Title: Time-varying model averaging
Authors: Yuying Sun - Academy of Mathematics and Systems Science, Chinese Academy of Sciences (China) [presenting]
Xinyu Zhang - Academy of Mathematics and Systems Science, Chinese Academy of Sciences (China)
Tae-Hwy Lee - University of California Riverside (United States)
Yongmiao Hong - Cornell University (United States)
Shouyang Wang - Academy of Mathematics and System Science, Chinese Academy of Sciences, (China)
Abstract: Structural changes often occur in economics and finance due to changes in preferences, technologies, institutional arrangements, policies, crises, etc. Improving the forecast accuracy of economic time series with the evolutionary behavior is a long-standing problem. Model averaging aims at providing an insurance against selecting a poor model. All existing model averaging approaches are designed with constant weights. Little attention has been paid to the time-varying model averaging, which is more realistic in economics under structural changes. A novel model averaging estimator is proposed which selects the smoothly time-varying weights by minimizing a local jackknife criterion. It is shown that the proposed time-varying jackknife model averaging (TJMA) estimator is asymptotically optimal in the sense of achieving the lowest possible local squared errors in a class of time-varying model averaging estimators, with allowing non-spherical errors. A simulation study and empirical application highlight the merits of the proposed TJMA estimator relative to a variety of popular estimators from constant model averaging and model selection.