JPM 27 min read

JPMorgan Chase & Co. (JPM) — Investment Tree v1

Stage 7 final essay. Bilingual companion: tree_v1_zh.md. Date: 2026-07-02 · Anchor price: $333.79† (derived from BVPS $128.38 × P/B 2.6x; live-quote R2 verification owed) · Trailing P/E: ~14.8-16x† · Forward P/E: ~14x† · P/B: ~2.6x† · Dividend yield: ~2.0%† Archetype: Money-center bank fortress balance sheet with cyclical-peak earnings and succession reshuffle — closest analogue is JPM's own 2021 post-COVID-stimulus earnings cycle, which normalized over 4-8 quarters through 2022-2023.

SOURCE QUALITY: Tier B (GENERATE mode) throughout. Direct WebFetch access to SEC EDGAR and jpmorganchase.com investor-relations documents returned HTTP 403 from cloud sandbox for every URL attempted this session. All figures relayed via secondary aggregators (StockTitan, GuruFocus, Motley Fool, CNBC, Nasdaq) citing the primary 10-K/10-Q/8-K filings, not direct primary-document confirmation. Known discrepancies flagged inline (†): efficiency ratio (53% vs. 54%), tangible book value per share ($104.39 vs. $108.87), diluted shares outstanding (2.72B vs. 2.79B), exact segment-level net income dollar figures, and the Category 2 (Capital Markets & Trading) revenue figure inconsistency flagged in taxonomy.md. R2 verification against primary filings is required before this tree is treated as final, particularly for any leaf where these specific figures are load-bearing.

0. Company Fundamentals — what JPMorgan Chase is and how it earns

Figures FY2025 (calendar 2025) / Q1 2026 unless noted. All Tier B unless marked.

What it is & how it earns. JPMorgan Chase is the largest bank holding company in the United States by assets (~$4.9T, Q1 2026), operating across four lines: Consumer & Community Banking (CCB, ~$14.4B FY2025 net income†, 32% ROE), Commercial & Investment Bank (CIB, 18-21% ROE), Asset & Wealth Management (AWM, 40% ROE on $4.8T AUM), and Corporate. It earns through net interest income on a ~$1.53T loan book funded by ~$2.68T of deposits, trading/Markets revenue (Fixed Income + Equities), investment-banking fees (#1 global ranking, ~8.4% wallet share), AWM management fees, and card/payments fees — net of the credit-cost line (provision for credit losses). FY2025 net revenue $182.4B, net income $57.0B, firmwide ROE 17% / ROTCE 20%. Q1 2026 stepped up sharply: net income $16.5B (+13% YoY), EPS $5.94 (+17% YoY), ROE 19% / ROTCE 23% — a 300bps ROTCE improvement in a single quarter that is this tree's central analytical question.

Cash-flow anatomy. As a regulated bank holding company, JPM's "capital generation" framing substitutes for a traditional FCF waterfall — excess capital above regulatory minimums is what funds capital return:

JPM returns the substantial majority of excess capital via dividend + buyback, consistent with its 14-consecutive-year dividend-growth track and consistent large-scale repurchase program.

Balance sheet & capital allocation. Standardized CET1 14.3% against an 11.5% requirement (~280bps buffer). 2026 CCAR/DFAST stress test held the Stress Capital Buffer at 2.5% through Sep 30, 2027, with projected 9-quarter cumulative pre-tax net income of $11.4B even under the Severely Adverse scenario (peak unemployment 10.0%, equity-market trough -58%) — direct evidence the fortress-balance-sheet framing survives real stress-testing, not just marketing language.

What drives it. Four distinct cyclical drivers, per taxonomy.md's custom axis: the rate cycle (NII), the capital-markets cycle (trading/IB fees), a low-cyclicality AUM/fee engine (AWM), and the credit cycle (provisioning). The tree resolves whether the market's willingness to price 23% ROTCE as durable correctly separates these four drivers — or is extrapolating a capital-markets-and-credit-cycle peak as a structural plateau.

