Appendix A: An Illustration of the Difficulties Involved in Benchmark Construction : Calculating Returns on the S&P 500

The construction of benchmarks have their own problems. Table A.1 shows the calculated return on the S&P 500 for various index fund managers and consultants in the United States. It indicates a 50 basis-point discrepancy in the return for the S&P 500. This is a result of these firms or companies either using alternative evaluation techniques and/or adopting different assumptions on the data used for the performance calculations.

Table A.2 shows the price-change-only returns as computed by two companies, Wells Fargo Investment Advisors ("WFIA") and S&P's. It reveals a cumulative discrepancy of 43 basis-points for the year. WFIA estimates that, of this difference, 11 basis-points resulted from WFIA's use of composite closing prices, while S&P's used NYSE closing prices. The remaining 32 basis-points were a result of differences in the treatment of restructuring distributions. 1988 was a year with many corporations restructuring; apparently WFIA and S&P differed on the treatment of many of them, as either dividends or capital distributions.

There are several reasons why calculating the return for a benchmark may be difficult. One problem is determining which investments to use. For example, a company might have multiple classes of equity outstanding. However, with only one class in the S&P 500, S&P adds the shares outstanding from all classes to compute the company's index weight. This produces a capitalisation that does not reflect the true value of that particular company in the index. Accruals of dividends, capital transactions and other cash flows are all issues to be considered in the calculation of returns.