The study looked at how the volatility of the Bangladesh stock market changes over time using GARCH type models. They found that the market's volatility characteristics vary in different periods: before, during, and after a crisis. The models showed changes in risk premium and persistence of volatility. During the crisis, there was a significant "low-yield associated with high-risk" trend, and a "leverage effect" was present in all periods. The market is not as mature as developed markets, with investors making irrational decisions based on risk and return. The threshold GARCH model was the most accurate, and the GARCH model was the most efficient for forecasting.