摘要：This research provides evidence of predictability and asymmetry in the returns and volatility of the two largest exchange-traded notes (ETNs), namely, JPMorgan Alerian MLP Index ETN (ticker: AMJ) and iPath DJ-UBS Commodity ETN (ticker: DJP). This study found that AMJ ETN has an intermediate memory based on the autoregressive fractionally integrated moving average (ARFIMA) model and the combined ARFIMA-fractionally integrated general autoregressive conditional heteroskedasticity (ARFIMA-FIGARCH) models. Long-memory properties also existed in the volatility structures of both the AMJ and DJP ETNs according to the ARFIMA-FIGARCH models making them predictable in the long-run, and violates Fama’s (1970) weak-form efficiency hypothesis. The combined ARFIMA-fractionally integrated asymmetric power autoregressive conditional heteroskedasticity (ARFIMA-FIAPARCH) models did not confirm the initial findings due to insignificant results. However, the gamma ( ) parameter of the ARFIMA-FIAPARCH models showed the presence of volatility asymmetry in the AMJ ETN, which means that negative shocks have relatively more impact than positive shocks on its volatility.
