Decomposing the structured product
The classic capital-protected note: a zero-coupon bond growing to principal at maturity, plus options bought with the accrued interest. At 5% for five years, 100 becomes 127.6 — that 27.6 is the option budget; if options are expensive, 'protection' costs opportunity in flat markets. Yield-enhancement products usually embed a short option (a far call): higher headline yield in exchange for capped upside and tail risk. The single question that decodes any product: who bought which risk from me?
- Vertical call spread (buy K₁, sell K₂): capped risk and capped premium — a measured directional view.
- Butterfly (buy K₁ and K₃, sell 2×K₂): a view on range; painful in breakouts.
- Collar (long stock + put + short call): two-sided capped returns; cheaper protection.
- Autocalls and barriers: path-dependent — they need multi-factor models and do not behave like their expiry profile.
Hidden costs and reading the profile
The expiry payoff is one slice: pre-expiry value depends on implied vol and time, and structured-product desks usually mark implied vol with a margin — so you can lose on the way to a good destination. Real costs: multi-leg execution spread, model margin, unwinding before maturity, and thin secondary liquidity. A product marketed as 'higher return with the same risk' has moved some unseen risk — tail, liquidity or model complexity.
Worked example: a collar on an equity book
- Book 100, beta 1; buy a 1y put at strike 95 (premium 3) and sell a 1y call at 115 (premium 1.5).
- Net hedge cost = 1.5 — cheaper than an outright put.
- Market falls 25%: book 75, put worth 20+ → total ≈ 95 − 1.5 = 93.5; loss capped at 6.5.
- Market rises 30%: book 130 but the short call absorbs above 115 → cap ≈ 113.5.
نتیجه: A collar does not remove risk: you buy a loss cap and sell an upside cap. Read the two-sided profile before the trade, not after the move.