AAPL › analyze
AAPL · Analyze from before
Date: 2026-09-10 | Price: $326.57 (+3.8% vs baseline's $314.58) | Mkt Cap: $4.77T
0. What This Updates
Baseline: Output/Stocks/Technology/AAPL/analyze-2026-08-27.md — WATCH, conviction 6.5,
fair value $258-305, entry $255-280, trim 36x fwd. kb.py find AAPL and a directory listing
confirm this is the only prior verdict-bearing report on AAPL — no second-newest report exists,
so there is no "survived a prior re-test" history to weight yet.
Event list since 2026-08-27 (what actually happened, driving what gets re-tested):
- CEO transition executed. Tim Cook stepped down as CEO effective 2026-09-01, becoming Executive Chairman; John Ternus (SVP Hardware Engineering, 25-yr Apple veteran — iPad, Mac, AirPods, Apple Silicon transition, Vision Pro) became CEO. Announced back in April 2026 (5 months' notice, not a surprise), unanimous board approval, described as the culmination of a long-planned succession. Arthur Levinson moved from non-executive chairman to lead independent director the same day. Apple Newsroom, Al Jazeera, ABC News, NBC News, Fox Business. Confirmed real — primary corporate source plus four independent major outlets agree on every detail (date, names, titles, board process).
- Market reaction: positive, muted, and already priced. Shares rose ~2.6% on 2026-09-01 (Ternus's first day) while the S&P 500 fell — a smooth-succession read, not a leadership-risk read. Yahoo Finance
- "Surprise and Shine" product event, 2026-09-09 — Ternus's first keynote as CEO. Launched iPhone 18 Pro/Pro Max (A20 Pro chip, variable-aperture 48MP camera, +$100 YoY to $1,199/$1,299) and iPhone Duo, Apple's first foldable ($1,999 for 256GB) — a new hardware category, not an ecosystem or business-model change. Huawei and Xiaomi launched competing foldables the same week (Mate XT2, 18 Fold) — Apple entering an already-contested category, not creating one. CNBC
- No new financial print. FY2026 closes 2026-09-26; Q4 FY26 has not reported. Every revenue/earnings/margin figure below is therefore the same TTM window as the baseline (Q4 FY25 through Q3 FY26) — this pass re-tests interpretation and price, not new fundamentals.
- Regulatory: no change in state. The Google-search antitrust remedies (final judgment 2025-12-05, before the baseline) already ruled out a full stop to the Google TAC payments to Apple; both DOJ/states (wanting harsher remedies) and Google (wanting lighter ones) have pending appeals with no ruling in this window. Google's own EU DMA compliance deadline (~late Sept 2026) is a Google obligation, not an Apple one. Net: same live, unresolved overhang as the baseline described — nothing resolved, nothing escalated.
- Analyst coverage: maintains, not re-ratings. Nine actions since baseline (TD Cowen, Rosenblatt, BofA, HSBC, Needham, Morgan Stanley, DA Davidson all reiterated ratings; BofA trimmed its target $380→$370). None cited the CEO transition as a reason to change a rating — itself a finding (see §2).
1. The Delta Ledger
No RETRACTED or SUPERSEDED rows this pass — nothing in the baseline was wrong, and no fact pattern reversed. The dominant movement is in the Price rows (spot ran further past the fair value band) and one genuinely NEW row (the CEO transition itself).
