The article explores how outliers in data can affect multivariate time series differently than in single-variable analysis. It shows that the impact of a multivariate outlier depends on both its size and the model's dynamic structure. This dynamic effect is unique to multivariate data and can create various types of outliers in individual components. By comparing results from single-variable and multivariate outlier detection, we can better understand the characteristics of outliers. The researchers use real examples to demonstrate their findings.