SP-Benchmark: the 56 questions

Yield Enhancement (SSPA code 1230). The memory coupon feature does not change the SSPA category — it modifies the coupon payment condition but the core payoff (short put embedded in a note) remains a yield enhancement structure.

Scoring: Deduct directionality points if model conflates memory coupon with capital protection.

1. Zero coupon bond (ZCB) — provides the capital repayment at maturity if not called 2. Down-and-in put on worst-of — the risk leg; client is short this 3. Series of digital calls (autocall triggers) — each observation date is a conditional early redemption 4. Coupon stream — funded by the premium received from selling the put and digital structure Worst-of feature increases premium vs. single-underlying due to correlation exposure (short correlation for issuer).

Scoring: Award max causal reasoning only if model explains *where the yield comes from* (put premium), not just lists the components.

Capital protection requires buying a ZCB to floor the principal, consuming most of the invested notional. Yield enhancement requires selling optionality (typically a put), which conflicts because protection removes the ability to monetize downside. The resolution approaches: (1) Partial capital protection (e.g. 90%) freeing budget for option premium, (2) longer tenor to grow ZCB value from lower present value, (3) participation structures with capped upside rather than coupon. There is no free lunch — the conflict is a direct consequence of put-call parity and interest rate levels.

Scoring: FM3 trap — model may incorrectly claim these are compatible without trade-off. Deduct directionality score if it misses the ZCB cost constraint.

BRC = long ZCB + short down-and-in put (barrier put). The structuring margin arises from: (1) pricing the barrier put using a vol surface that already includes skew (the bank prices at mid or offer, client pays offer), (2) funding spread — the ZCB yields the issuer's credit spread above risk-free, (3) bid-offer on the underlying hedge. The client receives a coupon that is less than the full put premium; the remainder is the bank's margin.

Scoring: Full marks only if model identifies the funding spread / credit spread component, not just the option bid-offer.

Correlation at the 90th percentile means the worst-of premium is compressed — the three underlyings are behaving like one, so the additional premium from worst-of (which comes from the investor being short correlation) is near its minimum. Proceeding is risky because: (1) if correlation mean-reverts lower during the product life, the issuer's short correlation position loses money, (2) the 10% coupon target may require pushing the barrier dangerously close to spot to compensate for the low correlation premium, compressing the margin below 1.5%. Alternative approaches: (1) Replace worst-of with a basket (average performance) — removes the correlation exposure entirely, but the coupon drops significantly; could be combined with a lower barrier to partially compensate, (2) Use a "best-of barrier, worst-of payoff" hybrid — barrier breach is only triggered if the best-performing underlying breaches, reducing barrier probability while maintaining worst-of coupon dynamics, (3) Add a lookback feature on the strike (set strike at the highest closing over the first month) — increases the effective moneyness of the short put, adding premium without changing the correlation exposure, (4) Extend the tenor to 18M — more time value in the options increases the available premium, (5) Accept a conditional coupon (memory) with a lower guaranteed minimum — reduces the issuer's fixed obligation. The desk should price all variants and present the correlation sensitivity (P&L impact of a 10-point correlation drop) alongside the coupon comparison.

Scoring: L4 — requires cross-book reasoning (structuring margin, correlation hedging, client suitability). Expect most models to recommend proceeding or to miss the correlation mean-reversion risk entirely.

Gamma near the barrier. As spot approaches the barrier, gamma spikes — the delta changes rapidly and becomes expensive to hedge dynamically. In a high-vol environment the probability of touching the barrier increases, amplifying this effect. Vega also increases overall option value but gamma near barrier is the dominant cost driver for the issuer's hedging P&L.

Replication: long ZCB (notional, 12M) + short 70% down-and-in put (European or daily barrier per termsheet). The 6% coupon is funded from: (1) put premium received (~main source), (2) issuer's funding benefit (ZCB issued at issuer's credit spread above LIBOR/SARON). If put premium alone exceeds 6%, the remainder is structuring margin. If rates are low, the ZCB funding contribution is minimal and the coupon is almost entirely option-derived.

A steeper skew increases the implied vol at the barrier strike (OTM put). The down-and-in put, which is effectively a low-strike put conditional on barrier breach, becomes more expensive. The issuer is short this put — their hedge cost increases. This either: (a) reduces the coupon offered to the client at the same barrier level, or (b) forces a higher barrier to maintain the same coupon. Net effect: the client receives a worse product (lower yield or higher risk) when skew steepens.

Scoring: FM3 trap — model may say "client benefits from steeper skew" by confusing issuer/client perspective.

For a down-and-in put: when spot is far above the barrier, higher vol increases the probability of breaching the barrier (increases option value → positive vega). When spot is near the barrier, higher vol increases the probability of both breaching *and* then recovering above strike (reducing expected payout → vega can flip negative). The sign change occurs in the region where the barrier is close to spot and the dominant vol effect shifts from "more likely to knock in" to "more likely to knock in and finish worthless." This is a known convexity trap in barrier hedging.

Scoring: Most models will fail this. FM3 is the primary target — directional confusion on vega sign.

The SRI is a 1–7 scale combining Market Risk Measure (MRM, derived from VaR of simulated returns) and Credit Risk Measure (CRM, based on issuer credit quality). For yield enhancement products, MRM is typically driven by the underlying equity vol and the non-linear payoff (barrier puts amplify tail losses in simulations), usually landing at MRM 3–5. CRM reflects the issuer's probability of default.

Daily closing observation is significantly cheaper than continuous (American) barrier observation because intraday breaches are ignored. Under continuous observation, the probability of touching the barrier is materially higher, especially for volatile underlyings. The adjustment is typically applied via a barrier shift — the effective continuous-equivalent barrier is set lower (e.g. Broadie-Glasserman-Kou approximation). For a 70% daily barrier, the continuous-equivalent might be ~68–69% depending on vol and tenor. Issuers pricing daily barriers offer higher coupons than continuous equivalents at the same stated barrier level.