I. One-sentence verdict

JPM's ~$333.79†/~2.6x P/B/23% Q1 2026 ROTCE combination prices the market's richest-in-peer-group premium on the assumption that this quarter's blended earnings print is a fair representation of durable forward earnings power, when the taxonomy-level decomposition shows a disproportionate share of the recent improvement came from the two most cyclically-sensitive lines in the P&L (capital-markets trading revenue and a credit-cost trough) rather than the core, flat-guided NII franchise — and the market has separately not priced the freshly-concrete Dimon-succession risk (Marianne Lake's June 25, 2026 exit) at all; the tree's own math finds current pricing offers negative expected value (-6.1%) and an unfavorable 0.78x risk/reward asymmetry, making this a WATCH position at 0% today, with a contingent 1-2% starter appropriate only if Q2 2026 earnings (July 14, 2026) confirm a partial-normalization path AND the price moves toward the $315-327 range this tree's own reverse-engineering identifies as the point where risk/reward turns favorable.

II. Company snapshot

JPMorgan Chase & Co. is the largest US bank holding company by assets (~$4.9T, Q1 2026), organized into three reportable operating segments — Consumer & Community Banking (CCB, retail/mass-affluent banking, cards, auto, home lending; ~32% ROE) — Commercial & Investment Bank (CIB, corporate/institutional lending, M&A advisory, underwriting, sales & trading; 18-21% ROE) — Asset & Wealth Management (AWM, $4.8T AUM, +18% YoY, 40% ROE, 36% pre-tax margin) — plus a Corporate segment. FY2025 net revenue $182.4B (managed basis), net income $57.0B, EPS $20.02.

Q1 2026 (reported April 14, 2026) was the print that set up this tree's central question: net income $16.5B (+13% YoY), EPS $5.94 (+17% YoY), ROE 19% / ROTCE 23% (up from FY2025's 20%), net revenue $49.8B reported / $50.5B managed (+10% YoY), driven substantially by record trading revenue (fixed income +21% YoY on top of FY2025's Equities +40% YoY). FY2026 NII guidance excluding Markets sits at ~$95B, described as "broadly flat YoY" — a deceleration signal directly beneath the trading-driven headline beat. Layered on top: a $50B buyback authorized effective July 1, 2026 and a 10% dividend hike to $1.65/quarter, both announced the same day (June 25, 2026) that Marianne Lake — the closest-watched internal frontrunner to succeed CEO Jamie Dimon — exited the firm after 25 years, elevating Doug Petno (CIB) and Troy Rohrbaugh (CCB) to co-President roles.

III. The five facts that drive everything

1. Q1 2026 ROTCE stepped up 300bps in a single quarter (20% FY2025 → 23%), driven substantially by Category 2 (capital-markets trading, Equities +40% YoY FY2025 / fixed income +21% YoY Q1 2026) — the two most cyclically-sensitive line items in the entire P&L. ✅B
2. 2026 NII guidance ex-Markets is "broadly flat" at ~$95B despite double-digit balance-sheet growth (loans +11% YoY, deposits +7% YoY) — the core, least-cyclical franchise is decelerating on a per-dollar-of-balance-sheet basis even as volume grows. ✅B
3. Provision for credit losses fell -24% YoY in Q1 2026 even as nonperforming assets rose to $10.0B — a leading/lagging-indicator divergence that neither bulls nor bears have yet explained. ⚠️B (single-quarter data; directionally anomalous)
4. Marianne Lake's June 25, 2026 exit removed the closest-watched Dimon-succession frontrunner, occurring within Dimon's own reported ~3-year internal timeline to Executive Chairman — the freshest, most concrete succession signal in years, and the market's own coverage (Fortune's "CEO factory" framing) is not currently treating it as a valuation-relevant risk. ✅B
5. JPM trades at the richest multiples among money-center peers (P/E ~14.8-16x, P/B ~2.6x) vs. BAC (~13.2x P/E), WFC (1.78x P/TBV), Citigroup (~1.04x P/B) — exactly at the moment its two most cyclical revenue lines are also at multi-year highs. ⚠️A — the framing is collectively assumed by the market, not actively defended against the cyclical-composition question.

IV. The H-0 thesis

H-0 (one sentence): The market is pricing JPMorgan Chase's Q1 2026 ROTCE (23%) as the durable new baseline for its record valuation premium, but simultaneous facts — a capital-markets-cycle-driven trading surge, a credit-cost line at a cyclical trough despite rising nonperforming assets, and a freshly reshuffled succession bench — suggest a portion of that premium rests on a cyclical peak and an unpriced leadership-transition risk, not solely on durable structural scale advantage.

