Time Series Modeling with Intervention Analysis to Evaluate of COVID-19 Impact on the Stock Markets in Indonesia and Global

The COVID-19 pandemic began in December 2019 and led to significant disruptions in global financial markets. This study investigates the impact of the pandemic on stock indices in Indonesia (IHSG), the United States (DJI), and South Korea (KOSPI) using intervention analysis with a step function, which is designed to model permanent shifts in time series data following external shocks. Unlike traditional models such as ARIMA that assume data continuity, intervention models, particularly those using step functions, are highly suitable for assessing long-term economic disruptions and structural breaks caused by pandemics. This research uses daily stock price index data from January 10, 2019, to May 8, 2020, obtained from Yahoo Finance. The step function identifies the point of sustained change triggered by the initial COVID-19 outbreak and subsequent market reactions. The analysis shows that the pandemic caused significant and persistent declines across all observed indices. IHSG recorded its sharpest drop on March 26, 2020, while DJI and KOSPI experienced similar downward trends from March to April 2020. The forecasting performance of the intervention model was excellent, with Mean Absolute Percentage Error (MAPE) values of 0.72% for IHSG, 0.87% for DJI, and 0.82% for KOSPI, demonstrating high accuracy in modeling stock market behavior during crisis conditions.