It is well known that, under some conditions, tests are asymptotically equivalent, and are optimal in the sense that the local power of these tests are equal to the maximum that can be achieved. However, in most practical situations, the density function f is unknown.

This paper investigates whether the systematic asymmetric behaviour of the US unemployment rate can be explained by the stock market. We consider threshold models to capture the asymmetric relationship between quarterly US unemployment rate and Dow Jones Industrial Average (DJ) stock returns. We...

An issue that arises in applications involving the ARCH-in-Mean (ARCH-M) model is whether or not the error variance is constant over time. A proper statistical formulation of this as a test of hypothesis presents two difficulties. First, the model does not satisfy the standard regularity...

Testing for linearity in time series models has been an active area of research [see Granger and Terasvirta (1993), Tong (1991)]. The authors consider a test for linearity against a particular regime switching model known as the smooth transition autoregressive (STAR) model.

Testing for linearity in time series models has been an active area of research [see Granger and Terasvirta (1993), Tong (1991)]. The authors consider a test for linearity against a particular regime switching model known as the smooth transition autoregressive (STAR) model.

This paper investigates whether the systematic asymmetric behaviour of the US unemployment rate can be explained by the stock market. We consider threshold models to capture the asymmetric relationship between quarterly US unemployment rate and Dow Jones Industrial Average (DJ) stock returns. We...

It is well known that, under some conditions, tests are asymptotically equivalent, and are optimal in the sense that the local power of these tests are equal to the maximum that can be achieved. However, in most practical situations, the density function f is unknown.

An issue that arises in applications involving the ARCH-in-Mean (ARCH-M) model is whether or not the error variance is constant over time. A proper statistical formulation of this as a test of hypothesis presents two difficulties. First, the model does not satisfy the standard regularity...