This paper studies the long-term dependence and the possible asymmetric behavior of the financial time series. Both can be modeled using a fractionally integrated autoregressive moving average time series model with threshold-type conditional heteroscedasticity, denoted as an ARFIMA–TGARCH...