Vector autoregressive moving-average models have been deemed impractical for large-scale applications due to their complex likelihood function and high computational costs. However, these models offer a significant advantage in capturing complex patterns with a limited number of parameters, making them a valuable tool for analyzing multivariate time series data. Researchers have now introduced an estimation method that overcomes the traditional limitations of VARMA models, enabling their application to high-dimensional data. This breakthrough has significant implications for fields such as finance, economics, and cybersecurity, where complex patterns and relationships must be identified and analyzed. The new estimation method allows for efficient and scalable analysis of large datasets, making it possible to uncover hidden patterns and relationships that were previously inaccessible. This matters to practitioners because it enables them to develop more accurate predictive models and gain deeper insights into complex systems1.