Simpler and Better? Revisit of a Time-Series Model of Forecasting in the Lodging Industry

The average hotel manager recognizes the criticality of forecasting. However, most managers are either frustrated by complex models researchers constructed or appalled by the amount of time and efforts to master the nuances of statistical theories. As hotel competition intensifies, managers require an effective method to forecast financial performance (Chon & Singh, 1993). Luckily, lodging industry data has underlying patterns and trends that repeat themselves daily, weekly, monthly and yearly (Wheaton & Rossof, 1998). This provides an opportunity for a time-series forecasting model for revenue managers, considering time-series models’ intuitive nature of assuming that the history is useful in predicting the future, specifically, the value of a financial performance variable can be a function of itself and other variables from the past (Weatherford & Kimes, 2003; Banker & Chen, 2006; Schmidgall, 2006). Moreover, time-series forecasting models, if constructed properly, can be very easy and intuitive to use in the lodging context. Considering annual forecast of sales is critical for budgeting, revenue management and control purposes, this paper focuses on how to forecast annual sales one year ahead using an easily applicable time-series model in the lodging industry and provides evidence of the model forecasting accuracy. Findings of this paper are timely and extremely valuable, especially considering the need for lodging companies to accurately forecast future sales in a time of decreasing demand.