Resumen de Functional Statistical Time Series Analysis of the Dividend Policy of Spanish Companies

Román Salmerón Gómez, María Dolores Ruiz Medina

• A functional statistical analysis of a data panel constituted by 33 companies in the IBEX-35, Spain, during the period 2006 to 2009, is achieved for the investigation of possible changes in the dividend policy. Empirical evidence of positive functional correlation of dividend policy changes with future changes of earnings per share is provided by the data panel studied. The functional estimation of the dividend annual increment per share, in all the companies of the sample, is obtained by implementing a functional version of Kalman filtering algorithm in an Autoregressive Hilbertian (ARH(1)) process framework