Abstract I propose a likelihood ratio test for fixed unit root against time switching unit root models. Random coefficients in the stochastic alternative effectively capture both expansion and contraction behaviors. My test is asymptotically optimal in the sense that it maximizes a weighted power function. The testing problem is challenging due to nonstationarity. A wide range of models that are popular in macroeconomics and finance can be applied to my test. The test is simple to implement, as it requires only the estimation of the parameters under the null hypothesis of constant parameters.Limit distribution of likelihood is derived.
Working in Progress
Estimation of Explosive Behavior of Cryptocurrency Prices Cryptocurrency Price Discrepancies and Financial Integration Event Studies of Private Placement and Lifting of Restricted Shares