Mispricing taxonomy: Cognitive bias × Lifecycle stage (per mispricing.md, primary). Recency bias treats the most recent quarterly ROTCE print as representative of durable forward earnings power without decomposing it into its structural, capital-markets-cyclical, and credit-cyclical components. Every headline (earnings release, CNBC recap, sell-side commentary) obtained this session cites the single blended ROTCE figure — the composition analysis that would separate a genuine structural improvement from a cyclical peak is not standard practice in how this print is being consumed.

Secondary mechanism: Structural blindness × Option value. No standard bank-valuation framework carries a distinct field for CEO-succession risk at a trust-dependent institution. The June 25, 2026 Lake exit — the first concrete, dated data point consistent with Dimon's own reported ~3-year timeline — is being processed by financial media (Fortune's "CEO factory" framing, June 30, 2026) as an institutional-strength story, not a valuation-relevant succession-risk story.

5 falsification conditions (per h0_thesis.md; Bull-thesis-confirmers, since H-0 is itself a bear-leaning thesis):

• FF1 Q2 2026 earnings (July 14, 2026) show NII ex-Markets tracking at/above the high end of guidance AND trading revenue growth continues at 21%+ YoY — no deceleration anywhere in the print
• FF2 Provisions continue declining or stabilize through Q3-Q4 2026 without NPA-growth acceleration — the Q1 divergence proves to be noise
• FF3 Basel III Endgame final rule adopted close to or more generous than the March 2026 re-proposal, on the expected timeline
• FF4 Sell-side/company commentary continues treating the Lake exit as routine turnover through 2+ more quarters
• FF5 ROTCE holds at or above 22-23% for 3+ consecutive quarters (not just one print)

V. Tree — five branches

H-0: Market prices JPM's Q1 2026 ROTCE (23%) as durable baseline for its record
     premium; simultaneous facts (trading-cycle peak + credit-cost trough +
     unpriced succession risk) suggest a cyclical peak, not a structural plateau
│
├── L1A — Earnings Mix & Cyclical Composition  ⚠️B PARTIALLY SUPPORTED (load-bearing, highest priority)
│   ├── 1.1 R2-verified segment data confirms Cat.2 > Cat.1/3 growth   ⚠️B directionally confirmed, dollar-magnitude R2-pending
│   ├── 1.2 Q2 2026 earnings show trading-revenue deceleration        ⊗ catalyst pending — July 14, 2026
│   ├── 1.3 Historical analogue (2021-23 cycle) base rate             ⚠️C training-knowledge interpolation
│   └── 1.4 FY2026 NII ex-Markets guidance reaffirmed/raised/cut      ⊗ catalyst pending — same window as 1.2
│
├── L1B — Credit-Cost Trajectory  ⊗ UNRESOLVED (single-quarter data insufficient)
│   ├── 2.1 NPA growth continues + provisions rise in response         ⊗ catalyst pending — Q2-Q3 2026
│   ├── 2.2 NPA increase concentrated in specific loan category        ⊗ evidence gap — R2 item
│   ├── 2.3 Allowance/loans trend vs. BAC/WFC/C                        ⊗ evidence gap — R2 item
│   └── 2.4 $50B buyback pace maintains regardless of credit signal    ⚠️B ambiguous — timing flagged, not resolved
│
├── L1C — Structural Moat Durability  ⚠️B WIDENING ON STRUCTURAL DIMENSION
│   ├── 3.1 National deposit share progresses toward 15% target        ⚠️B investment confirmed, outcome not yet shown
│   ├── 3.2 Basel III Endgame finalized close to re-proposed magnitude ⊗ catalyst pending — timing unconfirmed
│   ├── 3.3 Peer banks match JPM's technology/branch investment pace   ⊗ evidence gap — R2 item
│   └── 3.4 Private-credit/fintech disintermediation measurable        ⚠️C industry trend real, JPM-specific magnitude unconfirmed
│
├── L1D — Succession & Governance Risk  ✅B MISPRICING MECHANISM CONFIRMED LIVE
│   ├── 4.1 Further senior-executive departures occur                  ⊗ catalyst pending — forward-looking
│   ├── 4.2 Sell-side discusses succession as valuation-relevant       ✅B Fortune's framing confirms it's currently NOT priced
│   ├── 4.3 Dimon confirms/extends/contradicts ~3-year timeline        ⚠️C underlying sourcing itself is Tier C
│   └── 4.4 Petno/Rohrbaugh performance generates CEO-candidate read   ⊗ catalyst pending — too early
│
└── L1E — Valuation & Peer-Multiple Mechanism  ⚠️C UNFAVORABLE (illustrative)
    ├── 5.1 Reverse-DCF: 2.6x P/B requires ~cyclical-peak ROTCE         ⚠️C illustrative, unverified inputs — directionally H-0-consistent
    ├── 5.2 Consensus target reacts + discusses ROTCE durability        ⊗ catalyst pending — post-Q2 2026
    ├── 5.3 SOTP (CCB+CIB+AWM benchmarks) vs. blended-bank framing      ⚠️C partial support only — AWM mix-shift real but insufficient alone
    └── 5.4 Multiple compresses if L1A+L1B resolve H-0-consistent       ⊗ catalyst pending — contingent on prior leaves

