An alternative method for forecasting price volatility by combining models. Issue 6 (3rd July 2017)

• Record Type: Journal Article
• Authors: Gurung, Bishal; Singh, K. N.; Paul, Ranjit Kumar; Panwar, Sanjeev; Gurung, Biwash; Lepcha, Lawrence
• Is Part Of: Communications in statistics. Volume 46:Issue 6(2017)
• Page Start: 4627
• Page End: 4636
• Publication Date: 2017-07-03
• Subjects: CDC -- Combined models -- GARCH -- Kalman filter -- Nonlinear time-series model -- SV -- Volatility; 97M10; Mathematical statistics -- Periodicals; Mathematical statistics -- Data processing -- Periodicals; Digital computer simulation -- Periodicals; 519.5
• DOI: 10.1080/03610918.2015.1124115
• Languages: English
• ISSNs: 0361-0918

Abstract

ABSTRACT: In this article, we study the volatility in the monthly price series of edible oils in domestic and international markets using the two popular family of nonlinear time-series models, viz, Generalized autoregressive conditional heteroscedastic (GARCH) models and Stochastic volatility (SV) models. To improve the forecasts of the volatility process, we also propose a new method of combining the volatility of these two competing models using the powerful technique of Kalman filter. The individual models as well as the combined models are assessed on their ability to predict the correct directional change (CDC) in future values as well as other goodness-of-fit statistics. Further, forecasting performance are also evaluated by computing various measures to validate the proposed methodology.