New study reveals surprising pitfalls in analyzing tail dependence copulas.
Tail dependence copulas help us understand how variables are related in extreme situations. For certain types of copulas, the shape of the mathematical function near zero tells us about the strength of this relationship. The study shows that even if the function is smooth and changes slowly at zero, the relationship in the tails of the distribution may not become more independent. This challenges previous ideas about how these copulas behave.