International Journal of Hospitality Management 30 (2011) 756–758 Contents lists available at ScienceDirect International Journal of Hospitality Management journal homepage: www.elsevier.com/locate/ijhosman Research note The determinants of hotel room rates: Another visit with Singapore’s data Chew Ging Lee ∗ Nottingham University Business School, The University of Nottingham Malaysia Campus, Jalan Broga, 43500 Semenyih, Selangor, Malaysia article info Keywords: Volatility clustering Hotel room rates Terrorist activities abstract In this research note, volatility clustering modeling framework is used to examine the determinants of hotel room rates in Singapore. Using monthly data from January 1985 to June 2009, GARCH-M(1, 1) is identified as the appropriate model used to capture volatility clustering. The results suggest that total inbound tourists and economic performance have positive effects on hotel room rates. The main findings are (a) the occurrences of terrorist activities in the neighboring countries have negative impacts and (b) the volatility of hotel room rates has a positive effect, on hotel room rates. © 2010 Elsevier Ltd. All rights reserved. 1. Introduction The main focus of current theoretical literature related to hotel room rates is on developing an optimal hotel room rate model (see Pan, 2007; Collins and Parsa, 2006; Steed and Gu, 2005). The empir- ical researches in this area look at the factors influencing the pricing of hotel rooms. Israeli (2002) uses the data of 215 hotels in nine locations in Israel to analyze the impacts of star rating system and corporate affiliation on hotel room prices in different seasons for 1999 and 2000. Using linear regression model, his results suggest that the variation in hotel room rates is explained by the star rating system. However, the impact of corporate affiliation on the pricing of hotels is inconsistent depending on the location of hotels. Using a sample of 58 international tourist hotels classified by Taiwan’s Tourism Bureau in 2006, Hung et al. (2010) investigate the deter- minants of hotel room rates with OLS and quantile regression. They suggest that due to the skewed distribution of hotel room rates, quantile regression provides a more complete characterization of the hotel pricing determinants for the higher-price and lower-price quantile hotels. Other studies have used the hedonic pricing model to identify the determinants of hotel room rates (see Espinet et al., 2003; Monty and Skidmore, 2003). Espinet et al. (2003) estimate the hedonic pricing model for hotels in the southern Costa Brava area of Spain with panel data models where the monthly average of daily prices from May to October and from 1991 to 1998 is used as the dependent variable. The study of Monty and Skidmore (2003) focuses on the pricing of rooms in bed and breakfast accommoda- tions in Walworth County of Wisconsin. Their models are estimated with OLS. ∗ Tel.: +60 3 8924 8259; fax: +60 3 8924 8019. E-mail address: lee.chew-ging@nottingham.edu.my. Although different estimation methods have been used by researchers to identify the determinants of hotel room rates, lim- ited to my knowledge, none has used time series related estimation method. Therefore, the purpose of this research note is to provide another empirical modeling framework which can be used to iden- tify the factors that influence the hotel room rates in high frequency time series data, such as monthly or weekly data. Volatility cluster- ing frequently appears in high frequency data. Volatility clustering occurs if the variance of a time series data is high for extended peri- ods and then low for extended periods. The series exhibits this form of volatility is said to face conditional heteroscedasticity. Therefore, the assumption of identical and independent distribution is not appropriate in the presence of volatility clustering. Such volatil- ity clustering can be modeled with the autoregressive conditional heteroscedasticity (ARCH) model pioneered by Engle (1982) and extended methods developed by others, such as Bollerslev (1986) and Nelson (1991). Secondly, this research note will investigate the effect of terror- ist attacks in the neighboring countries on the hotel room rates of a destination. To investigate this, a destination which is free from terrorist attacks has to be identified. After careful consideration, Singapore is selected because it is free from terrorist attack, but may be subject to such an attack in the future. Terrorist activities take place frequently in the neighboring countries of Singapore, such as the Philippines, Indonesia and Thailand. For instance, the 2002 Bali bombings which occurred on 12 October 2002 in Bali, Indone- sia, killed 202 people and injured 240 people; a series of bombings happened in the south of Thailand; and, bombings against civilians and civilian property occurred mainly in the southern regions of the Philippines. The balance of this research note is organized as follows. In Section 2, the model and the modeling framework are discussed. The obtained results are also presented. Section 3 concludes this research note. 0278-4319/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhm.2010.09.010