where \(p\left(r_{t}\middle|r_{t-1}\right)\) is the change point
prior distribution containing our prior beliefs about the probability of
a regime shift. We assume constant prior probability for a shift defined
by the hazard rate parameter \(\lambda\),