The Silk Blog

One-Page Brief: AI Leadership Rotation Amid Contained Energy Risk – Spiderweb / Interconnected Market Implications (30–90 Days, as of July 18, 2026)

Core Thesis

Dominant: Large-cap AI leadership reasserts over small caps as breadth contracts, alongside a fading crude premium and range-bound yields — a "narrow-leadership, contained-macro" regime (~40% joint [uncalibrated]). Alternative: AI leadership fatigue (note the -11pp NDX and -7pp capex deltas) triggers broadening rotation into small caps and cyclicals (~30% [uncalibrated]). Key discriminator: whether NDX/IWM relative strength holds through the next NFP print and Q3 tech-capex guidance.

Markets Getting Stronger & Spiderweb Implications

• Large-cap AI proxies (NDX): Rotation base rate ~50% [outside view]; adjusted to ~58% over 4 days [n=1615] → ~52–55% at 30-day horizon (F4: -4pp horizon decay, -1pp capex-delta drag). But NDX-vs-IWM sub-signal fell to 46% (-11pp), so conviction is softening even as it leads.

Implications: Concentrated leadership tightens the web's dependency on a handful of names — strength is real but fragile, raising cascade risk if capex disappoints.

• Front-end rate stability (TNX): Range-bound base ~56% [n=1615]; firm front end anchors risk appetite.

Implications: Stable yields underwrite AI multiples; a break >4.60 (56% continuation [n=1119]) would sever this support.

Markets Getting Weaker & Spiderweb Implications

• Small caps (IWM/IJR): Mean-reversion-down base ~55%; adjusted to ~58% [n=1615] favoring large-cap rotation → ~52% at 30-day. Asymmetric payoff (-2% vs +0.5%).

Implications: Breadth contraction confirms narrow leadership — but is also the precondition for the alternative broadening thesis if it overshoots.

• Crude oil premium: Fade probability 65% [n=1615], supported by stable 0.38 risk score and active backchannels; geopolitical-only transmission low-cascade (60%).

Implications: Contained energy caps the inflation-to-rates channel, reinforcing yield stability.

The Connecting Spiderweb (Key Interconnections)

• Leverage point — AI capex/leadership (the -11pp/-7pp deltas): This is the highest-cascade signal. Its decay simultaneously threatens the rotation thesis (needs AI strength), the yield anchor (multiple compression pressures front-end demand), and the small-cap fade (weak AI = broadening).
• Non-linear risk (CT5): A capex miss is not linear. Concentrated ownership means a single guidance cut can trigger de-grossing, forcing hedge-fund unwinds that spill into IWM and NDX together — collapsing the rotation trade from both ends. Small input (one earnings line) → disproportionate breadth shock.
• Supporting: Crude fade → yield stability → AI multiple support forms a self-reinforcing loop; break any link and the loop reverses.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Crude fade ⊕ stable TNX ⊕ narrow AI leadership = "contained-macro, concentrated-risk" regime.
• Negation (¬) Scenarios:

| Scenario | Trigger | Implication |

|---|---|---|

| ¬Rotation | IWM outperforms NDX >2% | Broadening; capex-delta thesis wins |

| ¬Crude fade (F2) | Oil premium persists beyond 2× reversion window | Structural supply deficit, not statistical anomaly — invalidates fade positioning |

| ¬Yield range | TNX breaks >4.60 | Multiple compression cascades into AI names |

• Equivalence (∼): Small-cap breadth contraction ∼ leading indicator of concentration fragility (both measure narrowing participation).

Ideas for Thinking About the Spiderweb (Mental Models from Guardrails)

• Dominant lens — Forecasting.md (F2 Non-Stationarity): The core risk is regime shift, not dislocation. The -11pp/-7pp capex deltas may signal a stationarity break in AI leadership, not noise. Falsification test distinguishing decay-vs-noise: if NDX/IWM relative strength recovers >5pp within 10 trading days, treat as noise; if it continues declining, treat as regime shift and abandon rotation.
• Supporting — Critical Thinking.md (CT7 Conjunction Decay): The four-way ⊕ combination is correlated via shared risk-sentiment, so joint confidence bounds to ~40%, not naive multiplication.
• Supporting — Value.md: Asymmetric IWM payoff (-2% vs +0.5%) favors the fade even at modest edge.

Practical Prompts

• Track NDX/IWM relative strength over 25 trading-day window — if IWM outperforms NDX by >2%, rotation thesis invalidated; pivot to broadening.
• Monitor AI capex-delta signal over 10-day window — if it declines a further >5pp, treat as regime shift (F2) and exit large-cap rotation.
• Watch crude premium over 15-day window — if it persists above entry despite backchannel de-escalation, fade thesis invalidated (structural deficit).
• Track TNX vs 4.60 over 20-day window — if sustained break above, AI-multiple support severed; reduce concentrated exposure.

Markets are a single, homeostatic, arbitrage-driven neural net:

One-Page Brief: Large-Cap Reassertion Meets Soft-Data Rate Drift – Spiderweb / Interconnected Market Implications (30–90 days, as of July 17, 2026)

Core Thesis

Dominant: A firm-dollar, range-bound-rates regime supports large-cap AI leadership over small caps, with capital rotating toward liquidity and quality (~50% [uncalibrated at 30–90d; derived from 4-day signals]). Alternative: AI capex conviction cracks (AMD's -10pp delta) and soft data breaks TNX lower, triggering broad risk-off and USD unwind (~30%). Key discriminator: AMD stabilization above 495 vs. break below 480, and whether TNX holds the 4.45 pivot.

Markets Getting Stronger & Spiderweb Implications

• USD (short EURUSD): Base rate for USD-strength long signal 0.56 [n=2712], adjusted to 0.59 [n=1615] over 4 days given +0.4pp 7d slope and stable geo-risk. Decaying to ~52–55% at 30-day horizon (F4: -5pp horizon decay). Implications: A firm dollar tightens global financial conditions, pressuring EM and commodities — reinforcing the flight toward large-cap US tech.
• NDX vs IWM (AI leadership): Base rate 0.56 [n=2712] → 0.57 [n=1615] on low dispersion + capex thesis; ~52–56% at 30-day horizon (F4 decay -3pp). Implications: Low VIX (18 invalidates the entire leadership calibration in a single session).

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Range-bound TNX ⊕ USD strength ⊕ low VIX = "quality-concentration regime" favoring megacap tech.
• Negation (¬):

| Scenario | Trigger | Outcome |

|---|---|---|

| AMD breaks 480 | Capex doubt | NDX leadership thesis severed; rotation stalls |

| TNX breaks 4.45 | Soft data | USD unwinds, EURUSD short invalidated |

| Regime shift (F2) | VIX >18 persists >10 days | Extremes do NOT revert — signals structural risk-off, not dislocation; mean-reversion positioning invalidated |

• Equivalence (∼): AMD stabilization ∼ AI-capex conviction gauge; TNX pivot ∼ USD-strength enabler.

Ideas for Thinking About the Spiderweb (Mental Models)

Dominant lens — Forecasting.md (F4 Horizon Decay): Every probability here is 4-day calibrated [n=1615]. Mechanically, the 30–90 day horizon demands decaying 57–59% signals toward ~50–55%; treating short-horizon edges as durable is the primary failure mode. Supporting — Critical_Thinking.md (CT7 Conjunction Decay): The dominant thesis chains TNX-stable × USD-strong × NDX-leads; these are correlated via the shared rate/risk-sentiment driver, so joint confidence bounds to ~40–48%, not naive multiplication (~17%). Challenge hypothesis — Psychology.md: AMD's -10pp swing may be sentiment overshoot, not signal; falsification test — if AMD reclaims 495 within 5 trading days, the -6pp LJ3 bear adjustment was recency bias, not information.

Practical Prompts

• Watch AMD over 5 trading-day window — if it breaks below 480, AI-capex leadership thesis is invalidated and NDX/IWM long should be cut.
• Track TNX vs 4.45 over 10 trading-day window — if it breaks below on soft data, USD strength and EURUSD short are invalidated.
• Monitor NDX vs IWM relative performance over 25 trading-day window — if IWM outperforms by >2%, the concentration regime has inverted (guards against short-window bias).
• Watch VIX over 10 trading-day window — if it holds >18, all leadership calibrations void; shift to risk-off positioning.

Markets are a

Devil's AdvocateThe most likely reason for failure would be the thin margin of the probability estimates themselves — most of today's calls cluster in the 52-59% range, which historically maps to realized win rates of only 44-48% for the higher-volume signal tiers (n=111 and n=114), meaning the edge could easily evaporate if these estimates are even slightly overcalibrated. A second failure mode would be the interlocking nature of the theses: several positions all lean on the same twin assumptions of persistent large-cap tech leadership and a firm dollar anchored by a stable front-end rate structure, so if that regime cracks, correlated losses could stack rather than diversify. Finally, with an aggregate Brier score of 0.302 across a large sample (n=2428), the calibration is only modestly better than a coin flip, so the weakest, tightest-margin theses would be the first to disappoint if macro data surprises against the range-bound narrative.

Base rates: moderate signals 48% win [n=111], elevated signals 44% win [n=114], extreme outliers 64% win [n=14]Brier Score: 0.302 (Poor) [n=2428] | Drift: +0.008 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Late-Cycle Mean Reversion Meets Narrow AI Leadership – Spiderweb / Interconnected Market Implications (30–90 Days, as of July 16, 2026)

Core Thesis

Dominant: A tactical mean-reversion regime dominates — stretched single names (AMD) revert lower while a firm USD/front-end and narrow AI leadership persist, consistent with a contracting-breadth, low-geo-risk tape (~40% joint at 30-day horizon, correlated via shared risk-sentiment driver). Alternative: dispersion breaks upward, AI leadership broadens or fractures and reversion signals fail (~35%). Key discriminator: whether NDX/IWM dispersion widens with rising breadth (regime shift) or without (reversion holds).

Markets Getting Stronger & Spiderweb Implications

• USD (short EURUSD): Base rate for WATCH continuation ~58% [n=1615]; outside-view anchored, adjusted flat given no EU data catalyst. 4-day calibration → ~52-55% at 30-day (F4: -4pp horizon decay).

Implications: Firm USD reinforces front-end firmness (TNX) and pressures commodity/EM channels, tightening the reversion web.

• US front-end rates (TNX firm): 58% [n=1615], sharply revised from 72% (-14pp) — a large single-day delta signaling low conviction. Outside-view base for "firm then range."

Implications: Firm front-end + firm USD = classic late-cycle tightening backdrop; supports large-cap AI over rate-sensitive small caps (IWM).

Markets Getting Weaker & Spiderweb Implications

• AMD (tactical short): 62% [n=1615], up +7pp. Outside-view base rate 77% on 2σ+ (lookback per backtest_1042d), adjusted down to ~62% for WATCH status; skew -4% vs +1%. Decays to ~55-58% at 30-day (F4).

Implications: AMD reversion is the canary for the broader "stretched names revert" thesis; its failure would undercut the whole reversion leg.

• Broad small-caps (IWM, relative): NDX-over-IWM leadership at 56% [n=1615], outside-view base ~60% thesis.

Implications: Narrow leadership weakens breadth (already -10), which paradoxically supports reversion in laggards while concentrating risk.

