A local linear estimator of generalized impulse response (GIR) functions for nonlinear conditional heteroskedastic autoregressive processes is derived and shown to be asymptotically normal. A plug-in bandwidth is obtained that minimizes the asymptotical mean squared error of the GIR estimator. A...

Optimal bandwidths for local polynomial regression usually involve functionals of the derivatives of the unknown regression function. In the multivariate case, estimates of these functionals are not readily available, primarily because estimating multivariate derivatives is complicated. In this...

A nonparametric version of the Final Prediction Error (FPE) is proposed for lag selection in nonlinear autoregressive time series. We derive its consistency for both local constant and local linear estimators using a derived optimal bandwidth. Further asymptotic analysis suggests a greater...

This paper explores the implications of asset return predictability for long-term portfolio choice when return-forecasting variables are fractionally integrated. For important predictor variables, like the dividend-price ratio, and nominal and real interest rates, we estimate orders of...

This paper explores the implications of asset return predictability on long-term portfolio choice when return forecasting variables exhibit long memory. We model long memory using the class of fractionally integrated time series models. Important predictor variables for U.S. data, like the...

This paper explores nonlinear dynamics for the time series of the short term interest rate in the United States. The proposed model is an autoregressive threshold model augmented by conditional heteroskedasticity. The performance of the model is evaluated by considering its implications for the...

There has been recent evidence for long memory in the changes of foreign exchange spot rates that is captured by the fractionally integrated ARMA model. This paper extends these investigations in several directions. First, the estimation procedure allows for GARCH errors. Second, in addition to...

It is shown that the approximate Whittle estimator suffers from an inherent small sample estimation bias which is not present in other frequency domain maximum likelihood methods. A bias corrected approximate Whittle estimator is derived and compared with existing alternatives. Furthermore, the...