Total: 1 ✅ / 8 ⚠️ / 0 ✗ / 11 ⊗ across 20 leaves (5 branches × 4 leaves)
H-0 verdict: PARTIALLY SUPPORTED, ~58% confidence — genuinely evidenced but not yet tested against a second data point

VI. Key findings

Finding 1 — The ROTCE step-up is disproportionately a trading-cycle and credit-cost-trough story, not a core-franchise story

The 300bps improvement from FY2025's 20% ROTCE to Q1 2026's 23% coincides with FY2025 Equities revenue +40% YoY and Q1 2026 fixed income +21% YoY — both capital-markets-cycle-sensitive lines — while the core, least-cyclical NII engine is guided "broadly flat" for FY2026 excluding Markets, despite double-digit balance-sheet growth. This is the single most load-bearing fact in the tree. A 300bps firmwide ROTCE move in one quarter is a magnitude far more consistent with a cyclical swing than with the gradual, multi-quarter pace at which structural franchise improvements (branch-network maturation, technology-driven efficiency) typically show up. Neither bulls nor bears in consensus.md explicitly quantify this decomposition — the entire debate proceeds from the blended number.

Finding 2 — The credit-cost divergence is the cleanest, most quantifiable anomaly in the tree, and it remains genuinely unresolved

Provision for credit losses fell -24% YoY in the same quarter nonperforming assets rose to $10.0B. This is not, on its own, proof of an emerging credit-cycle turn — a single quarter's data cannot distinguish "genuine early-warning signal" from "loan-mix shift noise." But it is the kind of detail that recency-biased, headline-driven consumption of a beat-and-raise quarter would systematically miss, and it mechanically flatters the very ROTCE print the market is treating as durable. Q2-Q3 2026 data is required to resolve this, and this tree deliberately declines to force a premature verdict (Leaf 2.1: ⊗).

Finding 3 — The market is not pricing Dimon-succession risk at all, and this tree found direct evidence of that gap

Fortune's June 30, 2026 "CEO factory" coverage of the Lake exit frames JPM's deep succession bench as a talent-density asset ("the first stop for CEO recruiters") rather than engaging with the succession timeline as a valuation-relevant risk. This is not proof the risk is large — it is proof the risk is currently unpriced, which is exactly what the "structural blindness × option value" secondary mechanism predicts. No standard bank-valuation model carries a distinct field for CEO-transition risk, and the market's own coverage choice (talent-depth framing, not risk framing) is the clearest available evidence that this mechanism is live.

Finding 4 — The valuation math, even on illustrative Tier-C inputs, does not support the current premium requiring only a structural-floor ROTCE

A simplified reverse-DCF (Leaf 5.1) suggests the current ~2.6x P/B multiple is more consistent with requiring something close to the current cyclical-peak ROTCE (23%) than the plausible structural-floor estimate (high-teens) this tree's own decomposition points toward. The SOTP cross-check (Leaf 5.3, testing whether JPM deserves a conglomerate premium over a blended-bank multiple) finds AWM's genuinely high ROE provides partial, not complete, support — it does not fully explain the ~0.8-1.6x P/B premium gap over BAC/WFC/C that consensus.md names as the "orphan premium."

