Although conventional portfolio insurance strategies can virtually guarantee a minimal level of return (i.e., can guarantee that a portfolio will not lose more than a given amount in value), their ability to capture upside moves in the underlying equity market is less predictable. Historical one-year simulations run with 240 different starting months from 1963 through 1983 provide some indication of the cost of portfolio insurance in terms of forgone returns.

Given transaction costs of 0.5 per cent, insured portfolios with zero and –5 per cent floors (created by dynamically balancing stocks and Treasury bills) “cost,” respectively, 170 and 83 basis points per year, as measured by the differences between their average annual geometric returns and those of the S&P 500. Performance was considerably better when transaction costs were ignored.

The distributions of shortfalls and excesses indicate that the zero floor insured portfolio cost roughly as much on the upside (in terms of shortfall) as it saved on the downside (in terms of excess return above the market). However, this portfolio protected against down moves in the market of 20 to 40 per cent, while avoiding shortfalls of 20 per cent or more. The –5 per cent portfolio experienced virtually no shortfalls of 10 per cent or more.

The value of the studied insurance would seem to be its ability to protect the investor against dramatic (though rare) drops in market value. Furthermore, its performance relative to the market is likely to improve (in terms of both lower long-run cost and lower average shortfall), the more the investor is willing to lose (i.e., the lower the portfolio’s floor return). In general, however, the evidence does not indicate that a dynamically balanced, insured portfolio will over the long run outperform a static mix portfolio.