Volatility by means of Functional Data
- Alva Chavez, Kenedy Pedro
- Romo Urroz, Juan José
- Ruiz Ortega, Esther
Verlag: Universidad de Murcia. Departamento de Estadística e Investigación Operativa
ISBN: 978-84-691-8159-1
Datum der Publikation: 2009
Kongress: Congreso Nacional de Estadística e Investigación Operativa (31. 2009. Murcia)
Art: Konferenz-Beitrag
Zusammenfassung
We propose a method to predict the intra-day volatility for the whole day given past volatilities. We consider model diusion for the daily stock process to extract the volatility, means by functional principal components (FPC). The FPC scores are calculated by numerical integration and cross validation. The model to predict is an extension of the stochastic volatility model for one dimensional time series. This is a functional rst order autoregressive model for the volatility.