Finding 5 — Structural tailwinds are real and should not be conflated with the cyclical-composition question

This is not a "JPM is secretly weak" thesis. Basel III Endgame capital relief (~4.8% CET1 reduction for GSIBs, if finalized as re-proposed), the branch-network expansion pledge (targeting 15% national deposit share from 11.3%), and sustained technology investment (~$18-19.8B 2026 budget) are genuine, durable, non-cyclical tailwinds — confirmed by durability_test.md's 21/25 score with zero fatal flags. The tree's finding is narrower and more precise: these structural tailwinds justify some premium over peers, but the market appears to be extrapolating the current cyclical print as the baseline, rather than isolating the structural component the tailwinds actually support.

VII. Valuation analysis

(See peers.md, scenarios.md, and implied_prob.md for full breakdowns.)

Peer-group framing (money-center banks):

Median money-center peer multiple: roughly 1.0-1.8x book / 10.9-13.2x P/E. JPM at ~2.6x P/B / ~14.8-16x P/E is a full multiple-point-plus premium — the largest gap in the peer set is JPM-vs-Citigroup (~2.5x P/B differential).

SOTP framing (per assumption B1 — diversified-conglomerate reframing):

Read: AWM's disproportionate ROE is the one segment that unambiguously supports a conglomerate premium — but AWM is proportionally smaller within JPM (~20-25% of revenue) than an equivalent segment is within a pure-play wealth manager like Morgan Stanley, limiting how much this comparison alone can carry the full valuation argument (per Leaf 5.3). Unlike CAT (where SOTP trades below current price, implying the market has NOT yet fully re-rated) or AAPL (where SOTP supports the current price), JPM's SOTP provides only partial support for its current premium — the gap between "partial SOTP support" and "full 2.6x premium" is, per this tree's analysis, attributable to a combination of the cyclical-earnings-mix extrapolation (L1A) and the unpriced Dimon-premium/succession-risk bundle (L1D).

Reverse-DCF read (Leaf 5.1, Tier C/illustrative): at a 9-11% cost-of-equity assumption, the current 2.6x P/B multiple appears to require a durable ROTCE close to the current cyclical-peak print (low-to-mid 20s), not the structural-floor estimate (high-teens) this tree's own decomposition suggests is more durable. This is the single most direct (if least evidence-strong) confirmation of H-0's valuation-mechanism claim.

VIII. Scenario architecture

(See scenarios.md for full analytical breakdown and implied_prob.md for the reverse-engineering math.)

Probability-weighted expected return: -6.1% (my assigned probabilities, per implied_prob.md).

Asymmetry: 0.78:1 — UNFAVORABLE. Unlike the Bull scenario's narrower, four-condition-conjunction requirement, both Base and Bear require only partial or directional confirmation — which is why this tree assigns them 78% combined probability. The raw dollar spread confirms the same tilt: Bull upside ($51.21) is smaller than Bear downside ($65.29) even before probability-weighting. This is the least-favorable asymmetry of any ticker examined in this batch of trees — a materially different read from CAT's 1.0-1.4x or DELL's 1.05x.

Market-implied probabilities (reverse-engineered from current price): Bull ~40% / Base 50% / Bear ~10% — the market is assigning roughly 4x less probability to the Bear scenario than this tree's own analysis, which is the clearest quantitative expression of the mispricing this tree is built to test.

IX. What this means for position sizing

Position-sizing recommendation

• Current price ($333.79): 0% — WATCH ONLY. Negative expected value (-6.1%) combined with a sub-1.0x asymmetry ratio means this is not a "wait for confirmation before scaling up" situation (the CAT/DELL pattern) — it is a "the math argues against initiating at all" situation, at least until either the price moves or the evidence resolves favorably.
• Contingent starter (1-2%): appropriate only if BOTH (a) Q2 2026 earnings (July 14, 2026) confirm the Base-scenario partial-normalization path (not the Bull scenario — Bull firing would mean the stock is closer to fairly priced, not cheap) AND (b) price moves toward the $315-327 range identified in implied_prob.md Step 5 as the point where risk/reward reaches parity-to-favorable.
• Scale to 2-3%: only on confirmation of T3 (Basel III Endgame finalized close to re-proposed magnitude) AND T4 (succession bench stability confirmed over 2+ quarters) AND price remaining at or below the $315-327 entry band.
• Do not initiate, regardless of price, if RF1 fires (Q2 2026 trading-revenue deceleration below 15% YoY combined with NII tracking the low end of guidance) until the subsequent price reaction is observed — a confirmed H-0 combined with a price that has not yet adjusted would represent an even more unfavorable entry than today.
• Hard cap 3-4% even at full confirmation — this is a cyclical-macro-sensitive archetype, not a compounder-under-review; per K.3 position-sizing convention, this ticker's structural quality (21/25 durability) supports a higher cap than its current cycle-position risk should be sized to.

