|
|
 |
| |
|
|
|
Seasonal Autoregressive Integrated Moving Average with Exogenous Variables Intervention Analysis: Application to the South African Tourism Industry |
|
|
|
PP: 507-523 |
|
|
|
doi:10.18576/jsap/140402
|
|
|
|
Author(s) |
|
|
|
Amos M. Mphanya,
Sarah L. Mahlangu,
Thabiso E. Masena,
Sandile C. Shongwe,
|
|
|
|
Abstract |
|
|
| Tourism revenue forms an important component of many countries’ gross domestic product such that a drop in tourists arrivals causes a significant increase in unemployment in certain sectors of the economy, e.g. hotels, bed and breakfast, etc. While several economic sectors have fully recovered from the recent COVID-19 pandemic, in this paper, we use the famed time series analysis’ Box-Jenkins methodology to illustrate that the South African tourism accommodation income has not fully recovered from its negative effect. Additionally, the seasonal autoregressive integrated moving average with exogenous variables (SARIMAX) intervention model with a pulse function covariate vector incorporated through trial-and-error was used to fully model and quantify the negative ramifications the pandemic on the South African tourism accommodation income dataset. Using March 2020 as the intervention point, the South African tourism sector experienced a loss of ZAR 99,009 million in revenue in the 52-months intervention period from March 2020 to June 2024. More importantly, at the end of the study period (June 2024), the tourism accommodation income series had not recovered to its pre-intervention levels. The 1-year out-of-sample forecasts from the best fitting SARIMAX(1,1,2) (0,1,1)_12 intervention model estimates that the tourism accommodation income series will not recover to its pre-COVID-19 intervention levels by June 2025 if rescuing efforts are not taken to boost income within the sector. Such efforts include providing resources such as capital financing, skills development and mentorship for aspiring entrepreneurs within the tourism industry. |
|
|
|
|
|
|
|
