on momentum (+484.9%), on the Quality PASS, on positive free cash flow and on sovereign ownership. Borrow was not obtainable from any free source available here and no rate is invented. - Put spreads: rejected for the same reasons as short equity. A downside case with a 35% assessed probability against +485% momentum is not a tradeable short, it is a reason not to be long. - Front-month premium: rejected. The thesis resolves on external foundry revenue, a quarterly disclosure. Nothing settles inside a month. - Outright Jan-2027 $100 call at 12.82: rejected relative to the spread. The $130 strike sold sits above the entire plausible 12-month range except the extreme, so the give-up is small and the cost halves.
77.8% realised volatility on a $434.9bn market cap → HIGH volatility tier, notwithstanding the mega-cap. Conviction on the long: NONE (binding Criteria fails). Conviction on the option: LOW, and LOW × HIGH under the sizing grid rounds down to below minimum size.
Output: no position. No entry appended to trade_recommendations.jsonl.
| Test | Result |
|---|---|
| Equity liquidity for intended size and exit | PASS — $434.9bn |
| Options chain actually pulled before any structure named | PASS — 2026-07-29, Jan-2027, 15 strikes |
| Open interest and quoted size at the specific strikes | PASS — 4,166–18,022 OI, 2–9% spreads |
| IV versus realised | PASS — 1.12x |
| Borrow / short interest | NOT OBTAINABLE — no free source available here; no rate invented. Irrelevant on the long-only fork. |
To a long, ALL THREE required: 1. External foundry revenue above ~$1bn in a quarter (from $293m), or a named external 18A/14A customer with committed volume — this is what makes the terminal margin identifiable and it is the only route to clearing the Valuation Criteria on evidence rather than assumption; AND 2. Two consecutive quarters of consolidated operating margin above 15% — which would begin to establish a central tendency where there is currently a 46.6-point range; AND 3. A price at or below ~$50 (roughly the own six-year p75 EV/Sales on NTM revenue), which is where the required terminal margin falls back inside Intel's historical range.
Invalidation of the no-position stance in the other direction: a capex guide above ~$20bn for 2027 combined with external foundry revenue still below $500m a quarter would confirm the bear mechanism — but even then the position would remain no position, not a short, for the momentum and sovereign -ownership reasons in §1.
Next checkpoint: Q3 2026 results, late October 2026 (not announced — see
INTC_Catalyst_Calendar.md). The one line to read first is the external third-party foundry revenue
sentence in the segment note.
ORCL, INTC and DELL were underwritten as one batch and share an end-market. The 252-day pairwise
correlation matrix against existing book positions was not computed in this run — it is a
portfolio-book task and is flagged rather than fabricated. The prior should be high positive
correlation: all three re-rated on the same AI-capex narrative inside twelve months. The cluster
should compete for one slot, not three. Intel's Valuation FAIL means it should not be the name
that wins it.