What is Vector autoregression (VAR)?

Vector autoregression (VAR) is like having a group of friends who all influence each other’s choices every day.

Imagine you and your three best friends, let's call them Alex, Bella, and Charlie, decide to go to the park. Sometimes Alex brings snacks, which makes Bella laugh more, and that makes Charlie run faster on the swings. Each of you affects the others in different ways, and it creates a fun chain of events.

VAR is like keeping track of how these friends influence each other over time. Instead of just one friend (like in regular autoregression), VAR looks at several things happening at once, like snack-bringing, laughing, and swinging, all together.

Think of it as a diary that records what everyone did yesterday to predict what they might do today. If Alex brought snacks on Monday, Bella laughed more on Tuesday, and Charlie ran faster on Wednesday, the diary helps guess what will happen next.

This is useful when you want to see how different things, like prices, interest rates, or even your friends’ moods, affect each other over time in a fun, connected way.