Concentration-risk note (per K.3.4)

JPM's load-bearing macro factor is the capital-markets-and-credit cycle, not the AI-capex cycle — dashboards/ai_capex_cycle_overlay.md correctly tags this ticker uncorrelated alongside COST, F, WMT, CAT-adjacent-but-distinct, TSLA. This means JPM does NOT compound correlated exposure with the portfolio's NVDA/TSM/AMD/AJNMY/ASML/AVGO/MU AI-infrastructure sleeve. It DOES share a factor with any other financials names in the library (GS/MS if scaffolded, V, COIN) — specifically capital-markets-cycle and credit-cycle exposure. If V (payments network) and COIN (crypto-cycle) positions are already held, JPM would add a third distinct-but-overlapping cyclical-financial factor; Ming should acknowledge this in the decision-journal entry per K.3.4's explicit-naming requirement.

X. Triggers and red flags

(Full detail with 应对 playbooks in triggers_redflags.md.)

Triggers (confirm the market's current framing — i.e., falsify H-0):

• T1 (July 14, 2026): Q2 2026 earnings — trading revenue holds 20%+ YoY AND NII ex-Markets beats guidance
• T2 (Q2-Q3 2026): NPA growth stalls/reverses without provisions rising
• T3 (timing TBD): Basel III Endgame finalized close to re-proposed magnitude
• T4 (2-4 quarters): No further senior departures; succession treated as institutional strength
• T5 (2026-2027): National deposit share confirms progress toward 15% target

Red flags (confirm H-0):

• RF1 (July 14, 2026, HIGHEST PRIORITY): Trading revenue decelerates below ~15% YoY AND NII tracks low end of guidance
• RF2 (Q2-Q3 2026): Provisions rise sharply in response to continued NPA growth
• RF3 (ongoing): Further senior-executive departure
• RF4 (ongoing): Sell-side/company begins explicitly discussing succession as a valuation input
• RF5 (timing TBD): Basel III Endgame delayed or finalized less generous than re-proposed
• RF6 (any quarter): Peer bank (BAC/WFC/C/GS/MS) confirms comparable capital-markets-cycle deceleration

XI. Long-term holdability verdict

Per durability_test.md: aggregate score 21/25 (Medium-High durability, edge of High band). 0 fatal flags fired (Q3 capital allocation 4/5 ✅A, Q4 disruption survival 4/5 ✅B, balance-sheet survivability CET1 14.3% vs. 11.5% required with CCAR-validated resilience even under Severely Adverse stress).

Top 3 reasons supporting 5-10 year hold:

1. Money-center banking at JPM's GSIB scale is among the most durable, regulation-protected business models available — regulatory capital requirements structurally favor incumbents at this size
2. Capital allocation discipline is genuinely strong (14 consecutive years of dividend growth, large sustained buyback capacity, real organic reinvestment in branch network + technology + AWM)
3. Balance-sheet resilience is CCAR-validated even under a severe stress scenario — this materially de-risks the "disruption survival" and "fatal flag" dimensions relative to a typical durability test

Top 3 reasons against (entry-timing, not quality, concerns):

1. A material share of the recent ROTCE improvement is attributable to capital-markets-cycle-sensitive trading revenue and a credit-cost trough, not the core structural franchise — if this normalizes, the durable ROTCE level may sit meaningfully below the current 23% print
2. Private-credit and fintech disintermediation are real, gradual, structural competitive threats to both the CIB lending franchise and the CCB deposit franchise
3. The freshly-concrete Dimon-succession risk has no standard valuation-model field, is currently unpriced, and its eventual magnitude is unknowable in advance

Required catalysts for upgrade to high-conviction hold: Q2-Q3 2026 earnings confirming NII holds up while trading revenue normalizes gradually (not collapses); credit-cost divergence resolving as noise; Basel III Endgame finalized close to re-proposed magnitude; a clean, well-received succession-timeline confirmation.

