You are not currently logged in.
Access your personal account or get JSTOR access through your library or other institution:
Estimating Continuous-Time Processes Subject to Time Deformation: An Application to Postwar U.S. GNP
James H. Stock
Journal of the American Statistical Association
Vol. 83, No. 401 (Mar., 1988), pp. 77-85
Stable URL: http://www.jstor.org/stable/2288921
Page Count: 9
Preview not available
A class of time series models is presented in which variables evolve on a data-based rather than calendar time scale. The discrete calendar-time model thus obtained exhibits time-varying parameters and conditional heteroscedasticity. Using a procedure based on the Kalman filter, univariate models are estimated for postwar U.S. real gross national product (GNP) and short- and long-term interest rates. The results indicate significant time scale nonlinearities.
Journal of the American Statistical Association © 1988 American Statistical Association