The Connecting Spiderweb (Key Interconnections)

Leverage point — AI leadership dispersion (NDX vs IWM): This single signal cascades widest. If dispersion stays contained, AMD reverts (stretched name normalizes), breadth stays narrow-but-orderly, and firm USD/TNX coexist peacefully. Supporting connections: (1) Firm USD ⟶ pressures EURUSD lower and reinforces front-end; (2) Contracting breadth (-10) ⟶ mechanically raises reversion odds on laggards (SECONDARY 55%). Non-linear risk (CT5): narrow leadership is fragile — if the top 2-3 AI names stumble, index-level selling forces mechanical de-grossing and vol-target funds cut exposure disproportionately, turning an orderly -4% AMD reversion into a broad cascade far beyond linear expectation. Concentration converts small inputs into large outputs.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Firm USD ⊕ firm front-end ⊕ contracting breadth = late-cycle, low-vol reversion regime favoring quality large-cap over stretched/small.
• Negation (¬) Scenarios:

| Scenario | Trigger | Implication |

|---|---|---|

| ¬Reversion (regime shift, F2) | AMD/laggards stay stretched beyond 2× normal reversion window | Structural AI-capex re-rating, not anomaly — invalidates short |

| ¬USD firm | EU data surprises up | EURUSD short fails; front-end thesis weakens |

| ¬Narrow leadership | Breadth expands, dispersion rises | AI leadership fractures; bear case dominates |

• Equivalence (∼): AMD reversion ∼ SECONDARY WATCH mean-reversion (0.65) — same mechanism, correlated not independent.

Ideas for Thinking About the Spiderweb (Mental Models from Guardrails)

Dominant lens — Forecasting.md (F2 Non-Stationarity): The reversion thesis assumes a stationary distribution, but the TNX -14pp swing and AMD +7pp reveal an unstable prior. Falsification test distinguishing reversion vs regime shift: if AMD and WATCH laggards fail to revert within ~2× the historical mean-reversion window, treat as structural (supply/capex re-rating), not statistical. Supporting: HeuristicAlgebra.md for treating AMD and SECONDARY WATCH as correlated (CT7 — do not multiply as independent). Supporting: CriticalThinking.md (CT2) — all sigma claims must carry lookback windows; the 2σ+ base rate is only meaningful anchored to backtest_1042d's window.

Practical Prompts

• Track AMD reversion over 20 trading-day window — if AMD fails to close below entry by >2% and holds stretched, reversion leg is invalidated (F2 regime-shift signal).
• Watch NDX/IWM dispersion over 10 trading-day window — if IWM outperforms NDX by >2% on rising breadth, narrow-leadership thesis is invalidated.
• Monitor EURUSD over 15-day window — if it rallies >1.5% on EU data surprise, USD-firm leg is falsified, weakening the front-end connection.
• Watch TNX

Devil's AdvocateThe most likely reason for failure would be the reliance on lower-conviction signals, where the historical base rate shows win rates of only 44-46% across a large sample (n=228) — meaning these calls have historically resolved as coin-flips or worse despite positive mean returns driven by a few outliers. A second failure mode would stem from the mean-reversion thesis: with several positions clustered around 55-62% probability, an overall Brier score of 0.302 suggests the forecasting engine has been modestly overconfident, so a trending rather than reverting tape could see multiple correlated theses miss simultaneously. The weakest link is the correlated bet on both a rotation into large-cap tech leadership AND broad mean reversion — if a single macro catalyst (rates, dollar, or risk sentiment) moves against these, the losses could compound rather than diversify.

Base rates: moderate signals 46% win [n=114], elevated signals 44% win [n=114], extreme outliers 64% win [n=14]Brier Score: 0.302 (Poor) [n=2422] | Drift: +0.006 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Rate-Reversion Reflation Fade – Spiderweb / Interconnected Market Implications (30–90 days, as of July 15, 2026)

Core Thesis

Dominant: A low-panic mean-reversion regime pulls 10Y yields lower and the dollar softer, supporting a narrow tech bid while broad commodities stay contained (~48% joint, medium cone). Alternative: A hot-US-data / VIX-spike shock reverses all four legs simultaneously (~32% [uncalibrated]). Key discriminator: next CPI/NFP print and whether VIX holds below 20. Joint confidence is bounded, not multiplied: legs share one driver (risk-sentiment + rate path), so P(TNX lower) ~72% × P(EURUSD higher | rates lower) ~75% × P(tech holds | both) ~65% ≈ 35%, bounded up to ~45–50% given positive correlation.

Markets Getting Stronger & Spiderweb Implications

• 10Y Treasuries / lower yields (TNX): Base rate 77% [n=1686], adjusted to 72% [n=1615] for 4-day horizon; decaying to ~58–65% at 30-day (−10pp F4 horizon decay, +2pp momentum from +7pp daily delta). PRIMARY.

Implications: The keystone. Lower yields ease the dollar and lower tech discount rates — one signal feeding three.

• EURUSD long: Outside-view 77% base [n=1686], adjusted to 64% [n=1615] for dollar firmness; ~55–60% at 30-day. PRIMARY.

Implications: Dollar softening amplifies commodity containment ambiguity and eases EM funding stress.

• AMD / tech proxies: Momentum base 58% [n=128], adjusted to 55% on contracting breadth; the −7pp daily delta is a caution flag. PRIMARY, lowest conviction.

Implications: Narrow leadership rests on the rate leg holding; breadth contraction is the fragility.

Markets Getting Weaker & Spiderweb Implications

• Broad commodities (ex-ag): Contained, ~70% [uncalibrated] — treat as directional prior, not empirical. PRIMARY.

Implications: Contained commodities reinforce the disinflation narrative that justifies lower yields — a self-consistent loop that becomes circular if de-escalation reverses.

The Connecting Spiderweb (Key Interconnections)

Leverage point — TNX (lower yields). This single signal cascades furthest: it drives EURUSD (rate differential), tech valuations (discount rate), and validates the disinflation read on commodities. Cascade depth: 3+. If the rate reversion holds, the web is coherent; if it snaps, all four legs unwind together.

• Supporting: EURUSD strength is downstream of the yield move, not independent — correlation, not confirmation.
• Supporting: AMD's convexity structure means small rate/VIX inputs produce disproportionate option-P&L swings.
• CT5 Non-linear risk: A VIX break above 20 does not weaken tech linearly — it triggers convexity unwind + momentum reversal + safe-haven bid that simultaneously helps TNX but destroys the AMD leg, splitting the web. Small vol input, large dispersion output.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Lower yields ⊕ softer dollar ⊕ contained commodities = disinflation-carry regime favoring duration + quality tech.
• Negation (¬):

| ¬ Scenario | Trigger | Effect |

|---|---|---|

| ¬Rate reversion | Hot CPI/NFP | All legs unwind; regime shift |

| ¬Mean reversion (F2) | Yields sustain break >2× historical reversion window | Structural repricing, not dislocation — invalidates reversion positioning |

| ¬Tech leadership | VIX >20 | AMD convexity unwind despite rates |

• Equivalence (∼): EURUSD long ∼ short-USD ∼ implicit dovish-Fed bet — three framings of one trade.

Ideas for Thinking About the Spiderweb (Mental Models)

Dominant lens — Forecasting.md: The +7pp TNX delta and 4-day calibration must decay across the 30–90 day horizon (F4); do not treat 72% as durable — it is ~60% by day 30. F6 decomposition above makes each leg separately falsifiable. Supporting — Critical Thinking.md: CT7 conjunction discipline stops us multiplying correlated legs into false precision; the ~48% joint reflects shared drivers. Challenge — Scientific Method.md: The commodity leg is [uncalibrated] and circular with the disinflation thesis. Falsification test distinguishing genuine signal from narrative fit: if commodities fall while yields rise, the disinflation loop is broken and the web is confirmed spurious.

Practical Prompts

• Track 10Y yield over 20 trading-day window — if yields rise >25bp from entry, keystone thesis invalidated; unwind duration and tech together.
• Monitor EURUSD vs recent lows over 10-day window — if it breaks below defined-risk stop on hot US data, dollar-firmness alternative confirmed.
• Watch VIX over 10 trading-day window — if it closes >20, exit AMD convexity leg; regime split active.
• Compare commodities vs yields over 30-day window — if both rise together, disinflation loop falsified; abandon the reflation-fade frame.

Devil's AdvocateIF this forecast proves wrong, the most likely failure mode would stem from the moderate-conviction signals that anchor most of the book: the mid-tier alert-style setups have historically resolved as coin-flips at best (a 43% win rate across 113 samples), so any theme leaning on those probabilities would be the weakest link. A second plausible failure path would be the calibration gap implied by the current Brier score of 0.302 — a stated 72% conviction on the rates view may be systematically overconfident, and if the disinflation/lower-yield narrative reverses on a hawkish data surprise, the correlated rates-and-currency pair could disappoint together. The narrowest exposure — the tech/AI-capex leadership theme priced barely above a coin flip at 47-55% — would be the most fragile, since sub-50% base rates offer little margin if growth turns mixed.

Base rates: moderate signals 47% win [n=109], elevated signals 43% win [n=113], extreme outliers 64% win [n=14]Brier Score: 0.302 (Poor) [n=2416] | Drift: +0.006 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Range-Bound Rates & Narrow Tech Leadership – Spiderweb / Interconnected Market Implications (30–90 days, as of July 14, 2026)

Core Thesis

Dominant: Statistical extremes across rates, FX, and commodities revert toward means within a stable low-VIX regime, keeping the 10y range-bound and preserving narrow AI-led equity resilience (~48% joint over 30-day horizon [uncalibrated composite]). Factors are correlated via a shared "calm-regime + mean-reversion" driver, so joint is bounded, not naively multiplied: P(rate reversion) ~65% × P(commodity/FX reversion | rates calm) ~75% × P(AI leadership holds | both) ~70% ≈ 34%, floored to ~45–50% given the common driver. Alternative: regime shift — extremes persist as structural repricing, breaking reversion positioning (~30% [uncalibrated]). Key discriminator: whether VIX and realized rate vol stay compressed over the next 20 trading days.

Markets Getting Stronger & Spiderweb Implications

• ALB (long, reversion): Base rate 77% reversion-higher [n=1686], adjusted to 68% [n=1615] for ag/commodity context; F4-decayed to ~55–60% at 30-day horizon (-10pp horizon, +2pp calm regime).

Implications: A commodity-side reversion confirms the broad "extremes revert" thesis, reinforcing the SECONDARY mean-reversion signal (61%) that anchors the whole web.

• AMD (long vs equal-weight): WATCH momentum-continuation base 0.58 [n=128] → 62% thesis; but -10pp day-over-day delta flags fading conviction. Outside-view first: momentum persistence base ~58%.

Implications: Sole momentum (not reversion) leg — depends entirely on AI capex persistence (47%), the web's weakest structural pillar.

Markets Getting Weaker & Spiderweb Implications

• TNX (short / receive 10y): Reversion-lower base 77% [n=1686] → 65% [n=1615]; F4-decayed to ~55–62% at 30-day (-8pp horizon, -4pp bear-case overweight per LJ3).

Implications: Lower/range-bound yields are the keystone — they enable AI multiple support and cap contagion.

• EURUSD (short vs USD basket): Momentum-continuation-lower 0.58 anchor → 59%, weakest PRIMARY. Dollar firmness pressures EM and commodities against the ALB long — an internal tension in the web.

The Connecting Spiderweb (Key Interconnections)

Leverage point — TNX (rates): Rate direction cascades furthest. Range-bound yields (65%) → AI-duration multiples hold (supports AMD + capex) → USD stays firm but not disruptive (EURUSD short) → commodity reversion (ALB) proceeds without a rate-shock overlay.