| # | Claim (baseline) | Type | Status | Evidence this pass |
|---|---|---|---|---|
| 1 | FCF plateaued $99-111B over 4yr window, -3.9% 3yr CAGR; window capped at 4yr (data gap, not fabricated) | Trend | ✅ CARRIED | No new annual print since FY2025. Yahoo's window is still 4yr, roic.ai free tier still 2yr — the gap is unresolved for a second pass (see §6). |
| 2 | Buybacks+dividends = 107% of FY25 FCF, funded by securities drawdown, not new debt | Trend | ✅ CARRIED | No new annual cash-flow statement to re-test against; balance-sheet trend (below) is directionally consistent. |
| 3 | Debt/Assets deleveraging, 27.5%→22.0% through Jun-26 | State/Trend | ✅ CARRIED | Same quarter (Jun-26) is still the latest on the books; no new data point, but nothing contradicts it. |
| 4 | Shares outstanding declining ~2.5-3%/yr | Trend | 🔁 REFRESHED | 14.67B (Jun-26, baseline) → 14.59B (current, per Yahoo sharesOutstanding) — buybacks continued through the gap at a pace consistent with the stated annual rate. |
| 5 | Revenue TTM +16.4% YoY, iPhone-17-led | State | ✅ CARRIED | TTM revenue $466.82B — identical figure, same four quarters, no new print. |
| 6 | Earnings TTM +28.7% YoY, partly inflated by a Q3 FY26 tariff-refund one-off | State | ✅ CARRIED | TTM net income $128.93B — identical to baseline; the one-off character of the refund is unchanged (no new quarter to re-test it against). |
| 7 | FCF/share TTM ≈$9.25, corrected from the broken vendor snapshot field | State | ✅ CARRIED | Same underlying quarters; the vendor snapshot field is still broken this pull (freeCashflow reads $107.72B vs the summed $136.68B) — pitfall-yahoo-snapshot-fcf-field-diverges-from-quarter-sum still fires identically. |
| 8 | ROIC 59.8% (FY25) rising from 50.3% (FY24); ROE 171% is a buyback artifact, not a moat signal | Trend/Structural | ✅ CARRIED | No new annual print to move ROIC; the ROE-is-not-load-bearing reasoning is untouched by anything found this pass. |
| 9 | Services 28.1% of revenue at 75.6% GM, 42.4% of gross profit, fastest-growing segment | Structural | ✅ CARRIED | Same TTM quarters. |
| 10 | Moat = ecosystem switching costs (iMessage, Family Sharing, AirDrop, iCloud), not hardware specs | Structural | ✅ CARRIED | The Sep-9 event reinforces rather than undercuts this: Apple's new foldable competes on hardware in a category Huawei/Xiaomi already occupy — the switching-cost argument, not the hardware, is what the baseline said carries the moat, and nothing here contradicts that. |
| 11 | Real threat is regulatory (EU DMA, Google TAC), not competitive entry | Structural | ✅ CARRIED | No ruling moved in either direction this window (see event #5). Same live, unresolved risk. |
| 12 | AI = fast-follower strategy: on-device/PCC + China-specific Alibaba hedge | Structural | ✅ CARRIED, with an open question flagged | The Sep-9 event added "AI features" per press coverage, but no strategic redirection was announced — still an incremental, not a frontier, posture. Whether Ternus specifically accelerates this is genuinely unresolved — see §5/§6, not folded into this CARRIED status. |
| 13 | China: soft single quarter (Q3 FY26 miss), strong 9mo trend (+30% YoY) — no structural loss | State | ✅ CARRIED | No new quarter; unchanged. |
| 14 | Manufacturing concentration (Foxconn/China-heavy) is a live tariff exposure, evidenced by the Q3 tariff-refund tailwind | Structural | ✅ CARRIED | No new information moved this either way this window. |
| 15 | Evergreen on a 5-10yr view | Judgment | ✅ CARRIED | Adjudicated last, per §4 — nothing beneath it broke. |
| 16 | Regulatory moat-defense = compliance-and-negotiate, no structural insulation | Judgment | ✅ CARRIED, corroborated further | Cook's new role explicitly includes "engaging with policymakers around the world" (Apple Newsroom) — the company kept its chief regulatory relationship-holder in a policy-facing seat rather than losing that capability at the CEO transition. This is a small positive force on execution of the same posture, not a change in the posture itself. |
| 17 | Correct near-term multiple is priceEpsCurrentYear (not headline forwardPE), because Yahoo's forwardEps prices FY2027, not the closing FY2026 |