Swiss market default: dividends are typically passed through to the forward price (i.e. the product is priced on the dividend-adjusted forward). If the clause is missing and dividends exceed expectations, the forward drops — meaning the effective strike and barrier in forward-space are now higher relative to spot than originally priced. The client is negatively impacted: higher effective barrier probability and lower participation. The issuer, who hedges using the forward, is insulated. This is a documentation risk the client bears.

Scoring: FM3 trap — directional impact on client vs. issuer is the discriminating point.

Standard mechanism: the calculation agent (typically the issuer) adjusts strike, barrier, and ratio to preserve economic equivalence following the corporate action. For a 2:1 stock split: the strike is halved, the barrier is halved, and the ratio is doubled — so the dollar value of each threshold is unchanged. The adjustment follows the rules in the issuer's general terms (typically aligned with SIX corporate action procedures or ISDA equity definitions). Legal risk / ambiguity arises in: (1) extraordinary dividends — what constitutes "extraordinary" vs. ordinary is not always clear, and the adjustment methodology differs by issuer, (2) spin-offs and mergers — the replacement underlying or basket recomposition is at the calculation agent's discretion, creating potential conflict of interest (the issuer is both calculation agent and counterparty), (3) delisting — the termsheet may allow the issuer to substitute a comparable underlying or terminate the product early at "fair market value" (determined by the issuer), (4) timing mismatch — the adjustment may not coincide exactly with the ex-date, creating a window where the barrier level is economically misaligned. The core risk: the issuer as calculation agent has broad discretion, and the investor's recourse is limited to the general terms, which are drafted by the issuer.

Scoring: Award cross-domain bonus if model connects to D5 (lifecycle events) and identifies the calculation agent conflict of interest.

In bear markets, loss aversion dominates — clients pay a premium for downside protection (ZCB floor). In range-bound markets, clients perceive low probability of large moves, making the sold optionality in yield enhancement products feel "safe" while the coupon is attractive vs. cash. This is a behavioral finance pattern (prospect theory + recency bias) that the distribution desk exploits in product design timing.

Loss conditions: (1) Barrier breach AND underlying finishes below strike at maturity — client receives physical delivery of depreciated shares instead of notional. (2) Issuer default — client is an unsecured creditor. Communication standard: the yield must be contextualized against both risks. SSPA/FINMA practice requires the worst-case scenario (full loss of notional in extreme barrier breach + issuer default) to be explicitly disclosed. A RM framing the coupon without barrier breach probability is a mis-selling risk.

Contradiction: capital protection requires the ZCB floor (expensive, consumes notional), while the autocall coupon is funded by selling optionality. If you protect capital, there is no budget to sell meaningful optionality, so the autocall trigger either offers a trivial coupon or requires extreme tenor extension. Compromises: (1) soft protection (90–95% floor), freeing some budget; (2) very long tenor (5Y+) where ZCB is cheap; (3) lower autocall coupon, sometimes near cash rates — making the autocall feature nearly pointless; (4) conditional protection only (protection conditional on no barrier breach — which is not true capital protection). Full capital protection + meaningful autocall coupon is not achievable at normal rate levels.

Client-level risk: concentration of BRC sales means the bank's client base collectively holds a large short put position on the same underlyings. If the SMI corrects sharply and barriers breach, losses are correlated across the entire client base simultaneously — a mass suitability event. Clients who were each individually suitable for a single BRC may collectively represent a systemic mis-selling risk if the advisory desk was incentivized by distribution fees without portfolio-level monitoring. Hedging-level risk: the bank's derivatives desk is now short a large aggregate vega and gamma position on the same 5 underlyings. The concentration means: (1) the desk's own hedging activity (selling delta in a decline) amplifies downside pressure on the underlyings, (2) if other issuers have similar positions (likely — same popular underlyings), the hedging flow is correlated across the market, creating a feedback loop, (3) the desk may not be able to exit the vega position at mid-market if it needs to de-risk — the vol surface around those strikes/tenors will be distorted by the aggregate issuance overhang. Controls that should exist: (1) aggregate notional limits per underlying across all BRC issuance, (2) client concentration monitoring — what fraction of a client's portfolio is in BRCs on correlated underlyings, (3) hedging desk position limits with escalation triggers, (4) suitability re-assessment if aggregate product exposure exceeds a threshold of client AuM, (5) stress testing of the combined client P&L under a 2008-style scenario.

Scoring: Award cross-domain bonus if model connects to D6 (risk transfer / vega accumulation) and D1 (structuring incentives).

European barrier: observed only at maturity (or specific dates). American/continuous barrier: observed continuously throughout the product's life. European barriers are cheaper (lower knock-in probability), produce higher coupons at the same barrier level, but expose the client to a final-date-only risk. Continuous barriers are more expensive to the client but provide cleaner hedging for the issuer (no gap risk on a single observation date).

ISDA/termsheet fallback: typically the next good business day (Following or Modified Following convention). The hedging desk has positioned delta assuming the autocall triggers (or not) on the original date. A one-day delay means the autocall digital position is held one extra day — if the underlying is near the trigger level, this introduces overnight delta and gamma risk. If the product autocalls the next day, the hedge unwind is delayed; the desk is exposed to gap moves overnight. For clustered autocall dates across many products, this is a material risk.

Delta: has increased in magnitude (more negative for the short put position) — the issuer is more short delta as the put moves toward ATM. The desk is selling the underlying to maintain delta neutrality. Gamma: has increased significantly as the put approaches ATM — hedging costs are rising (frequent rebalancing). Vega: has increased — the put is now near-ATM and more sensitive to vol changes; the desk's short vega position is larger. Hedging actions: continuous delta rebalancing (selling into declining market), potentially buying back some barrier puts if P&L permits, monitoring barrier distance (15% from strike, barrier at 30% below strike → ~15% cushion remaining at current spot).

