--- canonical: true last_updated: 2026-08-14 session_id: 2026-08-14-block-027-close-block-028-prereads prior_session: 2026-08-12-block-027-partial current_block: 028 current_block_slug: fx-smile-positioning-signal block_status: ◐ partial (prereads) phase: 1 pre_block_probes_fired: 0 pre_block_probes_clean: 0 substrate_status: teach-delivered targets_pending: - target_1_smile_shape_read_for_carry - target_2_term_structure_forward_smile - target_3_block_027_four_cell_matrix_integration - target_4_pm_implications_leading_vs_lagging anki_queue: card_count: 0 cards: [] pre_reads: - source: castagna_fx_options_and_smile_risk title: "FX Options and Smile Risk (Ch. 3 + Ch. 4)" type: textbook time_min: 30 status: teach-delivered-2026-08-14 teach_artifact: blocks/_pre-read-teaches/028-fx-smile-positioning-signal-pre-read-teach-2026-08-14.md chapters: - "Ch. 3 (Dynamic Hedging and Volatility Trading) — smile phenomenology + smile exposures (3.7-3.8)" - "Ch. 4 (Volatility Surface) — building the surface + term structure (4.4, 4.7)" skip: - "Vanna-Volga hedging mechanics (Ch. 3.6)" - "SABR / Heston / LMUV calibration (Ch. 2.3-2.5, 4.3)" - "Exotic options pricing (Ch. 6-7)" - "Stochastic-vol arbitrage arguments (Ch. 4.1.1)" pre_read_substitution_note: "Pre-emptive teach-first (per macro-mentor teach-pre-reads.md). User opted to have substrate taught by mentor instead of reading. Verification gate at next open: 'did you do the read OR get the teach?'. 20-min time budget for teach; user opted for full content coverage over compression (2026-08-14)." anki_cards_pushed: - date: 2026-08-11 count: 15 deck: "Macro Study::Phase 1 — Components::FX" first_note_id: 1786448537065 last_note_id: 1786448537420 - date: 2026-08-12 count: 2 deck: "Macro Study::Phase 1 — Components::FX" note_ids: [1786534453742, 1786534817367] tag: user-flagged - date: 2026-08-14 count: 7 deck: "Macro Study::Phase 1 — Components::FX" note_ids: [1786707294917, 1786707400439, 1786707400474, 1786707400496, 1786707400520, 1786707400555, 1786707400571] breakdown: - "1786707294917: Block 027 three orthogonal filters (user-flagged, session:2026-08-14)" - "1786707400439-571: Block 028 Castagna Ch. 3-4 atomic Basic cards (source-castagna-ch3-4 tag)" carry_forwards: - block_027_three_orthogonal_filters_carry_risk_positioning_basis - block_027_funding_currency_selection_risk_off_correlation_filter - block_028_smile_three_instruments_atm_rr_butterfly - block_028_term_structure_contango_backwardation_signal - block_028_smile_as_forward_indicator_leading_cot_lagging - block_028_four_cell_matrix_integration_smile_replaces_or_complements_cot patterns_to_load: - pattern_11_mechanism_cold_recall - pattern_18_mechanism_not_calculation_probes - pattern_25_compact_output_discipline - pattern_23_frame_without_answer - pattern_26_multi_mechanism_conflation --- # Session 44 — Block 027 close + Block 028 pre-reads (2026-08-14) **Status:** Block 027 closed Working+. Block 028 pre-reads teach-delivered. ## What we did **Block 027 Targets 3-4 close (working through pending list from 2026-08-12):** *Target 3 — Carry unwind asymmetry:* - Probe 1: time-to-resolution read + cross-currency basis discriminator. Clean Working+ (positioning unwind = days-weeks self-resolving; funding unwind = weeks-months requires Fed). - Probe 2: PM trade construction by unwind type. Off on instrument/horizon — re-anchored: Unwind A = long the carry (FX spot/forward, days-weeks, finite forced sellers); Unwind B = long the basis (cross-currency basis swap, weeks-months, Fed is structural backstop). Different trades on different instruments because the mechanism that resolves them is different. - Probe 3: mechanism — finite vs infinite forced sellers, Fed-as-participant with institutional toolbox (Pattern #10). *Target 4 — PM implications:* - Probe 1: funding-currency selection — PM-A right, PM-B wrong on the correlation direction. JPY case = correlated failure mode (both legs move against you