Price | 🔁 REFRESHED | Same fields, same trap, new numbers: epsCurrentYear is unchanged at $8.82 (estimate stability); priceEpsCurrentYear moved 35.65x → 37.01x — the whole move is price, not earnings estimates. pitfall-vendor-forward-eps-is-the-wrong-fiscal-year still fires identically. |
| 18 | Graham IV near-zero weight — buyback-hollowed book (BVPS $7.36) makes the formula meaningless here | Judgment | ✅ CARRIED | BVPS unchanged at $7.36; reasoning unchanged. |
| 19 | DYT: yield (0.34%) below 5yr average (0.50%) reads as expensive | Price | ✅ CARRIED | Identical figures this pull — no dividend action in the window. |
| 20 | Bogle expected return ~9-11%/yr if multiple holds flat; compression risk given position near 52wk high | Judgment | 📉 DRIFTED | The compression risk named in the baseline moved toward materializing: price is now above the consensus mean target (see #21) for the first time, and closer to the trim ceiling. Not broken — still a live "if the multiple holds" thesis — but the cushion for that "if" shrank. |
| 21 | Consensus mean target ~$324.45 implied ~3% upside from $314.58 | State | 🔁 REFRESHED | Mean target now $323.86 (essentially flat — analysts did not chase the price) against a price of $326.57 — spot has crossed above the mean target for the first time in this file's history. Target dispersion also unchanged in shape: $245 (Barclays, bear) to $400 (Rothschild + TD Cowen, bull), median $335. |
| 22 | Insiders: steady, unbroken open-market selling, no offsetting buying | State | ✅ CARRIED | One new transaction in the window: General Counsel Newstead sold again on 2026-09-01 at $317.01 — her fifth consecutive scheduled sale since June, each at a higher price than the last ($307.49→$307.75→$310.95→$317.01). Consistent with a 10b5-1 plan, not a fresh signal. No new insider buying. No transactions yet from Ternus in the new-CEO seat (too soon to expect any). |
| 23 | Fair value $258-305 / Entry $255-280 / Trim 36x fwd | Price | 🔁 REFRESHED | Re-derived from scratch in §4 below: $259-307 / $259-278 / 36x fwd (unchanged multiple). The band itself barely moved (forward EPS estimate ticked $9.54→$9.58); what moved is the gap between spot and the band. |
| 24 | Verdict WATCH, conviction 6.5 | Judgment | 🔁 REFRESHED → WATCH, conviction 6.0 | See §4. |
NEW this pass (no baseline counterpart):
| # | Finding | Type | Status |
|---|---|---|---|
| 25 | CEO transition Cook→Ternus, executed cleanly, telegraphed 5 months in advance, market-neutral-to-positive reaction, Cook retained specifically for policy/regulatory relationships | Structural (governance) | 🆕 NEW — see §3 |
| 26 | Apple's first foldable device (iPhone Duo) and a richer iPhone 18 Pro price umbrella (+$100 YoY) launched under the new CEO | State | 🆕 NEW — too early to read as a data point either way (see §6) |
| 27 | PEG ratio 2.48-2.52 (Yahoo) — not computed in the baseline | Price | 🆕 NEW — corroborating color for "expensive," not load-bearing on its own |
2. How the Close Calls Were Decided
Did the CEO transition change any Structural claim? No — and this needed checking carefully rather than assuming it. The three plausible channels were tested: - Capital allocation discipline (buyback-heavy policy, claim #2): unchanged personnel — Kevan Parekh remains CFO, the officer who actually runs the capital-return program. No signal of a change in policy from either the transition announcement or the Sep-9 event. - AI strategy (claim #12): analysts explicitly framed Ternus as someone "expected to accelerate" Apple's AI push, given his hardware-integration background. That is a forward expectation, not yet an action — the Sep-9 keynote added incremental "AI features" language but no strategic redirection. Recorded as an open question (§6), not folded into a status change, because a single earnings-call cycle has not yet happened under the new CEO. - Regulatory posture (claim #16): this is the one channel with an actual, verifiable fact — Cook's new mandate explicitly retains the policymaker-relationship function. That is a small positive corroboration of continuity, weighed in, not overweighted (one data point, and the regulatory outcome itself hasn't moved either way).