Required disconfirms for downgrade: any of FF1-FF5 firing in the H-0-consistent (normalization) direction — most critical: Q2 2026 earnings (July 14, 2026) showing simultaneous trading-revenue deceleration AND NII tracking the low end of guidance.

Recommended position management: 0% at current price; re-evaluate for a 1-2% starter only if Q2 2026 confirms partial normalization AND price moves toward $315-327.

XII. Investment Scorecard (per MANUAL_en.md Part K.6 + K.10)

15-question scorecard (analytical-tree Q-list, Format B per K.3.5)

Verdict tally (M1 evidence-tier suffixes per K.3.6): 13 ✅ · 1 ⚠️ · 1 ✗ — Q1✅A, Q2✅A, Q3✅B, Q4✅B, Q5✗A (valuation genuinely unfavorable by this tree's own math), Q6✅A, Q7✅A, Q8✅B, Q9✅A, Q10⚠️B (multi-front competition), Q11✅B, Q12✅B, Q13✅A, Q14✅B, Q15✅A.

K.3.5 Weighted-score derivation

Applying the 4-tier weighting from MANUAL §K.3.5 (verdict values: ✅ = 1.0, ⚠️ = 0.5, ✗ = 0.0):

91.0% sits in the ≥85% high-conviction band — placing JPM alongside AJNMY/COST (91%). This is a genuinely correct reading of JPM's structural quality: the business, balance sheet, and moat trajectory are excellent by any GSIB standard, and Q5's full zero-weighting on the load-bearing tier already imposes the maximum available penalty the formula allows for the valuation problem — no further ad hoc haircut is needed on top of that, unlike CAT's case where the raw score didn't fully capture an analogous concern.

Critical distinction — do NOT read 91% as a buy signal. Per MANUAL_en.md's own framing: xii_score measures structural quality (what the scorecard says about business + balance sheet + moat + valuation-as-one-input-among-fifteen); h0 measures thesis confidence (58% here — partially supported, not yet tested against a second data point). The gap between JPM's 91% structural score and its 58% H-0 confidence, combined with the load-bearing Q5 ✗, is precisely why this tree's final verdict is WATCH/0% rather than a high-conviction buy: the company is excellent; the current price is not. This is analytically the same pattern CAT exhibited (91% raw → 87% operative, "high-conviction by structure, entry-timing discipline required") taken one step further — where CAT's price sat modestly above fair SOTP, JPM's own reverse-DCF and implied-probability math finds the current price offers negative expected value outright.

Scorecard summary

Final verdict: WATCH — 0% at current price; contingent 1-2% starter on Q2 2026 confirmation + price pullback

For Ming specifically:

• ❌ Do NOT initiate at $333.79 — negative expected value and unfavorable asymmetry per this tree's own reverse-engineering
• ⏳ Watch Q2 2026 earnings (July 14, 2026, 12 days from this tree's construction) — the single cleanest near-term test of the entire thesis
• ⚠️ If Q2 confirms partial normalization (Base scenario) AND price moves toward $315-327: consider a 1-2% starter
• ⚠️ If Q2 confirms structural durability (Bull scenario, FF1/FF5 firing): the stock is closer to fairly priced, not cheap — do not chase; re-evaluate the entire tree rather than assume "confirmation = buy"
• 🔻 If RF1+RF2 both fire (Bear scenario): wait for the subsequent price reaction before considering entry — a confirmed bear thesis without a price adjustment is a worse entry, not a better one
• 📅 Re-test on Q2 2026 earnings (~July 14, 2026), Q3 2026 earnings (~October 2026), and any Basel III Endgame final-rule announcement
• 🎯 Hard cap 3-4% even at full confirmation — cyclical-macro-sensitive archetype sizing discipline, not compounder-under-review sizing

The 2-minute pitch:

"JPMorgan Chase trades at ~$334 / ~2.6x book / 23% ROTCE — the richest multiple among money-center banks, right after a quarter where trading revenue and a credit-cost trough did most of the work in a 300bps ROTCE beat, while the core lending franchise is guided flat. The same week the bank announced a $50B buyback and a 10% dividend hike, its closest-watched CEO-succession candidate abruptly left after 25 years — and the market hasn't priced that risk at all. This isn't a bad-bank thesis: JPM's balance sheet is genuinely fortress-grade (14.3% CET1, passed CCAR's severe-stress test with room to spare) and its scale investments in branches, technology, and wealth management are real, durable tailwinds. But my own math on this stock says the current price offers negative expected value — the upside case requires four separate good things to keep happening at once, while the downside case only needs one or two things to normalize. Watch Q2 earnings on July 14; if the trading engine cools while the core franchise holds up, and the price drifts down toward $315-327, that's where this becomes a starter position, not before."