Scoring: Award cross-domain bonus if model connects to D6 (vega offload via new issuance or variance swap).

Client: nothing — the barrier was not breached per contract terms. No knock-in occurs. Issuer hedging desk: this is a significant event. The desk was hedging assuming the down-and-in put could knock in (hedging a conditional delta). When the barrier was touched intraday, the desk may have partially delta-hedged assuming knock-in (bought underlying on the way down, sold on recovery). This whipsaw trade is a realized hedging loss — pure gamma bleed. The client is unaffected contractually but the issuer absorbed real P&L damage. This is a known risk of daily-close barrier products.

Short vega. The issuer sold a structured note which embeds a short put to the client. The issuer is the counterparty — they are long the put (they sold it to fund the coupon, meaning they receive premium and deliver optionality). Wait — correction for precision: the issuer sells the BRC to the client. The client is long the BRC. The BRC contains an embedded short put (client is short put). The issuer, as the product manufacturer, is long the put from the client's perspective — BUT the issuer immediately hedges by selling the put in the market. Net hedged position: the issuer is short vega via the hedge (sold put to hedge = sold vol). The issuer profits if realized vol is lower than implied vol at pricing.

Scoring: This question has a subtlety — models often confuse the embedded position with the hedge. Award full causal reasoning only if the model correctly traces through both legs.

Internal mechanisms: (1) Cross-book vega transfer to flow desk if it is long vega from client vanilla options (typical in large banks), (2) vega netting against capital protection products (long vega) within the structured book. External mechanisms: (1) Sell variance swaps to hedge funds / asset managers (common), (2) Buy OTC vanilla options from market makers / prime brokers, (3) Issue capital protection products (long vega from the bank's perspective, offsets short vega from BRCs), (4) Listed options on liquid underlyings as a rough hedge. Typical external counterparties: hedge funds (variance swap buyers), market-making desks at other banks, asset managers with long vol mandates.

Internal vega transfer via an internal swap or book transfer: the structured desk buys vega internally from the flow desk at an agreed internal vol (usually mid-market). The flow desk books a P&L credit (sold vol at mid), the structured desk books a P&L debit (bought vol at mid). Net bank position: flat vega. P&L attribution: each desk is credited for managing its own book to flat. The transfer price (internal vol) is the key governance question — if set too low, the structured desk subsidizes the flow desk and vice versa. In practice, internal transfer pricing of vol is a significant source of inter-desk conflict and requires independent model validation sign-off.

Autocall wall: when many autocall observation dates coincide, the issuer has a large aggregate position in digital calls at the same strike/date. Near the observation date, as spot approaches the trigger, the aggregate delta of all these digitals creates a massive one-directional flow need. If spot is just below the trigger, the issuer is long delta (wants spot to go up to trigger autocall and unwind hedge). If spot crosses the trigger, delta flips sharply — the issuer must sell large delta quickly. This creates a predictable gamma squeeze pattern in the underlying around autocall dates. Vol surface impact: implied vol around the observation date tenor and trigger strike tends to be elevated/distorted by hedging demand, creating observable kinks in the term structure. Hedging problem: the concentration of digital exposure makes continuous rebalancing expensive and market-impact of hedging trades is non-trivial — the issuer's own hedging activity can move the market.

Structured products (SPPI-eligible notes) can often be classified differently under banking book treatment vs. variance swaps which are derivatives and sit in the trading book with full CVA/DVA and market risk capital charges. Structured products may benefit from: (1) lower regulatory capital consumption under certain internal model approaches, (2) the funding benefit of the note issuance (the client pays upfront, improving the bank's liquidity position), (3) no bilateral margining requirement (structured notes are not collateralized derivatives — no ISDA CSA applies for the client leg), (4) potential favorable treatment under SA-CCR vs. delta-equivalent variance swap notional. Net: the bank achieves similar economic short-vol exposure with better capital efficiency and a funding benefit.

Scoring: L4/ADV — expect most models to fail the regulatory capital angle. Partial credit for identifying the funding benefit alone.

Implied correlation for worst-of structures is typically higher than realized correlation — the market prices a correlation risk premium because investors are short correlation (they lose when correlation drops, as the worst-of payoff diverges from the basket average). Using flat realized correlation (lower than implied) means: (1) the desk underestimates the put premium from the worst-of feature → the desk is selling the embedded put too cheaply → the coupon offered to the client is too high for the risk taken, (2) the structuring margin appears wider than it truly is → when the desk hedges using market-implied correlation, the hedge cost exceeds what was priced in, (3) the book accumulates a hidden short-correlation position that is not properly reflected in the P&L or risk metrics. Direction of mis-pricing: the issuer is systematically underpricing the product (giving away value to the client). This is the opposite of what most people assume — the risk is not that the bank overcharges but that it undercharges and accumulates unhedged risk. The risk management fix requires: (a) moving to implied correlation calibrated from listed options or inter-dealer quotes, (b) reserving against the existing book's mark-to-market gap, (c) model validation sign-off on the correlation model — this is exactly the kind of model risk that SR 11-7 and FINMA 2017/1 are designed to catch.

Scoring: This is a model risk question disguised as a pricing question — award cross-domain bonus if model connects to D9 (regulatory / model validation requirements).

