Charles Explorer logo
🇬🇧

Structural breaks in panel data: large number of panels and short length time series

Publication

Abstract

The detection of the (structural) break or so called change point problem has drawn increasing attention from both theoretical and applied economic and financial research over the last decade. A large part of the existing research concentrates on the detection and asymptotic properties of the change point problem for panels with a large time dimension T.

In this article we study a different approach, i.e., we consider the asymptotic properties with respect to N (number of panel members) while keeping T fixed. This situation (N ? 8 but T being fixed and rather small) is typically related to large (firm-level) data containing financial information about an immerse number of firms/stocks across a limited number of years/quarters/months.

We propose a general approach for testing for the break(s) in this setup, which also allows their detection. In particular, we show the asymptotic behavior of the test statistics, along with an alternative wild bootstrap procedure that could be used to generate the critical values of the test statistics.

The theoretical approach is supplemented by numerous simulations and extended by an empirical illustration. In the practical application we demonstrate the testing procedure in the framework of the four factors CAPM model.

In particular, we estimate breaks in monthly returns of the US mutual funds during the period January 2006 to February 2010 which covers the subprime crises.