We study quasi-maximum likelihood (QML) breakpoint estimation for
 covariance regime switches in multivariate time series. We move beyond
 the classical framework and show that regime switches can be detected
 as soon as the signal-to-noise ratio is high enough. We identify a
 quantitative global recovery threshold that compares signal separation
 between regimes to signal fluctuations within regimes, and show its
 sharpness via an explicit counterexample.  We also discuss further
 developments in breakpoint detection in high dimensions.

 Joint work with Hubeyb Gurdogan (UCLA)