(1) Cash settlement: the investor receives a cash amount. For a BRC, cash settlement applies if the barrier was never breached OR if the barrier was breached but the underlying recovered above the strike at maturity (depending on termsheet terms). The cash amount is typically 100% of nominal plus any final coupon. (2) Physical delivery: the investor receives shares of the underlying (or worst-performing underlying in a worst-of). For a BRC, physical delivery applies if the barrier was breached AND the underlying closes below the strike at maturity. The number of shares delivered equals the nominal divided by the strike price (the "ratio"). The investor bears the loss — the delivered shares are worth less than the nominal invested. Some termsheets allow the issuer to choose between cash and physical delivery (issuer's option), which can disadvantage the investor in certain tax scenarios. The settlement type is determined by the final fixing price relative to strike and barrier conditions.

Settlement: barrier not breached → put did not knock in → client receives full notional in cash (100%). The below-strike position is irrelevant because the barrier condition was not met. Issuer hedge book T-1: the issuer holds a short down-and-in put position that will expire worthless tomorrow (barrier never triggered). Delta of a near-expiry unbreached barrier put approaches zero rapidly. The desk is unwinding any residual delta hedge (buying back underlying sold during the product life). P&L: issuer retains the full put premium received at inception minus hedging costs — this is the ideal scenario for the issuer.

Product A (European barrier): barrier observation is final fixing only. The underlying is 5% above strike at maturity → barrier is NOT breached (assuming barrier is below strike, which is standard). Client receives 100% of nominal in cash plus final coupon. Full return of capital. Product B (American barrier): barrier was breached 3 months ago — the knock-in condition is permanently triggered regardless of subsequent recovery. At maturity, the underlying is 5% above strike → since the underlying is above strike, even though the barrier was breached, the put expires out-of-the-money. Client receives 100% of nominal in cash plus coupons. In this specific case, both clients receive 100% — the recovery above strike saves Product B's client. The asymmetry emerges when the underlying is between the barrier and strike at maturity: Product A client receives 100% (barrier never observed), Product B client receives physical delivery at a loss (barrier was breached, underlying below strike). The key insight: the American barrier client carries "memory" of the breach — they are exposed to any final-day outcome below strike, while the European barrier client only cares about the final day relative to the barrier. This path-dependency is the fundamental pricing difference and explains why American barriers offer higher coupons.

Scoring: FM3 trap — model may incorrectly state Product B always loses when barrier was breached, missing that recovery above strike still produces full redemption.

The client is FX-protected by the quanto feature — they receive CHF regardless of USD/CHF moves. The FX risk has been transferred to the issuer. The issuer must hedge the quanto adjustment, which involves: buying CHF/selling USD forward in proportion to the equity delta, and adjusting dynamically as equity delta changes. The quanto adjustment modifies the effective forward on the equity (correlation between equity and FX enters the pricing). Typically the issuer hedges with a quanto swap or adjusts the replication using FX-equity correlation from the vol surface. In practice, the issuer charges for this FX risk via the quanto adjustment in the pricing (client receives a lower coupon or worse terms than a non-quanto equivalent). The FX risk ultimately resides on the issuer's hedging book.

1. Capital Protection (e.g. 1100 — Capital Protection Note with Participation) 2. Yield Enhancement (e.g. 1230 — Barrier Reverse Convertible) 3. Participation (e.g. 1300 — Tracker Certificate) 4. Investment Products with Additional Credit Risk (e.g. 1410 — Credit-Linked Note) 5. Leverage Products (e.g. 2100 — Warrant) The first four are classified as "Investment Products"; the fifth is "Leverage Products." The SSPA taxonomy is the European standard (adopted by EUSIPA) and updated monthly. A new product type is created when three issuers each launch at least three products of the same type.

Scoring: Deduct factual accuracy if model confuses categories or invents codes. The credit risk category (4th) is commonly missed.

Discount Certificate = long underlying + short ATM call. The investor buys the underlying at a discount funded by the call premium, but upside is capped at the strike. BRC = long ZCB + short down-and-in put (with barrier). The BRC investor sells a conditional put rather than a vanilla call. The BRC typically offers a higher coupon because: (1) the barrier creates a conditional payoff with discontinuity, making it harder to hedge (wider bid-offer for the issuer), (2) the short put is OTM and further leveraged by skew — investors receive more premium for bearing tail risk, (3) in multi-underlying (worst-of) BRCs, the correlation risk adds further premium. The Discount Certificate has a simpler, more continuous payoff (covered call writing) with no gap risk.

Scoring: FM1 trap — model may describe both as "selling optionality" without distinguishing the specific option type. Award cross-domain bonus if model connects to D2 (vol skew) or D6 (hedging cost).

Capital Protection Note (1100): protection is unconditional — the ZCB floor guarantees minimum redemption (typically 90–100% of nominal) regardless of what the underlying does. Protection applies at maturity irrespective of path. Bonus Certificate (1320): protection is conditional on the barrier NOT being breached during the product's life. If the barrier is hit, the bonus mechanism disappears and it becomes a Tracker Certificate, exposing the client to full downside. The Bonus Certificate is classified under Participation (not Capital Protection) because the protection is conditional and can be lost entirely via barrier breach. SSPA defines Capital Protection as requiring unconditional minimum redemption of at least 90% of nominal at maturity. Conditional protection does not qualify.

Scoring: Core discriminator is unconditional vs. conditional protection. Deduct points if model conflates the two.

As the underlying approaches the barrier, the Bonus Certificate's delta increases sharply (becomes more negative for the issuer's hedge, as the product's value is increasingly sensitive to whether the barrier is hit). Gamma spikes because delta is changing rapidly — the product's sensitivity is highly non-linear near the barrier. For the issuer: (1) the down-and-in put component is near-barrier, creating the same gamma spike as any barrier option, (2) the bonus mechanism creates a discontinuity at the barrier — above it the product pays the bonus level, below it becomes a tracker — this jump in value at the barrier creates massive pin risk, (3) hedging requires frequent delta rebalancing in a zone where the underlying oscillates near the barrier, generating realized hedging losses (gamma bleed). The product is most expensive to hedge precisely in this zone because the expected hedging cost scales with gamma × realized vol.