Net read on the transition: it is real, it is resolved smoothly, and the market had already priced it as a non-event by the time this file was written (analyst maintains outnumbered re-ratings 9-to-0 in the window, and the CEO-day stock pop happened before this analysis, not as new information within it). The transition earns a NEW row, not a rewrite of any existing Structural claim, because nothing about what Apple sells, who buys it, or why it's hard to displace changed on 2026-09-01.
Self-check on the fair-value anchor (per
Knowledge/Playbook/principle-repeated-fair-value-above-spot-is-method-failure): this is now
the second consecutive file where spot sits above the top of the fair-value band ($314.58
vs $305 top in August; $326.57 vs $307 top now), and the gap widened. That pattern — in the
opposite direction from the principle's TTD case, where FV kept sitting above a falling price —
is the mirror failure mode: a band that's persistently too conservative for a business whose
own fundamentals (ROIC, Services mix) are still improving, not deteriorating. The principle's
test is whether the anchor was "earned under a growth regime that has ended" — for TTD, yes;
for Apple, no, ROIC and the Services mix shift are still moving in the band's favor, not
against it. On that basis the 27-32x anchor is not widened this pass — but if the band's top is
breached a third consecutive time at the next re-analysis without a fundamental reason
(a real margin or ROIC improvement, not just multiple expansion), the anchor itself should be
audited rather than re-published unchanged a fourth time.
3. Thesis Persistence and Conviction Delta
Persistence: every Structural and Trend claim in the ledger (10 of 10: #1-4, #8-12, #14) survived as CARRIED or REFRESHED. That is 100% persistence against a real, confirmed governance event — the textbook "business held, only the multiple moved" pattern the framework flags as worth naming explicitly.
Conviction: 6.5 → 6.0. The specific rows driving the move: - Down: #17 (near-term multiple 35.65x→37.01x on flat EPS estimates), #20 (Bogle compression risk moved from hypothetical toward live), #21 (spot crossed above consensus mean target for the first time), #23 (gap between spot and the trim ceiling narrowed from ~$28/8.3% to ~$18/5.4%). All four are the same underlying fact — price ran 3.8% ahead of unchanged earnings estimates — counted once as "the valuation cushion shrank," not four times. - Flat/neutral: #25 (CEO transition) — confirmed real, resolved cleanly, no evidence it changes the business, and the market had already finished pricing it before this file was written. It removes a hypothetical overhang the baseline never actually priced (the baseline had no leadership-risk discount to begin with), so resolving it cleanly adds no fresh upside either — net contribution to conviction is zero, named explicitly rather than silently rolled into the score. - Not moved: all ten Structural/Trend rows, because nothing tested against them broke.
This is a small, named move, not a re-rating of the business — the business case is unchanged; the price got 3.8% less attractive with no offsetting fundamental improvement to justify it.
4. Verdict — Updated Valuation, Entry, Trim
Two questions, re-asked:
- Is this a good business? Unchanged: yes, unambiguously. Every Structural and Trend claim in the ledger survived untouched. ROIC (59.8%, rising), the Services mix shift, the buyback engine, and the ecosystem moat are all exactly as strong as the baseline found them — this pass found no new information on the business itself, only on the price and the person in the corner office.
- Has the market priced that in? More so than in August. TTM/current-year multiple is now 37.0x (was 35.65x), the stock trades above its own analyst consensus mean target for the first time in this file's history, and it sits 94.6% of the way up its 52-week range ($226.65- $344.57), closer to the trim ceiling than at the last check.
Fair value: $259 - $307 (27x-32x × $9.58 FY27 fwd EPS — anchor unchanged, EPS estimate up $0.04 from baseline; see §2 for why the anchor was not widened this pass).
Entry: $259 - $278 (27-29x FY27 fwd EPS, a genuine margin of safety — essentially unchanged from the baseline's $255-280, the $1-2 wobble is rounding on the EPS estimate, not a new view).
