Abstract/Summary

Despite the arguments that can be made against using the CAPM, it is very widely used in regulation. In particular, it is relied upon in setting utility prices in each of Australia, New Zealand and the United Kingdom, and also features in this context in Germany. In addition, UK competition authorities make use the CAPM to assess profitability in the case of “market investigations”. All of these applications require beta as an input into the CAPM, but the beta estimates typically vary depending on frequency of the returns data used. However, there is little evidence on how estimates of beta depend on the return frequency used in their estimation and in particular, their relationship to any firm characteristics. This study examines the evidence for Australia, Germany and the UK and broadly shows that longer frequency betas have superior characteristics for regulatory purposes in these countries. We find that differences in beta can be explained by size and liquidity factors. Our conclusions are unequivocal and have important policy implications for regulatory use of the CAPM, as they imply that low frequency beta estimates should always be preferred to high frequency beta estimates.