Risk types most relevant (per MANUAL_en.md Part K.4):

• Valuation risk (richest-in-peer-group multiple at a cyclical-earnings peak; this tree's own math finds negative EV at current price)
• Cyclical risk (capital-markets-cycle and credit-cycle exposure is the entire L1A/L1B analytical core of this tree)
• Competition risk (multi-front: GS/MS on Markets/IB, private credit on CIB lending, fintech/neobanks on CCB deposits)
• Regulatory risk (Basel III Endgame finalization timing/magnitude still unconfirmed; ongoing CFPB/Epstein-related litigation)
• Correlated-factor risk (shares capital-markets/credit-cycle exposure with any other financials names in the library — V, COIN, and any future GS/MS/BAC/WFC/C trees — per K.3.4)

"When NOT to buy" anti-pattern check (per MANUAL_en.md Part K.5):

• ❌ NOT buying because of a single news headline (this tree explicitly weighs a beat-and-raise quarter against a same-week succession reshuffle and a credit-cost anomaly — multiple converging signals, not one)
• ✅ NOT chasing a fast-rising stock — JPM has not had a dramatic recent momentum spike comparable to CAT's +54% YTD or TOTDY's +18% single-day move; this is a valuation-discipline concern, not a momentum-chasing one
• ❌ NOT an "AI stock" theme attachment — this tree explicitly classifies JPM as cycle_exposure: uncorrelated, i.e., NOT riding the AI-capex narrative
• ❌ Not a "looks cheap because the price is low" trap — analysis is multiple-based (P/B, P/E), not share-price-based
• ❌ Not brand-familiarity bias — the analysis is grounded in segment-level cyclicality decomposition (taxonomy.md), not "JPMorgan is a great company" vibes
• ✅ This tree's discipline is specifically resisting the opposite anti-pattern — the temptation to buy a "best-in-class" company reflexively because its structural quality (91% xii_score) is undeniable, without separately checking whether the current price offers positive expected value. The 91%-quality / negative-EV combination is exactly the scenario K.5 exists to guard against.

Net: 0 anti-pattern flags on the buy-side hype dimensions; 1 self-imposed discipline flag (resisting quality-bias-driven entry at an unfavorable price) — the tree's own verdict is WATCH, not buy, specifically because it applies this discipline to itself.

XIII. What's NOT in this tree (deferred / documented but not built)

• Stage 2 supplementary: historical-analogue.md (2021-2023 bank-earnings-cycle analogue is discussed inline in leaves.md Leaf 1.3 and frameworks.md, but not built as a standalone file)
• Stage 4 supplementary: premortem-steelman.md (the bear case is steelmanned inline in consensus.md and Section VI Finding 2, but not built as a standalone file)
• Sources catalog: sources.md — full bibliography of aggregator sources (StockTitan, GuruFocus, CNBC, Motley Fool, Nasdaq, Fortune, Banking Dive, Axios, Holland & Knight/Freshfields regulatory summaries) — deferred to next refresh
• R2 primary-filing verification — direct SEC EDGAR/company-IR access was blocked (403) throughout this session; the R2 verification items listed in dashboard.md are the highest-priority follow-up before this tree should be treated as fully load-bearing for a real capital-allocation decision
• Stage 6 living: update_{YYYY-MM-DD}.md files — first scheduled immediately after Q2 2026 earnings (~July 14, 2026)
• External research packets: ChatGPT review packet + Codex review packet — not built this session

Last updated 2026-07-02 (Routine C, Stage 3-7 build). Source quality: Tier B (GENERATE mode) throughout — primary-filing R2 verification owed. K.3.6 evidence-strength suffix convention applied to all leaf verdicts. Next refresh: immediately after Q2 2026 earnings (~July 14, 2026) — the single highest-priority refresh trigger in this tree.

"Don't read news; update your tree." — 90s.PM.Investing