Tension: SSPA classifies AMCs as Tracker Certificates (participation, 1:1 tracking), but the discretionary element means the payoff is not purely determined by the initial underlying composition — the manager changes it. This blurs the line between a structured product (fixed payoff formula) and a collective investment scheme (CIS/fund). Regulatory risks: (1) under Swiss KAG/FIDLEG, if the AMC resembles a fund (pooled assets, active management, open-ended), it could be re-classified as a collective investment scheme, triggering licensing requirements, (2) the absence of a prospectus obligation (if sold to professional clients) means less disclosure than equivalent funds, (3) FINMA may challenge the classification if the management discretion is broad. Economic risks specific to AMCs: (1) no separation of assets — investor bears full issuer credit risk (unlike a fund with segregated assets), (2) manager conflicts of interest (no independent custodian), (3) fees embedded in the spread may be opaque, (4) liquidity depends entirely on the issuer's market-making commitment, (5) no NAV reporting standard like funds. These risks do not exist for static trackers where the composition is fixed at inception.

Yes, BVV2 Art. 53 allows pension funds to invest in structured products, typically classified under alternative investments (Art. 53 Abs. 3 lit. a). The key barrier was cost transparency: BVV2 Art. 48a requires pension funds to disclose all management costs in their financial statements. Until 2019, structured products were considered cost-intransparent because the embedded costs (structuring margin, hedging costs, distribution fees) were not separately disclosed. In 2019, the OAK (Oberaufsichtskommission) recognized the SSPA's cost transparency framework (Richtlinie), which requires issuers to provide a PRIIP KID and daily cost data on the SSPA website. This made structured products cost-transparent under BVV2 and opened the door for pension fund investment.

Pension funds are classified as "professional clients" (professionelle Kunden) under FIDLEG Art. 4. Practical implications: (1) No prospectus requirement — the exemption under Art. 36 Abs. 1 lit. a FIDLEG means the issuer need not publish a full prospectus if the offer is directed exclusively at professional clients, (2) No mandatory BIB (Basisinformationsblatt) — under Art. 58 FIDLEG, the BIB obligation applies only to retail (Privatkunden), not professional clients, (3) Waiver of suitability/appropriateness testing — under Art. 13 Abs. 3 and Art. 20 Abs. 2 FIDLEG, the financial service provider can forgo these checks for professional clients unless there are doubts about the client's understanding. However: the SSPA's cost transparency framework separately requires issuers to provide a PRIIP KID (European equivalent) to pension funds, creating a higher disclosure standard than FIDLEG alone mandates. Pension funds can also opt-in to retail protection (stronger protection) or opt-out to institutional status (weaker protection) under Art. 5 FIDLEG.

The Golden Rule: from a macroeconomic perspective, the pension fund capital stock is efficient when capital returns (r) equal economic growth (w). From an actuarial perspective, the guarantee requires r>w (return above wage growth) on risk-free assets — but this has been violated in Switzerland since ~2012 as risk-free rates fell below wage growth. The implication: pension funds cannot meet their obligations through risk-free bonds alone and must take risk. Justification for structured products: (1) Asymmetric structuring allows risk-taking with downside control — capital protection products (convex payoff) act as portfolio insurance, yield enhancement products (concave payoff) act as systematic risk premium harvesting, (2) Strategic structuring for ALM: improving cash flow matching, tail risk management, and dynamic asset allocation without manual rebalancing, (3) Selective risk exposure: structured products enable access to emerging markets, commodities, and alternative themes with controlled risk budgets (e.g., capital-protected exposure to technology or ESG themes), (4) Duration management via fixed-income structures paired with credit risk (CLN). The Handbook emphasizes that the real value for pension funds lies NOT in retail-style complex payoffs (multi-barrier worst-of) but in institutionally tailored solutions for ALM and risk budgeting.

Scoring: FM3 trap — model may argue pension funds should avoid all risk, missing the Zimmermann argument that risk-free investing is itself a risk to the system.

Sharpe Ratio is based on mean-variance optimization, which assumes symmetric return distributions. Structured products are inherently asymmetric: capital protection creates positive skewness (limited downside, unlimited upside → convex), yield enhancement creates negative skewness (limited upside, unlimited downside → concave). Using Sharpe Ratio for yield enhancement products systematically overstates risk-adjusted performance because the low measured volatility hides the tail risk (negative skew). Conversely, for capital protection products, Sharpe Ratio understates performance because the high volatility includes desirable upside movements. Alternative measures: (1) Value-at-Risk (VaR) and Expected Shortfall (ES/CVaR) — capture tail losses, with ES being preferable as it considers the average loss beyond VaR, (2) Return-to-Expected-Shortfall Ratio (RTS) — replaces Sharpe's standard deviation with expected shortfall, giving correct rankings for asymmetric payoffs, (3) Higher-moment measures incorporating skewness and kurtosis — Zimmermann's research shows that capital protection reduces VaR much more effectively than yield enhancement (VaR drops from 23.6% to 13.4% vs. 20.1%), while Sharpe Ratio misleadingly suggests the opposite. This is directly relevant to pension fund reporting under BVV2 and OAK requirements.

The TER of a structured product is defined as the difference between the issuance price (100%) and the fair price of its components (the sum of the theoretical values of the embedded options, bonds, etc.). It encompasses: (1) net margin (issuer profit), (2) production and distribution costs (staff, technology, sales, advertising, documentation), (3) risk management costs (hedging P&L, which is uncertain and only fully known at maturity). The TER is NOT the same as the issuer margin: the issuer margin is only the net profit component, which is lower than the TER. The TER includes all costs the issuer must cover, many of which are passed through to operational expenses. The SFI 2015 study found median TERs of: Tracker Certificates ~0.3% p.a., Capital Protection ~0.6% p.a., Bonus Certificates ~1.0% p.a., Discount Certificates ~1.4% p.a., Barrier Reverse Convertibles ~1.7% p.a.

