| Summary: | This paper attempts to explore the relationship between four macroeconomic variables on Bursa Stock Exchange using the Machine Learning method. Quarterly data has been used from 2015 quarter 1 until 2022 quarter 4 for all the variables like, overnight policy rate, industrial production index, consumer sentiment index, and unemployment rate. Then, the machine learning method, SHapley Additive Explanation (SHAP) was used to calculate the impact value between stock price and macroeconomic variables. After calculating the impact value, technical indicators were used to test the contribution of macroeconomic variables across the 13 sectors.
Results showed all variables evolve differently in the different phases of the economic cycle. The value of the impact of each sector on the index of industrial production is different. Some sectors show a positive impact value on the overnight policy rate, while the rest of the sectors show the opposite. During the strong economy, almost all sectors show divergence. However, during the pandemic, most of the sectors have had an almost neutral or positive impact on the unemployment rate. Whereas, the consumer sentiment index has an almost neutral impact value on all sectors.
|
|---|