DOI:
关键词: Autoregressive integrated moving average 、 Box–Jenkins 、 Integer 、 Econometrics 、 Business cycle 、 Applied mathematics 、 Order of integration 、 Autoregressive fractionally integrated moving average 、 Time series 、 Economics 、 Component (UML)
摘要: The issue in this paper is to analyse the business cycle frequencies US real output. However, instead of using classical approaches based on linear and non-linear models, we use a specification fractional cyclical integration, which Gegenbauer processes. We apply procedure that permits us test roots with integer orders integration at fixed over time thus, it approximate length cycles. results, first differenced data, show cycles have duration about four years half, an order higher than 0 but smaller 0.5, being thus stationary component long memory behaviour. Comparing model those ARIMA (and ARFIMA) via simulations structure can better describe features data.