The cost hierarchy reflects hedging complexity and risk transfer: (1) Tracker Certificates (~0.3%): simplest to hedge — delta-one product, minimal optionality, near-static hedge. Costs mainly cover funding spread and operational overhead. Comparable to ETF costs. (2) Capital Protection (~0.6%): requires a ZCB (simple funding) plus a vanilla call (well-understood hedge). Moderate hedging costs. (3) Bonus Certificates (~1.0%): contain a down-and-in put with barrier, creating path-dependency and gamma exposure near the barrier. Hedging is more complex than vanilla options. (4) Discount Certificates (~1.4%): contain a short call with pin risk near the cap. In multi-underlying variants, correlation hedging adds cost. (5) BRCs (~1.7%): most complex to hedge — barrier puts with potential worst-of features, skew exposure, correlation risk in multi-underlying variants, and significant gamma/vega near the barrier. The issuer's hedging P&L is most uncertain, requiring larger risk buffers priced into the product. Additionally, BRCs are the most popular retail product in Switzerland, allowing issuers to maintain wider margins due to demand. The general principle: the more exotic the optionality, the more expensive the hedging → the higher the TER.

Gap 1 (100% − 98.5% = 1.5%): the difference between issuance price and IEV. This represents the issuer's disclosed sales and distribution fees (production, marketing, advisor compensation). It is known and fixed at issuance. Gap 2 (98.5% − 97.5% = 1.0%): the difference between IEV and theoretical price. This is the risk management / hedging cost component. It is fundamentally different because: (1) It is NOT a fixed cost — it is the issuer's expected hedging P&L, which is stochastic. The issuer may make or lose money on this component depending on realized vs. implied vol, realized vs. expected correlation, hedging friction, and gap events. (2) It functions as a risk premium paid by the investor to the issuer for bearing the hedging risk — analogous to an insurance premium. (3) It can only be definitively known at product maturity, after all hedging trades are unwound. (4) The issuer may actually lose money on this component in volatile markets (hedging costs exceed the embedded premium), making the TER as defined by the SFI study a lower bound on the true cost in some scenarios. This is why the SFI study cautions that the TER they define does not equal the issuer margin — the margin is lower and uncertain.

TER alone is misleading for several reasons: (1) TER decomposition matters — Product X's lower TER could mean a tighter structuring margin (genuine cost savings) or it could mean the issuer is pricing the hedging component too aggressively (underestimating future hedging costs, which creates issuer risk if the hedge P&L turns negative). (2) Funding spread asymmetry — if Issuer A has a wider credit spread (lower credit quality), the ZCB component costs less in present value, freeing more budget. The lower TER may be partially funded by the investor bearing more issuer credit risk — effectively a hidden cost transfer. (3) Secondary market quality — Issuer B may provide tighter bid-offer spreads in secondary trading, better market-making commitment (higher QQM score on SIX), and more frequent price updates. The ongoing cost of exiting Product X at wide bid-offer could exceed the TER savings. (4) Execution quality — the effective fill price at launch may differ from the indicated TER due to timing, delta hedging slippage, and the issuer's vol marking practices. (5) Coupon vs. TER trade-off — a higher TER product may offer a lower coupon at the same barrier, which could be more appropriate for the fund's risk budget. The committee should compare: secondary market liquidity metrics (QQM), issuer credit quality (credit spread / COSI availability), coupon-adjusted return net of TER, and the historical variance between stated TER and realized cost (available from SSPA's cost transparency platform).

Scoring: Award cross-domain bonus if model connects to D11 (COSI / issuer credit quality as hidden cost dimension).

COSI is a collateralization framework developed by SIX and SSPA (introduced 2009). Issuers deposit liquid securities with SIX SIS (a custodian) as collateral. The collateral is assessed daily against the market value of outstanding products. If the product price moves, the issuer must adjust the collateral accordingly. If the issuer defaults, the deposited collateral serves as security for the investor's claims. Key points: (1) COSI protects against issuer default risk only — NOT against market risk (losses from underlying performance), (2) available only for products listed on SIX Swiss Exchange, (3) depending on price movements, issuers must supplement collateral as needed (margin call mechanism), (4) COSI has been extended to foreign issuers since 2012. COSI was a Swiss innovation that gained international recognition.

The COSI BRC from Issuer B, despite the higher credit spread, may be preferable because the collateral mechanism largely neutralizes Issuer B's higher default risk. The standard BRC from Issuer A exposes the investor to unsecured creditor status in default — even with a tighter spread, the investor bears full issuer risk. Residual risks with COSI: (1) Collateral gap risk — in a rapid default, the collateral may not fully cover the product's value if there is a lag in the daily adjustment process, (2) collateral quality risk — the deposited securities themselves may lose value in a systemic crisis (correlation between issuer default and collateral value), (3) operational risk in the SIX SIS process, (4) market risk remains entirely unaffected by COSI, (5) the COSI BRC may have a slightly lower coupon because the issuer's funding benefit is reduced (they must post collateral, which has an opportunity cost). The comparison highlights a key principle: issuer credit risk should NOT be confused with product market risk. Many investors conflate the two.

Scoring: FM3 trap — model may incorrectly state COSI eliminates all risk, or confuse market risk with credit risk.

