ABSTRACT
Seasonal heteroskedasticity refers to variation in uncertainty or volatility that occurs in a seasonal pattern across calendar years. In this chapter, we examine and compare two different approaches to modeling seasonal
K12089 Chapter: 2 page: 37 date: February 14, 2012
K12089 Chapter: 2 page: 38 date: February 14, 2012
Modeling and
heteroskedasticity: the seasonal specific models introduced by Proietti (2004) and an extension of the airline model proposed by Bell (2004). We examine the use of likelihood ratio tests with the models to test for the presence of seasonal heteroskedasticity, and the use of model comparison statistics (AIC) to compare the models and to search among alternative patterns of seasonal heteroskedasticity. We apply the models and tests to the U.S. Census Bureau monthly time series of regional housing starts and building permits. For some of these series, there is a clear reason to expect seasonal heteroskedasticitythe variable effects of winter weather on the activity surrounding new construction.