• Supporting: AI capex (47%) is the fragile dependency — if it fails, AMD's momentum leg collapses independent of rates.
• Supporting: EURUSD short and ALB long are partially contradictory (strong dollar headwinds commodities); the web tolerates this only in a calm regime.
• CT5 Non-linear risk: A sharp yield break (TNX spikes rather than reverts) triggers non-linear duration de-risking — AI leaders unwind disproportionately, margin-driven forced selling amplifies far beyond a linear rate→equity mapping. Small rate input → outsized equity output.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Rate reversion ⊕ commodity reversion ⊕ AI persistence = "calm mean-reversion regime with narrow leadership."
• Negation (¬):

| ¬ Scenario | Condition | Implication |

|---|---|---|

| ¬Reversion (regime shift) | Extremes persist beyond 2× normal reversion window | Structural repricing, not anomaly — invalidates TNX/ALB reversion legs (F2) |

| ¬AI persistence | Capex guidance cuts | AMD collapses; rates reversion insufficient to hold equities |

| ¬Dollar firmness | USD basket softens | EURUSD short fails but ALB long strengthens |

• Equivalence (∼): ALB reversion ∼ TNX reversion — both draw on the same ~77% mean-reversion base rate and calm-regime discount.

Ideas for Thinking About the Spiderweb (Mental Models from Guardrails)

Dominant lens — Forecasting.md: This thesis rests on short-horizon (4-day) calibrations stretched to 30–90 days. F4 horizon decay is the governing risk — every headline probability must be discounted ~8–10pp, and the 47% AI capex signal (declining -3pp) is below coin-flip, so the equity-resilience pillar is not empirically supported. Supporting — HeuristicAlgebra.md: use conditional (not independent) composition given the shared calm-regime driver. Challenge hypothesis — ScientificMethod.md: the regime-shift alternative is a competing hypothesis; falsification test = if realized 10y vol exceeds its 30-day mean by >2σ, reversion framing is rejected in favor of structural repricing.

Practical Prompts

• Watch TNX 10y yield over 20 trading-day window — if yield breaks above +2σ (30-day mean) rather than reverting, rate-reversion keystone is invalidated; de-risk AI-duration longs.
• Watch AMD vs equal-weight index over 10 trading-day window — if AMD fails to outperform by >1%, the momentum leg (already -10pp) is falsified.
• Watch **A

Devil's AdvocateIF this forecast turns out to be wrong, the most likely failure mode would be that the moderate-conviction signals in this batch — the category historically winning only around 44% of the time across a large 112-trade sample — fail to clear their coin-flip base rate, dragging down a cluster of positions that are correlated through a single macro narrative. A second failure path would be over-reliance on the mean-reversion thesis: the elevated per-position probabilities (59–68%) sit well above the observed win rates for comparable setups, so a regime in which rates trend rather than stay range-bound would allow one broken assumption to cascade across multiple linked trades simultaneously. Finally, with an overall Brier score of 0.302 indicating only modest calibration, the weakest scenario would be that the stated probabilities are systematically too confident, meaning the realized hit rate lands closer to the historical 44–49% band than the 60%+ implied here.

Base rates: moderate signals 49% win [n=105], elevated signals 44% win [n=112], extreme outliers 64% win [n=14]Brier Score: 0.302 (Poor) [n=2410] | Drift: +0.006 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Statistical Extremes Meet Range-Bound Regime – Spiderweb Implications (30–90 Day Horizon, as of July 13, 2026)

Core Thesis

Dominant: A cluster of >2σ signals (AMD, ALB, TNX) is resolving toward mean reversion inside a range-bound, restrictive-policy regime — favoring tactical reversion trades over trend extension (~48% joint, correlated). Alternative: AI capex persistence keeps large-cap tech leadership intact, invalidating the reversion-into-range thesis (~30% [uncalibrated]). Key discriminator: whether AMD's +14pp probability jump reflects genuine downside skew or a crowded, one-sided AI de-risking that snaps back.

Markets Getting Stronger & Spiderweb Implications

• ALB (long reversion): Base rate 77% [n=1686] → 70% [n=1615] inside-view, decaying to ~58–62% at 30-day horizon (-8pp F4 horizon decay, -2pp geo-uncertainty drag). Signal favors reversion higher off >2σ (6-day lookback) dislocation.

Implications: A materials/commodity-linked bounce loosens the "everything reverts down" tension; strengthens the range-bound (not breakdown) reading of the web.

• USD vs EUR (tactical long): Outside view 56% [n=2712] → 57% inside-view, effectively coin-flip at 30-day horizon (~52–55% after decay).

Implications: Restrictive front-end policy anchors USD strength, reinforcing the TNX range-bound and pressuring risk proxies (AMD).

Markets Getting Weaker & Spiderweb Implications

• AMD (put protection / reduced long): Base rate 77% [n=1686] → 72% [n=1615]; horizon-adjusted ~60% (-10pp F4, -2pp AI-support offset). Downside skew (-4% vs +1% expected range), +14pp daily jump.

Implications: AMD is the fulcrum — AI-proxy de-risking transmits to the entire tech-leadership thesis; the +14pp spike is the single most information-dense move in the set.

• TNX (yield pullback via duration): Base rate 77% [n=1686] → 65% [n=1615], ~55–60% at horizon (-8pp F4, -4pp restrictive front-end drag). Range-bound, not breakout.

Implications: Capped yields support duration but a restrictive front end limits the rally — the ceiling on the whole reversion regime.

The Connecting Spiderweb (Key Interconnections)

Leverage point — AMD's +14pp jump (58→72%). This is the highest-cascade signal: AI-proxy sentiment drives (a) tech leadership (SECONDARY capex 50%), (b) VIX/mean-reversion cluster, and (c) risk-on/risk-off USD demand. A resolved AMD downside validates the range-bound thesis across all four PRIMARYs simultaneously — which is precisely why the factors are correlated, not independent.

• Supporting: Restrictive front end (TNX/USD) caps upside, keeping the system range-bound rather than trending.
• CT5 non-linear risk: A crowded one-sided AI de-risk could invert non-linearly — if AMD's put demand unwinds, a short-covering squeeze cascades through correlated semis and lifts the entire tech complex, overwhelming the modest ALB/TNX reversion signals. Small positioning input → disproportionate index output.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): AMD-weak ⊕ TNX-range ⊕ USD-strong = restrictive, range-bound consolidation regime (~48% joint, correlated via shared risk-sentiment driver; not the ~16% naive product).
• Negation (¬): (1) AI capex re-accelerates → AMD reverts up, breaking the down-reversion cluster. (2) F2 disequilibrium: if ALB/AMD extremes persist beyond 2× historical reversion window (>12 days), this signals structural regime shift (supply/AI-demand), invalidating mean-reversion positioning. (3) Front-end policy softens → TNX breaks range lower, USD long fails.
• Equivalence (∼): AMD-put-protection ∼ VIX-reversion; both express the same "extended momentum pauses" bet.

Ideas for Thinking About the Spiderweb (Mental Models)

Dominant lens — Forecasting.txt (F4 horizon decay, F2 non-stationarity): every headline probability here is a 4-day calibration; the mechanism decaying these to ~55–62% at 30 days is the core adjustment governing this brief. F2 forces the explicit question: dislocation or regime shift? Supporting — HeuristicAlgebra.txt: the correlated ⊕ combination prevents overstating joint confidence. CriticalThinking.txt (CT7 conjunction): challenges the thesis via a falsification test — if AMD, ALB, TNX resolve independently (divergent directions) within 15 days, the shared-driver assumption is false and the ~48% joint collapses.

Practical Prompts

• Watch AMD implied downside skew over 10 trading-day window — if AMD rallies >4% against put positioning, reversion-down thesis is invalidated; implies AI-leadership regime (¬1).
• Track ALB reversion over 6 trading-day window — if ALB fails to close its >2σ gap, reversion signal is stale, tighten stop.
• Monitor TNX range over 25-day window — if yields break the range decisively (breakout, not pause), range-bound regime and duration thesis are falsified (F2 regime shift).
• Track EURUSD over **20

Devil's AdvocateThe most likely reason for failure would be a calibration gap between the confidence levels attached to today's themes (57-72%) and the historical reality that the two highest-volume signal tiers have resolved at only 44-49% win rates across roughly 217 trades — meaning stated edges have tended to run well ahead of realized outcomes. A second failure mode would emerge if the mean-reversion thesis on statistically extended moves is overwhelmed by a persistent-momentum regime; the modest average returns (+0.50% to +0.87%) leave little cushion, so a single strong trend day against the range-bound assumption could erase the thin expected value. The overall Brier score of 0.302 across 2,404 forecasts signals meaningful probabilistic imprecision, so the weakest links here would be the near-coin-flip theses where a 50-57% assumption offers almost no margin for error.

Base rates: moderate signals 49% win [n=105], elevated signals 44% win [n=112], extreme outliers 64% win [n=14]Brier Score: 0.302 (Poor) [n=2404] | Drift: +0.003 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Falling Yields Meet Semiconductor Momentum – Spiderweb Implications (30–90 Days, as of July 12, 2026)

Core Thesis

A low-VIX, low-panic regime supports risk-on continuation (tech/semis momentum) while yields mean-revert lower, easing duration stress. Dominant: coordinated "soft-landing risk-on" — falling yields + AI capex resilience lift SOXX and cyclical laggards like ALB (~40% joint [uncalibrated, see decomposition]). Alternative: yields fall on growth scare, not benign disinflation, decoupling bonds from equities (~30%). Key discriminator: whether TNX decline coincides with credit-spread widening (growth scare) or stable spreads (benign).

Markets Getting Stronger & Spiderweb Implications

• TNX (long bonds / short yields): Mean-reversion base rate ~62% for rate extremes [outside view]; adjusted to 65% [n=1615, backtest_1042d] on -15bp vs +8bp skew over 4-day window. Decaying to ~55% at 30-day horizon (-10pp F4 horizon decay).

Implications: Lower yields relax discount rates on long-duration tech — directly feeds AMD/SOXX and AI-capex thesis. Highest-leverage node.

• AMD / SOXX (momentum): Base rate for momentum continuation in low VIX ~55%; 58% [n=1615] here. Decays to ~50% at 30-day.

Implications: Reinforced by AI capex persistence (SECONDARY, 53%) and falling yields — a self-reinforcing loop while VIX stays suppressed.

• ALB (long on downside): +12pp delta (56%→68%) is the strongest single mover. Base rate mean-reversion ~64% [outside view]; 68% [n=1615]. Decays to ~58% at 30-day.

Implications: Lithium/materials strength requires the low-panic commodity regime (SECONDARY, 63%) to hold; a cyclical-recovery confirmation if it co-moves with semis.

Markets Getting Weaker & Spiderweb Implications

• EURUSD (fade EUR strength): -2pp delta (57%→55%), weakest conviction. Base rate ~54%; 55% [n=1615], asymmetric (0.8% up vs 1.2% down). Decays to ~52% at 30-day — near coin-flip.

Implications: A softer EUR is dollar-supportive, a mild headwind to commodities (ALB) and EM risk appetite — a tension within the risk-on web.

The Connecting Spiderweb (Key Interconnections)

Leverage point — TNX: Falling yields is the master node. It cascades into (1) SOXX/AMD via lower discount rates, (2) AI-capex financing conditions, and (3) commodity carry (ALB). Nearly every bullish signal traces back to the rate path.

• Supporting: Low VIX + contained geo-energy transmission (63%) is the permission slip keeping the momentum loop intact.
• Non-linear risk (CT5): If yields fall because of a growth scare, the TNX→SOXX link inverts — bonds rally while equities fall, and low VIX becomes fragile. A single credit-spread gap could trigger disproportionate de-risking (VIX regime break, forced deleveraging in crowded semi longs), converting a benign 15bp move into a cascade.
• EURUSD counter-thread: Dollar strength partially offsets the commodity leg, capping ALB upside.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): TNX↓ ⊕ low-VIX ⊕ AI-capex = "duration-relief risk-on." Joint P (correlated via shared VIX/rate driver): P(TNX reverts) 65% × P(SOXX up | yields fall) ~70% × P(ALB up | both) ~65% ≈ 30%, bounded to ~38–45% given shared regime driver — NOT independent multiplication.
• Negation (¬):

| Scenario | Trigger | Implication |

|---|---|---|

| Growth-scare yields | Yields fall + spreads widen | Bond/equity decouple; risk-on invalidated |

| Regime shift (F2) | ALB/commodity extremes persist beyond 2× reversion window | Structural supply deficit, not anomaly — mean-reversion longs fail |

| VIX break | Geo transmission un-contained | Low-panic assumption void; momentum unwinds non-linearly |

• Equivalence (∼): ALB reversion ∼ TNX reversion — both bet on statistical pull-back within a stationary regime; both fail identically under regime shift.

