Title: Inference on time series with systematically missing data
Authors: Fulvia Marotta - Queen Mary University of London (United Kingdom) [presenting]
Abstract: Unevenly spaced observations is a fundamental issue in econometric time series analysis. There has been considerable focus on parametric estimation methods permitting systematically missing data or observations missing at random times. We suggest a new model fitting approach for such data. The basic idea is to correct the periodogram for missing data so that it enables standard parametric Whittle estimation. After missing data have been accounted for, the dynamics parameters of a time series can be estimated with parametric rate and confidence intervals constructed. The main advantage of such modelling is its simplicity and easiness of use in applications. We investigate performance of the method with simulated and empirical data.