Start Date

8-1-2011 9:45 AM

End Date

8-1-2011 10:30 AM

Track

1. Track 1 – Formal Paper Presentation

Subject Area

Finance and Economics

Faculty Member

Dr. Tianshu Zheng tzheng@iastate.edu

Abstract

This study utilizes common time series forecasting methods to determine which of several simple, popular time series forecasting techniques was the best predictor of the decline in United States weekly RevPAR as the lodging industry entered its severe downturn in 2009. This study identifies the strong seasonality and trend components contained in historic U.S. weekly RevPAR data and utilized that data to test various moving average, exponential, and seasonal forecasting methods. The study clearly identified that seasonal forecasting methods such as Holt-Winters are far superior for use with this dataset than other methods and that among the various seasonal forecasting methods that multiplicative forecasting methods are somewhat superior to additive forecasting methods in working with this dataset.

Keywords

RevPAR, Time Series, Lodging Industry, Forecasting



Share

COinS
 
Jan 8th, 9:45 AM Jan 8th, 10:30 AM

Forecasting RevPAR in a Declining Market: An Application of Time Series Forecasting Techniques to U.S. Weekly RevPAR Data

This study utilizes common time series forecasting methods to determine which of several simple, popular time series forecasting techniques was the best predictor of the decline in United States weekly RevPAR as the lodging industry entered its severe downturn in 2009. This study identifies the strong seasonality and trend components contained in historic U.S. weekly RevPAR data and utilized that data to test various moving average, exponential, and seasonal forecasting methods. The study clearly identified that seasonal forecasting methods such as Holt-Winters are far superior for use with this dataset than other methods and that among the various seasonal forecasting methods that multiplicative forecasting methods are somewhat superior to additive forecasting methods in working with this dataset.