Ideas for Thinking About the Spiderweb (Mental Models)

• Forecasting.md (dominant lens): Every probability here is a 4-day backtest calibration [n=1615] stretched to a 30–90 day horizon. Apply F4 decay rigorously — the headline 68% ALB is realistically ~58% at 30 days. Treat the confidence cone as medium and widen it further at the far horizon.
• Critical_Thinking.md (challenge): The thesis assumes benign disinflation. Competing hypothesis: yields fall on growth deterioration. Falsification test — track HY credit spreads; if they widen >30bp while TNX falls, the benign interpretation is falsified and the entire risk-on web collapses.
• Heuristic_Algebra.md (support): Enforces conjunction decay — the four PRIMARY signals share a VIX/rate driver, so joint confidence must be bounded (~40%), not naively multiplied to a misleading ~14%.

Practical Prompts

1. Watch **TNX vs HY cred

Devil's AdvocateThe most likely reason for failure would be an over-reliance on stated probabilities (55–68%) that outrun the historical hit rate of the underlying signal tiers, where the two largest sample buckets have only converted 44–49% of the time across roughly 100+ observations each — meaning several of today's coin-flip-adjacent themes (the 53% and 55% calls) could easily disagree with their assigned edge. A second failure mode would be correlated breakdown: if the 'calm regime' assumption cracks, the bond-friendly rates view, the tech-leadership resilience theme, and the contained-energy thesis could all miss together, since they implicitly share one macro backbone. Finally, a current Brier score of 0.302 across ~2,398 forecasts signals only modest calibration, so the higher-conviction probabilities here may be somewhat overstated relative to demonstrated accuracy.

Base rates: moderate signals 49% win [n=105], elevated signals 44% win [n=112], extreme outliers 64% win [n=14]Brier Score: 0.302 (Poor) [n=2398] | Drift: -0.002 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Low-Vol Mean Reversion Meets Small-Cap Momentum – Spiderweb Implications (30–90 Day Horizon, as of July 11, 2026)

Core Thesis

Dominant: A benign low-VIX regime lets mean reversion in beaten-down materials/FX (ALB, EURUSD) coexist with small-cap momentum (IJR), producing broad but shallow equity breadth (~40% joint, [uncalibrated] at 30-day horizon). Alternative: The low-vol regime is a compressed-convexity trap where a vol spike >17 breaks both the momentum and reversion legs simultaneously (~35%). Key discriminator: VIX behavior — sustained sub-17 confirms dominant; a single close >17 invalidates IJR and pressures reversion legs.

Markets Getting Stronger & Spiderweb Implications

• IJR / Small-caps (PRIMARY): Base rate 0.58 [n=1615, backtest_1042d], momentum continuation over 4-day horizon; Russell +4.1% (30d) with low VIX. Decompose: P(VIX stays 17 falsifier), underwrites the risk score behind EURUSD, and provides the "low vol confirmation" that lifted ALB's base rate. Every thesis leg is a conditional child of this one node.

• Supporting: EURUSD weakness → materials (ALB, LIT) reversion via dollar channel.
• Supporting: IJR breadth → risk-on sentiment → EURUSD de-escalation loop (reflexive, self-reinforcing while calm holds).
• CT5 Non-Linear Risk: The short-convexity tail (52%) means a small VIX move >17 is not linear. Compressed positioning implies a single vol shock triggers simultaneous IJR stop-outs, dollar bid, and ALB reversion failure — a synchronized unwind where three "independent" 56–58% legs collapse together because they share one driver. Margin de-risking amplifies beyond linear expectation.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Low-VIX ⊕ small-cap momentum ⊕ USD-fade = "benign broadening" regime — but leg correlation (shared VIX driver) means joint confidence bounds to ~35–45%, not the naive product.
• Negation (¬) Scenarios:

| Negation | Condition | Implication |

|---|---|---|

| VIX >17 | Vol regime breaks | All three PRIMARY legs fail simultaneously |

| EURUSD 17 on any single day, the entire spiderweb (IJR, EURUSD, ALB) is invalidated as a synchronized unwind.

• Watch ALB over 5 trading-day window — if it fails to stabilize above recent low, the -9pp decay confirms reversion-leg failure, not noise.

Devil's AdvocateIF this forecast turns out to be wrong, the most likely failure mode would be that the moderate-conviction signals underperform their stated ~56-58% probabilities, since the broader base rate for this tier of signal has historically resolved closer to a coin flip (44-49% win rates across ~217 samples) — meaning the calibration gap between the assumed edge and the realized hit rate is the primary vulnerability. A second failure path would be that the thesis leans on a calm-volatility regime persisting; if that low-volatility backdrop breaks, the several theses tied to broad equity breadth could correlate and fail together, converting what looks like diversification into a single concentrated bet. Finally, with an overall Brier score of 0.301, the historical record suggests these probability estimates carry meaningful miscalibration, so returns clustering near zero rather than the modeled convexity would be the quietest but most probable way this is wrong.

Base rates: moderate signals 49% win [n=105], elevated signals 44% win [n=112], extreme outliers 64% win [n=14]Brier Score: 0.301 (Poor) [n=2392] | Drift: -0.011 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Risk-On Rotation on Fragile Foundations – Spiderweb / Interconnected Market Implications (30–90 days, as of July 10, 2026)

Core Thesis

Dominant: A low-intensity risk-on regime — small-cap momentum, oversold mean reversion (ALB/EURUSD), and stable-to-lower real yields reinforcing each other (~40% joint over horizon [uncalibrated aggregation]). Alternative: Extremes fail to revert; a regime shift keeps EURUSD/commodities dislocated as small-cap momentum stalls (~35%). Key discriminator: whether TNX holds range and VIX cone stays contained — a real-yield break-out invalidates the risk-on cascade.

Markets Getting Stronger & Spiderweb Implications

• IJR (small-caps vs large-caps): Momentum base rate 58% [n=1615, backtest_1042d] over 4 days → ~50-55% at 30-day horizon (-6pp F4 horizon decay). Skew favorable (+2.1% vs -0.8%).

Implications: Small-cap leadership signals broad risk appetite, pulling semis (AMD) and cyclical credit along — a breadth amplifier for the web.

• ALB (mean reversion): Outside-view base rate 65% [n=1615, backtest_1042d] at 4 days → ~52-58% at 30-day (-8pp F4 decay, -3pp commodity non-stationarity). Convexity via defined stop 4 days then reverts | Exit rule triggers; breadth thesis weakens |

• Equivalence (∼): ALB reversion ∼ EURUSD rebound — both are 2σ-decliner mean-reversion plays sharing one sentiment engine.

Ideas for Thinking About the Spiderweb (Mental Models)

Dominant lens — Heuristic Algebra (heuristic_algebra.md): The four positions are not additive; they combine through a shared risk-sentiment operator, so joint confidence must use conditional (not independent) probability — hence the ~40% joint bound despite individual 57-65% legs. Supporting — Forecasting (forecasting.md): F4 horizon decay mandates discounting all 4-day base rates toward ~50-55% at the 30-90 day window; F2 warns the commodity extremes may be structural. Challenge test — Critical Thinking (critical_thinking.md): The competing regime-shift hypothesis is distinguished by one test — if ALB fails to revert within 2× its historical window while VIX stays low, the dislocation is structural, falsifying mean reversion.

Practical Prompts

• Watch TNX real yields over 20 trading-day window — if yields break above range on inflation data, risk-on web is invalidated; unwind ALB/IJR/EURUSD together.
• Watch **IJR vs large-caps over 10

Devil's AdvocateIF this forecast proves wrong, the most likely failure mode would be that the stated probabilities (57-65%) are meaningfully more optimistic than the historical base rates justify — lower-conviction signals in this framework have resolved at just 42-48% win rates over 200+ recent trades, so the largest cluster of positions here is effectively coin-flip territory dressed up as an edge. A second failure mode would be correlated breakdown: several of the positions lean on the same underlying themes (a risk-on regime, a weaker reserve currency, small-cap outperformance, and a commodity rebound), so if the macro backdrop shifts the wrong way, these do not fail independently but cascade together. Finally, the overall calibration score of 0.301 suggests the model's probability estimates carry real error, meaning the aggregate 'medium confidence' framing could overstate the true reliability of today's basket.

Base rates: moderate signals 48% win [n=104], elevated signals 42% win [n=111], extreme outliers 64% win [n=14]Brier Score: 0.301 (Poor) [n=2384] | Drift: -0.014 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Reversion & De-Escalation Regime – Spiderweb of Rate, FX, and Commodity Interconnections (30–90 Day Horizon, as of July 09, 2026)

Core Thesis

Dominant: A coordinated mean-reversion regime — lower 10Y yields, higher EURUSD, lithium rebound, small-cap weakness — driven by a shared "de-escalation + statistical dislocation unwind" mechanism (~35% joint, correlated). Alternative: yields grind higher and risk-premium stays elevated as extremes reflect regime shift, not dislocation (~30% [uncalibrated]). Key discriminator: whether TNX reverts lower within the 4-day calibration window and holds through 20-day; a break higher invalidates the linked bond/FX/commodity leg. Joint note: factors share a single risk-sentiment/rates driver, so conditional (not independent) chaining applies.

Markets Getting Stronger & Spiderweb Implications

• ALB (lithium): Reversion-up. Base rate 77% [n=1686] → 72% [n=1615, source: backtest_1042d]: −5pp inside-view (stable geo, no supply shock). Decomposed: P(dislocation unwind) ~85% × P(no fresh supply glut | unwind) ~85% ≈ 72%.

Implications: A commodity rebound corroborates the de-escalation branch (52%), reinforcing risk-currency strength (EURUSD) and lower safe-haven bond demand.

• EURUSD: Long, falsified below 1.12. Base rate 77% → 63%: −14pp for USD momentum drag and horizon decay. F4: 63% over 4d → ~52–58% at 30-day.

Implications: Weaker USD lowers imported inflation pressure, easing yield-upside risk — a self-reinforcing loop with the TNX leg.

• Bonds (via lower TNX): See leverage point below.

Markets Getting Weaker & Spiderweb Implications

• IJR (small caps): Tactical short/avoid. Base rate 77% for 2σ+ [n=1686] → 60% given LJ3 bear-case overweight on contracting breadth. F4: 60% (4d) → ~50–55% at 30-day.

Implications: Small-cap weakness contradicts a clean risk-on read, flagging internal fragility even as EURUSD/ALB rise — the web's tension point.

• TNX (higher-yield bear case): LJ3 overweights continuation risk; if realized, cascades against every long leg here.

The Connecting Spiderweb (Key Interconnections)

Leverage point — TNX (10Y yields). This is the single highest-cascade signal: its direction sets bond prices, USD carry (EURUSD), commodity discount rates (ALB), and small-cap financing (IJR). Reversion lower (65% [n=1615], base 77% → −12pp momentum/horizon) validates the full web; a break higher unwinds it.

