Structured Products: You Are Lending to a Bank

RISK-FORWARD, and this is the most important sentence in the article: a structured product is generally a debt obligation of its issuer. Whatever the payoff formula promises, the holder is an unsecured creditor of the institution that issued the note — so if the issuer fails, the formula becomes irrelevant and the holder joins the queue of unsecured claimants. "Capital protection" means the issuer has promised to return capital. It does not mean the capital is protected from the issuer's failure, and that distinction has been the difference between a full return and a partial recovery. These products are marketed heavily to European retail investors. This article explains the mechanics and the exposures. It names no issuer, product, or distributor and recommends nothing.

The construction is genuinely clever and genuinely opaque: the payoff can be shaped almost arbitrarily, which is what makes these products attractive to sell and difficult to evaluate. The note or certificate issued by a bank is the common form and the one this article describes; a minority are structured as bank deposits, which may fall within a deposit-protection scheme, or as collateralised vehicles, and a reader should check which form they hold because the answer determines who owes them the money. This pillar applies Pillar 17's option material rather than re-deriving it.

How the payoff is built, and what it costs

Take the canonical example: a five-year note offering 90% capital protection and 60% participation in an index. The issuer constructs this from two components. Most of the money buys a zero-coupon bond — a debt instrument maturing at the protected amount in five years, which is where the "protection" comes from and why the protection is a promise from the issuer rather than a property of the structure. The remainder buys call options on the index, which is where the participation comes from. Three consequences follow directly from that construction, and all three are arithmetic. The participation rate is limited by what the option budget buys. A 60% participation exists because 60% is what the leftover money could purchase — so the terms of these products are set by option pricing rather than by an issuer's generosity, and a more attractive headline in one dimension is paid for in another. The holder receives no dividends. This is the single most consequential omission and it is almost never prominent: an index participation tracks the index price, and the dividends the underlying companies pay go to the issuer, not the holder. Over five years at a 2% dividend yield, that is roughly 10.4% of compounded return the holder never sees. And the costs are embedded rather than charged. There is no expense ratio; the issuer's margin, distribution fees, and hedging costs are built into the terms — which is the invisible-cost problem in its purest form, since a cost that appears in no fee table cannot be compared to one that does. Four common structures, in ascending order of hazard. Capital-protected notes: as above. Yield-enhancement or reverse convertibles: a high coupon in exchange for accepting downside in the underlying below a barrier — the holder is selling an option and receiving the premium as a "coupon", which is the money-arriving-is-not-income pattern the portal has now identified four times. Autocallables: notes that redeem early if a condition is met, which caps the good outcome while leaving the bad one open, and reinvestment then happens at whatever terms exist later. And leveraged or barrier certificates, where a threshold breach can extinguish value entirely — the most hazardous configuration in this pillar, and one where a temporary breach can permanently destroy the position even if the underlying subsequently recovers.

The payoff table, computed — and the finding it produces

Here is the canonical note against holding the index directly, over five years. The third column is the index price return; the fourth adds reinvested dividends at 2% annually, which is what a holder of the actual index would receive.

Four findings, and the third is the one that should change how a reader reads these products. The protection is real in a severe fall. At −50%, the note holder loses 10% where a direct holder loses 44.8% — that is a genuine and substantial benefit, and any honest treatment must say so. The cost of it rises with the outcome. At +100% the note delivers 60.8 percentage points less than the index with dividends. And in the mild-decline case the "protected" product actually underperforms. At −10%, the note loses 10% while the index with dividends loses 0.6% — so the holder of a capital-protected note did worse than an unprotected direct holding in a falling market. That is not a paradox: the dividends the holder forfeited more than covered the fall. Protection against a 10% decline is worth less than five years of dividends, and the product charges for it in both directions. And the protection only binds at maturity. Selling early means accepting a market price that reflects rates, volatility, and issuer credit standing, so the protected amount is a maturity feature rather than a floor on the position's value throughout. What this article will not say is that these products are never appropriate. An investor who genuinely wants defined downside in a specific window, understands they are lending to the issuer, and accepts the dividend forfeit is making an informed choice, and the payoff shaping is a real capability that cannot be replicated cheaply by a retail investor assembling options themselves. What it will say is that the holder is an unsecured creditor, that the costs are embedded and uncomparable, that the dividend forfeit is usually the largest single cost and rarely disclosed prominently, and that in a mild decline the protected product can lose more than the thing it was protecting against.