Straightforward manipulation show that the sum of the sample autocorrelation function over the lags 1 to T 1 equal to -0.5 for all time series of length T. There are some alternatives in the literature however those are not common in practice. In this paper, we provide a new approach for estimating the autocorrelation function. This approach uses singular spectrum analysis which is a non-parametric technique for time series analysis. The paper utilises a simulation study to illustrate the performance of the new approach. The results suggest that further improvement to the sample autocorrelation is possible and the new method provides an attractive alternative to the classical approach.
