Abstract
As a common practice in portfolio management based on the mean-variance model, the daily means and variances of stock returns are usually estimated using daily historical data and then converted to longer periods of time depending on needs. However, since means and variances may show different behaviors over non-trading days, this paper shows the necessity of employing different time scale conversion formulas for means and variances. In particular, we examine the issues arising from the conversion of 1-day returns into estimates of longer period means and investigate the impact of this conversion on capital allocation decisions and portfolio performance evaluations.