The paper investigates a time series prediction problem. It constructs a pointwise prediction model for a set of time series. These time series have high variance and high covariance. Paper introduces space of pairwise distances between time series and analyses its properties. In this space, the pairwise distance matrix is predicted in its dynamics. The time series values are reconstructed at the next time moment with this matrix. The authors propose several methods for the pointwise prediction of time series space using the reconstruction reconstructed space of the pairwise distance matrices. We prove the existence of multiple time series values satisfying the same pairwise distance matrix. It presents two algorithms based on the use of matrices constructed over different time intervals using pairwise correlation. The paper derives an explicit view of the reconstructed values through the pairwise correlation matrix. Also, it derives an evaluation of the quality of the reconstruction when noise is added to the pairwise distance matrices. The novelty of the method is that the prediction is not done in the original space but in the space of pairwise distances.