论文信息 - Structural Time Series Models

Structural Time Series Models

1 Trend and Cycle Decomposition y t = t + t where y t is an n 1 vector and t and t represent trend and cycle components respectively. This decomposition into components is not unique. Beveridge and Nelson (1981) and Stock and Watson (1988) derive the following decomposition: y t = C(L)" t = C(1)" t + (1 L)C (L)" t Integrating up gives: y t = C(1) 1 X i=0 " ti | {z } + C (L)" t | {z } trend cycle NB: (1) The 'cycle'here could be white noise or a …rst order AR process y t = y t1 + t (2) Innovations in trend and cycle are perfectly correlated.

Beveridge-Nelson-Stock-Watson Decomposition | Beveridge-Nelson-Stock-Watson Decomposition

[1] Mark W. Watson,et al. Univariate detrending methods with stochastic trends , 1986 .

[2] Farshid Vahid,et al. Common Trends and Common Cycles , 1993 .

[3] J. Stock,et al. Testing for Common Trends , 1988 .

[4] Andrew Harvey,et al. Trends and Cycles in Macroeconomic Time Series , 1985 .

[5] R. Engle,et al. Testing for Common Features , 1990 .

[6] Siem Jan Koopman,et al. Stamp 5.0 : structural time series analyser, modeller and predictor , 1996 .

[7] Andrew Harvey,et al. Forecasting, Structural Time Series Models and the Kalman Filter , 1990 .

[8] C. Nelson,et al. A new approach to decomposition of economic time series into permanent and transitory components with particular attention to measurement of the ‘business cycle’☆ , 1981 .