Beta as a measure of the systematic risk of a security has won widespread acceptance among financial analysts. Proof of this acceptance lies in the fact that several investment information services provide investors with estimates of beta. Allegedly, the more sophisticated the statistical technique, the more accurate the estimate. More sophisticated estimates of beta can be obtained either by buying them from an information service or by calculating them directly. In either case, the costs involved in getting “better betas” can be considerable.

This article investigates, for one portfolio construction rule, the advisability of investing in better betas. If naive betas are used in constructing a portfolio, the rate of return on this portfolio will suffer from an additional element of uncertainty stemming from errors in beta measurement. However, this risk can be largely eliminated by increasing the size of the portfolio. This approach may prove more cost-effective than an investment in better betas.