Abstract
This paper investigates the dynamical behavior of logistic self-exciting threshold autoregressive model. The existence and stability of the equilibria of the skeleton are studied. The complex dynamics, bifurcations and chaos are displayed by computing numerically the largest Lyapunov exponents and sensitive dependence on initial conditions. Simulated data is generated to discuss the effect of changes in parameters of observed values and to indicate that LSTAR model can be used to model chaotic time series observations.