|
Scenario-based forecasting of beverage sales in Vietnam’s F&B sector using ARIMA and exploratory factor analysis |
|---|---|
| รหัสดีโอไอ | |
| Title | Scenario-based forecasting of beverage sales in Vietnam’s F&B sector using ARIMA and exploratory factor analysis |
| Creator | Thi Ngoc Nhung Nguyen |
| Contributor | Sarinpas Kornwattanakit, Advisor |
| Publisher | Thammasat University |
| Publication Year | 2568 |
| Keyword | Sales forecasting, ARIMA, Exploratory factor analysis (EFA), Product strategies, Events strategies, Advertisement, Scenario-based forecasting |
| Abstract | This study investigates sales forecasting in Vietnam’s Food and Beverage (F&B) industry by integrating traditional time-series forecasting with a scenario-based adjustment framework that reflects the influence of multiple internal and external factors. Monthly revenue data from five representative companies—VINAMILK, Chuong Duong Beverage JSC, SABECO, HABECO, and Ha Long Beer—were collected to capture long-term demand patterns, seasonal behavior, and forecast uncertainty.Prior to forecasting, stationarity testing was conducted on all revenue time series to ensure their suitability for ARIMA modeling, with appropriate differencing applied where necessary. Exploratory Factor Analysis (EFA) was then employed to reduce thirteen influencing variables into five statistically coherent factor groups: product-related, promotion-related, external-related, internal-related, and consumer-related factors. These factor groups provide a structured representation of the key drivers affecting sales performance.Baseline forecasts were generated using ARIMA models estimated in IBM SPSS Statistics, capturing historical trends and seasonal patterns without incorporating strategic or environmental shocks. Scenario-based forecasts were subsequently constructed by adjusting the ARIMA baseline according to the average influence level of each EFA-derived factor group. In addition, the study examined the isolated effects of three key strategic drivers—Product Strategies, Events Strategies, and Advertisement—to evaluate their directional and magnitude-specific impacts on future revenue projections.The empirical results indicate that the influence of factor groups varies considerably across firms and product categories. Promotion-related and consumer-related factors often generate strong positive effects, particularly through marketing activities, events, and shifts in consumer sentiment. External-related factors, associated with macroeconomic conditions, market volatility, and unexpected disruptions, tend to exert negative or destabilizing impacts, especially for beer products. Product-related and internal-related factors generally contribute smaller but more stable effects, reflecting incremental improvements in product attributes and operational conditions.Overall, the findings demonstrate that reliance on ARIMA forecasting alone is insufficient to capture the complex demand dynamics of Vietnam’s F&B industry. By combining EFA-based factor grouping with scenario-adjusted time-series forecasting, this study proposes a more adaptive and realistic forecasting framework that supports strategic planning, demand management, and risk mitigation in a volatile and competitive market environment. |