in risk-off). User got PM-A right but the slip was on *which correlation direction is bad* — re-anchored. - Probe 2: Scenario 1 (global risk-off) vs Scenario 2 (BRL-specific) P&L. Off on Scenario 2 — re-anchored: Scenario 2 = single-leg loss only, JPY leg unaffected = hold/add not cut. The slip was multi-mechanism conflation (Pattern #26): "carry trade losing money" collapses three distinct failure modes (rate-differential reversal / single-name spot move / correlated both-legs-same-direction failure). - Probe 3: incoming data ranking. VIX up + equities down + EM down = most threatening (global risk-off channel = correlated failure mode). IMM COT 95th percentile = contingent amplifier (positioning crowdedness). Brazil CB hawkish hike = least threatening (BRL-positive). - Probe 4: contingent amplifier question. VIX +5pt overnight + already crowded COT. User correctly identified: cross-currency basis as discriminator (Block 023/026 substrate) + spot move to confirm direction. Working+ on substrate. - Probe 5: four-cell matrix integration. Pre-cut state = Cell 2 (carry-to-risk 0.7 + positioning 90th percentile crowded). Post-cut = tightened Cell 2 (carry-to-risk compressed + no funding-stress signal = single-name deterioration not full unwind). Default PM action: trim further. The slip: user framed as Cell-1 ("still high carry") but positioning was already crowded pre-event, which is the Cell-2 defining condition. **Synthesis — three orthogonal filters:** 1. Carry-to-risk (rate-diff vs realized-vol → base size) 2. Positioning (IMM COT crowdedness → timing risk + stop tightness) 3. Cross-currency basis (stable = positioning unwind, blow-out = funding unwind requiring Fed) Funding-currency selection (positive risk-off correlation = correlated failure mode) is the upstream gate. ## Block 028 pre-reads teach **Castagna "FX Options and Smile Risk" Ch. 3 + Ch. 4, ~15 min teach:** Substrate: - Three smile instruments (ATM / RR / butterfly) and what each tells you (level / direction / tail) - Why the smile exists (RND ≠ lognormal + dealer inventory + carry bias) - Term structure (contango = carry-friendly, backwardation = pre-positioning signal) - Vanna-Volga = dealer convention for hedging smile as a surface **Synthesis — FX smile as forward indicator for Block 027 four-cell matrix:** - COT = lagging (weekly print), single-instrument - Smile = leading (minutes-to-hours repricing), cross-instrument - Both aligned (Cell 2 + smile backwardation) = max-confidence fade setup **Probe queue (4 items, deferred to next session open):** 1. Smile-shape mechanism (RR term-structure timing) 2. Discriminator (contango flat RR — signal or not) 3. Block 027 four-cell matrix + smile integration PM action + exit triggers 4. Live-data probe (USD/JPY 1m/1y 25-delta RR + ATM term structure) ## Anki cards pushed (Session 44, 7 total) - **1786707294917** — Block 027 three orthogonal filters (user-flagged) - **1786707400439-571** — Block 028 Castagna Ch. 3-4 atomic Basic (6 cards) ## State for next session - **Block 027 closed Working+** (Targets 1-4 across Sessions 43-44) - **Block 028 pre-reads taught**, targets pending. Probe queue at next session open. - **Verification gate at next open:** "did you do the read OR get the teach?" per Pattern #13a - **Resume file:** this document. ## Directive for next session open 1. Probe queue fires 4/4 at next open. Walk Targets 1-4 in order. 2. Pattern #11 + #18 + #25 + #23 + #26 active. 3. Compact output, no streaming deliberation. 4. Frame each probe without pre-loading the answer. 5. At Block 028 close: assign Block 029 pre-read with full chapter/section specificity. ## User feedback applicable next session - 20-min teach budget honored; full content coverage maintained (no scope cut per user direction 2026-08-14) - Multi-mechanism conflation (Pattern #26) fired in Target 4 — labels collapse distinct channels - Compact output + no streaming deliberation (Pattern #25 + user preference)