ARTICLE TEMPLATE Demand forecasting in supply chain: The impact of demand volatility in the presence of promotion ARTICLE HISTORY Compiled October 1, 2019 Mahdi Abolghasemi a , Richard Gerlach b , Garth Tarr c , Eric Beh a a School of Mathematical and Physical Sciences, The University of Newcastle, NSW, Australia; b The university of Sydney Business School, Sydney, NSW, Australia; c School of Mathematics and Statistics, The university of Sydney, Sydney, NSW, Australia ABSTRACT The demand for a particular product or service is typically associated with different uncertainties that can make them volatile and challenging to predict. Demand un- predictability is one of the managers’ concerns in the supply chain that can cause large forecasting errors, issues in the upstream supply chain and impose unnecessary costs. We investigate 843 real demand time series with different values of coefficient of variations (CoV) where promotion causes volatility over the entire demand se- ries. In such a case, forecasting demand for different CoV require different models to capture the underlying behavior of demand series and pose significant challenges due to very different and diverse demand behavior. We decompose demand into baseline and promotional demand and propose a hybrid model to forecast demand. Our results indicate that our proposed hybrid model generates robust and accurate forecast across series with different levels of volatilities. We stress the necessity of decomposition for volatile demand series. We also model demand series with a num- ber of well known statistical and machine learning (ML) models to investigate their performance empirically. We found that ARIMA with covariate (ARIMAX) works well to forecast volatile demand series, but exponential smoothing with covariate (ETSX) has a poor performance. Support vector regression (SVR) and dynamic lin- ear regression (DLR) models generate robust forecasts across different categories of demands with different CoV values. KEYWORDS Demand volatility; Promotions; Forecasting models; Robust forecasts. 1. Introduction Demand is one piece of the important information that can be shared and used in supply chain management (SCM). Demand sharing and demand forecasting are extremely helpful for supply chain managers since it provides a great source of information for planning and decision making. Demand forecasting is the basis for a lot of managerial decisions in the supply chain such as demand planning [Narayanan et al.], order fulfilment [Narayanan et al.], production planning [26], and inventory control [64]. It is usually difficult to carry out forecasting with a desired level of precision because of the volatility and varying uncertainties involved [41, 66]. Demand volatility inherently exists due to the consumers’ behavior that is constantly changing [72]. Various variables such as promotion, weather, market trends, and season may have an impact on consumers behavior and contribute to demand volatility [30]. Promotion, in particular, is a very common practice in the retailing industry that can make demand arXiv:1909.13084v1 [stat.AP] 28 Sep 2019