A Generalized Asymmetric Conditional Autoregressive Range Model

• Activity Number: 344 - Methods in Financial Econometrics
• Type: Contributed
• Date/Time: Wednesday, August 5, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Isuru Ratnayake* and V A A. Samaranayake
• Companies: Missouri University of Science and Technology and Missouri University of Science and Technology
• Keywords: Volatility Modeling; CARR Models; ACARR; Price Range; Time Series

Abstract

The Conditional Autoregressive Range (CARR) model is an alternative to the Generalized Autoregressive Conditionally Heteroscedastic (GARCH) formulation. The former models the price range while the latter focus on modeling the price returns. An Asymmetric CARR (ACARR) model was introduced to allow for separate modeling of upward and downward ranges. This formulation, however, ignores any feedback from one type of range to the other. The Feedback Asymmetric Conditional Autoregressive Range (FACARR) was introduced in 2017 to remedy this drawback. The FACARR, however, limits this cross feedback to past ranges and do not include past conditional means. The proposed Generalized Asymmetric Conditional Autoregressive Range Model (GFACARR) removes this limitation and allows the upward range model to include both past upward and downward ranges as well both past conditional means of upward and downward ranges, and a similar formulation when modeling the conditional mean of the downward range. The proposed model is more in line with the multivariate CARR model. The use of the GFACARR model is illustrated by its application to several price series, including the S&P 500.