• Supporting: EURUSD strength ← lower yields compress USD carry advantage.
• Supporting: ALB reversion ← lower discount rates + de-escalation risk-premium relief.
• CT5 Non-linear risk: A disorderly yield spike (fiscal supply shock, failed auction) could trigger cross-asset margin de-risking — small-cap liquidation amplifying beyond linear expectation, dragging EURUSD and ALB down simultaneously despite their independent bull cases. Small input (one weak auction) → disproportionate correlated drawdown.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Lower TNX ⊕ de-escalation (52%) ⊕ USD unwind = "soft-landing reversion" regime favoring bonds, EUR, lithium.
• Negation (¬):

| Scenario | Trigger | Effect |

|---|---|---|

| Yield regime shift | TNX rises past 20-day, no reversion | Invalidates bond/EUR/ALB legs (F2 disequilibrium) |

| De-escalation fails | Geo re-escalation | USD/haven bid, ALB rolls over |

| Persistent commodity extreme | ALB stays stretched beyond 2× reversion window | Signals structural deficit, not anomaly — kills reversion positioning |

• Equivalence (∼): EURUSD reversion ∼ TNX reversion — both express the same USD-carry/rate-differential mechanism; treat as one bet, not two.

Ideas for Thinking About the Spiderweb (Mental Models from Guardrails)

Dominant lens — Forecasting.md (F2, F4): The thesis rests on mean reversion, but F2 demands testing whether 2σ+ extremes are dislocations or regime shifts. Mechanism supporting thesis: 77% historical reversion base rate [n=1686]. Falsification test distinguishing regimes: if TNX and ALB both fail to revert within 2× the historical window (~8 days), reclassify as structural — abandon reversion positioning. Supporting: Heuristic_Algebra.md clarifies that TNX/EURUSD are ∼equivalent (shared driver), preventing double-counting correlated bets. Critical_Thinking.md (CT7) enforces conjunction decay — the ~35% joint figure reflects correlation, not naïve multiplication of four 60–72% legs.

Practical Prompts

• Watch TNX vs 30-day mean over 20 trading-day window — if yields rise above the 30-day mean and hold >5 sessions, the reversion web is invalidated; de-risk bond/EUR/ALB longs.
• Watch EURUSD over 10 trading-day window — if it breaks below 1.12, long thesis falsified (per input); flip

Devil's AdvocateIf this forecast proves wrong, the most likely failure mode would be over-reliance on mean-reversion at stretched statistical extremes during a period when the underlying signals have historically resolved little better than a coin flip — the mid-tier conviction buckets have shown only 43-48% win rates across roughly 200 samples, so several of the 60-72% probability estimates here may be optimistically calibrated relative to realized outcomes. A second failure channel would be correlated exposure: the bond, currency, and commodity themes all lean on the same de-escalation-and-normalization narrative, meaning a single macro surprise (a fresh geopolitical shock or a hawkish rate repricing) could invalidate multiple positions at once. Finally, with an overall Brier score near 0.30, the calibration edge is modest, so the confident sub-branch resting on only a handful of high-conviction historical cases carries meaningful small-sample fragility.

Base rates: moderate signals 48% win [n=104], elevated signals 43% win [n=111], extreme outliers 64% win [n=14]Brier Score: 0.301 (Poor) [n=2378] | Drift: -0.013 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Lithium Reversion Meets Small-Cap Momentum – Spiderweb / Interconnected Market Implications (30–90 Day Horizon, as of July 08, 2026)

Core Thesis

Dominant: A "risk-on reversion" regime where oversold lithium rebounds, small-caps extend, and 10Y yields drift lower on geopolitical de-escalation — a mutually reinforcing risk-appetite cycle (~40% joint, correlated). Alternative: Supply-shock/regime-shift path where commodity extremes persist and yields back up on inflation or hotspot escalation (~30% [uncalibrated]). Key discriminator: whether 10Y yields fall ≥5bp with stable VIX (confirms de-escalation) vs. back up on CPI/geopolitical shock. Conjunction: P(lithium reverts) ~70% × P(IWM holds | risk-on) ~58% × P(yields drop | de-escalation) ~65% would imply ~26% if independent, but shared risk-sentiment driver lifts joint to ~38–42%.

Markets Getting Stronger & Spiderweb Implications

• Lithium (ALB, -1.98σ vs 30-day mean): Mean-reversion base rate 76% [n=1615, backtest_1042d, 4-day] → ~62% at 30-day (F4: -10pp horizon decay, -4pp regime uncertainty). Decompose: P(no fresh supply shock) ~80% × P(reversion | no shock) ~78% ≈ 62%.

Implications: Lithium recovery signals broader industrial-metals rebound (SECONDARY 65%), feeding EV/clean-energy sentiment and small-cap cyclicals — a node connecting commodities to equity breadth.

• Small caps (IWM/IJR, +2σ vs 30-day mean ALERT): Momentum-continuation base rate ~55% [n=1615] → ~52% at 30-day given stretched entry.

Implications: Small-cap strength depends on financial breadth and lower yields — most fragile node, exposed if rate path reverses.

Markets Getting Weaker & Spiderweb Implications

• 10Y Treasury yields: Mean-reversion-lower base rate 65% [n=1615, 4-day] → ~55% at 30-day (F4 decay + CPI risk).

Implications: Falling yields lubricate the entire risk-on web — supporting small-cap valuations and commodity carry. But this is conditional on de-escalation (58%), the web's weakest structural assumption.

The Connecting Spiderweb (Key Interconnections)

Leverage point (CT4): 10Y yields — the single signal cascading into the most nodes. Lower yields simultaneously (a) support stretched small-cap multiples, (b) weaken USD to aid commodity/lithium reversion, and (c) confirm the de-escalation narrative. If yields fall on de-escalation, three signals align.

• Supporting: Lithium reversion ⊕ industrial-metals rebound reinforce cyclical risk appetite feeding IWM.
• Supporting: De-escalation lowers term premium → yields drop → risk assets bid.
• Non-linear risk (CT5): A hotspot escalation is a small input with disproportionate output — it simultaneously spikes yields (safe-haven flip or inflation fear), shocks commodities upward (breaking reversion), AND crushes small-cap breadth. One escalation event can invalidate all three PRIMARY theses at once via correlated unwinding, not linear degradation.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Lithium reversion ⊕ de-escalation ⊕ falling yields = coherent risk-on reflation regime.
• Negation (¬) Scenarios:

| Scenario | Trigger | Effect |

|---|---|---|

| Regime shift (F2) | Lithium extreme persists past 2× reversion window | Structural supply deficit, not anomaly — invalidates long |

| Hot CPI | Inflation surprise | Yields back up, breaks leverage point |

| Escalation | Hotspot strike | Correlated risk-off, all PRIMARYs fail |

• Equivalence (∼): Small-cap ALERT ∼ late-cycle momentum stretch — analogous to prior +2σ breadth episodes that often round-tripped.

Ideas for Thinking About the Spiderweb (Mental Models)

Dominant lens: Forecasting.md — F4 horizon decay is central: every signal is calibrated on 4-day holds, so the 30–90 day brief must discount ~10–15pp. Mechanism: reversion edges dissipate as new information arrives; the thesis holds only if fresh catalysts sustain it. Supporting: Heuristic_Algebra.md — the ⊕ of three correlated risk-on signals must not be multiplied independently (CT7); shared sentiment driver bounds joint confidence. Challenge: Critical_Thinking.md — competing regime-shift hypothesis (F2): if lithium fails to revert within 20 days AND industrial metals stay >2σ, this falsifies statistical-dislocation framing and confirms structural deficit.

Practical Prompts

• Watch 10Y yield over 20 trading-day window — if yields rise >15bp on CPI/escalation, leverage point fails and full risk-on web is invalidated.
• Watch ALB vs. broad commodity basket over 15-day window — if ALB fails to close >2% of the -1.98σ gap, mean-reversion thesis weakens.
• Watch IWM vs. SPX over 10 trading-day window — if breadth momentum ### Devil's AdvocateIF this forecast turns out to be wrong, the most likely failure mode would be that the elevated probabilities assigned to the highest-conviction themes (up to 76%) overstate their true edge — the mid-tier signal tier this forecast leans on has historically converted at only 44% with n=107, meaning coin-flip outcomes are common and the stated confidence may not be earned. A second likely failure path would be the mean-reversion thesis on extreme statistical outliers: while deeper-extreme signals have shown a 64% hit rate, that reading rests on a thin sample of just 14 observations, so the apparent strength could be statistical noise rather than a durable pattern. Finally, the overall calibration backdrop (Brier score of 0.301) suggests the forecasting engine is only modestly better than chance, so clustering multiple correlated themes — commodity reversion, small-cap breadth, and geopolitical de-escalation — risks a single macro shock unwinding several positions at once.

Base rates: moderate signals 50% win [n=98], elevated signals 44% win [n=107], extreme outliers 64% win [n=14]Brier Score: 0.301 (Poor) [n=2372] | Drift: -0.011 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Reversion Regime Under Test – Spiderweb / Interconnected Market Implications (30–90 days, as of July 07, 2026)

Core Thesis

Dominant: A broad 4-day mean-reversion regime is signaling — small-caps underperform AI leaders, EUR and duration revert higher, and industrial commodities firm relative to gold — consistent with an orderly risk-sentiment stabilization (~40% at 30-day horizon). Alternative: These are not temporary dislocations but early regime-shift signals (breadth contraction + fading commodity delta suggesting late-cycle rotation, ~30%). Key discriminator: whether 10Y yields revert lower on soft data (thesis) or hold/rise on hot data (regime shift).

Joint confidence decomposition (correlated, not independent — all share a single risk-sentiment driver): P(IWM/NVDA reverts) ~68% [n=1615] × P(EURUSD higher | risk stable) ~62% × P(yields lower | both) ~60% → naive product ~25%, bounded up to ~38–42% given shared driver reduces conjunction penalty. Note calibration: 68% base is a 4-day figure [n=1615, backtest_1042d]; at 30-day horizon → ~52–58% (−10pp F4 horizon decay, −3pp breadth-contraction drag).

Markets Getting Stronger & Spiderweb Implications

• NVDA / AI proxies (relative to IWM): Reversion base rate 68% [n=1615, 4-day], decaying to ~55% at 30-day. Implications: Large-cap leadership persistence pulls index breadth narrower — strengthens the "concentration" node while weakening small-cap and cyclical breadth.
• EURUSD: Outside-view reversion base ~62% [n=1615], adjusted to ~56% at 30-day given DXY stability assumption. Implications: A firmer EUR ratifies the lower-yield node (dollar-yield feedback) and supports commodity-currency second-order strength.
• Crude vs Gold (relative value): 55% [n=1615], down from 59% (−4pp) — momentum fading. Implications: Weak-but-positive; only modestly reinforces the "no-escalation, orderly reflation" web.

Markets Getting Weaker & Spiderweb Implications

• IWM small-caps: Reversion skewed to downside (−2.8% vs +0.9%), ~68% [n=1615, 4-day]. Implications: Breadth contraction node — the SECONDARY ALERT (55%) confirms consolidation, pressuring cyclical risk appetite.
• 10Y yields: Positioned lower, 60% [n=1615, 4-day] → ~52% at 30-day. Implications: Falling yields are supportive for duration and EUR, but weakening = the load-bearing assumption most at risk from hot data.

The Connecting Spiderweb (Key Interconnections)

Leverage point — 10Y yields (CT4). Yields cascade into the most nodes: lower yields support EURUSD (dollar weakness), sustain AI-leader multiples (duration-sensitive growth), and validate the orderly-reversion regime. A single hot CPI/payroll print inverts all three simultaneously.

