Predictability of Return in Pakistan Stock Market through the Application of the Threshold Quantile Autoregressive Models

Stock market behavior is a contentious matter among researchers in the field of finance. In literature, various conventional and behavioral explanations exist to the real-life stock market behavior. This study considered and incorporated all three schools of thought on the matter and applied a nonlinear model namely a threshold quantile autoregressive model as a contribution to exploring the behavior of the Pakistan stock market from 2000 to 2018. The findings of the study indicate that autocorrelation exists in the KSE 100 index and has a significant impact on both higher and lower regimes. The results also point out that the investors overreact and underreact in different states of the stock market. During the examination of the impact of stock characteristics and behavioral factors on the existence of stock market autocorrelation. It is concluded based on empirical evidence that these factors cause a significant impact on autocorrelation in the index. The study is of the view that behavioral biases are among the prime reasons for violation of efficient market behavior and need further exploration.