Table 4

Forecast point estimates and mean squared errors

TimeSETARARObserved
2016q10.096−0.0620.137
2016q2−0.1020.043−0.163
2016q3−0.172−0.097−0.093
2016q40.1930.0970.104
2017q1−0.009−0.0030.022
2017q20.1480.002−0.072
2017q3−0.14−0.029−0.233
2017q40.1080.0390.115
Forecast RMSE0.0990.13 
Mean absolute error0.0780.099 

Note(s): AR = autoregressive model, SETAR = selfexciting threshold autoregressive model, RMSE = root mean square error

or Create an Account

Close subscription notice
Close access options