• Supporting: EURUSD strength is largely derivative of the yield node — not an independent bet.
• Supporting: Fading crude/gold delta (−4pp) is the early crack — commodity conviction eroding while equity/FX signals hold.
• Non-linear risk (CT5): IWM downside + breadth contraction creates cascade potential — if small-cap liquidation triggers systematic de-grossing, the AI-leader "safe" leg could gap lower with small-caps (correlation → 1 in stress), collapsing the pair trade's assumed hedge. Small input (one liquidation) → disproportionate cross-asset output.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Lower yields ⊕ firmer EUR ⊕ AI leadership = "orderly-reflation-with-narrow-breadth" regime.
• Negation (¬):

| Scenario | Trigger | Implication |

|---|---|---|

| ¬Yield reversion | Hot data | Entire web inverts; EUR/AI legs fail |

| ¬Reversion (regime shift, F2) | Commodity extremes persist >2× normal reversion window | Signals structural, not statistical — mean-reversion positioning invalidated |

| ¬Hedge decorrelation | Systematic de-gross | IWM/NVDA both fall; pair-trade edge collapses |

• Equivalence (∼): EURUSD long ∼ short-USD ∼ long-duration — three expressions of one yield-driven bet (avoid double-counting conviction).

Ideas for Thinking About the Spiderweb (Mental Models from Guardrails)

Dominant lens — Forecasting (Forecasting.md): F4 horizon decay is the core discipline here — every 4-day base rate [n=1615] must be discounted for the 30–90 day window; treat headline probabilities as decaying, not static. Mechanism: the reversion edge is a short-holding artifact. Supporting — Critical Thinking (CriticalThinking.md): CT7 conjunction decay flags the correlated-driver trap; the four PRIMARIES are not independent votes but one risk-sentiment bet. Falsification test distinguishing thesis from regime-shift: does 10Y revert lower on the next soft print? If yes → reversion; if yields rise on soft data → regime shift confirmed. Supporting — Value (Value.md): The fading crude/gold delta (−4pp) tests whether the reflation leg carries real edge or is decayed noise.

Practical Prompts

• Watch

Devil's AdvocateThe most likely reason for failure would be that this forecast leans heavily on mean-reversion themes across equities, currencies, rates, and commodities simultaneously — and if a persistent trend or macro regime shift dominates instead, several of these correlated bets could fail together rather than diversifying risk. Historically the moderate-conviction signals underpinning much of this book have delivered only a 44-50% win rate (n=96-104), meaning roughly half of such calls do not resolve favorably even when the modest positive mean returns (+0.93-0.96%) hold. With an overall Brier score of 0.301 (n=2366) reflecting meaningful calibration slippage, the probability estimates clustered in the 55-68% range may be optimistic, and the weakest sub-theses at 48-55% offer little margin above a coin flip if volatility spikes or a directional catalyst overwhelms the reversion setups.

Base rates: moderate signals 50% win [n=96], elevated signals 44% win [n=104], extreme outliers 64% win [n=14]Brier Score: 0.301 (Poor) [n=2366] | Drift: -0.011 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Breadth Rotation Fade & The Disinflation Handoff – Spiderweb Implications (30–90 days, as of July 06, 2026)

Core Thesis

Dominant: A cluster of statistical extremes (IWM outperformance, EURUSD undershoot, 10Y yields elevated) mean-reverts together as a single "risk-recalibration" regime driven by disinflation and stable-rate expectations (~52% joint [uncalibrated, see decay below]). Alternative: The extremes persist as a genuine regime shift — sticky rotation + structural yield repricing (~30%). Key discriminator: NFP surprise and China demand data — a beat re-anchors yields higher and validates persistence over reversion.

Markets Getting Stronger & Spiderweb Implications

• EURUSD (long, mean-reversion): Base rate for -1.9σ reversions ~64% over 4 days [n=1615]; anchored to outside-view, adjusted to ~58% at 30-day horizon (−6pp F4 horizon decay). Implications: A EUR bounce is the mirror image of the softer-USD, lower-yield thesis; it reinforces the disinflation web and eases financial conditions globally.
• AI proxies (NVDA/AMD) vs IWM (relative): Concentration re-asserts as small-cap breadth fades. Base rate 77% at 2σ+ [n=1615]. Implications: Capital rotates back into mega-cap quality, tightening breadth and validating the IWM-short leg.

Markets Getting Weaker & Spiderweb Implications

• IWM relative strength (short): +2.06σ vs 30-day AI-proxy mean; reversion base rate 72% [n=1615, 4-day] → ~62% at 30-day (−10pp F4). Implications: Breadth contraction (SECONDARY 47%, −2pp delta) removes the risk-on prop; small-caps/financials lose leadership.
• 10Y yields (position lower): +1.51σ vs 30-day mean; 62% reversion [n=1615] → ~54% at 30-day (−8pp F4). Implications: Lower long-end supports duration and the EURUSD reversion — same disinflation impulse.
• Industrial commodities (crude/copper, short): 59% [n=1615], anchored to 58% disinflation-continuation base rate. Implications: China demand weakness cascades into reflation trades and EM FX.

The Connecting Spiderweb (Key Interconnections)

Leverage point — 10Y yields. Long-end yields cascade into the most nodes: lower yields → weaker USD (EURUSD long confirms) → duration bid → mega-cap growth re-rating (AI proxies over IWM). Yields are the hinge on which four of six signals swing.

• Supporting: EURUSD reversion is largely a dependent variable of the yield move, not independent — treat as correlated, not additive.
• Supporting: Industrial commodity weakness feeds disinflation, closing the loop back to yields.
• Non-linear risk (CT5): An NFP beat is a small input with disproportionate output — it simultaneously invalidates the yield, EURUSD, and IWM-fade legs at once (shared driver), risking a correlated stop-out cascade rather than four independent losses. Crowded reversion positioning amplifies gap risk beyond linear expectation.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Lower yields ⊕ EUR reversion ⊕ commodity softness = "orderly disinflation, growth-quality leadership" regime.
• Negation (¬) Scenarios:

| Scenario | Trigger | Effect |

|---|---|---|

| Regime-shift (F2) | Commodity/yield extremes persist beyond 2× reversion window | Signals structural yield repricing/supply deficit — invalidates all mean-reversion legs |

| NFP beat | Hot labor print | Yields extend, USD firms, IWM-short and EUR-long both fail together |

| Breadth genuine | Financials/healthcare confirm | IWM rotation is durable, not statistical |

• Equivalence (∼): IWM-fade ∼ EURUSD-long — both are single-driver (yield/disinflation) bets wearing different clothes; do not size as diversified.

Ideas for Thinking About the Spiderweb (Mental Models)

Dominant lens — Heuristic Algebra (heuristic_algebra.md): The four PRIMARY legs are correlated via the yield hinge, so CT7 forbids independent multiplication. Mechanism: shared disinflation driver means joint confidence bounds to ~48–55%, not the ~17% naive product. Supporting — Forecasting (forecasting.md): F4 horizon decay demands the 4-day calibrations be discounted for the 30–90 day cone; every headline probability above is decomposed accordingly. Challenge lens — Scientific Method (scientific_method.md): The regime-shift hypothesis (F2) competes directly; falsification test — if the commodity/yield z-scores fail to compress by >50% within a 20-day window, reject mean-reversion in favor of structural repricing.

Practical Prompts

• Watch 10Y yield z-score vs 30-day mean over 20 trading-day window — if yields fail to compress below +0.75σ, the yield-hinge thesis is invalidated and EUR/IWM legs should be cut.
• Track IWM vs NVDA/AMD relative over 10 trading-day window — if IWM outperforms by >2%, breadth-persistence (regime shift) is confirmed, invalidate the short.
• Monitor EURUSD over 7-day window — if EUR fails to reclaim -1.0σ, USD-strength regime dominates; abandon reversion.

Devil's AdvocateThe most likely reason for failure would be that this forecast leans heavily on mean-reversion of statistical extremes (setups stretched 1.5σ to 2σ from equilibrium), yet the historical base rates for medium-conviction signals show only 44-50% win rates across the two largest sample buckets (n=104 and n=96), meaning roughly half of comparable setups have historically resolved against the thesis despite positive average returns of ~+0.95%. A second failure path would be that the interlocking calls — small-cap relative weakness, currency mean-reversion, lower long-end yields, and softer industrial commodities — are all correlated bets on the same macro regime (breadth contraction plus cooling demand), so a single surprise such as a stronger-than-expected jobs print or resilient China demand could invalidate several positions simultaneously rather than independently. Finally, with an overall Brier score of 0.301, the model's calibration is only modestly better than a coin flip, so the stated 59-72% probabilities may be optimistically wide relative to demonstrated forecasting accuracy.

Base rates: moderate signals 50% win [n=96], elevated signals 44% win [n=104], extreme outliers 64% win [n=14]Brier Score: 0.301 (Poor) [n=2360] | Drift: -0.014 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Breadth Rotation Meets Disinflation – Spiderweb of Interconnected Market Implications (30–90 Day Horizon, as of July 05, 2026)

Core Thesis

Dominant: A synchronized "soft-disinflation + breadth rotation" regime — small caps/financials/healthcare broaden as industrial commodities and USD soften, easing AI concentration risk (~40% joint, correlated). Alternative: Regime shift where commodity/FX extremes are structural, not statistical, and AI mega-cap leadership persists (~30%). Key discriminator: whether crude/copper slopes stabilize (disinflation genuine) versus resume falling on demand destruction (recessionary, not rotational).

Joint math (correlated via shared risk-sentiment/disinflation driver, NOT independent): P(IWM breadth) ~59% × P(commodity softness | breadth) ~62% × P(USD softens | both) ~65% ≈ 24% naive, bounded up to ~40% given all three share a single de-escalation/disinflation catalyst.

Markets Getting Stronger & Spiderweb Implications

• IWM / financials / healthcare (breadth): Base rate for extreme-concentration mean reversion ~60% [n=1615], adjusted to 59% for 4-day; decays to ~52–58% at 30-day (F4: -4pp horizon decay). Implications: Broadening lowers index fragility, redistributes flows from mega-cap AI, and validates the disinflation-supports-cyclicals loop.
• EURUSD / short USD (EURUSD −1.78σ vs ~1000d mean): Outside-view USD reversion base ~55%, adjusted to 56% [n=1615]; ~50–55% at 30-day. Implications: Softer USD eases EM/commodity-importer stress and reinforces the disinflation web via cheaper non-USD funding.

Markets Getting Weaker & Spiderweb Implications

• Industrial commodities (copper/crude/PDBC −1.84σ vs ~1000d mean, crude −26.7%): Base rate for downside continuation at extremes ~55%, raised to 58% [n=1615] on slope. Implications: Falling industrials feed disinflation → front-end easing → duration-sensitive sector support, but risk masking demand destruction (see negation).
• Concentrated mega-cap AI (NVDA/AMD): Short leg of breadth trade; ~59% [uncalibrated on single-name basis]. Implications: Compresses concentration risk premium; a crowded unwind could transmit non-linearly (below).

The Connecting Spiderweb (Key Interconnections)

Leverage point — Industrial commodity slope. This is the highest-cascade signal: it drives (1) disinflation persistence, which (2) enables front-end easing (51%), which (3) supports duration-sensitive breadth sectors (financials/healthcare, 48%), while (4) softer commodities align with a softer USD (56%). One signal cascades into four.

• Supporting: Breadth expansion (48%) and USD softness (56%) are downstream confirmations, not independent bets — correlation inflates apparent conviction.
• CT5 Non-Linear Risk: If the AI short leg coincides with a crowded mega-cap unwind, forced de-grossing could spike volatility disproportionately, dragging IWM down WITH mega-caps (correlation → 1 in stress), collapsing the rotation thesis into broad risk-off rather than clean rotation.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Commodity softness ⊕ USD softness ⊕ front-end easing = "goldilocks disinflation" regime supporting breadth without full risk-off.
• Negation (¬):

| Scenario | Trigger | Implication |

|---|---|---|

| Structural deficit | Commodities stay −1.8σ beyond 2× reversion window | Regime shift, NOT anomaly — invalidates commodity-short mean reversion (F2) |

| Demand destruction | Crude falls further AND IWM falls with it | Recession, not rotation — breadth thesis fails |

| USD safe-haven bid | EURUSD reverses on escalation | GMT4 de-escalation branch (70%) breaks; full risk-off |

• Equivalence (∼): Current breadth setup ∼ post-concentration-peak broadenings, but only if disinflation (not recession) drives it.

