workflow-derivatives-structured-products

# Structured Products Pricing and Decomposition ## What this skill covers A pipeline for pricing and decomposing structured investment products into their constituent components. Every structured note is treated as a combination of a zero-coupon bond (or coupon-bearing bond) and one or more embedded derivative positions. Pricing uses the appropriate tool for each component type: `structured_note_pricing` for standard catalog products, `exotic_product_pricing` for non-standard payoffs, `monte_carlo_simulation` for path-dependent features, and `convertible_bond_pricing` for equity-linked convertible structures. The output establishes a fair-value breakdown that supports both client disclosure and internal risk reporting. ## Workflow ### Phase 1 — Product Classification 1. Classify the structured note by payoff type: - **Principal-protected note (PPN)**: zero-coupon bond + long call. - **Reverse convertible**: coupon bond + short put (investor is short the downside). - **Autocall / Autocallable**: callable structure with conditional coupons and barrier observation. - **Range accrual**: coupon accrues only on days the underlying closes within a defined range. - **Convertible note**: bond with equity conversion option (long call on issuer equity). - **Custom / exotic**: any payoff not covered above; treated via Monte Carlo. 2. Document: underlying(s), notional, tenor, coupon / participation rate, barrier levels, call schedule, and protection level. ### Phase 2 — Funding Leg (Bond Floor) 3. Source the issuer's credit spread: - Call `credit_spreads` or `fmp_quote` for the issuer's CDS or bond yield spread over the risk-free curve. 4. Compute the bond floor (present value of the capital repayment and any fixed coupons): - Use `bond_pricer` with the issuer's all-in discount rate (risk-free rate + credit spread) and the scheduled cash flows. - For principal-protected notes: bond floor = PV of par repayment at maturity. The residual budget = issue price minus bond floor = maximum option budget. ### Phase 3 — Embedded Derivative Valuation 5. **Vanilla components** (calls, puts, spreads): call `option_pricer` per leg using Black-Scholes or binomial as appropriate. Source volatility from `implied_volatility` or the vol surface from `workflow-derivatives-vol-surface`. 6. **Autocall and barrier features**: call `structured_note_pricing` with the product type set to `"autocall"` or the relevant catalog identifier. Required inputs include barrier level, observation frequency, conditional coupon, and early redemption premium. 7. **Exotic / path-dependent payoffs**: call `exotic_product_pricing` for Asian options, barrier options, or lookback features embedded in the note. 8. **Monte Carlo path-dependent pricing**: call `monte_carlo_simulation` with the payoff function specified as the simulation rule. Minimum 10,000 paths; document path count, seed, and distributional assumptions (log-normal for equity; Hull-White for rate underlyings). 9. **Convertible equity-linked structures**: call `convertible_bond_pricing` using the CRR binomial tree with the specified conversion ratio, call/put provisions, and credit spread. Extract: bond floor, conversion premium, CB delta, and investment vs parity value. ### Phase 4 — Decomposition and Fairness Check 10. Assemble the component price table: - Issue price = Bond floor + Embedded option value + Distributor margin. - Compute implied distributor margin: Issue price minus (bond floor plus fair option value). - Benchmark: distributor margin > 3% of notional for a 1-3 year structured note is a red flag for retail fairness. 11. Call `sensitivity_matrix` with the underlying spot ±10% / ±20% and implied vol ±5 vol points to produce the structured note's price sensitivity grid. ### Phase 5 — Investor Net Economics 12. Compute the investor's breakeven: - For capital-at-risk products: the underlying must close above the barrier at maturity for the investor to recover principal; compute the barrier return required. - Effective yield: if the investor simply reinvested in the risk-free rate over the same tenor, what terminal value would they receive? Compare to the expected payout from `monte_carlo_simulation` (mean path outcome). 13. Document the investor's economic position versus a simple bond + direct equity investment alternative. ## Output Format ### Product Classification and Terms | Field | Value | |---|---| | Product type | — | | Underlying | — | | Notional | — | | Tenor | — | | Issue price | — | | Coupon / participation rate | — | | Barrier level | — | | Capital protection level | — | ### Component Decomposition | Component | Type | Tool Used | Fair Value | % of Issue Price | |---|---|---|---|---| | Bond floor | Zero-coupon bond | `bond_pricer` | — | — | | Embedded option(s) | Call / put / barrier | `option_pricer` / `exotic_product_pricing` | — | — | | Monte Carlo correction | Path-dependent feature | `monte_carlo_simulation` | — | — | | Distributor margin | Residual | Implied | — | — | | **Issue price (total)** | | | — | 100% | ### Monte Carlo Output (if applicable) | Metric | Value | |---|---| | Path count | — | | Seed | — | | Mean payout | — | | 5th percentile payout | — | | 95th percentile payout | — | | Probability of capital loss | — | ### Convertible Bond Metrics (if applicable) | Metric | Value | Source | |---|---|---| | Bond floor | — | `convertible_bond_pricing` | | Conversion premium | — | `convertible_bond_pricing` | | CB delta | — | `convertible_bond_pricing` | | CB category | Investment / Balanced / Busted | `convertible_bond_pricing` | ### Sensitivity Matrix: Note Price vs Spot and Vol (5 × 5 grid — output from `sensitivity_matrix`) ### Investor Economics Comparison | Scenario | Structured Note Payout | Simple Bond + Equity Payout | |---|---|---| | Bear (underlying −30%) | — | — | | Base (underlying flat) | — | — | | Bull (underlying +30%) | — | — | ### Tool-Call Traceability | # | Tool | Key Inputs | Output | |---|---|---|---| ## Quality Gates - [ ] Product correctly classified before any pricing tool is called. - [ ] Bond floor computed with issuer credit spread included; not priced off the risk-free curve alone. - [ ] Embedded option fair value computed with a tool call; not estimated from the residual. - [ ] Distributor margin computed; flagged if > 3% of notional for retail products. - [ ] Monte Carlo path count ≥ 10,000; seed and distributional assumptions documented. - [ ] Convertible bond: bond floor, conversion premium, and delta all extracted from `convertible_bond_pricing`. - [ ] Sensitivity matrix produced: underlying spot ± 20% × implied vol ± 5 pts. - [ ] Investor breakeven and capital-loss probability documented for capital-at-risk products. - [ ] Every number in output maps to a row in the traceability table. ## Related Skills - `workflow-derivatives-option-pricing` — vanilla component pricing for each embedded option leg. - `workflow-derivatives-vol-surface` — vol surface inputs for exotic and barrier option components. - `workflow-derivatives-swaps` — interest rate swap components embedded in structured rate notes. - `corp-finance-analyst-derivatives` — agent body with CB benchmarks (balanced: 20-40% premium, delta 0.4-0.6; busted: >60% premium, delta <0.3) and Monte Carlo conventions.