# EWMA parameter estimation

In the context of the portfolio allocation problem, i have calculated one step ahead volatility forecasts on a rolling window basis with DCC GARCH while re-estimating the parameters every 25th observation. I want to do the same using EWMA. The initial parameter that I used was 0.94 (for daily return data from riskmetrics). Now I am confused about whether I need to recalculate EWMA parameters the way I did for DCC GARCH or if I should just use 0.94 for each one step ahead forecast. If I need to re-estimate the parameters, how should I go about it? Thanks