Ideas for Thinking About the Spiderweb (Mental Models from Guardrails)

Dominant lens — Forecasting.md: The three PRIMARY signals share one driver, so conjunction decay (CT7) is severe; treat joint confidence as ~40%, not the ~24% naive product nor the ~59% headline. Horizon decay (F4) further erodes all three from 4-day calibration to 30-day. Supporting — Scientific Method.md: The competing regime-shift hypothesis is falsifiable — if commodities hold extremes past the 2× reversion window, disinflation-rotation is rejected in favor of structural deficit. Supporting — Value.md: Breadth rotation only "pays" if broadening reflects earnings/valuation catch-up, not a liquidity-driven mean reversion that reverts again.

Practical Prompts

• Track IWM vs QQQ relative return over 20 trading-day window — if IWM fails to outperform by >1.5%, breadth-rotation thesis invalidated, revert to concentration regime.
• Watch front-month crude and copper slope over 15-day window — if both stabilize/rebound above −1σ (30-day mean), commodity-short leg falsified; disinflation loop weakens.
• Monitor EURUSD over 10 trading-day window — if it breaks below prior lows on USD safe-haven bid, de-escalation branch broken, exp

Devil's AdvocateIF this forecast proves wrong, the most likely failure mode would be a continuation of the very concentration and momentum trends the thesis bets against — a breadth-rotation call has historically been fragile, and the moderate-conviction signal tier backing several of these views carries only a 44% historical win rate across a large sample (n=104), meaning less than half of similar setups resolved favorably. A second likely failure path would be the mean-reversion bet on stretched commodity and currency readings (around -1.8σ): extreme statistical outliers can extend further before snapping back, so timing a reversal on a short 4-day horizon is a weak link, and the current Brier score of 0.301 signals only modest calibration edge. Finally, the clustered nature of these themes — commodity disinflation, breadth expansion, and currency reversal are all correlated bets on the same regime shift — means one wrong macro read could cause several positions to fail together rather than diversify the risk.

Base rates: moderate signals 52% win [n=92], elevated signals 44% win [n=104], extreme outliers 64% win [n=14]Brier Score: 0.301 (Poor) [n=2354] | Drift: -0.011 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Breadth Rotation Meets Commodity Disinflation – Spiderweb / Interconnected Market Implications (30–90 Days, as of July 04, 2026)

Core Thesis

Dominant: A benign rotation regime is underway — equity breadth (small caps, financials, healthcare) broadens away from mega-cap AI concentration while industrial-commodity disinflation persists without triggering risk-off (~40% joint, correlated). Alternative: The disinflation signal is a demand-weakness warning that eventually drags cyclicals and breadth down with it (~30% [uncalibrated]). Key discriminator: whether financials/small-caps hold their outperformance while copper stays soft — divergence confirms benign rotation; convergence-to-downside confirms the demand-scare alternative.

Joint confidence note: P(breadth outperforms) ~61% [n=1615, 4d] × P(commodity softness persists | breadth holds) ~58% — these share a "soft-landing / restrained-inflation" driver, so not independent. Correlated bound ≈ 40%, decaying from 4-day calibration to ~30–38% at 30-day horizon (−8pp F4 horizon decay, −3pp medium confidence cone).

Markets Getting Stronger & Spiderweb Implications

• Small caps / financials (IWM, RUT): +2.07σ vs broad (30-day lookback). Outside view: breadth long-signals hit 56% base [n=2712], adjusted to ~61% [n=1615, 4d], decaying to ~52–57% at 30-day.

Implications: Rotation drains relative bid from mega-cap AI, steepens the "everything-but-Nvidia" trade, and rewards duration-sensitive rate cyclicals if disinflation holds.

• Precious metals / gold (vs industrial commodities): relative strength as industrials sit at −1.84σ (30-day).

Implications: Gold firm + copper weak is the classic disinflation-without-recession tell, feeding the benign breadth case.

Markets Getting Weaker & Spiderweb Implications

• Industrial commodities (PDBC, copper): −1.84σ (30-day mean). Base rate favors mean reversion (77% anchor on extremes), but at WATCH level momentum-continuation dominates: ~56% [n=1615, 4d] down-continuation, ~50–54% at 30-day.

Implications: Softness relieves cost pressure (supports margins, breadth, duration) — but the same signal is ambiguous: cheap copper can mean tame inflation or fading demand.

• Mega-cap AI (NVDA, MSFT): +6pp deterioration in relative standing (55%→61% breadth-favored).

Implications: Concentration unwind is the pressure valve; a sharp AI drawdown could invert the "benign" framing into index-level risk-off.

The Connecting Spiderweb (Key Interconnections)

Leverage point — industrial-commodity disinflation. This single signal cascades furthest: it (a) validates the soft-landing narrative enabling breadth rotation, (b) supports duration without risk-off (SECONDARY, dir+), and (c) determines whether gold's bid is defensive or benign.

• Supporting: Breadth expansion depends on disinflation staying "good news" — cyclicals need cheap inputs, not vanishing demand.
• Supporting: AI-concentration unwind provides capital that flows into the broadening cohort — a fungibility handoff, not net selling.
• Non-linear risk (CT5): If copper's −1.84σ flips from "disinflation" to "demand collapse," the breadth long and AI short both break simultaneously — correlated positions unwind non-linearly as the shared soft-landing premise dissolves, converting a rotation into a de-grossing cascade far larger than any linear copper→equity beta implies.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Breadth-strength ⊕ commodity-softness ⊕ AI-unwind = benign rotation regime (soft landing).
• Negation (¬) Scenarios:

| Negation | Condition | Implication |

|---|---|---|

| ¬Rotation | AI reasserts leadership | Breadth longs bleed; concentration risk returns |

| ¬Mean-reversion (F2) | Copper stays ### Devil's AdvocateThe most likely reason for failure would be that the central bet on breadth expansion overtaking mega-cap leadership never materializes within the compressed 4-day windows — a mean-reversion thesis that historically has hovered near coin-flip reliability, with the mid-tier conviction signals in this range showing only a 44-52% win rate despite modest positive mean returns of roughly +1%. A second failure mode would stem from the several stacked positions all leaning the same structural direction (rotation away from concentration, continued commodity softness), meaning a single macro surprise — a risk-off shock or a commodity supply squeeze — could turn correlated bets into correlated losses rather than diversified ones. Finally, with an overall Brier score of 0.301 signaling only modestly-better-than-chance calibration, the clustering of stated probabilities in the 56-61% band leaves little margin if the true edge is thinner than assumed.

Base rates: moderate signals 52% win [n=92], elevated signals 44% win [n=104], extreme outliers 64% win [n=14]Brier Score: 0.301 (Poor) [n=2346] | Drift: -0.009 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

One-Page Brief: Dollar Softening Meets Selective Breadth Rotation – Spiderweb / Interconnected Market Implications (30–90 Days, as of July 03, 2026)

Core Thesis

Dominant: A synchronized mean-reversion regime — softer USD, stabilizing industrial commodities, and gradual AI-to-breadth rotation — reflecting a benign disinflationary reflation of laggards (~40% [uncalibrated, decayed from short-horizon inputs]). Alternative: Statistical extremes are early signals of a genuine regime shift (structural dollar strength + AI-concentration entrenchment), where reversion fails (~35% [uncalibrated]). Key discriminator: whether EURUSD and PDBC hold reversion gains beyond the 10-day mark, or fade back toward extremes — confirming dislocation vs. structural break.

Markets Getting Stronger & Spiderweb Implications

• EURUSD (long, -2.32σ vs 30-day mean): Outside view base rate 64% over 4 days [n=1247]; decayed to ~52–58% at 30-day horizon (F4): -8pp horizon decay, +2pp stable-geo tailwind. Decompose: P(USD softens) ~70% × P(EUR captures move | USD soft) ~80% ≈ 56%.

Implications: A softer dollar is the master valve — it eases global financial conditions, supports commodities priced in USD, and reduces the earnings drag that favors defensive mega-caps.

• Industrial commodities / PDBC (long, -1.8σ vs 30-day mean): Base 77% [n=1686] mean-reversion, adjusted to 60% inside-view for capped upside; ~50–55% at 30-day horizon (F4 decay). Direction bounded — a stabilization rebound, not a disinflation reversal.

Implications: Commodity firming underwrites the reflation-of-laggards story, feeding directly into financials/materials breadth.

• Small caps / financials / healthcare breadth: 45% [n=1615]; sectors showing +8.5%–12.3% over 30d. Outside view base for sustained breadth expansion ~50%, trimmed on -3pp daily deltas.

Implications: Beneficiary node, not driver — depends on dollar and commodity signals holding.

Markets Getting Weaker & Spiderweb Implications

• IWM/AMD/SOXX vs broad (short pair, +2.07σ vs 30-day mean): Base rate 77% [n=1686] pullback; adjusted 65% [n=1615] for 4-day hold, skew -2.5% vs +0.8%. Note: 4-day calibration — do NOT extend unchanged to 30-day horizon (F4).

Implications: Tactical froth correction within a still-expanding breadth regime — a local, not systemic, weakening.

• AI concentration (NVDA/MSFT): 55% relative weakness (down from 58%, -3pp); base case 60% over 30 days.

Implications: Capital-rotation source funding the breadth trade. The negative delta trio (-3pp across breadth, disinflation, AI) signals fading conviction across the whole reflation web.

The Connecting Spiderweb (Key Interconnections)

Leverage point — EURUSD / USD direction. The dollar cascades furthest: it drives commodity pricing (PDBC), risk-sentiment breadth (small caps/financials), and the relative appeal of AI mega-caps vs. cyclical laggards. All four PRIMARY signals share this single risk-sentiment/dollar driver — meaning they are correlated, not independent (CT7): do not multiply naively.

• Supporting: Commodity stabilization → financials/materials breadth → validates AI rotation.
• Non-linear risk (CT5): A sudden USD spike (surprise Fed hawkishness or safe-haven bid) could reverse all four signals simultaneously — the shared driver makes correlated unwind non-linear: a 1σ dollar move could trigger >2σ pain across the crowded rotation, as forced de-grossing cascades through leveraged small-cap and commodity longs.

Heuristic Algebra Applications (⊕, ¬, ∼)

• Combination (⊕): Soft USD ⊕ commodity firming ⊕ AI de-concentration = benign reflation-of-laggards regime.
• Negation (¬): (1) USD strengthens → entire web unwinds. (2) Regime-shift (F2): If PDBC/EURUSD extremes persist beyond 2× typical reversion window, this is structural supply deficit / dollar bull-market — invalidating all mean-reversion positioning, not a temporary anomaly. (3) AI concentration re-accelerates → breadth fails despite soft dollar.
• Equivalence (∼): IWM +2.07σ froth ∼ AI-concentration crowding — both are extension trades vulnerable to the same sentiment reversal.

Ideas for Thinking About the Spiderweb (Mental Models)

Dominant lens — Forecasting.md: Every signal here is 4-day-calibrated; applying it to 30–90 days demands explicit horizon decay (F4) and conjunction discipline (CT7). The shared dollar driver forbids independent multiplication — joint confidence stays ~35–45%, not the ~14% naive product. Supporting — Critical Thinking.md (CT2/F2): sigma extremes are only meaningful with lookback windows and a falsification test for regime shift — the discriminator is persistence past the reversion window. Supporting — Data Integrity.md: the uniform -3pp daily deltas warrant checking whether this is genuine conviction fade or a single correlated data-feed adjustment.