The 1.5% spread can be decomposed into: (1) Funding spread (~60-100bp): the lower-rated issuer funds at a higher credit spread. The ZCB component in the product is issued at the issuer's own funding rate. A higher funding rate means the ZCB is cheaper in present value, freeing more budget for the coupon. This is counterintuitive: worse issuer credit → higher coupon. (2) Credit risk premium (~50-90bp): the investor demands compensation for bearing the higher issuer default risk, which manifests as a higher coupon. Direction of causality: worse credit quality → higher funding spread → more budget for option premium → higher coupon. This creates a perverse incentive structure: the highest-coupon products often come from the lowest-quality issuers. Pre-Lehman, many investors chased high coupons without recognizing they were harvesting issuer credit risk premium. The coupon is NOT purely option-derived — a material portion represents compensation for bearing issuer default risk. This is why COSI products and products from high-quality issuers systematically offer lower coupons at identical barrier/underlying parameters.

Scoring: Award cross-domain bonus if model connects to D1 (structuring margin decomposition) or D4 (mis-selling risk).

Mechanism of loss: structured products are debt securities issued by a bank. The investor's claim is against the issuer, not against the underlying assets. When Lehman defaulted, the underlying equities in many BRCs and capital protection notes were fine — some were even above their strike prices. But the product's value dropped to the recovery rate of Lehman's unsecured debt (~8-12 cents on the dollar in most jurisdictions), because the promise to pay was Lehman's obligation. The ZCB component (which provides capital protection) is funded by the issuer's own credit — if the issuer defaults, the ZCB is worthless. Investors in "capital protected" products lost nearly everything despite the name. Why COSI would not fully prevent losses in a systemic crisis: (1) Collateral correlation — in a systemic crisis, the collateral deposited by the issuer (typically liquid securities) loses value simultaneously with the issuer's deterioration. If the issuer posts equities as collateral and the equity market crashes alongside the issuer's default, the collateral may cover only 60-80% of the product's value. (2) Timing gap — daily collateral assessment means a sudden weekend or overnight default (Lehman collapsed over a weekend) creates a gap between last collateral call and default. (3) Wrong-way risk — if the issuer's collateral includes its own bonds or assets correlated with its creditworthiness, the collateral degrades precisely when needed most. (4) Operational risk — in a systemic crisis, the SIX SIS process for liquidating collateral and distributing proceeds faces queuing, legal challenges, and potential cross-border complications. (5) Aggregate exposure — if multiple issuers default simultaneously, the collateral pool may be insufficient system-wide. COSI significantly reduces issuer risk in idiosyncratic defaults but provides incomplete protection in systemic events.

Scoring: Expect most models to get the basic issuer default mechanism but miss the systemic limitations of COSI. Award full marks only if model identifies wrong-way risk and collateral correlation.

A CLN combines an issuer's bond with a sold credit default swap (CDS) on a reference entity. The investor bears TWO credit risks: (1) the issuer of the note (as with any structured product), and (2) the reference entity. If the reference entity experiences a credit event (default, restructuring), the CLN's redemption is reduced. If no credit event occurs, the investor receives the coupon and full principal at maturity. Difference from a standard corporate bond: (1) a bond exposes you to one credit risk (the bond issuer); a CLN exposes you to two (note issuer + reference entity), (2) CLNs can be constructed on entities that have no CHF bonds outstanding — they are a diversification tool for bond portfolios, (3) CLN coupons can exceed equivalent bond yields when there is a positive basis (CDS spread > bond spread), (4) credit events in CLNs follow ISDA-defined procedures, providing more standardized resolution than bond default processes.

With negative CHF rates, CHF-denominated bonds had negative yield-to-maturity — investors were paying to hold them. CLNs offered an alternative: by combining the issuer's note with a sold CDS on a chosen reference entity, the CLN could generate a positive yield in CHF even when equivalent bonds yielded negatively. The positive basis opportunity: "Basis" = CDS spread minus bond spread for the same reference entity. When the basis is positive (CDS spread > bond spread), an investor can capture this difference via a CLN. Post-SNB floor removal, basis was positive for many European and US corporates because: (1) bond markets were distorted by ECB QE (compressed bond spreads), (2) CDS markets reflected truer credit risk, creating a wedge. Additional advantage: CLNs allowed Swiss pension funds to gain CHF-denominated credit exposure to large US/European corporates that had no CHF bonds outstanding — impossible via direct bond investment. The SFI study (Opportunity 5) specifically cited this as a structured product use case, noting CLNs are more liquid than underlying bonds and follow standardized ISDA default procedures.

Scoring: FM3 trap — model may confuse direction of the basis trade or misunderstand that the investor is selling protection (taking credit risk) via the CDS.

Credit risk layers: (1) First-to-default risk across the 5 reference entities — if ANY one entity defaults, the CLN suffers a loss. This is a first-to-default basket structure. (2) Issuer risk — mitigated but not eliminated by COSI (residual gap risk on collateral). (3) Correlation between the 5 reference entities — the key pricing variable. Correlation effect on coupon: (1) Low correlation → higher coupon. With low correlation, the probability that at least one entity defaults is significantly higher than any individual default probability (diversification of default risk is low). The investor is taking more risk → receives more premium. (2) High correlation → lower coupon. If all entities are highly correlated (e.g., same sector), they either all default or none do — the basket behaves more like a single-name CLN, reducing the spread vs. single-name. The investor is short correlation: they benefit (receive higher coupon) when the market prices low correlation. This is analogous to the worst-of correlation exposure in equity BRCs. The pension fund must evaluate whether the coupon adequately compensates for the first-to-default risk, which is substantially higher than the average individual default probability.

