Abstract
Overnight forecasting is a crucial challenge for revenue managers because of the uncertainty associated between demand and supply. However, there is limited research that focuses on predicting daily hotel demand. Hence, this paper evaluates various models’ of traditional time series forecasting performances for daily demand at multiple horizons. The models include the seasonal naïve, Holt–Winters (HW) triple exponential smoothing, an autoregressive integrated moving average (ARIMA), a seasonal autoregressive integrated moving average (SARIMAX) with exogenous variables, multilayer perceptron (MLP) artificial neural networks model (ANNs), an sGARCH, and GJR-GARCH models. The dataset of this study contains daily demand observations from a hotel in a US metropolitan city from 2015 to 2019 and a set of exogenous social and environmental features such as temperature, holidays, and hotel competitive set ranking. Experimental results indicated that under the MAPE accuracy measure: (i) the SARIMAX model with external regressors outperformed the ANN-MLP model with similar external regressors and the other models, in every one horizon except one out of seven forecast horizons; (ii) the sGARCH(4, 2) and GJR-GARCH(4, 2) shows a superior predictive accuracy at all horizons. The results performance is evaluated by conducting pairwise comparisons between the different model’s distribution of forecasts using Diebold–Mariano and Harvey–Leybourne–Newbold tests. The results are significant for revenue managers because they provide valuable insights into the exogenous variables that impact accurate daily demand forecasting.
Highlights
Over the past several years, the rapid development of information technology has been instrumental in the growth of demand in the hospitality sector
Our results show that the multilayer perceptron neural network model, including exogenous variables, is not favorable compared to the seasonal ARIMAX (SARIMAX) model, including the same exogenous variables such as temperature, holidays, competitive set ranking, or the generalized ARCH (GARCH) models
The findings indicate that forecasting with the Holt–Winters Seasonal Exponential Smoothing including a trend component and a seasonal component state-space additive model and additive errors with a no Box–Cox transformation was the best model minimized the Akaike Information Criterion (AIC) statistic with parameterization results as α = 0.9590, β = 0.0046, γ = 0.0324, and AIC = 6178.4534
Summary
Over the past several years, the rapid development of information technology has been instrumental in the growth of demand in the hospitality sector. The new marketplace is becoming more competitive, which causes pricing pressure on the traditional service industries as the market supply increases With this increase in prominent destinations supply, and the number of new accommodation listings beyond the traditional purchasing options (e.g., sharing accommodation), new challenges on overnight demand related to forecasting have been created. These new business models brought dramatic changes to the sales processes. Forecasting demand as an essential function for the hospitality industry stakeholders has developed new interesting forecasting models To this point, accurate forecasting is a critical component that affects revenue maximization when selecting a specific model, the model choice within forecasting is crucial in itself. A reduction of the forecasting error might generate incremental revenue
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
