Do Structured Products Improve Portfolio Performance? A Backtesting Exercise

We consider a laboratory where bootstrapped synthetic structured products (convertible bonds, reverse convertibles, or barrier reverse convertibles) are added to a 60\% stock and 40\% bond portfolio. By using market data on the underlying assets which are stocks and bonds we directly inherit a realistic dynamic. We show that including structured products in the 60/40 portfolio results in a lower return and lower risk-adjusted performances regardless of what type of structured product is considered. Adding structured products to a portfolio impacts the distributional properties, rendering the distribution less Gaussian. By computing the opportunity costs of including structured products in the 60/40 portfolio, based on a Taylor approximation around the expected utility, we demonstrate that the opportunity cost is negative for all structured products and for reasonable levels of risk-aversion. This finding implies disutility for investors who choose to include structured products in their portfolio. For more complicated opaque structured products without explicit close-form solutions but with stale prices and transaction costs, we expect further performance deterioration.