Trim: 36x fwd (≈$345 at current forward EPS — unchanged multiple; the dollar level drifts up mechanically with the EPS estimate per CLAUDE.md's rule, now within 5.4% of spot vs 8.3% at the last check).
Verdict: WATCH. Conviction 6.0/10 (down from 6.5). Still a Great business at a price the framework's own quadrant calls "wait, not buy" — now with slightly less room between spot and that judgment than a month ago, and a leadership change that resolved as a non-event rather than as either a new risk or a new reason for optimism.
Key risks (updated): - Valuation cushion continues to shrink with no fundamental catalyst behind the move — the stock is being bid up on multiple expansion, not on any new print (there hasn't been one since August). - Regulatory overhang (EU DMA, Google TAC/antitrust appeals) is unresolved and unchanged — still targets the highest-margin, fastest-growing Services line specifically. - New: whether Ternus's tenure brings any change to Apple's AI posture or capital-allocation discipline is a genuinely open question that this pass could not test — the first real evidence arrives with his first earnings call as CEO (Q4 FY26, ~late Oct 2026). - iPhone Duo (Apple's first foldable) enters a category Huawei and Xiaomi already occupy — no evidence yet either way on reception or sell-through; too early to score.
Upgrade condition (unchanged from baseline): a pullback into the $259-278 entry zone — plausible on post-earnings digestion, a China miss, or a broader multiple reset — converts this to a much higher-conviction name without the business needing to change at all.
5. What Is Genuinely New
- The CEO transition itself (claim #25) — the central question this pass was commissioned to answer. Verdict: confirmed real, resolved cleanly, does not change the thesis. See §2-3.
- Apple's first foldable device and a richer iPhone Pro price ladder, launched as the new CEO's first product event (claim #26) — a hardware milestone, not yet a data point on demand or margin; the framework has nothing to test here until sales data exists.
- PEG ratio (2.48-2.52) as a new corroborating "expensive" data point (claim #27) — not load- bearing, consistent with everything else in the Price rows.
6. What This Pass Did NOT Test
- The 5-8yr FCF/ROIC CAGR window is still capped short of the framework's target — Yahoo's free window remains ~4yr, roic.ai's free tier remains capped at 2yr. This is the second consecutive pass carrying this same data gap forward unresolved (first flagged in the 2026-08-27 baseline). If a paid roic.ai tier or a 10-K-sourced longer series is available before the next re-analysis, that is the specific thing to go get — flagging it twice without closing it is exactly the "keeps deferring the same question" pattern this command is meant to surface.
- Whether Ternus changes AI strategy or capital-allocation policy — genuinely unresolved. Nothing in the Sep-9 event or the transition announcement gives a testable answer either way. First real evidence: the Q4 FY26 earnings call, his first as CEO.
- iPhone Duo / iPhone 18 reception and sell-through — the product launched the day before this file was written. No sales, review-cycle, or channel-check data exists yet to score it.
- The Google-search antitrust appeal timeline — both sides have pending appeals with no scheduled ruling found in this window. Carried forward as an unresolved, undated overhang, not estimated.
- This pass leaned on WebSearch and vendor snapshots for the CEO-transition and analyst-action facts; it did not independently open Apple's 8-K or proxy filing on the transition. Corroboration came from one primary corporate source (Apple Newsroom) plus four independent major outlets in full agreement — treated as sufficient, but a single-document primary-source read was not done.
Report by Financebot. Data as of 2026-09-10. Baseline: analyze-2026-08-27.md. Sources:
.mcp/fin.py, Yahoo Finance MCP (get_stock_info, get_recommendations, insider transactions),
WebSearch (Apple Newsroom, Al Jazeera, ABC News, NBC News, Fox Business, CNBC, Yahoo Finance,
MacRumors — see inline citations above). No Knowledge/ or Watchlist writes performed — scoped to
this report only per orchestrator instruction; shared-file updates (Watchlist conviction/trim,
Watchlist.zh.json, kb.py index) are handled centrally.