In the empirical finance literature most frequently monthly returns are applied for measuring fund performance or testing market efficiency. We propose a new return calculation method, the daily recalculated monthly returns which has not been used in academic studies for asset pricing purposes. We argue that our method outperforms daily and monthly return calculations in the case of Hungarian mutual funds when only short time series are available. Daily recalculated monthly returns induce the best fitting property of the market model while the time series remain sufficiently long to derive asymptotic tests even when we work on a one-year-long time series. Using our method the estimated parameters and the
s are very close to the results obtained when using monthly returns which are considered a good working approximation.