Some robust sequential procedures for the detection of structural breaks in the Capital Asset Pricing Model (CAPM) are proposed and studied. Most of the existing procedures for this model are based on ordinary least squares (OLS) estimates.
Here we propose a class of cumulative sum (CUSUM)-type procedures based on M-estimates and partial weighted sums of M-residuals. The theoretical results are accompanied by a simulation study that compares the proposed procedures with those based on OLS estimates.
An application to a real data set is also presented.