Practical Prompts

• Track EURUSD over 10 trading-day window — if

Devil's AdvocateThe most likely reason for failure would be over-reliance on statistical mean-reversion in extreme outliers: several of these calls assume that stretched readings (+2.07σ, -2.32σ, -1.8σ) will snap back, yet the historical base rates show that moderate-conviction signals have won only 46-52% of the time across roughly 200 trades, barely better than a coin flip. A second failure mode would be the interconnected 'breadth rotation' thesis — the assumption that leadership shifts away from concentrated mega-cap themes toward broader sectors — which is expressed at only 45-49% probability in its supporting components, meaning the theme is currently a below-even bet and could easily reverse if concentration persists. Finally, with an overall Brier score of 0.301, the model's probability estimates carry meaningful calibration error, so the stated 55-65% confidence levels may be somewhat optimistic relative to realized outcomes.

Base rates: moderate signals 52% win [n=92], elevated signals 46% win [n=102], extreme outliers 64% win [n=14]Brier Score: 0.301 (Poor) [n=2340] | Drift: -0.007 | Recalibration: TRIGGERED

Markets are a single, homeostatic, arbitrage-driven neural net: any local shock is transmitted globally because prices are information, capital is fungible, and every participant is watching every other participant.

Opportunity

PRIMARY

IWM/AMD/SOXX vs broad indices at 2σ+: Short-term fade of ALERT equity moves into defensives/healthcare

65%

PRIMARY

Industrial commodities (lithium/copper/PDBC): Avoid long energy/industrials; selective shorts capped via options

63%

PRIMARY

USD strength vs EUR (EURUSD -2.07σ): Long EURUSD or short USD vs basket on reversion; de-escalation branch overweight

61%

PRIMARY

AI concentration (NVDA/MSFT) vs breadth: Long Russell/financials vs NVDA; compounds via repeated 4d trades

58%

SECONDARY

Commodity disinflation persistence: Supports equity breadth, softer industrials/energy, duration assets

53%

SECONDARY

Equity breadth expansion (small caps/financials/healthcare): Mega AI underperforms relative; aligns with thesis base case

48%

Opportunity

PRIMARY

PDBC and industrial commodities: Short industrial exposure or long defensives; 72% per 4-day trade [n=1686], updated +7pp on new data confirming low transmission

72%

PRIMARY

IWM/russell2000 vs NVDA/AI leadership: Long IWM/AMD basket vs short NVDA; 67% per 4-day trade [n=128], updated +3pp on sigma signals

67%

PRIMARY

EURUSD after -2.03σ: Long EURUSD; 64% per 4-day trade [n=1686], updated +10pp from prior on base rate anchor and stable regime

64%

SECONDARY

Commodity disinflation persistence (PDBC -2.02σ, broad -9.9% 30d): supports equity breadth and defensives per base case, limited by GMT4 de-escalation paths

61%

SECONDARY

Equity breadth continuation (IWM +2.31σ, AMD/TSM +2.5-2.7σ vs NVDA lag): favors small caps/financials/healthcare over mega AI, updating +3pp on new sigma data

57%

SECONDARY

EURUSD mean reversion after -2.03σ: flatter USD path vs developed FX, updating +10pp from base rate 77% [n=1686] and stable geo

54%

Opportunity

PRIMARY

PDBC and industrial commodities: long selective commodities vs short energy

65%

PRIMARY

IWM/russell2000 vs NVDA/AI leadership: long IWM/AMD basket vs short NVDA

64%

PRIMARY

EURUSD after -2.37σ: long EURUSD

64%

SECONDARY

Commodity disinflation persistence: supports equity breadth and duration-sensitive defensives

49%

SECONDARY

Equity breadth continuation (IWM/AMD/TSM vs NVDA): broadens leadership per active thesis

52%

SECONDARY

EURUSD mean reversion after -2.37σ: flattens dollar path vs developed FX

54%

Opportunity

PRIMARY

PDBC and industrial commodities: Fade rebounds in energy/industrials; long defensives and breadth

62%

PRIMARY

IWM/russell2000 vs NVDA/AI leadership: Long IWM/financials/healthcare, relative short mega-cap AI

59%

PRIMARY

EURUSD: Long EURUSD on further dips below 1.14 with tight stop

70%

SECONDARY

Commodity disinflation persistence: Supports equity breadth, defensives, and flatter USD; updated +7pp on crude -21.3% 30d, PDBC -2.26σ and sentiment drop

53%

SECONDARY

Equity breadth continuation (IWM/AMD vs NVDA): Favors small caps, financials, healthcare; aligns with high dispersion and thesis base case

50%

SECONDARY

EURUSD mean reversion after -2.62σ: Flatter dollar vs developed FX; updated +5pp on sigma signal strength

60%

Opportunity

PRIMARY

PDBC and industrial commodities: Long selective commodities (copper/crude) with defined 4-day exit for convexity

70%

PRIMARY

IWM/russell2000 vs NVDA/AI leadership: Rotate long IWM/financials/short concentrated AI for 30d horizon via repeated 4d trades

59%

PRIMARY

EURUSD: Long EURUSD vs USD with tight 2d stop for lead-time edge before saturation

65%

SECONDARY

Mean reversion after 2σ+ ALERT signals: Drives near-term commodity and FX stabilization supporting disinflation breadth

61%

SECONDARY

Equity breadth continuation (IWM/AMD outperformance): Reinforces base case 60% favoring small-caps/financials/healthcare over mega AI

49%

SECONDARY

Commodity disinflation persistence (PDBC/crude softness): Caps energy upside and supports defensives without full risk-off

55%

Opportunity

PRIMARY

Financials (SOXX/IWM/AMD): short for mean-reversion edge over 4d hold

62%

PRIMARY

PDBC/crude: long bias, stabilization over 4 days

60%

PRIMARY

EURUSD: long for 4d relief bounce

67%

PRIMARY

Short-term rates (^IRX): position for lower yields

65%

SECONDARY

Mean reversion after 2σ+ ALERT extensions: pullback in financials, relief in EURUSD and commodities

61%

SECONDARY

Geopolitical de-escalation equilibria (ME/Ukraine): contained risk premia, lower commodity vol

44%

Opportunity

PRIMARY

SOXX/IWM/AMD: short extensions or buy vol on any continued rally

60%

PRIMARY

PDBC/crude: long dips in energy with tight stop below recent lows

62%

PRIMARY

EURUSD: long EURUSD on break above 1.145 targeting 1.16

62%

PRIMARY

^IRX short-term rates: position for lower short rates if no hawkish data

63%

SECONDARY

Mean reversion after 2σ+ ALERT extensions: pullback in rates, semis, small-caps and EURUSD after extensions

55%

SECONDARY

Geopolitical de-escalation equilibria: limits cascade to oil and risk premia, supports commodity rebound

45%

Opportunity

PRIMARY

SOXX/IWM/AMD: tactical short or hedge into 4-day window

62%

PRIMARY

EURUSD: long EURUSD with defined 4-day exit

64%

PRIMARY

PDBC/crude oil: long commodities basket on any further dip

67%

PRIMARY

^IRX/short-term rates: short short-term rates or T-bill proxies

70%

SECONDARY

Mean reversion after ALERT extensions: drives near-term stabilization in rates, FX, commodities after 2-3σ moves

61%

SECONDARY

Geopolitical de-escalation equilibria: weak transmission to commodities and contained safe-haven bids

47%

Opportunity

PRIMARY

SOXX / IWM / AMD: short extension bias over 4d hold; skewed to downside -2.5% vs +1% expected range

68%

PRIMARY

EURUSD: long EURUSD for 4d hold if yield trigger met

60%

PRIMARY

PDBC / crude oil: long oversold commodities with capped downside via options

59%

PRIMARY

short-term rates (^IRX / 2Y): position for lower yields if data soft

62%

SECONDARY

Mean reversion after ALERT (2-3σ) extensions: primary near-term anchor for rates, FX, commodities

58%

SECONDARY

Geopolitical de-escalation equilibria (Ukraine talks, contained Middle East proxies): limits oil and safe-haven cascade

47%

Opportunity

PRIMARY

SOXX: short bias on extended rally with defined risk

62%

PRIMARY

crude_oil: long from WATCH levels with cheap convexity

54%

PRIMARY

EURUSD: short EURUSD or long USD

56%

PRIMARY

IWM: short small caps on extension

60%

SECONDARY

Mean reversion after ALERT extensions in tech/small caps: pressures financials and correlated crypto 4d

53%

SECONDARY

Geopolitical de-escalation equilibria via diplomacy: caps energy upside and vol expansion

51%

Opportunity

PRIMARY

SOXX: short semis or buy vol for 4-day hold targeting 2-4% pullback

64%

PRIMARY

lithium: long lithium futures for mean-reversion edge over next 4 days

57%

PRIMARY

crude_oil: long crude with tight stop for 4-day convexity

59%

SECONDARY

Mean reversion from ALERT sigma in tech and small caps: modest 1-3% equity pullback over 4 days

55%

SECONDARY

Geopolitical de-escalation equilibria in Ukraine and Middle East: caps commodity volatility and supports risk assets with low transmission

58%

SECONDARY

Commodity rebound from WATCH lows on muted supply fears: stabilization in oil and lithium reducing input cost pressures

49%

Opportunity

PRIMARY

SOXX/IWM vs mega-cap AI: long semis/small-caps for rotation edge, exit on any VIX spike >20

57%

PRIMARY

PDBC/crude after -1.96σ/-28.9% 30d move: tactical long commodities for reversion, invalidated if crude <74

63%

PRIMARY

BTC relative to equity rotation: long BTC on dips with tight stop below 61000

55%

SECONDARY

Momentum continuation in semis/small-caps at ALERT: broadens equity gains beyond mega-caps per thesis

48%

SECONDARY

Mean reversion in broad commodities after -1.96σ: limits vol/oil cascade supporting disinflation

54%

SECONDARY

Contained geopolitical risk with stable regime: reduces transmission to energy and risk premia

55%

Opportunity

PRIMARY

PDBC/commodities: Long commodities on dips for mean-reversion edge, exit on saturation per GMT6

68%

PRIMARY

SOXX/AMD/IWM: Long semis/small-caps on any 2d pullback for 4-day hold per calibration

55%

PRIMARY

BTC: Long BTC on equity confirmation, cheap convexity via options

54%

SECONDARY

Mean reversion in broad commodities after -2.13σ move: Supports disinflation base case with soft prices not extending

58%

SECONDARY

Momentum continuation in semis/small-caps at ALERT levels: Drives rotation from mega-caps toward cyclicals per thesis

47%

SECONDARY

Contained geopolitical risk limiting oil/vol cascade: Low transmission per GMT1-5 with de-escalation branch ~70%

61%

Opportunity

PRIMARY

PDBC: long broad commodities for 4-day hold, asymmetric +3.2% vs -1.1% expected range

70%

PRIMARY

SOXX/AMD: tactical long semis with defined stop, 4-day edge compounds via repeated trials

57%

PRIMARY

IWM: long small caps as expression of rotation, convex to thesis upside case

60%

PRIMARY

BTC: avoid uncompensated short convexity, watch for de-escalation branch per GMT4

52%

SECONDARY

Mean reversion in broad commodities after -2.13σ ALERT move: reinforces disinflation base case and supports cyclical rotation

60%

SECONDARY

Momentum continuation in semis and small caps at ALERT levels: broadens equity leadership per active macro thesis

48%