Sovereign CLNs differ from corporate because: (1) Sovereigns cannot "default" in the traditional corporate sense — they can restructure, repudiate, or impose moratoriums, but there is no bankruptcy court. The credit event determination relies entirely on ISDA definitions, which historically were ambiguous for sovereign restructuring. (2) Voluntary vs. involuntary restructuring — the 2012 Greek PSI was structured as a "voluntary" bond exchange (investors were offered new bonds at ~53.5% of face value). ISDA's Determinations Committee initially ruled this was NOT a credit event because participation was voluntary. Only when Greece retroactively inserted Collective Action Clauses (CACs) forcing holdout bondholders to accept the exchange did ISDA declare a credit event (March 2012). (3) This exposed the gap: a sovereign can impose massive losses on bondholders through "voluntary" exchanges that technically avoid triggering CDS/CLN payouts, leaving CLN investors unprotected despite real economic loss. (4) The 2014 ISDA definitions addressed this by: introducing "Governmental Intervention" as a new credit event type (binding exchange offers, currency redenomination, haircuts imposed by law), standardizing the auction process for sovereign credit events, and clarifying that CAC-triggered restructuring is a credit event. For a CLN investor: the underlying ISDA definitions directly determine whether a loss is covered. A CLN written under pre-2014 definitions on a sovereign that imposes a "voluntary" exchange could leave the investor bearing losses on the reference bond without triggering the CLN's credit event protection. The contract law, not the economics, determines the payoff.

Scoring: Expect most models to handle the basic sovereign/corporate distinction but miss the Greek PSI voluntary/involuntary nuance and the 2014 definitional fix. Award cross-domain bonus if model connects to D3 (documentation risk / which ISDA definitions apply).

A Dual Currency Note (DCN) is a short-term (typically 1–6 month) yield enhancement product where the investor deposits one currency and sells an FX option. The investor receives an enhanced yield (above money market rates) in exchange for accepting the risk that the principal is repaid in the weaker of two currencies at maturity. Embedded optionality: the investor is short a put on the investment currency (or equivalently short a call on the alternate currency). If the investment currency depreciates beyond the strike, the investor receives repayment in the depreciated currency at the pre-agreed strike rate. The enhanced yield is funded by the FX option premium. DCNs are popular among private banking clients for FX management and yield enhancement on cash positions.

FX forward hedge: the forward rate reflects the interest rate differential between CHF and USD. When CHF rates are negative and USD rates are positive, the forward points are in the pension fund's favor: they sell USD forward at a rate ABOVE spot (the forward premium compensates for holding the lower-yielding CHF). The implicit "cost" is actually a positive carry: the fund earns the interest rate differential through the forward premium. Structured FX overlay: could use options (protective puts on USD/CHF) which cost premium but allow participation in favorable moves, or BRC-like structures on the FX pair. Comparison: (1) forwards are cheapest and provide certainty but eliminate all FX upside, (2) options preserve upside but cost premium (reducing the hedge's net benefit), (3) structured overlays can be tailored (e.g., participating forwards with barriers) to balance cost vs. protection. Post-SNB floor removal, the SFI study noted that USD 1-year rates via FX swap from a -0.75% CHF base were ~2.7%, creating a significant positive carry opportunity for qualified investors with USD accounts — a direct market distortion exploitable via structured products.

Scoring: FM3 trap — model may state "hedging always costs money" without recognizing the carry benefit of CHF/USD hedging when rate differentials favor CHF.

Physical delivery is impractical for most commodities (storage, perishability). Only precious metals (gold, silver) have low enough convenience yield and storage costs to be held physically. All other commodities use futures-based indices (e.g., UBS Bloomberg CMCI). Roll yield impact: (1) Contango (futures price > spot): when the futures curve is upward-sloping, rolling from expiring to next-month contracts means buying at a higher price — negative roll yield. The Tracker Certificate systematically underperforms spot. (2) Backwardation (futures price < spot): rolling into cheaper contracts produces positive roll yield — the Tracker outperforms spot. This means a Tracker Certificate on WTI crude in contango can show losses even when spot oil prices are flat, because each monthly roll erodes value. The CMCI (Constant Maturity Commodity Index) was designed to address this by diversifying across the entire futures curve rather than just the front month, reducing but not eliminating roll yield impact. For pension funds, this makes commodity structured products materially different from equity products: the return driver includes basis risk and roll yield, not just directional spot moves.

Cross-asset worst-of is more dangerous because: (1) Regime-dependent correlation — equity/FX/commodity correlations are unstable and regime-dependent. In normal markets, Nestlé/EUR-CHF/Gold may have low or moderate correlation (~0.1-0.3). In a crisis (e.g., 2015 SNB floor removal), correlations can spike or flip sign. EUR/CHF moved 20% in minutes while equities dropped and gold spiked — the worst-of mechanism would have been triggered by the FX leg alone. (2) Non-comparable volatility scales — equity vol (~15-25%), FX vol (~5-15%), and commodity vol (~15-30%) operate on different scales. A 25% decline in Nestlé, a 25% move in EUR/CHF, and a 25% decline in gold represent entirely different probability events. The barrier at 75% may be reasonable for the equity but extremely deep for the FX pair (EUR/CHF rarely moves 25%) or relatively shallow for gold (which has moved 25%+ in several years). The investor is effectively selling puts with very different moneyness across asset classes. (3) Correlation model breakdown — standard Gaussian copula or constant-correlation models used for pricing assume: (a) stable correlation structure, (b) symmetric dependence — but cross-asset tail dependence is highly asymmetric (assets correlate more in crashes than rallies), (c) no jump risk — FX pairs subject to central bank intervention exhibit discontinuous moves that violate the diffusion assumption underlying most pricing models. The pricing model systematically underestimates the probability of the worst-of barrier breach during stress events. (4) Hedging complexity — the issuer must hedge across three different market microstructures (equity, FX, commodity) with different liquidity profiles, trading hours, and margin requirements. The cross-gamma between asset classes is model-dependent and poorly observed in the market. The investor is being compensated for complexity that the pricing model cannot fully capture — the 8% coupon is likely inadequate for the true tail risk.

Scoring: This is the hardest question in D13. Expect most models to identify the correlation instability but miss the non-comparable volatility scales and the asymmetric tail dependence. Award cross-domain bonus if model connects to D6 (hedging across asset classes) and D2 (correlation model limitations).