Forecasting of lead-time demand variance: Implications for safety stock calculations

•We propose three strategies to estimate the variance of the lead-time demand.•Demand is assumed to be autocorrelated and lead-times are stochastic.•We derive analytical results under an ARMA(1,1) demand process.•We consider both SES and the MMSE forecasting method.•We show the underperformance of the classical strategy for positive autocorrelation.