VARMA Models: Beyond Simplicity

VARMA, or Vector AutoRegressive Moving Average, models are a versatile tool for understanding how multiple time series interact. Despite their potential, VAR models, a simpler subset, often steal the show in real-world applications. Is it because VAR models are just easier to use, or are users missing out on what VARMA models have to offer? This review aims to explore the strengths and capabilities of VARMA models.First, let's dive into the challenges of identifying these models. Both classic and modern methods have their place. Estimation, specification, and diagnosis follow similar paths. The real power comes when we discuss practical uses, like Granger Causality, forecasting, and structural analysis. These areas showcase how VARMA models can outperform their simpler VAR counterparts. Recent advancements and extensions of VARMA models are making them more accessible. As we look to the future, there's potential for these models to fill gaps where VAR models fall short.